문제 정의: 기존의 파력(Wave Energy) 변환 장치는 경제적으로 경쟁력이 부족하며, 건설 및 유지보수 비용이 높음.
목표: 기존 방파제(Breakwater) 구조를 활용하여 파력 에너지를 효율적으로 수집할 수 있는 OBREC(Overtopping Breakwater for Energy Conversion) 장치의 성능을 분석.
접근법: FLOW-3D 기반 CFD(Computational Fluid Dynamics) 시뮬레이션을 통해 실험 데이터를 검증하고, 파도 월류량(overtopping discharge)을 예측.
연구 방법
OBREC 개요 및 기존 연구
OBREC는 전통적인 방파제에 저수조(Reservoir) 를 결합하여 월류하는 파도를 저장하고, 낮은 수두(low-head) 터빈을 통해 전력을 생산하는 개념.
2012~2014년 Aalborg 대학에서 실험을 수행하여 가능성을 입증.
수치 모델링
FLOW-3D를 활용하여 RANS(Reynolds-Averaged Navier-Stokes) 방정식 및 VOF(Volume of Fluid) 기법을 사용한 자유 표면 계산 수행.
기존 실험 데이터를 바탕으로 JONSWAP 스펙트럼을 적용한 파도 환경을 구성.
격자 수렴(Mesh Convergence) 분석
7가지 메쉬 크기 비교 → 연산 비용과 정확도의 균형을 고려하여 최적의 메쉬 크기(0.005m)를 선정.
주요 결과
수치 시뮬레이션 vs 실험 데이터 비교
월류량(overtopping discharge)에 대한 시뮬레이션 결과가 실험값과 높은 일치도를 보임.
단, 수치 모델이 부드러운 방파제 표면을 가정하여 실험보다 다소 높은 월류량 예측.
수치 시뮬레이션 vs 이론 공식 비교
기존 연구(Vicinanza, 2014)에서 제안한 월류량 예측 공식과 비교 → 유사한 경향성을 보이며 검증됨.
저수조 크기(Rr)가 증가할수록 월류량이 감소하는 경향 확인.
다른 연구와의 비교
Kofoed(2002), EurOtop(2007), Van der Meer(1998) 등의 기존 월류 모델과 비교하여 일관된 결과 도출.
통계 분석 결과, 실험 대비 수치 시뮬레이션의 월류량 예측 오차는 약 6% 이내로 양호한 성능을 보임.
결론 및 향후 연구
FLOW-3D 기반 CFD 시뮬레이션이 OBREC의 초기 설계 검토에 효과적임을 입증.
실험 대비 비용이 낮고 신속한 예측이 가능, 초기 설계 최적화에 유용함.
향후 연구에서는 방파제 표면 거칠기 및 다공성(Porosity) 요소 추가 등을 통해 더욱 정밀한 모델 개선 필요.
연구의 의의
이 연구는 기존의 실험적 접근법을 CFD 시뮬레이션으로 보완하여, OBREC와 같은 파력 에너지 변환 시스템의 설계 최적화 및 경제성 향상을 위한 새로운 방향을 제시했다는 점에서 의의가 있다.
Reference
D.Vicinanza, P. Contestabile, J. Quvang Harck Nørgaard, and T. Lykke Andersen, “Innovative RubbleMound Breakwaters for Overtopping Wave Energy Conversion,” Journal of Coastal Engineering, Vol. 88,pp. 154–170, 2014.
Muzathik.A,Wan Nik WB, Sulaiman O, Rosliza R, Prawoto Y, “Wave Energy Resource Assessment andReview of the Technologies,” International Journal of Energy and Environment, Vol. 2, no. 6, pp. 1101–1112, 2011.
S.Ding, D.Han, and Y.Zan, “The Application of Wave Energy Converter in Hybrid Energy System,” TheOpen Mechanical Engineering Journal , Vol 8. pp. 936–940, 2014.
D.Vicinanza, J.Harck Nørgaard, P.Contestabile, and T.Lykke Andersen, “Wave Loadings Acting onOvertopping Breakwater for Energy Conversion,” Journal of Coastal Research., Vol. 29, no. 65, pp. 1669–1674, 2013.
D.Vicinanza;, “Innovative Beakwaters Design for Wave Energy Conversion B,” in Coastal EngineeringProceeding, 2012, pp. 12–24.
D.Vicinanza, P.Contestabile, C.Luppa, L.Cavallaro, E.Foti, “Innovative Rubble Mound Breakwaters forWave Energy Conversion,” Journal of Special Conversion System, pp. 86–95, 2015.
P.Contestabile, V.Ferrante, E.Di Lauro, and D.Vicinanza, “Prototype Overtopping Breakwater for WaveEnergy Conversion at Port of Naples,” in Proceedings of the Twenty-sixth (2016) International Ocean andPolar Engineering Conference, 2016, no. June 26 to July 1, pp. 616–621.
N.Olsen, “Computational Fluid Dynamics in Hydraulic and Sedimentation Engineering,” NorwegianUniversity of Science and Technology, 1999.
J. P. Kofoed, “Model Testing of the Wave Energy Converter Seawave Slot-Cone Generator”,CivilEngineering Department, Aalborg University, 2005.
Y. Chen, “Numerical Modeling on Internal Solitary Wave propagation over an Obstacle using FLOW 3D,Master Thesis” National Sun Yat-sen University, 2012.
P.Prinos, M.Tsakiri, O.Souliotis, “ Numerical Simulation of the Wos and the Wave Propagation Along aCoastal Dike,” Coastal Engineering Proceeding, 2012.
T.W. Hsu, J.W. Lai, and Y.J. Lan, “Experimental and Numerical Studies on Wave Propagation Over CoarseGrained Sloping Beach,” Coastal Engineering Proceeding, 2011.
B.H.Choi, E.Pelinovsky, D.C.Kim, I.Didenkulova, and S.B.Woo, “Two and Three DimensionalComputation of Solitary Wave Runup on Non-Plane Beach,” Journal of Nonlinear Process in Geophysic,Vol. 15, no. 2006, pp. 489–502, 2008.
F. Dentale, G.Donnarumma, E.P.Carratelli, V.Giovanni, P.Ii, and F.Sa, “A New Numerical Approach to theStudy of the Interaction between Wave Motion and Roubble Mound Breakwaters,” Latest Trends in Engineering, Mechanics,Structures and Engineering Geology, pp. 45–52, 2014.
B. W. Nam, S. H. Shin, K. Y. Hong, and S. W. Hong, “Numerical Simulation of Wave Flow over the SpiralReef Overtopping Device,” in Proceedings of the Eighth (2008) ISOPE Pacific/Asia Offshore MechanicsSymposium, 2008, pp. 262–267.
Y. H. Yu and Y. Li, “Reynolds-Averaged Navier-Stokes Simulation of the Heave Performance of a TwoBody Floating-Point Absorber Wave Energy System,” Journal of Computer Fluids, vol. 73, pp. 104–114,2013.
D.Vicinanza, “Structural Response of Seawave Slot-cone Generator ( SSG ) from Random Wave CFDSimulations,” Proceedings of the Twenty-fifth (2015) International Ocean and Polar EngineeringConference no. June, 2015,pp.985-991.
J. P. Kofoed, “Wave Overtopping of Marine Structures: Utilization of Wave Energy,” PhD Thesis, AalborgUniversity, 2002.
Van Der Meer and C.J.M.Stam, “Wave Runup on Smooth and Rock Slope,” Deflt Hydraulic, 1991.
D.Vicinanza, “Structural Response of Seawave Slot-cone Generator ( SSG ) from Random Wave CFDSimulations,” Proceedings of the Twenty-fifth (2015) International Ocean and Polar EngineeringConferenceno. June, 2015,pp.985-991
J.W.Van Der Meer, “Wave Run-Up and Overtopping,” in Chapter 8 In: Pilarczyk, K.W. (Ed.), Seawalls,Dikes and Revetments., Balkema, Rotterdam, 1998.
EurOtop, Wave Overtopping of Sea Defences and Related Structures: Assessment Manual. 2007.
L.Victor, J.W.Van Der Meer, and P.Troch, “Probability Distribution of Individual Wave OvertoppingVolumes for Smooth Impermeable Steep Slopes with Low Crest Freeboards,” Journal of CoastalEngineering, Vol. 64, pp. 87–101, 2012.
해상풍력 발전 기초는 파랑 조건에 의해 주변 유동이 크게 교란되어 세굴(Scour) 현상이 발생할 수 있다.
기초의 안정성 확보를 위해 세굴 현상을 정확하게 예측하는 것이 필수적이다.
본 연구는 Flow-3D를 활용하여 해상풍력기초(모노파일 및 삼각대 파일) 주변의 세굴현상을 수치해석하였다.
연구 방법
모형 설정 및 입력 조건
해상풍력기초 형상: 직경이 다른 모노파일(예: D = 5.0 m, d = 1.69 m)과 동일한 직경의 모노파일, 그리고 삼각대 파일 형식을 대상으로 분석.
경계조건: 상류경계에 관측 유속(약 1.066 m/s) 및 극치파랑 조건을 적용하여 세굴현상을 평가.
난류 모형: LES 모형과 RNG 모형을 각각 적용하여 세굴 깊이 및 분포에 미치는 영향을 비교.
수치 해석 기법
Flow-3D 모형을 이용하여 3차원 유동해석을 수행.
FAVOR 기법과 VOF(Volume of Fluid) 방법을 사용해 복잡한 경계와 자유 표면을 정확히 재현.
메쉬 독립성 및 민감도 분석을 통해 계산의 신뢰성을 확보함.
주요 결과
모노파일 분석:
서로 다른 직경의 모노파일에서는 최대 세굴심이 약 4.13 m로 나타났으며, 동일한 직경의 모노파일에서는 하강류가 증가하여 최대 세굴심이 약 7.13 m로 증가함.
이는 동일 직경 모노파일에서 유속이 더욱 빨라지며, 세굴 현상이 심화됨을 시사함.
삼각대 파일 분석:
상류 경계조건으로 관측 유속과 극치파랑 조건을 각각 적용한 결과, 극치파랑 조건에서는 최대 세굴심이 약 1.3배 정도 더 깊게 발생함.
난류 모형 비교:
LES 모형을 적용한 경우, 세굴심이 일정 시간이 경과하면 평형상태에 도달함.
반면, RNG 모형은 전체 해석 영역에서 계속해서 세굴현상이 발생하여 평형상태에 도달하지 않음.
따라서 해상풍력기초 세굴 해석에는 LES 모형과 극치파랑 조건의 적용이 타당함.
결론 및 향후 연구
해상풍력기초에 대한 세굴현상 분석에서는 동일 직경 모노파일보다 서로 다른 직경의 파일 형식이 기초 안정성 측면에서 유리할 수 있음.
LES 난류 모형과 극치파랑 조건을 적용하는 것이 실제 세굴현상을 더 정확하게 예측할 수 있음을 확인함.
향후 연구에서는 다양한 해양 파랑 조건 및 추가 난류 모형 비교를 통해 보다 정밀한 세굴예측 모델을 개발할 필요가 있음.
Reference
W. J. Yang, S. U. Choi. “Three- Dimensional NumericalSimulation of Local Scour around the Bridge Pier usingLarge Eddy Simulation”, Journal of KWRA, vol. 22, no.4-B, pp. 437-446, 2002.
C. G. Yeo, J. E. Lee, S. O. Lee, J. W. Song. “SensitivityAnalysis of Sediment Scour Model in Flow-3D”,Proceedings of KWRA, pp. 1750-1754, 2010.
M. H. Oh, O. S. Kwon, W. M. Jeong, K. S. Lee.“FLOW-3D Analysis on Scouring around Offshore WindFoundation”, Journal of KAIS, vol. 13, no. 3, pp.1346-1351, 2012.DOI: http://dx.doi.org/10.5762/KAIS.2012.13.3.1346
A. Stahlmann, T. Schlurmann, “Physical Modeling ofScour around Tripod Foundation Structures for OffshoreWind Energy Converters”, Proceedings of 32ndConference on Coastal Engineering, Shanghai, China,no. 32, pp. 1-12, 2010.
Flow Science. Flow-3D User’s Manual. Los Alamos,NM, USA, 2016.
KEPRI. 『Test Bed for 2.5GW Offshore Wind Farm atYellow Sea』 Interim Design Report(in Korea), 2014.
Germanischer Lloyd. Guideline for the Certification ofOffshore Wind Turbines. Hamburg, Germany, 2005.
B. M. Sumer, J. Fredsøe, The Mechanics of Scour in theMarine Environment. World Scientific Publishing Co.Pte. Ltd. 2002.
S. J. Ahn, U. Y. Kim, J. K. Lee. “Experimental Studyfor Scour Protection around Bridge Pier by Falling-FlowInterruption”, Journal of KSCE, vol. 19, no. II-1, pp.57-65, 1999.
V. Yakhot, S. A. Orszag, S. Thangam, T. B. Gatski, C.G. Speziale, “Development of turbulence models forshear flows by a double expansion technique”, Physicsof Fluids, vol. 4, no. 7, pp. 1510-1520, 1992.DOI: https://doi.org/10.1063/1.858424
문제 정의: 배관 부품 제조 공정에서 중력 모래 주형을 이용한 주조는 용융 금속의 복잡한 열 전달 및 응고 과정으로 인해 결함(예: 기공, 수축 결함)이 발생할 수 있어 생산 효율과 제품 품질에 영향을 준다.
목표: CFD 기법(특히 FLOW 3D CAST v5.03)을 활용하여 실제 생산 라인과 동일한 주형 및 내부 챔버 형상을 기반으로 용융 금속의 충진, 응고 및 냉각 단계를 해석하고, 다양한 타설 온도와 러너 설계가 주조 결함에 미치는 영향을 평가하는 데 있다.
연구 방법
CFD 시뮬레이션
프로그램 및 기법: FLOW 3D CAST v5.03 사용, Volume of Fluid (VOF) 방법을 통해 용융 금속의 자유 수면을 추적.
난류 모델: 두 방정식 k–ε 모델을 채택하여 난류 효과를 반영.
모델 형상: 실제 생산 라인의 주형과 내부 챔버 형상을 그대로 반영.
주요 변수 및 조건
타설 온도: 다양한 타설 온도(예: 1329°C, 1529°C)를 적용하여 유동 속도, 응고 시간 및 결함 발생에 미치는 영향 평가.
러너 설계: 러너의 크기와 수가 용융 금속의 흐름 및 결함 위치에 어떤 영향을 미치는지 분석.
메쉬 독립성 및 시간 단계
여러 메쉬 크기를 비교하여 계산 정확도와 효율성을 확보함(예: 250,000 요소 사용).
주요 결과
충진 및 응고 해석: CFD 시뮬레이션을 통해 용융 금속이 주형 내에서 충진되는 과정과 이후 응고 및 냉각 단계가 상세하게 재현되었음.
타설 온도의 영향:
높은 타설 온도(1529°C)는 용융 금속의 유동을 빠르게 하며, 반면 응고에는 더 긴 시간이 소요됨.
낮은 타설 온도(1329°C)에서는 유동 속도가 다소 느리고, 응고 과정이 상대적으로 빠르게 진행됨.
러너 설계의 효과: 다양한 러너 각도 및 구조 변경 시도에도 불구하고, 현재 연구에서는 러너 설계가 기공 결함(캐비티) 감소에 큰 영향을 미치지 않음.
전체 공정 소요 시간: 충진, 응고, 냉각 단계 각각의 소요 시간이 계산되어 생산 공정 개선에 활용 가능함.
결론 및 향후 연구
CFD 기법은 중력 모래 주형을 이용한 배관 부품 주조 공정에서 용융 금속의 충진, 응고 및 냉각 단계를 효과적으로 해석할 수 있음을 보여준다.
타설 온도가 용융 금속 유동 및 응고 거동에 결정적인 영향을 미치며, 이로 인해 주조 결함 발생이 달라짐을 확인하였다.
향후 연구에서는 시뮬레이션 결과와 실험 데이터를 비교 검증하고, 결함 발생 원인 및 위치에 대한 추가 분석을 통해 생산 공정의 최적화를 도모할 예정이다.
Reference
Deev V.B., Ponomareva K.V., Prikhodko O.G.,and Smetanyuk S.V., Influence of temperaturesof melt overheating and pouring on the quality ofaluminum alloy lost foam castings. RussianJournal of Non-Ferrous Metals, 2017. 58(4): p.373-377.
Ayar M.S., Ayar V.S., and George P.M.,Simulation and experimental validation fordefect reduction in geometry varied aluminumplates cast using sand casting. Materials Today:Proceedings, 2020. 27: p. 1422-1430.
Borikar G.P., and Chavan S.T., Optimization ofCasting Yield in Multi-cavity Sand Moulds ofAl-alloy Components. Materials Today:Proceedings, 2020. 28: p. 819-824.
Zhao H.-D., Bai Y.-F., Ouyang X.-X., and DongP.-Y., Simulation of mold filling and predictionof gas entrapment on practical high pressure diecastings. Transactions of Nonferrous MetalsSociety of China, 2010. 20(11): p. 2064-2070.
Li Y., Liu J., Zhong G., Huang W., and Zou R.,Analysis of a diesel engine cylinder head failurecaused by casting porosity defects. EngineeringFailure Analysis, 2021. 127.
Dou K., Lordan E., Zhang Y., Jacot A., and FanZ., A novel approach to optimize mechanicalproperties for aluminum alloy in High pressuredie casting (HPDC) process combiningexperiment and modeling. Journal of MaterialsProcessing Technology, 2021. 296.
Pulisheru K.S. and Birru A.K., Effect of pouringtemperature on hot tearing susceptibility of AlCu cast Alloy: Casting simulation. MaterialsToday: Proceedings, 2021.
Pulivarti S.R. and Birru A.K., Effect of MouldCoatings and Pouring Temperature on theFluidity of Different Thin Cross-Sections ofA206 Alloy by Sand Casting. Transactions of theIndian Institute of Metals, 2018. 71(7): p. 1735-1745.
Li Y., Liu J., Zhang Q., and Huang W., Castingdefects and microstructure distributioncharacteristics of the aluminum alloy cylinderhead with complex structure. Materials TodayCommunications, 2021. 27.
Sarkar C., Gawande S.H., and Keste A.A.,Design optimization of precision casting forresidual stress reduction. Journal ofComputational Design and Engineering, 2016.3(2): p. 140-150.
Gopalan R. and Prabhu N.K., Oxide bifilms inaluminum alloy castings – a review. MaterialsScience and Technology, 2013. 27(12): p. 1757-1769.
Luo L., Xia H.-Y., Luo L.-S., Su Y.-Q., Cai C.-J., Wang L., Guo J.-J., and Fu H.-Z., Eliminatingshrinkage defects and improving the mechanicalperformance of large thin-walled ZL205A alloycastings by coupling traveling magnetic fieldswith sequential solidification. Transactions ofNonferrous Metals Society of China, 2021.31(4): p. 865-877.
Nandagopal M., Sivakumar K., andSengottuvelan M., Process parameteroptimization to reduce cold metal defect inferrous casting using Taguchi technique andregression analysis. Materials Today:Proceedings, 2021. 45: p. 7917-7921.
Yang W., Luo Z., Zou Z., Zhao C., and You Y.,Modelling and analysis of bubble entrapment bysolidification shell in steel continuous castingconsidering bubble interaction with a coupledCFD-DBM approach. Powder Technology,2021. 390: p. 387-400.
Loon W.L., Reddy S.S., and A.K. R.P., CFDsimulation of direct chill casting process ofmagnesium alloy billets. Journal ofManufacturing Processes, 2019. 45: p. 447-454.
YIN J., Numerical modeling of the centrifugalcasting process. 2016, KTH ROYALINSTITUTE OF TECHNOLOGY.
Reicher A., Numerical Analysis of Die-CastingProcess in Thin Cavities Using LubricationApproximation. 2012, University of WisconsinMilwaukee.
Yuwen X.-X., Chen L., and Han Y.-J.,Numerical Simulation of Casting Filling ProcessBased on FLUENT. Energy Procedia, 2012. 17:p. 1864-1871.
Kermanpur A., Mahmoudi S., and Hajipour A.,Numerical simulation of metal flow andsolidification in the multi-cavity casting molds ofautomotive components. Journal of MaterialsProcessing Technology, 2008. 206(1-3): p. 62-68.
Vijayaram T.R., Sulaiman S., Hamouda A.M.S.,and Ahmad M.H.M., Numerical simulation ofcasting solidification in permanent metallicmolds. Journal of Materials ProcessingTechnology, 2006. 178(1-3): p. 29-33.
Reikher A. and Pillai K.M., A fast simulation oftransient metal flow and solidification in anarrow channel. Part I: Model developmentusing lubrication approximation. InternationalJournal of Heat and Mass Transfer, 2013. 60: p.797-805.
Reikher A. and Pillai K.M., A fast simulation oftransient metal flow and solidification in anarrow channel. Part II. Model validation andparametric study. International Journal of Heatand Mass Transfer, 2013. 60: p. 806-815.
Wang H., Djambazov G., Pericleous K.A.,Harding R.A., and Wickins M., Modelling thedynamics of the tilt-casting process and the effectof the mould design on the casting quality.Computers & Fluids, 2011. 42(1): p. 92-101.
Aravind S., Ragupathi P., and Vignesh G.,Numerical and experimental approach toeliminate defects in al alloy pump- crankcaseprocessed through gravity die casting route.Materials Today: Proceedings, 2021. 37: p.1772-1777.
Barot R.P. and Ayar V.S., Casting simulation anddefect identification of geometry varied plateswith experimental validation. Materials Today:Proceedings, 2020. 26: p. 2754-2762.
Promtong M., Kasemjirapatara C., Srithep P.,Masoodi Y., Namchanthra S., Priyadumkol J.,and Suvanjumrat C., Investigation ofAerodynamic Performance of Four PotentialAirfoils for a Formula SAE Car: A 2D ValidationStudy, in The 34th Conference of the MechanicalEngineering Network of Thailand. 2020: 15 – 17July 2020, Prachuap Khiri Khan, Thailand.
Promtong M., Khunsri K., TeachapanitvittayaK., Trakulkumlue T., Watechagit S., andSuvanjumrat C., Experimental and NumericalInvestigations into the Natural Convection of HotGas in a Vertical Smoking Oven: A ValidationStudy, in The 34th Conference of the MechanicalEngineering Network of Thailand. 2020: 15 – 17July 2020, Prachuap Khiri Khan, Thailand.
Promtong M., Cheung S., Yeoh G., Vahaji S.,and Tu J., CFD investigation of sub-cooledboiling flow using a mechanistic wall heatpartitioning approach with Wet-Steamproperties. The Journal of ComputationalMultiphase Flows, 2018. 10(4): p. 239-258.
Promtong M., Cheung S., and Tu J., Numericalmodelling of subcooled boiling flow based onmechanistic approach: A validation study usingwet steam (IAPWS) as working fluid properties.Lecture notes in engineering and computerscience, 2016.
Promtong M. and Tekasakul P., CFD study offlow in natural rubber smoking-room: I.Validation with the present smoking room.Applied Thermal Engineering, 2007. 27(11-12):p. 2113-2121.
Hirt C.W. and Nichols B.D., Volume of Fluid(VOF) method for the dynamics of freeboundaries. Journal of Computational Physics,1981. 39(1): p. 201-225.
Aniszewski W., Ménard T., and Marek M.,Volume of Fluid (VOF) type advection methodsin two-phase flow: A comparative study.Computers & Fluids, 2014. 97: p. 52-73.
문제 정의: 데이터 집약적 애플리케이션의 증가로 인메모리 컴퓨팅에 대한 관심이 증대되었으며, 전통적인 2D 크로스바 설계는 저항 및 커패시턴스 기생 요소로 인해 성능 한계에 직면하고 있다.
목표: Boolean 함수를 3D 나노 크로스바 설계로 자동 합성하는 첫 번째 프레임워크인 FLOW-3D를 제안하여, 반둘레(semiperimeter)를 최소화하고, 면적, 에너지 소비, 지연 시간 등의 측면에서 기존 2D 도구보다 우수한 성능을 달성하는 것이 목적이다.
연구 방법
기본 아이디어 및 문제 정의
Boolean 함수의 합성을 위해 BDD(Binary Decision Diagram)와 3D 크로스바 사이의 유사성을 활용.
BDD의 노드와 에지에 해당하는 3D 크로스바의 금속 와이어와 멤리스터를 적절히 매핑하는 문제를 “L-labeling 문제”로 정의하고, 이를 ILP(정수 선형 계획법)로 최적 해결한다.
L-labeling 단계: 각 노드에 대해 할당 가능한 금속 층의 범위를 결정하고, 인접 층 간의 연결 제약(에지 제약 및 노드 제약)을 만족하도록 레이블링 수행.
크로스바 할당: 레이블링 결과를 바탕으로 실제 3D 크로스바 구조를 구성하여 Boolean 함수를 구현하는 하드웨어 디자인을 도출.
성능 평가
제안된 FLOW-3D 프레임워크는 2D 크로스바 기반의 기존 합성 도구와 비교하여, 반둘레, 면적, 에너지 소비, 지연 시간에서 각각 최대 61%, 84%, 37%, 41%의 개선 효과를 보임.
RevLib 벤치마크를 통해 실험적으로 평가되었으며, 3D 크로스바 설계의 효율성과 성능 향상을 입증하였다.
주요 결과
자동 합성 도구 제안: Boolean 함수를 3D 크로스바 설계로 자동 합성하는 최초의 프레임워크를 제안.
최적화 성능: FLOW-3D는 ILP 기반 L-labeling 문제 해결을 통해 3D 크로스바의 반둘레를 최소화하고, 면적 및 전력 소비를 현저히 감소시킴.
비교 평가: 기존 2D 기반 합성 도구 대비, 제안된 프레임워크는 에너지 효율과 응답 속도 면에서 우수한 성능을 나타냄.
결론 및 향후 연구
제안된 FLOW-3D 프레임워크는 3D 나노 크로스바를 이용한 흐름 기반 컴퓨팅에서 Boolean 함수 합성을 효율적으로 수행할 수 있음을 입증.
향후 연구에서는 더 복잡한 회로 및 대규모 데이터셋에 대한 확장성과, 다양한 하드웨어 제약 조건을 고려한 추가 최적화 기법이 연구될 필요가 있다.
Reference
John Backus. 1978. Can programming be liberated from the von Neumann style?CACM 21, 8 (1978), 613–641.
ABC Berkeley. 2009. A system for sequential synthesis and verification.
Debjyoti Bhattacharjee et al. 2020. CONTRA: area-constrained technology mapping framework for memristive memory processing unit. In ICCAD’20. 1–9.
Dwaipayan Chakraborty et al. 2017. Automated synthesis of compact crossbarsfor sneak-path based in-memory computing. In DATE’17. IEEE, 770–775.
IBM ILOG Cplex. 2020. IBM ILOG CPLEX Optimization Studio 20.1. InternationalBusiness Machines Corporation (2020). https://www.ibm.com/products/ilog-cplexoptimization-studio
Hadi Esmaeilzadeh et al. 2011. Dark silicon and the end of multicore scaling. InISCA’11. IEEE, 365–376.
Miao Hu et al. 2018. Memristor-based analog computation and neural networkclassification with a dot product engine. Advanced Materials 30, 9 (2018), 1705914.
Sumit Jha et al. 2016. Computation of boolean formulas using sneak paths incrossbar computing. US Patent 9,319,047.
Shahar Kvatinsky et al. 2013. Memristor-based material implication (IMPLY)logic: Design principles and methodologies. IEEE VLSI 22, 10 (2013), 2054–2066.
Shahar Kvatinsky et al. 2014. MAGIC—Memristor-aided logic. IEEE TCAS-II 61,11 (2014), 895–899.
Shuangchen Li et al. 2016. Pinatubo: A processing-in-memory architecture forbulk bitwise operations in emerging non-volatile memories. In DAC. IEEE, 1–6.
Yibo Li et al. 2018. Review of memristor devices in neuromorphic computing:materials sciences and device challenges. J. Phys. D: Appl. Phys 51, 50 (2018),503002.
Jiale Liang, Stanley Yeh, S Simon Wong, and H-S Philip Wong. 2013. Effect ofwordline/bitline scaling on the performance, energy consumption, and reliabilityof cross-point memory array. ACM JETC 9, 1 (2013), 1–14.
Thomas N Theis and H-S Philip Wong. 2017. The end of moore’s law: A newbeginning for information technology. CiSE 19, 2 (2017), 41–50.
Sven Thijssen et al. 2021. COMPACT: Flow-Based Computing on NanoscaleCrossbars with Minimal Semiperimeter. In DATE’21. IEEE, 232–237.
Alvaro Velasquez and Sumit Jha. 2014. Parallel computing using memristivecrossbar networks: Nullifying the processor-memory bottleneck. In IDT’14. IEEE,147–152.
Alvaro Velasquez and Sumit Jha. 2015. Automated synthesis of crossbars fornanoscale computing using formal methods. In NANOARCH’15. IEEE, 130–136.
Alvaro Velasquez and Sumit Jha. 2016. Parallel boolean matrix multiplication inlinear time using rectifying memristors. In ISCAS’16. IEEE, 1874–1877.
Alvaro Velasquez and Sumit Jha. 2018. Brief announcement: Parallel transitiveclosure within 3d crosspoint memory. In SPAA’18. 95–98.
Robert Wille et al. 2008. RevLib: An online resource for reversible functions andreversible circuits. In ISMVL’08. IEEE, 220–225.
Cong Xu et al. 2015. Overcoming the challenges of crossbar resistive memoryarchitectures. In HPCA’15. IEEE, 476–488.
Culvert Outlet Scouring의 영향 매개변수 예측 및 최적화: FLOW-3D와 서로게이트 모델링을 활용한 연구
연구 배경
문제 정의: 박스형 수로(culvert) 출구에서 발생하는 침식(scouring)은 구조물 설계에 중요한 영향을 미친다.
목표: 침식 깊이와 위치를 예측하여 구조적 실패를 방지하고, 설계를 최적화하는 새로운 방법론을 제안한다.
접근법: 수치 모델링(FLOW-3D)과 Box-Behnken 설계 기법을 이용한 서로게이트 모델링을 결합.
연구 방법
FLOW-3D:
Reynolds 평균 Navier-Stokes 방정식을 기반으로 유체 흐름 시뮬레이션을 수행.
침식 예측을 위해 RNG 난류 모델을 사용.
Box-Behnken 설계:
세 가지 주요 변수: 유량(Flow Discharge, QQQ), 수로 기울기(Slope, SSS), 토양 입자 크기(d50d_{50}d50).
총 15개 모델을 통해 변수와 침식 깊이 및 위치 간 상호작용 분석.
민감도 분석:
각 변수의 변화가 결과(침식 깊이와 위치)에 미치는 영향을 정량화.
최적화:
침식 깊이 및 위치를 최소화하거나 최대화하기 위한 설계 변수의 조합 도출.
주요 결과
모델 성능:
침식 깊이 예측 정확도: R2=0.931R^2 = 0.931R2=0.931
침식 위치 예측 정확도: R2=0.969R^2 = 0.969R2=0.969
민감도 분석:
유량 증가: 침식 깊이와 위치에 선형적(또는 비선형적) 영향을 미침.
기울기 증가: 일정한 비선형 패턴 관찰.
토양 입자 크기 증가: 복잡하고 비선형적인 패턴 확인.
최적 설계:
침식 깊이 최소화: 유량과 토양 입자 크기를 낮게, 기울기를 높게 설정.
침식 위치 최대화: 유량, 토양 입자 크기, 기울기의 조합을 조절.
결론
FLOW-3D와 서로게이트 모델링: 침식 예측과 최적화에 효과적인 도구로 확인.
설계 최적화 가능성: 구조적 침식 문제를 예방하기 위해 설계 단계에서 주요 변수의 영향을 정밀히 평가.
향후 연구 제안: 추가적인 변수 도입 및 데이터를 통한 모델 개선.
이 논문은 수치 해석과 통계적 설계 접근법을 결합하여 수로 설계 문제를 해결하는 새로운 방법론을 제시하며, 향후 관련 연구에 중요한 기초 자료를 제공할 수 있습니다.
Reference
Abt SR, Ruf JF, Doehring FK (1985) Culvert slope efects on outlet scour. J Hydraul Eng 111(10):1363–1367. https://doi.org/10.1061/(ASCE)0733-9429(1985)111:10(1363)
Ahmad N, Mohammed TA, Melville BW, Ali FH, Yusuf B (2017) Modelling the efect of sediment coarseness on local scour at wide bridge piers. Pertanika J Sci Technol 25(1):191–200
Amini A, Parto AA (2017) 3D numerical simulation of fow feld around twin piles. Acta Geophys 65(6):1243–1251. https://doi.org/10.1007/s11600-017-0094-x
Amini A, Solaimani N (2018) The efects of uniform and nonuniform pile spacing variations on local scour at pile groups. Mar Georesour Geotechnol 36(7):861–866. https://doi.org/10.1080/1064119X.2017.1392658
Bui DT, Shirzadi A, Amini A, Shahabi H, Al-Ansari N, Hamidi S, Ghazvinei PT et al (2020) A hybrid intelligence approach to enhance the prediction accuracy of local scour depth at complex bridge piers. Sustainability (switzerland). https://doi.org/10.3390/su12031063
Chiew Y-M, Lim S-Y (1996) local scour by a deeply submerged horizontal circular jet. J Hydraul Eng 122(9):529–532. https://doi.org/10.1061/(ASCE)0733-9429(1996)122:9(529)
Elsebaie IH (2013) An experimental study of local scour around circular bridge pier in sand soil. Int J Civ Environ Eng IJCEE-IJENS 13(1):23–28
Falconer RA (1991) Computational fluid dynamics. By M. B. ABBOTT and D. R. BASCO. Longman, 1989. 425 pp. J Fluid Mech 229:689–689. https://doi.org/10.1017/S0022112091213233
Flow Science I (2012) FLOW-3D, version 12.0 user’s manual. In: I Flow Science (ed). FLOW-3D [Computer software]. Flow Science, Inc, Santa Fe. https://www.fow3d.com
Galán Á, González J (2019) Efects of shape, inlet blockage and wing walls on local scour at the outlet of non-submerged culverts: undermining of the embankment. Environ Earth Sci 79(25):1–16. https://doi.org/10.1007/s12665-019-8749-3
Günal M, Günal AY, Osman K (2019) Simulation of blockage efects on scouring downstream of box culverts under unsteady fow conditions. Int J Environ Sci Technol 16(9):5305–5310. https://doi.org/10.1007/s13762-019-02461-w
Hariri-Ardebili MA, Seyed-Kolbadi SM, Noori M (2018) Response surface method for material uncertainty quantifcation of infrastructures. Shock Vib 2018:1–14. https://doi.org/10.1155/2018/1784203
Hosseini R, Fazloula R, Saneie M, Amini A (2018) Bagged neural network for estimating the scour depth around pile groups. Int J River Basin Manag 16(4):401–412. https://doi.org/10.1080/15715124.2017.1372449
Kincl M, Turk S, Vrečer F (2005) Application of experimental design methodology in development and optimization of drug release method. Int J Pharm 291(1):39–49. https://doi.org/10.1016/j.ijpharm.2004.07.041
Klun M, Kryžanowski A, Schnabl S (2016) Application of response surface method in analysis of hydraulic structures. Acta Hydrotech 29(51):103–123
Kocaman S (2014) Prediction of backwater profles due to bridges in a compound channel using CFD. Adv Mech Eng 6:905217–905217. https://doi.org/10.1155/2014/905217
Mehnifard MA, Dalfardi S, Baghdadi H, Seirfar Z (2015) Simulation of local scour caused by submerged horizontal jets with Flow-3D
numerical model. Desert 20(1):47–55 Najafzadeh M (2016) Neurofuzzy-based GMDH-PSO to predict maximum scour depth at equilibrium at culvert outlets. J Pipeline Syst Eng Pract 7(1):06015001
Najafzadeh M, Kargar AR (2019) Gene-expression programming, evolutionary polynomial regression, and model tree to evaluate local scour depth at culvert outlets. J Pipeline Syst Eng Pract 10(3):04019013
Nariman NA, Hussain RR, Msekh MA, Karampour P (2019) Prediction meta-models for the responses of a circular tunnel during earthquakes. Undergr Space 4(1):31–47. https://doi.org/10.1016/j.undsp.2018.06.003
Othman Ahmed K, Amini A, Bahrami J, Kavianpour MR, Hawez DM (2021) Numerical modeling of depth and location of scour at culvert outlets under unsteady fow conditions. J Pipeline Syst Eng Pract 12(4):04021040
Pasma SA, Daik R, Maskat M, Hassan O (2013) Application of boxbehnken design in optimization of glucose production from oil palm empty fruit bunch cellulose. Int J Polym Sci 2013:1–8. https://doi.org/10.1155/2013/104502
Sorourian S (2015) Study of blockage efects on scouring pattern downstream of box culverts. (PhD), University of Technology Sydney, Retrieved from https://opus.lib.uts.edu.au/handle/10453/44198
Soulsby R (1997) Dynamics of marine sands: a manual for practical applications. Telford, London
Iranian Journal of Science and Technology, Transactions of Civil Engineering
Taha N, El-Feky MM, El-Saiad AA, Fathy I (2020) Numerical investigation of scour characteristics downstream of blocked culverts. Alex Eng J 59:3503–3513
Wang L, Melville BW, Guan D (2018) Efects of upstream weir slope on local scour at submerged weirs. J Hydraul Eng 144(3):04018002.
Xiong W, Cai CS, Kong B, Kong X (2016) CFD simulations and analyses for bridge-scour development using a dynamic-mesh updating technique. J Comput Civ Eng 30(1):04014121.
날카로운 정상부를 가진 삼각형 허들(Sharp-Crested Triangular Hump) 위의 유동 특성 수치 모델링
연구 배경
문제 정의: 수리 구조물의 성능 및 수면 프로파일을 정확히 예측하는 것은 실험적으로 어렵고 비용이 많이 듦.
목표: CFD(Computational Fluid Dynamics)를 활용하여 삼각형 허들 위의 유동 특성을 보다 효율적이고 정확하게 분석.
접근법: FLOW-3D 기반 시뮬레이션을 수행하여 실험 데이터와 비교 검증.
연구 방법
삼각형 허들(Weir) 개요
위어(Weir)는 개수로에서 유량 조절과 방류 역할을 수행하는 중요한 수리 구조물.
본 연구에서는 크기가 50 cm × 30 cm × 7 cm인 Sharp-Crested Triangular Hump 모델을 사용.
수치 모델링
FLOW-3D를 사용하여 RANS(Reynolds-Averaged Navier-Stokes) 방정식과 VOF(Volume of Fluid) 방법을 적용.
FAVOR(Fractional Area-Volume Obstacle Representation) 기법을 사용하여 메쉬 내 장애물 영향을 반영.
총 1,920,000개의 격자 셀을 사용하여 시뮬레이션 수행.
실험 설정
Universiti Teknologi PETRONAS(UTP)의 수리 실험실에서 실험 수행.
30cm 폭, 60cm 높이, 10m 길이의 플룸(flume)에서 실험 진행.
4가지 유량 조건(30, 51.3, 75.3, 31 m³/h) 및 경사 조건(0, 0.006, 0.01)으로 실험 설계.
주요 결과
수치 시뮬레이션 vs 실험 데이터 비교
수치 시뮬레이션과 실험 결과 간의 차이는 4~5% 이내로 매우 높은 정확도를 보임.
수면 프로파일, 평균 유속, 프로우드 수(Froude Number) 등이 실험과 잘 일치.
유동 특성 분석
프라우드 수(Froude Number) 변화:
상류(Upstream)에서는 Froude Number < 1.0 → 서브크리티컬(Subcritical) 흐름.
하류(Downstream)에서는 Froude Number > 1.0 → 슈퍼크리티컬(Supercritical) 흐름.
유속(Flow Velocity) 변화:
하류로 갈수록 유속 증가, 삼각형 허들이 흐름을 방해하면서 압력 변화를 유발.
수심(Flow Depth) 변화:
상류에서는 높은 수심 유지, 하류에서는 급격한 감소 확인.
수치 시뮬레이션의 유용성
FLOW-3D가 삼각형 허들 및 수리 구조물의 유동 해석에 효과적임을 확인.
기존의 실험적 접근보다 비용이 낮고 신속한 설계 검토 가능.
결론 및 향후 연구
FLOW-3D 기반 CFD 시뮬레이션이 삼각형 허들의 유동 해석 및 설계 최적화에 효과적임을 검증.
실험 데이터와 비교했을 때 높은 정확도(오차 4~5%)를 나타내며, 초기 설계 검토에 유용함.
향후 연구에서는 다양한 난류 모델(k-ε, RNG, LES) 적용 및 추가적인 수리 구조물 연구가 필요.
연구의 의의
이 연구는 수리 구조물의 유동 해석을 위해 CFD 시뮬레이션을 실험적으로 검증하여, 위어 및 삼각형 허들 설계의 최적화 및 성능 예측을 위한 신뢰성 높은 방법론을 제시했다는 점에서 큰 의미가 있다.
Reference
R.L. France, Wetland Design: Principles and Practices for Landscape Architects andLand-Use Planners, WW Norton and Company, 2003.
A. Ferrari, SPH simulation of free surface ow over a sharp-crested weir, Adv. WaterResour. 33 (3) (2010) 270–276.
R. Singh, D. Manivannan, T. Satyanarayana, Discharge coefficient of rectangularside weirs, J. Irrig. Drain. Eng. 120 (4) (1994) 814–819.
S.M. Borghei, M.R. Jalili, M. Ghodsian, Discharge coefficient for sharp-crested sideweir in subcritical ow, J. Hydraul. Eng. 125 (10) (1999) 1051–1056.
A. Al-Shukur, M. Al-jumaily, Z. Shaker, Experimental investigation of flow characteristics over crump weir with Different conditions, Saudi J. Eng. Technol. 2 (10)(2017) 373–379.
K.S. Erduran, G. Seckin, S. Kocaman, S. Atabay, 3D numerical modelling of owaround skewed bridge crossing, Eng. Appl. Comput. Fluid Mech. 6 (3) (2012)475–489.
A.S.I. Abdurrasheed, K.W. Yusof, H.B. Takaijudin, A. Ab, B.S. Iskandar, Effects ofbackwater on hydraulic performance evaluation of rainsmart modules in sustainable drainage systems, in: International Conference on Water Resources, Langkawi,Malaysia, November, 2018, pp. 27–28.
P.G. Chanel, J.C. Doering, Assessment of spillway modeling using computationaluid dynamics, Can. J. Civ. Eng. 35 (12) (2008) 1481–1485.
S. Dehdar-Behbahani, A. Parsaie, Numerical modeling of ow pattern in damspillways guide wall. Case study: balaroud dam, Iran, Alexandria Eng. J. 55 (1)(2016) 467–473.
A. Parsaie, A.H. Haghiabi, A. Moradinejad, CFD modeling of ow pattern in spillwaysapproach channel, Sustain. Water Resour. Manag. 1 (3) (2015) 245–251.
H.K. Versteeg, W. Malalasekera, An Introduction to Computational Uid Dynamics:the infinite Volume Method, Pearson education, 2007.
A.S.I. Abdurrasheed, K.,W. Yusof, H.B. Takaijudin, A.A. Ghani, M.M. Muhammad,A.T. Sholagberu, Advances and challenging issues in subsurface drainage moduletechnology and BIOECODS: a review, in: MATEC Web of Conferences, 203, EDPSciences, 2018, 07005.
S.Y. Kumcu, Investigation of ow over spillway modeling and comparison betweenexperimental data and CFD analysis, KSCE J. Civil Eng. 21 (3) (2017) 994–1003.
M. Darw, F. Moukalled, M. Luca, Finite Volume Method in Computational FluidDynamics: an Advanced Introduction with OpenFOAM an Matlab, Springer, 2015.
H. Zahabi, M. Torabi, E. Alamatian, M. Bahiraei, M. Goodarzi, Effects of geometryand hydraulic characteristics of shallow reservoirs on sediment entrapment, Water10 (12) (2018) 17–25.
G. Li, X. Li, J. Ning, Y. Deng, Numerical simulation and engineering application of adovetail-shaped bucket, Water 11 (2) (2019) 242.
D.C. Lo, J.S. Liou, S. Chang, Hydrodynamic performances of air-water flows ingullies with and without swirl generation vanes for drainage systems of buildings,Water 7 (2) (2015) 679–696.
T. Cebeci, Turbulence models and their application: efficient numerical methodswith computer programs, Springer Science and Business Media, 2003.
B. Mohammadi, O. Pironneau, Analysis of the k-epsilon turbulence model, 1993.
S. Patankar, Numerical Heat Transfer and Uid Ow, CRC press, 1980.
L. Choufu, S. Abbasi, H. Pourshahbaz, P. Taghvaei, S. Tfwala, Investigation of flow,erosion, and sedimentation pattern around varied groynes under Differenthydraulic and geometric conditions: a numerical study, Water 11 (2) (2019) 235.
S. Haun, N.R.B. Olsen, R. Feurich, Numerical modeling of ow over trapezoidalbroad-crested weir, Eng. Appl. Comput. Fluid Mech. 5 (3) (2011) 397–405.
댐 방수로(Spillway) 안내벽의 유동 패턴 수치 모델링: 이란 Balaroud 댐 사례 연구
연구 배경
문제 정의: 댐 방수로의 안내벽(Guide Wall)은 흐름 패턴을 조절하는 중요한 구조물로, 최적의 형상을 설계하면 방수로의 성능을 향상할 수 있음.
목표: Balaroud 댐의 방수로 안내벽에 대해 물리적 및 수치적 모델링을 수행하여, 최적의 안내벽 형상을 도출.
접근법: CFD(Computational Fluid Dynamics) 소프트웨어인 FLOW-3D를 활용하여 다양한 안내벽 설계를 비교 분석.
연구 방법
모델링 개요
AutoCAD를 이용하여 3D 모델 생성 후 FLOW-3D로 내보내기(STL 파일 형식).
1:110 축척의 실험실 모델을 구축하고 실험 결과와 수치 해석을 비교.
수치 모델링 과정
격자 생성(Meshing): 다양한 해상도로 수치 해석을 진행.
경계 조건 설정: 유입 및 유출 조건을 설정하고 난류 모델 선택.
난류 모델 비교
K-epsilon, RNG K-epsilon, LES(Large Eddy Simulation) 모델을 비교.
RNG K-epsilon 모델이 가장 적합한 결과를 보임.
세 가지 안내벽 설계 평가
모델 1: 유동 분리가 심하게 발생하여 부적합.
모델 2: 접근 채널에서 교차파(Cross Waves) 형성.
모델 3: 최소한의 유동 분리 및 교차파 제거 → 최적의 설계로 선정.
주요 결과
모델 3이 가장 우수한 성능을 보이며, 교차파 발생을 최소화하고 유량을 원활하게 전달.
유량-수위 곡선(Rating Curve) 분석을 통해 모델 3이 다른 설계보다 효율적임을 확인.
FLOW-3D의 RNG K-epsilon 난류 모델이 유동 패턴 해석에 가장 적합.
결론 및 향후 연구
수치 모델링과 물리적 실험을 결합하여 최적의 안내벽 형상을 도출.
최적 설계(모델 3)를 통해 방수로 성능을 개선하고, 수력 구조물의 안전성을 향상 가능.
향후 연구에서는 다양한 유입 조건과 추가적인 설계 변수를 고려하여 더욱 정밀한 최적화를 수행할 필요.
이 연구는 댐 방수로 안내벽 설계의 최적화를 목표로 하며, 수치 해석 기법을 활용한 CFD 기반 설계 검증 방법론을 제시한다는 점에서 의의가 있다.
Reference
M.C. Aydin, CFD simulation of free-surface flow overtriangular labyrinth side weir, Adv. Eng. Softw. 45 (1) (2012)159–166.
M.C. Aydin, M.E. Emiroglu, Determination of capacity oflabyrinth side weir by CFD, Flow Meas. Instrum. 29 (2013) 1–8.
H. Babaali, A. Shamsai, H. Vosoughifar, Computationalmodeling of the hydraulic jump in the stilling basin withconvergence walls using CFD codes, Arab. J. Sci. Eng. 40 (2)(2015) 381–395.
T. Cebeci, Turbulence Models and Their Application: EfficientNumerical Methods with Computer Programs, Horizons Pub,2004.
P.G. Chanel, An Evaluation of Computational Fluid Dynamicsfor Spillway Modeling, University of Manitoba, 2008 (Master ofScience).
J. Chatila, M. Tabbara, Computational modeling of flow overan ogee spillway, Comput. Struct. 82 (22) (2004) 1805–1812.
C. Chinnarasri, D. Kositgittiwong, P.Y. Julien, Model of flowover spillways by computational fluid dynamics, Proc. ICE –Water Manage. (2014) 164–175.
R. Ettema, Hydraulic Modeling Concepts and Practice, ASCE,2000.
D. Gessler, CFD modeling of spillway performance, ImpactsGlob. Clim. Change (2005) 1–10.
G. Guyot, H. Maaloul, A. Archer, A Vortex modeling with 3DCFD, in: P. Gourbesville, J. Cunge, G. Caignaert (Eds.),Advances in Hydroinformatics, Springer, Singapore, 2014, pp.433–444.
W.H. Hager, M. Pfister, Hydraulic modelling – an introduction:principles, methods and applications, J. Hydraul. Res. 48 (4)(2010) 557–558.
D. Kim, J. Park, Analysis of flow structure over ogee-spillway inconsideration of scale and roughness effects by using CFDmodel, KSCE J. Civ. Eng. 9 (2) (2005) 161–169.
R. Maghsoodi, M.S. Roozgar, H. Sarkardeh, H.M.Azamathulla, 3D-simulation of flow over submerged weirs,Int. J. Model. Simul. 32 (4) (2012) 237.
B. Mohammadi, O. Pironneau, Analysis of the K-epsilonTurbulence Model, Wiley, 1994.
F. Moukalled, L. Mangani, M. Darwish, The Finite VolumeMethod in Computational Fluid Dynamics: An AdvancedIntroduction with OpenFOAM and Matlab, SpringerInternational Publishing, 2015.
Y. Nakayama, R.F. Boucher, Computational fluid dynamics, in:Y.N.F. Boucher (Ed.), Introduction to Fluid Mechanics,Butterworth-Heinemann, Oxford, 1998, pp. 249–273.
A. Parsaie, A. Haghiabi, Computational modeling of pollutiontransmission in rivers, Appl. Water Sci. (2015) 1–10.
A. Parsaie, A. Haghiabi, Predicting the longitudinal dispersioncoefficient by radial basis function neural network, Model.Earth Syst. Environ. 1 (4) (2015) 1–8.
A. Parsaie, A. Haghiabi, A. Moradinejad, CFD modeling offlow pattern in spillway’s approach channel, Sustain. WaterResour. Manag. 1 (3) (2015) 245–251.Figure 14 Head discharge curve of the model (3).Figure 16 The rating curve of the models for the guide wall.Figure 15 Cross section of the flow through the guide wall at theflood return period 1000 year.472 S. Dehdar-behbahani, A. Parsaie
A. Parsaie, H. Yonesi, S. Najafian, Predictive modeling ofdischarge in compound open channel by support vector machinetechnique, Model. Earth Syst. Environ. 1 (2) (2015) 1–6.
S. Patankar, Numerical Heat Transfer and Fluid Flow, Taylor& Francis, 1980.
W. Rodi, Turbulence Models and Their Application inHydraulics, Taylor & Francis, 1993.
M. Suprapto, Increase spillway capacity using labyrinth weir,Proc. Eng. 54 (2013) 440–446.
H.K. Versteeg, W. Malalasekera, An Introduction to Computational Fluid Dynamics: The Finite Volume Method, Pearson Education Limited, 2007.
J.-B. Wang, H.-C. Chen, Improved design of guide wall of bank spillway at Yutang Hydropower Station, Water Sci. Eng. 3 (1) (2010) 67–74.
J. Wang, H. Chen, Experimental study of elimination of vortices along guide wall of bank spillway, in: Advances in Water Resources and Hydraulic Engineering, Springer, BerlinHeidelberg, 2009, pp. 2059–2063.
문제 정의: 샤프 크레스트 위어는 수로에서 유량 측정과 조절을 위해 가장 널리 사용되는 구조물이다.
목표: CFD(Computational Fluid Dynamics) 기법을 활용하여 샤프 크레스트 위어 위의 유동 특성을 분석하고 방출 계수(Discharge Coefficient)를 예측.
접근법: FLOW-3D를 사용하여 수치 해석을 수행하고 실험 데이터와 비교.
연구 방법
위어 특성 및 방출 계수(Cd) 분석
기존 실험 연구를 기반으로 방출 계수 CdCdCd 추정식을 개발.
다양한 유량 및 위어 높이 조합을 사용하여 최적의 방출 계수 관계식 도출.
FLOW-3D 기반 수치 모델링
VOF(Volume of Fluid) 기법을 적용하여 자유 수면을 해석.
RNG k−ϵk-\epsilonk−ϵ난류 모델을 사용하여 난류 흐름을 해석.
FAVOR(Fractional Area-Volume Obstacle Representation) 기법을 활용하여 격자 내 장애물 표현.
격자 수렴 분석
다양한 해상도의 격자를 비교하여 최적의 계산 비용과 정확도를 확보.
주요 결과
수치 모델링 vs 실험 데이터 비교
방출 계수(Cd) 예측값과 기존 실험값 간의 오차 범위가 ±3% 이내로 매우 높은 정확도를 보임.
Cd는 Ht/tw(총 수두 대비 위어 높이)와 강한 상관관계를 가짐.
유동 특성 분석
유량 변화에 따른 방출 계수:
유량이 증가할수록 방출 계수가 점진적으로 감소하는 경향 확인.
위어 주변의 속도 및 압력 분포 분석:
위어 크레스트에서 유동이 가속되면서 속도 증가 및 압력 감소 현상 관찰.
위어 하류에서 수압이 낮아지며 유동 패턴이 변화.
FLOW-3D의 유용성
FLOW-3D는 실험 대비 비용이 낮고 신속한 설계 검토 가능.
다양한 위어 형상 및 유량 조건에서 적용 가능성이 높음.
결론 및 향후 연구
FLOW-3D 기반 CFD 시뮬레이션이 샤프 크레스트 위어의 방출 계수 예측 및 유동 분석에 효과적임을 입증.
실험 결과와 비교했을 때 높은 정확도(오차 ±3%)를 나타내며, 초기 설계 검토에 유용함.
향후 연구에서는 다양한 위어 형상 및 추가적인 난류 모델 적용(k-ω, LES 등)을 통해 더욱 정밀한 해석이 필요.
연구의 의의
이 연구는 샤프 크레스트 위어의 유동 특성을 CFD 기반으로 해석하여 설계 최적화 및 방출 계수 예측의 신뢰성을 향상시켰다는 점에서 의미가 크다.
Reference
Bhallamudi, S.M. and M.H. Chaudhry, 1994. Computation of Flows in Open Channel Transitions. Journal of Hydraulic. Research, 30(1): 77-93.
Fritz, H.M. and H.W. Hager, 1998. Hydraulics of Embankment Weirs. Journal of Hydraulic Engineering,
ASCE., 124(9): 963-971.
Hargreaves, D.M., H.P. Morvan, N.G. Wright, 2007. Validation of the Volume of Fluid Method for Free
Surface Calculation. Engineering Applications of Computational Fluid mechanics, 1(2): 136-147.
Kindsvater, C.E., R.W. Carter, 1957. Discharge Characteristics of Rectangular Thin-Plate Weirs. Journal of Hydraulic Engineering, ASCE., 14: 1-36.
King, H.W. and E.F. Brater, 1963. Handbook of Hydraulics, 5th Edition, McGraw-Hill Book Company, New York.
Patankar, S.V., 1980. Numerical Heat Transfer and Fluid Flow. McGraw-Hill Book Company, New York.
Raju, K.G.R., G.L. Asawa, 1977. Viscosity and Surface Tension Effects on Weir Flow. Journal of Hydraulic Engineering, ASCE., 103: 1227-1231.
Rouse, H., 1950. Engineering Hydraulics. Proceedings of the Fourth Hydraulics Conference, Iowa Institute of Hydraulic Research, John Wiley and Sons, Inc., New York.
Sarginson, E.J., 1972. The Iinfluence of Surface Tension on Weir Flow. Journal of Hydraulic Research, 10:431-446.
레이저 빔 형상이 용접 품질과 금속 혼합에 중요한 역할을 하며, 적절한 코어-링 직경 비율 설정이 필요.
향후 연구에서는 더 다양한 빔 형상과 용접 속도 조건을 고려하여 최적 설계 도출.
연구의 의의
이 연구는 CFD 기반 다중 물리 모델링을 활용하여 레이저 빔 형상의 금속 혼합 및 용접 품질에 미치는 영향을 체계적으로 분석하였으며, EV 배터리 제조에서 신뢰성 높은 용접 기술 개발을 위한 기초 데이터를 제공한다.
Reference
Brand, M.J., Schmidt, P.A., Zaeh, M.F., Jossen, A., 2015. Welding techniques for battery cells and resulting electrical contact resistances. J. Energy Storage 1, 7–14.
Chianese, G., Franciosa, P., Sun, T., Ceglarek, D., Patalano, S., 2022. Using photodiodes and supervised machine learning for automatic classification of weld defects in laser welding of thin foils copper-to-steel battery tabs. J. Laser Appl. 34, 042007.
Chianese, G., Jabar, S., Franciosa, P., Ceglarek, D., Patalano, S., 2022. A multi-physics CFD study on the part-to-part gap during remote laser welding of copper-to-steel battery tab connectors with beam wobbling. Procedia CIRP 111, 484–489.
Daligault, J., Dal, M., Gorny, C., Coste, F., Fabbro, R., 2022. Combination of Eulerian and ray-tracing approaches for copper laser welding simulation. J. Laser Appl. 34, 042042.
Das, A., Li, D., Williams, D., Greenwood, D., 2018. Joining Technologies for Automotive Battery Systems Manufacturing. World Electr. Veh. J. 9 (2), 22.
Drobniak, P., Otto, A., Vázquez, R.G., Arias, R.M., Arias, J.L., 2020. Simulation of keyhole laser welding of stainless steel plates with a gap. Procedia CIRP 94, 731–736.
Hao, Y., Chen, N., Wang, H., Carlson, B.E., Lu, F., 2021. Effect of zinc vapor forces on spattering in partial penetration laser welding of zinc-coated steels. J. Mater. Process. Technol. 298, 117282.
Huang, W., Wang, H., Rinker, T., Tan, W., 2020. Investigation of metal mixing in laser keyhole welding of dissimilar metals. Mater. Des. 195, 109056.
Huang, W., Cai, W., Rinker, T.J., Bracey, J., Tan, W., 2023. Effects of laser oscillation on metal mixing, microstructure, and mechanical property of Aluminum–Copper welds. Int. J. Mach. Tools Manuf. 188, 104020.
Hummel, M., Schöler, C., Häusler, A., Gillner, A., Poprawe, R., 2020. New approaches on laser micro welding of copper by using a laser beam source with a wavelength of 450 nm. J. Adv. Join. Process. 1, 100012.
IEA, 2022: Global EV Outlook 2022 – Securing supplies for an electric future.
Jabar, S., Barenji, A.B., Franciosa, P., Kotadia, H.R., Ceglarek, D., 2023. Effects of the adjustable ring-mode laser on intermetallic formation and mechanical properties of steel to aluminium laser welded lap joints. Mater. Des. 227, 111774.
Jeong, Y.B., Jo, H.R., Park, H.J., Kato, H., Kim, K.B., 2020. Mechanical properties and microstructural change in (Cu–Fe) immiscible metal matrix composite: Effect of Mg on secondary phase separation. J. Mater. Res. Technol. 9 (6), 15989–15995.
Kogel-Hollacher, M., 2020. The full potential of Photonics in e-mobility: an overview. Laser Use 97, 22–23.
Kogel-Hollacher, M., Nicolay, T., Reiser, J., 2022. Going Green – Laser Welding and Advanced Sensor Technology for Electric Vehicles.
Kumar, N., Masters, I., Das, A., 2021. In-depth evaluation of laser-welded similar and dissimilar material tab-to-busbar electrical interconnects for electric vehicle battery pack. J. Manuf. Process. 70, 78–96.
Li, J., Cai, Y., Yan, F., Wang, C., Zhu, Z., Hu, C., 2019. Porosity and liquation cracking of dissimilar Nd:YAG laser welding of SUS304 stainless steel to T2 copper. Opt. Laser Technol. 122, 105881.
Lin, R., Wang, H., Lu, F., Solomon, J., Carlson, B.E., 2017. Numerical study of keyhole dynamics and keyhole-induced porosity formation in remote laser welding of Al alloys. Int. J. Heat. Mass Transf. 108 (A), 244–256.
Liu, Z., Wang, X., Hang, K., 2019. Enhancement of trapping efficiency by utilizing a hollow sinh-Gaussian beam. Sci. Rep. 9, 10187.
Lu, F., Li, X., Li, Z., Tang, X., Cui, H., 2015. Formation and influence mechanism of keyhole-induced porosity in deep-penetration laser welding based on 3D transient modeling. Int. J. Heat. Mass Transf. 90, 1143–1152.
Perez Zapico, E., Ascari, A., Dimatteo, V., Fortunato, A., 2021. Laser dissimilar welding of copper and steel thin sheets for battery production. J. Laser Appl. 33 (1), 012016.
Prieto, C., Vaamonde, E., Diego-Vallejo, D., Jimenez, J., Urbach, B., Vidne, Y., Shekel, E., 2020. Dynamic laser beam shaping for laser aluminium welding in e-mobility applications. Procedia CIRP 94, 596–600.
Rinne, J., Nothdurft, S., Hermsdorf, J., Kaierle, S., Overmeyer, L., 2020. Advantages of adjustable intensity profiles for laser beam welding of steel copper dissimilar joints. Procedia CIRP 94, 661–665.
Rinne, J., Nothdurft, S., Hermsdorf, J., Kaierle, S., Overmeyer, L., 2022. Multivariate parametric study on laser beam welding of copper lap joints using adjustable intensity profiles. Procedia CIRP 111, 415–419.
Sadeghian, A., Iqbal, N., 2022. A review on dissimilar laser welding of steel-copper, steel-aluminum, aluminum-copper, and steel-nickel for electric vehicle battery manufacturing. Opt. Laser Technol. 146, 107595.
Shi, R.P., Wang, C.P., Wheeler, D., Liu, X.J., Wang, Y., 2013. Formation mechanisms of self-organized core/shell and core/shell/corona microstructures in liquid droplets of immiscible alloys. Acta Mater. 61 (4), 1229–1243.
Sokolov, M., Franciosa, P., Sun, T., Ceglarek, D., Dimatteo, V., 2021. Applying optical coherence tomography for weld depth monitoring in remote laser welding of automotive battery tab connectors. J. Laser Appl. 33 (1), 012028.
Wagner, J., Heider, A., Ramsayer, R., Weber, R., Faure, F., Leis, A., Armon, N., Susid, R., Tsiony, O., Shekel, E., Graf, T., 2022. Influence of dynamic beam shaping on the geometry of the keyhole during laser beam welding. Procedia CIRP 111, 448–452.
Zwicker, M.F.R., Moghadam, M., Zhang, W., Nielsen, C.V., 2020. Automotive battery pack manufacturing – a review of battery to tab. J. Adv. Join. Process. 1, 100017.
300 gal/min 이하에서는 실험과 예측이 잘 일치, 350~500 gal/min 구간에서 약간의 오차 발생 (버블 생성이 원인으로 추정됨).
결과 및 결론
고정 메쉬 기법의 장점:
이동 및 변형형 메쉬 기법보다 효율적이며, 이동 객체 간의 거리 제한이 없음.
충돌 처리 가능.
유체와 객체의 상호작용을 보다 정밀하게 반영 가능.
실험 결과와 비교:
밸브 개폐 시뮬레이션에서 실험 결과와 높은 일치도를 보임.
고유량(>300 gal/min)에서 약간의 차이가 존재하지만, 이는 실험 조건(버블 발생 등)으로 설명 가능.
향후 연구 방향
다양한 공학적 응용(자동차, 항공, 유압 시스템 등)에 적용하여 성능 검증.
더욱 복잡한 이동 객체 및 다중 상호작용 모델 확장.
이 논문은 기존의 이동형 메쉬 기법의 한계를 극복하고, 복잡한 유체-구조 상호작용을 효율적으로 모델링할 수 있는 새로운 CFD 기법을 제안한다는 점에서 큰 의미가 있다.
Reference
C. W. Hirt and J. M. Sicilian, “A porosity technique for the definition of obstacles in rectangular Cell Meshes”, Proc. Fourth International Conf. Ship Hydro., National Academic of Science, Washington, DC, Sept. 1985.
G. Wei, “A general moving object model”, Technical Report, Flow Science Inc., 2005.
H. Goldstein, P. Charles and J. Safko, Classical Mechanics (Addison Wesley, Washington, 2002).
Tailings dams are structures that store both tailings and water, so almost all tailings dam accidents are water related. This paper investigates a tailings dam’s failure pattern and damage development under flood conditions by conducting a 1:100 large-scale tailings dam failure model test. It also simulates the tailings dam breach discharge process based on the breach mode using FLOW-3D software, and the extent of the impact of the dam failure debris flow downstream was derived. Dam failure tests show that the form of dam failure under flood conditions is seepage failure. The damage manifests itself in the form of flowing soil, which is broadly divided into two processes: the seepage stabilization phase and the flowing soil development damage phase. The dam failure test shows that the rate of rise in the height of the dam saturation line is faster and then slower. The order of the saturation line at the dam face is second-level sub-dam, third-level sub-dam, first-level sub-dam, and fourth-level sub-dam. The final failure of the tailings dam is the production of a breach at the top of the dam due to the development of the dam’s fluid damage zone to the dam top. The simulated dam breach release results show that by the time the dam breach fluid is released at 300 s, the area of over mud has reached 95,250 square meters. Local farmland and roads were submerged, and other facilities and buildings would be damaged to varying degrees. Based on the data from these studies, targeted measures for rectifying hidden dangers and preventing dam breaks from both technical and management aspects can be proposed for tailings dams.
1. Introduction
1.1. Research Status
The mud wastewater containing tailings will be discharged after metal and non-metal mine beneficiation. Tailings slurry contains mercury, arsenic, and other heavy metal ions, both resources and pollution sources [1]. The tailings dam is a dam body formed by the accumulation and rolling of the tailings after the mine selects the useful components [2]. It is of great significance to research the dynamic stability of tailings reservoirs for mine safety production, protection of downstream life and property safety, and the surrounding environment [3]. Tailings dams, an important source of danger if an accident, are bound to people’s lives and property [4]. In 2008, a dam break accident occurred in the 980 ditch tailings pond of Shanxi Xinta Mining Co., Ltd., Yuncheng, China, resulting in 281 deaths and 33 injuries. The direct economic loss was as high as CNY 96,192,000 [5]. On the afternoon of 30 April 2006, the tailings dam of Zhen’an Gold Mine in Shaanxi Province was constructed. The accident caused 17 people to disappear, five people were injured, and 76 houses were destroyed [6,7].
In many tailings reservoir accidents, due to the lack of flood discharge capacity of flood discharge structures in the reservoir area, flood overtopping, tailings dam break, and other phenomena occur occasionally [8,9]. In this regard, scholars in related fields have performed much research and achieved certain results. Chen Zhang et al. [10] established three-dimensional and two-dimensional finite element models. The seepage field of the project under different operating conditions was simulated, and the safety factor under different operating conditions was obtained by combining the seepage field with the stable surface. The influence of the length of the dry beach and the upstream slope ratio on the seepage and stability of the tailings dam was determined. Sánchez-Peralta et al. [11] took a dry tailings pond in Colombia as the research object, studied the movement characteristics of dam break debris flow with different water contents, and obtained the relationship between the length and width of dam break debris flow movement. Changbo Du et al. [12] studied and analyzed the influence of reinforcement on tailings dam and the change law of pore water pressure and internal pressure of the dam body after mud discharge. The pore water pressure and internal earth pressure of the accumulation dam after grouting gradually increased with time. Reinforcement can greatly reduce the pore water pressure and internal pressure of reinforced dams. Gregor Petkovšek et al. [13] proposed a dam break model EMBREA-MUD to calculate the water and tailings outflow of the tailings reservoir and the corresponding break growth. Weile Geng et al. [14] conducted experimental research on the settlement deformation and mesostructure evolution of unsaturated tailings under continuous load. The results showed that the mesostructure deformation of unsaturated tailings with different moisture contents under load was the same and could be divided into four stages: pore compression, elastic deformation, structure change, and further compaction. Alan Lolaev et al. [15] developed a method to determine the tailings filtration and secondary consolidation coefficient in the process of alluvial according to the physical conditions, density, and water phreatic, and a mathematical model to calculate the consolidation time. Kun Wang et al. [16] proposed a multidisciplinary program to simulate the dam break runoff of hypothetical tailings reservoirs on the downstream complex terrain using UAV photogrammetry and smooth particle hydrodynamics (SPH) numerical method. Rawya M. Kansoh [17] studied the influence of the earth-rock dam’s structural parameters on the dam failure process. Kehui Liu et al. [18] studied the microscopic characteristics of hydraulic erosion of reinforced tailings dams and revealed the influence of different reinforcement spacing on the critical start-up speed of tailings particles. It shows that the smaller the reinforcement spacing, the greater the critical start-up speed of the reinforced tailings samples. Luca Piciullo et al. [19] proposed a regression analysis that considers the functional relationship between the release amount and the characteristics of the tailings dam, such as height and water storage (i.e., dam factor). The effects of construction type, filling material, and failure mode on the release amount were also evaluated, as well as the failure frequency of the tailings dam as a function of the construction method. Tailing dams built using upstream construction methods are more prone to failure and are more susceptible to static and dynamic liquefaction. Chunhui Ma et al. [20] pointed out that a reasonable construction schedule and flexible waterproof material are key features of impervious bodies for dams with significant deformation. When the dam deformation becomes stable, consideration should be given to secondary treatment of the impervious body to enhance dam safety. Fukumoto et al. [21] used finite element software to simulate the seepage failure process caused by seepage. Alibek Issakhov et al. [22] combined the k-ω turbulence model to study this process numerically. The VOF (volume of fluid) method was used to simulate the fluid movement behind the tailings dam during the break-up of the fluid and the riverbed landscape. Yonas B. Dibike et al. [23] A two-dimensional hydrodynamic and component transport model was used to study the effect of OS tailings release on the water quality and sediment quality of LAR by simulating sediments and related chemicals. It was concluded that the tailings release location was different; 40% to 70% of the sediments and related chemicals were deposited on the riverbed of the 160 km study section, while the remaining sediments and related chemicals left the study area in the first three days after the release event. Research conducted by Xiaofei Jing et al. [24] investigated the overflow characteristics of tailings dams reinforced with steel bars. During the overflow process, they measured dam displacements, saturation lines, and internal stresses. The study demonstrated that the erosion resistance of tailings dams significantly improves with an increase in the number of reinforcement layers. Abdellah Mahdi et al. [25] studied the potential consequences of a hypothetical oil sand tailings dam failure. For this reason, a non-Newtonian dam–dam model with a viscoplastic rheological relationship is used. The model can reproduce the flood and water level changes in downstream lakes (due to destructive waves). The simulation study of oil sand tailings overflow proves the importance of considering the non-Newtonian characteristics of tailings. Naeini et al. [26] used SIGMA/W and QUAKE/W software to analyze the high-middle line tailings dam’s dynamic response and permanent deformation and evaluated the dam’s performance. Mohammad Reza Boroomand [27] used the numerical analysis method to analyze the earth dam’s seepage under the uncertainty of geotechnical parameters and analyzed the seepage of the earth dam under the condition of uncertainty of geotechnical parameters. Sumin Li et al. [28] simulated and analyzed the hazard range, degree, and spatial state of sediment flow after the dam break and obtained the influence of sand flow velocity, flow depth, and impact on the downstream villages in the disaster area. The feasibility of the expansion and heightening of the tailings dam project was demonstrated, and the disaster risk levels of different spatial locations in the downstream villages were obtained through simulation. Through experiments, Kong et al. [29] studied the influence parameters of tailings dams under seepage. They concluded that the particle size gradation, non-uniformity coefficient, and water content of tailings sand were the main factors affecting the critical hydraulic gradient. It is concluded that the seepage failure gradient with suitable gradation, uniform particles, and suitable water content is significantly higher than that with poor gradation, uneven particles, and poor water content.
Flood overtopping and seepage failure account for 80% of the total accidents, and these two failure forms mainly occur in flood season and are closely related to water. Therefore, it is necessary to explore the tailings dam failure mode, development process, and the impact on the downstream after dam failure under flood conditions to ensure its safe operation. Based on the engineering background of a tungsten mine tailings dam in Ganzhou City, Jiangxi Province, a 1:100 physical model test was carried out to explore the dam break form and failure development process of the tailings dam under flood conditions. The FLOW-3D fluid simulation software was used to solve the influence of the tailings dam on the downstream after the dam break, and the change law of the flow area, velocity development, and submerged depth of the dam break fluid during the flood discharge process was analyzed. Finally, reasonable prevention and remediation suggestions are proposed for the hidden dangers of tailings dams.
The innovation of this paper is to determine the dam break mode and dam break position of the tailings dam under flood conditions by constructing a physical model, which provides a basis for simulating the influence of dam break on the downstream of the dam body.
1.2. Research Flowchart
Figure 1. Research flowchart.
2. Design and Construction of Large Physical Models
2.1. Overview of the Prototype Tailings Dam
The prototype tailings dam is a tungsten tailings dam located in a narrow valley running north–south in Ganzhou City, Jiangxi Province, China. The downstream of the tailings accumulation dam is farmland, dormitory buildings, mountain roads, etc., and the valleys in the downstream are relatively open. Figure 2 shows an overhead view of the prototype tailings dam. The tailings dam is built using the upstream damming method. The initial dam is a clay core wall weathering material dam located at the mouth of the northern valley. The bottom elevation is 262.0 m, the top elevation is 284.0 m, the dam height is 22 m, the upstream and downstream slope ratio is 1:2.5. The design average external slope ratio of the tailings accumulation dam is 1:5, the average slope of the tailings deposition beach is 5%, the design final tailings accumulation dam elevation is 368.00 m, the total dam height 106.00 m, with a complete storage capacity of 1550 × 104 m3 and a service life of 65 years. The present top elevation of the stacked dam is about 315 m, the height of the dam is 53 m, the accumulated storage capacity is about 559 × 104 m3, and the average external slope ratio of the stacked dam is 1:4.9. The reservoir is currently a fourth-class reservoir, with a flood protection standard of one in 200 years. At a later stage, it will be a second-class reservoir with a flood protection standard of 1000 years.
Figure 2. Top view of a tailings dam.
2.2. Selection of Model Sand
To ensure the relative reliability of the test results, the selected dam materials are properly relaxed to meet the primary conditions of similar main influencing factors. The model test focuses on the agglomeration effect of particle movement during the deformation of the dam body. For the selection of model sand, the initial dam is built with silty clay, and the accumulation dam adopts the mine prototype tailings. Figure 3 is the particle size distribution curve of the model sand. According to the particle size distribution curve, two quantitative indexes of soil particles can be determined: non-uniformity coefficient Cu and curvature coefficient Cc. Cu and Cc can jointly determine the gradation of soil. The expressions of the two are:
Figure 3. Cumulative distribution curve of particle size.
The calculated Cu and Cc of the model sand are 2.29 and 0.84, respectively. It is generally believed that the sand soil with Cu < 5 or Cc outside 1~3 belongs to the poorly graded soil, so the model sand is determined to be the poorly graded soil. If the seepage failure occurs in the dam, the development mode of seepage failure can be predicted by some parameters of the soil, that is, whether the soil is piping or flowing soil. According to the non-uniform coefficient discrimination method proposed by the former Soviet Union scholar Istomina, it is preliminarily judged that the model sand is a flowing soil-type soil.
2.3. Construction of the Dam Failure Model
The dam break process of a tailings reservoir involves many aspects, such as hydraulics, mud and sand dynamics, and soil mechanics. It involves many disciplines and is highly complex, which leads to the similarity relationship of model tests.
Therefore, we must put aside the generality of similarity and focus on the similarity of critical elements. This experiment uses the engineering background of a tungsten mine tailings dam in Jiangxi Province, China. The similarity criterion is appropriately relaxed, and the accumulation effect of particle movement during the deformation of the dam body is emphasized to construct the physical model.
Under the condition of geometric similarity, the physical model test of the 1:100 large-scale tailings dam is carried out according to the level of the second-class reservoir of the prototype tailings dam. The prototype range of the tailings dam is 1200 m × 700 m, and the model size is 12 m × 7 m. The model mainly comprises bedrock, a dam body, an observation system, and a water supply circulation system. The specific steps are as follows: According to the topographic map data provided by the mine, the three-dimensional model of the prototype tailings dam is established using Civil-3D modeling software (ver.2018) according to the size of the actual tailings dam (Figure 4). Then, several vertical sections are cut out in the model with the east–west direction as the standard line, and the points on each vertical section are taken equidistantly to extract the elevation value of each point on each vertical section. The model is intended to build a model with an elevation of 230 m in the actual terrain. The steel frame structure of the bedrock is made based on the elevation of each point on the vertical section. The square steel pipe is used as the bedrock support. Each steel pipe corresponds to the elevation of its relative point in proportion. Finally, the waterproof cloth is covered on the steel frame group and fixed to obtain a complete view of the bedrock terrain. Figure 5 is the completed mountain steel frame group, and Figure 6 is the complete bedrock after laying waterproof cloth.
Figure 4. Three-dimensional model of the tailings dam.
Figure 5. Mountain support structures.
Figure 6. A complete view of bedrock.
The initial dam is piled up with silty clay. In the process of stacking the initial dam, two PVC pipes with holes in the wall body and tightly wrapped with permeable geotextiles are symmetrically buried at the bottom of the initial dam to simulate the drainage pipe. A valve is installed at the outlet end of the two drainage pipes to control the drainage speed. The sub-dam uses the pipeline method commonly used in the mine to simulate the ore drawing. An ore drawing main pipe is introduced from the slurry pool to start the ore drawing from the model’s right side, and a valve is set in the main pipe to control the flow rate of the square ore. When the pulp flows into the tailings pond, the tailings will be layered and precipitated under hydraulic screening. After precipitation, the ore is suspended when the tailings reach the target dam height. Start to build the next sub-dam, use the layered filling method to build the dam body to the design elevation, and use the vertical line method to control the elevation when building the dam. (Figure 7) is the construction of the second sub-dam. Four pore water pressure gauges are buried in the dam construction process to monitor the position of the saturation line of the dam body. The four pore water pressure gauges’ positions are arranged along the dam body’s central axis. They are located directly below the dam crest of the first-, second-, third-, and fourth-level accumulation dams. They are named as site 1, site 2, site 3, and site 4 (Figure 8).
Figure 7. Second-stage sub-dam stacking.
Figure 8. Buried pore water pressure gauges.
Due to the need to supply a large amount of water for the test, a water tower was placed on the site (Figure 9), and a return water collection system was designed to achieve a water supply cycle (Figure 10). The observation equipment of the test (Figure 11) uses a trinocular camera and a high-definition camera to record the dam break process of the tailings dam.
Figure 9. Water tower.
Figure 10. Water supply circulation system.
Figure 11. Experimental observation system.
3. Tailings Dam Break Model Experiment
The dam failure mode under specific flood conditions is characterized by permeation damage, manifested as soil erosion. Through analysis of experimental phenomena and data, the development process of dam failure is elucidated, revealing the variation patterns of pore water pressure at different locations and the saturation line of the dam body.
3.1. Dam Failure Experiment
The test was carried out by intermittently injecting water into the reservoir to simulate flood conditions, keeping the flow rate stable during the injection, and keeping the drainage pipe open during the whole test. The beginning of the water injection was taken as the beginning of the test, and the entire dam break test lasted 448 min. It can be roughly divided into two stages, each accounting for one-half of the total length. Figure 12 shows a typical picture of the damage to the dam during the test. Figure 13 shows a timeline of the test damage development. The specific tailings dam damage development process is as follows:
Figure 12. Dam failure model tests.
Figure 13. Timeline of the development of the flowing soil destruction.
The first stage is seepage stabilization: the overflow water is clear, the dam’s surface is stable, and there is no movement of particles. At 138 min of the test, the contact zone between the right end of the second sub-dam and the bedrock began to seep first (Figure 12a). The seepage water flows along the contact zone between the dam body and the bedrock and overflows down the dam face. There are two reasons for the seepage here. The first is fine cracks in the contact area between the soil and the bedrock, which provides a breakthrough for the seepage water. The second is based on calculating the data collected by the pore water pressure gauge. It can be seen that the saturation line at this time escapes on the slope of the second sub-dam, where the dam surface overflows. Subsequently, the second-stage sub-dam continued to seep, and the overflow area gradually expanded and merged with the second-stage sub-dam dam surface. At 174 min, the third-stage sub-dam began to overflow on the left side (Figure 12b). At this time, according to the collected data, it can be calculated that the buried depth of the saturation line has been exposed to the third-level sub-dam. At 190 min, the first sub-dam also overflowed (Figure 12c). Then, the sand boiling point appears at the right end of the first-stage sub-dam, and the soil particles fluctuate obviously with the overflow water. The sand boiling causes the soil particles to be continuously taken out of the soil body. At 203 min, the dam surfaces of the first, second, and third sub-dams have all become swampy. The second stage is the development and failure stage of the flowing soil: seepage deformation occurs continuously, and more earthwork is lost. At 236 min, the first flow soil damage happened at the right end of the second sub-dam (Figure 12d). The failure form is flow slip. At 239 min, a second flow soil damage occurred on the left side of the first sub-dam (Figure 12e). The flow-slipping soil will form a pit that evolves into a breach, making the seepage velocity and seepage flow faster and larger. Then, the pit part of the soil slides, and the seepage water erodes the downstream dam surface. At 248 min, two erosion ditches have been formed in the flow soil failure area on both sides of the dam (Figure 12f,g). The erosion gully on the left side is located at the junction of the right side of the first-order sub-dam and the bedrock. The critical hydraulic gradient is lower, the dominant flow develops more rapidly, the sand is wrapped violently, and the subsequent seepage damage is more likely to occur. The erosion gully produces more water flow to scour multiple branches on the dam’s surface. The right erosion ditch is located at the junction of the secondary dam and the bedrock. At this time, the erosion ditch has developed to a certain depth, and the sand boiling point has reached 6. The flowing soil migrates downward under the action of overflow water. The flowing water will bring the fine particles to the downstream area. The coarse particles will be accumulated to form a ‘filter layer‘ to block the overflow water channel. The seepage pressure on both sides of the filter layer gradually increases. A new seepage channel will be formed when the seepage pressure on one side reaches the critical value. At 292 min, the flow soil damage eroded to the third sub-dam and further developed upstream along the boundary. Part of the erosion gully’s inner wall soil is washed away underwater, and the internal wall forms holes and expands upward until the upper part forms a suspended surface. When the shear strength of the upper soil is greater than the shear strength of the soil, it will collapse and continue to repeat the next round of erosion. At this time, the left-flowing soil does not develop to the upstream failure but to the proper lateral erosion, and the right side flushes out a new channel due to the obstruction of the ‘filter layer‘. At 303 min, the third flow soil failure occurred on the left side of the third sub-dam (Figure 12h). Because of the increase in overflow water and the acceleration of water flow, the right scouring area opens the downstream channel at the particle deposition, and fine particles are continuously taken out of the dam by seepage water. The overflow water also washes away the ‘filter layer‘ on the left side. After that, the first sub-dam eroded to the deep, and the dam surface failure area did not expand. There is a hydraulic–gravity erosion cycle in the flow soil damage area of the second-stage sub-dam, which extends to the upstream and the middle of the dam body. With the increase in the erosion damage area of the water flow, the more the sand boiling point, the faster the seepage damage, and the erosion area of the lower section continues to expand, and the water flow in the erosion gully is large and fast. The flowing soil failure zone of the third-stage sub-dam has not yet formed a penetrating failure path and is in the initial stage of erosion. At 348 min, the fourth-stage sub-dam overflowed (Figure 12i). At 448 min, the flow soil was eroded to the fourth sub-dam (Figure 12j). The flow soil damage area is eroded to the fourth sub-dam, which is regarded as the whole dam damage. It is measured that the depth of the collapse area is about 12 cm, and the width is about 80 cm. It should be noted that although the dam body has undergone a large area of seepage failure, the dam body has not yet experienced an unstable landslide. The tailings dam finally broke because the dam body soil damage zone developed to the top of the dam to produce a breach.
3.2. The Change Rule of the Saturation Line
Figure 14 is about the change curve of the saturation line. At the beginning of the test, as the upstream water level rose, the saturation line rose rapidly despite the drain being in a normal discharge condition. After the lifting of the head has ceased, the rate of the upward lifting of the saturation line becomes significantly slower due to the hysteresis effect. Then, a certain depth of burial is maintained. In the middle and late stages of the test, most of the dam had become saturated, and the soil matrix suction had weakened. When water is again stored in the reservoir, the saturation line will again lift, but at a reduced rate compared to the initial period. If the reservoir level is no longer raised, the saturation line tends to fall after a period of time. By approximately 270 min into the test, the dam face had already developed a certain size of the flow damage zone, and it was no longer meaningful to discuss the depth of the saturation line.
Figure 14. Variation curve of saturation line.
In the previous study [30], a two-dimensional finite element model of the tailings dam, chosen from the central axis of the three-dimensional tailings dam model, was used to analyze the distribution of saturation line in the tailings dam under flood conditions. The numerical simulation results show that when the upstream water head rose to 125 m (Figure 15), the saturation line intersected with the first and fourth-level accumulation dams and was exposed throughout the dam surface. The variation law of the saturation line obtained by the numerical simulation is consistent with the experimental phenomenon; that is, the saturation line increases with the rise of the reservoir water level, and the order of the dam surface exposure is the second-stage sub-dam, the third-stage sub-dam, the first-stage sub-dam, and the fourth-stage sub-dam. According to the simulation results, the displacement of the dam body does not change greatly, and the plastic strain zone does not appear on the slope and crest of the dam body, and there is no penetration. It can be judged that the tailings dam model does not have deep slip when the water level is about to overflow; that is, the skeleton structure of the dam body is stable. Combined with the physical model test, before the saturation line of the dam body reaches the dam surface of the fourth-level sub-dam, the tailings dam has undergone seepage failure, but the dam body has not undergone structural instability. The results of numerical simulation are consistent with the phenomenon of physical model test. After that, with the development of flowing soil, the damaged area of the dam body continues to extend to the top of the dam, which will eventually cause the breach of the dam top and cause the flood discharge of the dam.
Figure 15. Saturation line distribution of the tailings dam under 125 m water level.
4. Impact Analysis after Dam Break and Prevention Suggestions
Based on the results of the physical model experiment, it can be inferred that the tailings dam failure was triggered by seepage failure. This means the area of flowing soil gradually eroded upstream until a breach was created at the top of the dam, and the reservoir fluid poured downstream. Therefore, an erosion damage trench was set up on the model for the dam breach calculation in FLOW-3D (ver. 9.3), extending from the top of the initial dam to the top of the dam, and the shape was simplified to a semi-cylinder. Figure 16 shows a model of the tailings dam after completion of the pre-treatment.
Figure 16. FLOW-3D 3D calculation model.
4.1. Dam Failure Test Results
An overview of the area downstream of the tailings dam is shown in Figure 17. The downstream area is dominated by the production facilities (red and yellow line areas in the figure), staff accommodation buildings (pink line area), the road around the mountain (blue curve), villages (green line area), and scattered agricultural land.
Figure 17. Aerial view downstream of tailings dam.
4.1.1. Overflow Area
Figure 18 shows the change in the extent of fluid inundation at 60 s, 120 s, 180 s, 240 s, and 300 s as calculated by the software, with the fluid in blue in the figure. As can be seen from the diagram, the breached fluid was rapidly released downstream in a short period and, by 300 s, covered the entire flat area downstream, with an overflow area of approximately 95,250,000 square meters. Farmland and roads in the area will be flooded, and production facilities and residential buildings will also be affected. In addition, emergency escape plans can be challenging to implement successfully at short notice. It is thus clear that in the event of a breach of this tailings dam, it would be a major accidental disaster.
Figure 18. Time-course diagram of mud area.
4.1.2. Flooding Depth
Figure 19 shows a cloud of the distribution of flooding depth at 60 s, 120 s, 180 s, 240 s, and 300 s. Due to the lower topography in the eastern part of the downstream area, the fluids that wash down first collect in the east and then spread westwards. As can be seen from the graph, the maximum inundation depth is always located in the eastern part of the lower reaches near the initial dam. The mudslide did not affect the northern area due to the terrain’s advantage; when the situation was urgent, people could be evacuated along the northwest-facing road to the north.
Figure 19. Cloud map of flooding depth.
4.1.3. Flow Rate Analysis
The flow velocity during the release process reflects the magnitude of the fluid impact. Figure 20 shows the flow velocity clouds during the dam breach release process at 60 s, 120 s, 180 s, 240 s, and 300 s. Due to inertia, the fluid emerges from the breach. It rapidly completes the transformation from potential energy to kinetic energy in the trench eroded by the flowing soil, with the flow velocity reaching a maximum. In addition, there is some leakage around the dam at the junction of the tailings dam and the mountain. After the fluid is flushed off the tailings dam, the average flow velocity decreases due to the diffusion principle and frictional forces. In general, the flow of emissions increases and then falls.
Figure 20. Cloud map of flow rate.
Three points, A, B, and C, are selected in the flow direction of the release to analyze the fluid’s flow velocity characteristics, specifically during the dam failure process. The three points are located at the top of the initial dam, the foot of the initial dam, and the downstream area adjacent to the tailings dam (Figure 21). Figure 22 shows the variation in flow rate over time at three points. Overall, the flow velocities at points A, B, and C are successively reduced as the flow path develops. From the point of view of the flow velocity at a single point, it does not increase to a peak all at once but has an undulating, phased variation. At about 30 s, the overflow velocity starts to appear at the three points, after which the trend is a cyclic process of “increase-smooth or decrease” because the increase in flow velocity does not coincide with the expansion of the breach, which, in turn, determines the flow velocity of the discharge. The flow rate increases accordingly when the breach expands and becomes deeper again. After several cycles of this until the breach is no longer extended, the flow rate at points A, B, and C all fall during the last 30 s of the figures and will return to zero as the flooding stops.
Figure 21. Flow rate reference points.
Figure 22. Flow rate time history diagram.
The analysis of the variation in the flow rate of the release shows that the debris flow impacts downstream in a segmented manner. Therefore, the decrease in flow velocity should not be regarded as the end of the entire dam break, nor should blindly carry out the aftermath of the accident at this stage, but should wait for a longer period to observe and confirm so as not to cause more damage.
4.2. Recommendations for Prevention and Management
4.2.1. Technical Measures
Dam surface treatment: According to the seepage characteristics of the tailings dam, to prevent overflow water and rainwater from scouring the shoulder and face of the dam and to collect the seepage water, a shoulder drainage ditch should be installed along the junction of the dam with the slopes of the two banks, and a face drainage ditch should be installed on the face of the dam. Moreover, the downstream slope of the dam can be mulched, turfed, and, if necessary, reinforced by stone pitching at the foot of the dam.
Additional seepage facilities: Combined with the model test results, it is clear that control of the saturation line of the tailings dam should be a top priority for safety management. To effectively control the depth of the saturation line, additional drainage facilities can be provided in the form of a combined horizontal drainage pipe and a vertical shaft connected to the end of the horizontal drainage pipe. In addition, the vertical drainage pipe should be raised with the height of the stockpile dam and pumped out periodically.
4.2.2. Management Measures
Routine inspection and maintenance: Besides monitoring various safety indicators such as the tailings dam saturation line and dry beach length, the person responsible for safety should regularly inspect the dam body for cracks, collapses, and surface erosion. They should also ensure that the slope protection is intact and that the drainage facilities are clear of blockages, siltation, or waterlogging. Check for seepage, pipe surges, or flowing soil, focusing on the junction between the dam and the hills on either side, and be vigilant for changes in seepage flow and turbidity. If a potential problem is identified, the cause must be immediately determined, and remedial action must be taken to prevent it.
Ensure excellence in flood management, including pre-flood preparation, response during flooding, and post-flood rescue work.
5. Conclusions
The reservoir’s water level had not yet crested before the dam was damaged. In other words, the cause of dam failure under flood conditions is seepage failure, which manifests itself in the form of flowing soil. Before the flow soil is destroyed, the dam surface will produce overflow, water accumulation, sand boiling, and other phenomena. The phenomenon of the dam failure test shows that the flow soil damage starts at the weak point of the dam at the junction with the bedrock. These areas have a high saturation gradient and are more prone to local damage. In the early stage of soil flow failure, multiple sand boiling points were generated on the dam surface. With the development of seepage, collapsible cracks appeared on the dam surface one after another, forming erosion ditches. In the middle stage of soil failure, the failure area is widened. The soil cycle undergoes the process of erosion–gravity erosion, and the ‘filter layer’ will slow down the failure rate to a certain extent. In the later stage of flow soil damage, the flow water damage area began to penetrate, and the erosion intensified until the whole dam body was damaged. Therefore, when the sand boiling point is generated, and the collapsible cracks appear on the dam surface, these can be used as a warning sign of seepage failure.
The buried depth of the saturation line becomes shallow with the increase in the upstream water head. And, the rate of increase is first fast and then slow. After the lifting head is stopped, the saturation line will still rise slightly for a period of time due to the lag effect. If the reservoir water level is not replenished for a long time, the saturation line will be reduced under normal drainage. The order of the saturation line escaping from the dam surface is the second sub-dam, the third sub-dam, the first sub-dam, and the fourth sub-dam. It can be seen that before the flood, it is necessary to check and repair the drainage facilities to ensure their suitable operation. During the flood season, all measures should be taken to enhance the flood discharge, reduce the saturation line, and avoid the seepage damage of the tailings dam.
The results of the FLOW-3D hydrodynamic simulation software show that the breach fluid was rapidly discharged within a short period, covering the entire flat area downstream by 300 s. The local farmland and roads were submerged, and the rest of the construction facilities were also damaged to a certain extent. Therefore, it will be a major disaster once the tailings dam breaks. The rapid development of the dam breach mudslide and the short release time make it impractical to organize the evacuation of people when the release occurs. Therefore, in combination with the mechanism of tailings dam failure, targeted measures for potential remediation and dam failure prevention can be proposed from both technical and management aspects.
The innovation point is to use a large-scale physical model test to study the dam break mode of tailings dam under flood conditions. By monitoring the internal changes of the tailings reservoir under flood conditions, the stage of seepage failure of the dam body can be judged, which can serve as an early warning for the subsequent break of the tailings dam. The experimental process and experimental results of the model can provide a reference for the changes in tailings reservoir under flood conditions under real working conditions so as to correspond to the changes of tailings reservoir fluid under flood conditions under real working conditions. Provide guidance for staff to monitor changes in tailings ponds. The determination of dam break position and dam break mode by model test provides a basis for simulating the influence of tailings dam break on the downstream. The use of a steel frame structure to build a tailings dam model can cover the entire tailings dam terrain more comprehensively and economically and can more comprehensively analyze the entire dam break process of the tailings dam. Compared with the local tailings dam similarity simulation and on-site exploration, it is more profound and comprehensive, which has practical significance for the safety of the tailings reservoir. The defect is that there is a prototype of the model, and it cannot be used for all tailings mines. The actual situation needs to be analyzed in detail. In addition, according to the tailings pond model test, it can be expected that the tailings pond model can be used to study the useful mineral components in the recovery reservoir, which has practical significance for environmental protection and resource recovery.
References
Yue, T.; Wu, X. Effect of magnetic starch on the clarification of hematite tailings wastewater. IOP Conf. Ser. Earth Environ. Sci. 2018, 121, 032051.
Jin, W.; Wei, Z. Research progress on land reclamation and environmental management benefit estimation of non-ferrous metal tailings reservoir. Environ. Sci. Res. 2019, 8, 1304–1313. (In Chinese)
Kang, Z.; Han, Q.; Wang, S. umerical model of dynamic stability of a proposed tailings dam To be studied. Min. Res. Dev. 2015, 35, 59–62.
Zhou, K.; Liu, F.; Hu, J.; Gao, F. Research on dam-break disaster chain and chain-breaking disaster reduction control technology of tailings reservoir. Disasterology 2013, 28, 24–29. (In Chinese)
Chen, C.; Zhao, Y.; Jiang, L. Review on the current research of dam break of tailings pond. Min. Res. Dev. 2019, 39, 103–108. (In Chinese)
Zhou, Z.; Li, X. Causes and ecological environment impact assessment of tailings dam break accident in China. Met. Mine 2012, 11, 121–124. (In Chinese)
Zhou, L.; Huang, Y. Discussion on the causes and countermeasures of tailings reservoir accident. China Manganese Ind. 2016, 34, 37–39. (In Chinese)
Yang, M.; Song, Z. A tailings reservoir based on flood calculation and flood routing analysis. China Met. Bull. 2020, 5, 263–264. (In Chinese)
Zhang, X.; Song, Z.; Zhang, Z. Example analysis of a tailings pond flooded with a dam failure. Eng. Constr. 2020, 52, 44–48. (In Chinese)
Zhang, C.; Chai, J.; Cao, J.; Xu, Z.; Qin, Y.; Lv, Z. Numerical simulation of seepage and stability of tailings dams: A case study in Lixi, China. Water 2020, 12, 742.
Sánchez-Peralta, J.A.; Beltrán-Rodríguez, L.N.; Trujillo-Vela, M.G.; Larrahondo, J.M. Flows of liquefied filtered tailings: Laboratory-scale physical and numerical modeling. Int. J. Civ. Eng. 2020, 18, 393–404.
Du, C.; Liang, L.; Yi, F.; Niu, B. Effects of geosynthetic reinforcement on tailings accumulation dams. Water 2021, 13, 2986.
Petkovšek, G.; Hassan, M.A.A.M.; Lumbroso, D.; Roca, M. A two-fluid simulation of tailings dam breaching. Mine Water Environ. 2021, 40, 151–165.
Geng, W.; Wang, W.; Wei, Z.; Huang, G.; Jing, X.; Jiang, C.; Tian, S. Experimental study of mesostructure deformation characteristics of unsaturated tailings with different moisture content. Water 2021, 13, 15.
Lolaev, A.; Oganesyan, A.; Badoev, A.; Oganesyan, E. Tailing dams formation algorithm. Arab. J. Geosci. 2020, 13, 974.
Wang, K.; Yang, P.; Yu, G.; Yang, C.; Zhu, L. 3D numerical modelling of tailings dam breach run out flow over complex terrain: A multidisciplinary procedure. Water 2020, 12, 2538.
Kansoh, R.M.; Elkholy, M.; Abo-Zaid, G. Effect of shape parameters on failure of earthen embankment due to overtopping. KSCE J. Civ. Eng. 2020, 24, 1476–1485.
Liu, K.; Cai, H.; Jing, X.; Chen, Y.; Li, L.; Wu, S.; Wang, W. Study on hydraulic incipient motion model of reinforced tailings. Water 2021, 13, 2033.
Piciullo, L.; Storrøsten, E.B.; Liu, Z.; Nadim, F.; Lacasse, S. A new look at the statistics of tailings dam failures. Eng. Geol. 2022, 303, 106–657.
Ma, C.; Gao, Z.; Yang, J.; Cheng, L.; Chen, L. Operation performance and seepage flow of impervious body in blast-fill dams using discrete element method and measured data. Water 2022, 14, 1443.
Fukumoto, Y.; Ohtsuka, S. 3-D direct numerical model for failure of non-cohesive granular soils with upward seepage flow. Comput. Part. Mech. 2018, 5, 443–454.
Issakhov, A.; Borsikbayeva, A. The impact of a multilevel protection column on the propagation of a water wave and pressure distribution during a dam break: Numerical simulation. J. Hydrol. 2021, 598, 126–212.
Dibike, Y.B.; Shakibaeinia, A.; Droppo, I.G.; Caron, E. Modeling the potential effects of Oil-Sands tailings pond breach on the water and sediment quality of the Lower Athabasca River. Sci. Total Environ. 2018, 642, 1263–1281.
Jing, X.; Chen, Y.; Williams, D.J.; Serna, M.L.; Zheng, H. Overtopping failure of a reinforced tailings dam: Laboratory investigation and forecasting model of dam failure. Water 2019, 11, 315.
Mahdi, A.; Shakibaeinia, A.; Dibike, B.Y. Numerical modelling of oil-sands tailings dam breach runout and overland flow. Sci. Total Environ. 2020, 703, 134–568.
Naeini, M.; Akhtarpour, A. Numerical analysis of seismic stability of a high centerline tailings dam. Soil Dyn. Earthq. Eng. 2018, 107, 179–194.
Boroomand, M.R.; Mohammadi, A. Evaluation of earth dam leakage considering the uncertainty in soil hydraulic parameters. Civ. Eng. J. 2019, 5, 1543–1556.
Li, S.; Yuan, L.; Yang, H.; An, H.; Wang, G. Tailings dam safety monitoring and early warning based on spatial evolution process of mud-sand flow. Saf. Sci. 2020, 124, 104–579.
Kong, X.; Liu, Y.; Li, X.; Li, R.; He, Y. Analysis and experimental study of mechanical factors of permeation damage of a uranium tailings pond dam. Eng. Saf. Environ. Prot. 2019, 45, 30–36. (In Chinese)
He, W.; Chen, H.; Zheng, C. Experimental study on dam break of a tailings dam under flood conditions. J. Saf. Environ. 2022, 22, 3126–3134. (In Chinese)
This study aims to investigate the impact of laser beam shaping on metal mixing and molten pool dynamics during laser beam welding of Cu-to-steel for battery terminal-to-casing connections. Four beam shapes were tested during LBW of 300 µm Cu to 300 µm nickel-plated steel. Both experiments and simulations were used to study the underlying physics. A CFD model was firstly calibrated against experiments and then deployed to explore the effect of the increasing ring-to-core diameter, as well as a tandem laser spot configuration. The study showed that metal mixing is influenced by the keyhole dynamics and collapse events, but also there is an intricate interplay between keyhole geometry, fluid dynamics via Marangoni forces and buoyancy forces. Notably, the buoyance forces due to the different densities of steel and Cu, along with the recoil pressure contribute to the upward flow of steel towards Cu, and hence impact meaningfully the material mixing. The study pointed-out that the selection of a custom ring-to-core diameter and ring-to-core power is a decision with a trade-off between the need of stabilising the keyhole dynamics and the need to reduce the mixing. Findings indicated that 350 µm ring and 90 µm core with 30% of ring power (weld configuration C3) resulted in more stable dynamics of the keyhole, with significant reduction of collapse events, and ultimately controlled migration of steel towards Cu. Additionally, the pre-heating approach with the tandem beam only led to local fusion of Cu and no significant improvement in keyhole stability was observed.
1. Introduction
The push towards net-zero mobility is globally influencing industrial strategies in the automotive sector as reported by IEA (2022). Manufacturers are introducing new vehicles by replacing internal combustion engines with hybrid or fully electric powertrains. The battery pack is a critical component for un-interrupted supply of electricity to e-drives and other electrical systems in electric vehicles (EV). A battery pack typically consists of several battery modules that are electrically connected in series and parallel based on the desired power and capacity requirements (Zwicker et al., 2020). Battery modules hold the battery cells that store the electrical charge and supply it on-demand to the electrical systems. Electrical connections play a critical role in the entire process of battery pack manufacturing since joints with different electrical resistance may result in uneven current loads that can affect the overall performances of the battery system (Kumar et al., 2021). Joining of dissimilar materials is the most deemed since it complements the properties of the individual materials and allows to develop functionally efficient connections. Joints in EV battery pack involve low-thickness materials (typically 0.3–1 mm) and the welding process is normally performed in lap or fillet configuration. Depending upon design and functional requirements as well as manufacturing costs, research has shown that the following combinations of materials are the most regarded: aluminium (Al) to copper (Cu), steel to Al, Al to steel, Cu to steel (Das et al., 2018). Connections between Cu and steel have gained much attention in EV applications for joining cells in battery modules. For example, in the cylindrical format, the negative terminals are made of Cu and are generally connected to the steel casing of the cell (Sadeghian and Iqbal, 2022). Several joining processes have been studied for Cu-to-steel welding and they include wire bonding, micro-spot welding, ultrasonic welding, micro-TIG welding, electron beam welding and Laser Beam Welding (LBW) (Zwicker et al., 2020). LBW is an attractive option and has recently gained popularity due to advances in versatile methods for laser beam delivery and associated sensors technology for quality control and process monitoring that make LBW comparatively affordable (Kogel-Hollacher, 2020). Brand et al. (2015) demonstrated that LBW is a suitable process for joining battery terminals since it allows the lowest electrical resistance and the highest joint strength, when compared to micro-spot welding and ultrasonic welding; also, it is potentially applicable to any cell configuration and dissimilar metal combinations. Despite the benefits of LBW, opening and maintaining a stable molten pool on the Cu-side is challenging when using LBW with infrared sources. The absorptivity of Cu at ambient temperature is approximate 5% and increases with rising temperature, and it suddenly jumps up when the melting temperature is reached. A problem with this is that when fusion of the material does happen, a surplus of energy flows through it, which can vaporise the material and create spatters, as well as pores inside the joint. These defects can reduce the electrical conductivity of the joint. At first sight, the solution to the low coupling efficiency of Cu is to switch from infrared sources to visible sources. The absorption increases drastically up to 60% when using visible sources. Green (515 nm) or blue (450 nm) lasers have been investigated by Kogel-Hollacher et al. (2022) and proved that lower power needed for same penetration achievable with infrared lasers and less thermal damage to enamel and insulators. Hummel et al. (2020) experimentally evaluated and proved the beneficial effects of blue laser during laser micro-welding of Cu, and achieved high welding speed with low input power. Nonetheless, compared to infrared lasers, the higher cost, lower plug efficiency and lower beam quality of visible lasers, push practitioners towards the use of multi-kW infrared sources at very high brightness for Cu welding. In addition to the challenge posed by the laser beam coupling to the Cu, the welding of Cu to steel presents a series of problems. First, they are quite different in terms of physical properties such as density, melting points and thermal expansion and make defect-free welding difficult. Second, although Cu-Fe alloys are completely miscible in the stable liquid state and do not form brittle intermetallic compounds, the system shows a wide metastable miscibility gap at an undercooling level. The liquid phase separation occurs as the liquid cools in the miscibility gap resulting in the supersaturation of one or both liquids. Jeong et al. (2020) has shown that increasing the content of Fe tends to improve the mechanical properties of alloys but reduce electrical conductivity and ductility. Chen et al. (2013) proved that the toughness and fatigue strength of the joint decreases with the increase in the amount of molten Cu into the steel. Thus, melting of Cu was suggested to be kept at a minimum. Third, excessive penetration of Cu in grain boundaries of steel may result in cracks in the heat affected zone and fusion zone, and ultimately reducing structural performance of the joint. Therefore, to reduce these issues, controlling the mixing of Cu and steel in the molten pool is quite important for producing sound joints. Laser beam shaping is gaining popularity since it holds the promise to control cooling rates and thermal gradients in and around the molten pool. This theoretically leads to a tailored material response to the heat input both spatially and temporally. A tailored power density profile (Fig. 1 shows typical power density profiles obtained via adjustable ring-mode laser) is generated via adequate insertion of optical components (specially coated lenses of silica substrate) in the optical chain of the welding head; or by electro-optical switching multiple laser beams generated in the laser source itself and enabled by beam combiners with optical phased array. Research has confirmed a positive effect of the laser beam shaping on the control of the weld profile and keyhole stabilization with suppression of spatters and significant reduction of porosity in the weldments. Caprio et al. (2023) investigated the use of beam shaping and beam oscillation to weld 0.2 mm Ni-plated steel sheets in lap joint configuration, which are materials commonly involved in cell to busbar connections. Sokolov et al. (2021) employed the ARM laser coupled with Optical Coherent Tomography (OCT) in Al-to-Cu thin sheets and observed that the use of combined core and ring-shaped laser beams reduced the fluctuations of the keyhole, improved the stability, and ultimately the accuracy of OCT measurements. Rinne et al. (2022) studied the effect of different power distributions between the inner core and outer ring-shaped laser beams on spatter ejection and penetration depth during welding of Cu sheets. Wagner et al. (2022) investigated and proved the influence of dynamic beam shaping on the geometry of the keyhole during welding of Cu by varying the patterns of the intensity distribution in longitudinal and transversal direction. Prieto et al. (2020) implemented dynamic laser beam shaping with infinite pattern and assessed quality of weld seam in 0.8 mm Al thin-sheet and observed that tailored beam with shape frequency over 10 kHz enables welding speed up to 18 m/min with stable keyhole.
Fig. 1. Example of laser beam shapes obtained via an adjustable ring-mode laser.
Despite the benefits, laser beam shaping introduces new set of parameters and finding the optimal combination of number of beams, shape of beams (multiple spots, C-spot, ring-core spots, pyramid, infinity, spiral shapes, etc. (Prieto et al., 2020)) can be expensive and time consuming since it may require dedicated equipment, expertise and experimental setups. In this context, multi-physics computational fluid dynamics (CFD) enable simulations of the process to reproduce mechanisms which are difficult to observe with in-situ investigations. With the raise of computational power and multi-core computing on high performance clusters, advanced simulations of LBW processes are now a close reality. Huang et al. (2020) developed a CFD model in FLOW-3D WELD® to study the metal mixing during linear laser welding of 200 µm Al to 500 µm Cu with different levels of laser power and velocity of the laser spot. They analysed the contribution of recoil pressure and Marangoni effect on the overall mixing process. Chianese et al. (2022) developed a multi-physics model using FLOW-3D and FLOW-3D WELD® to investigate the effect of part-to-part gap in LBW of Cu-to-steel thin sheets with beam wobbling. They showed that the presence of part-to-part gap and mixing mechanism between parent metals are linked, and the occurrence of part-to-part gap influences the temperature and velocity fields in the molten pool resulting in different mixing mechanisms. However, they did not implement any strategies for weld improvement. Drobniak et al. (2020) and Buttazzoni et al. (2021) implemented CFD multi-physics simulations of 1 mm-thick stainless steel plates with adaptive mesh refinement to predict the shape of the weld seam in presence of part-to-part gap, and they predicted the effect on the process of secondary laser beams with different shapes to optimize the weld quality. Recently, Huang et al. (2023) combined experimental approach and CFD simulations in FLOW-3D WELD® to reveal the effect of oscillation frequency and amplitude on fluid-flow and metal mixing during laser welding of 200 µm Al to 500 µm Cu with circular beam wobbling implemented. Additionally, they implemented a Scheil solidification model to predict the phase distributions in the welds based on the predicted thermo-solute conditions. While significant research has been already developed using linear laser welding or laser welding with wobbling for joining of dissimilar materials, a clear understanding of metal mixing and dynamics of the keyhole during Cu-to-steel welding with beam shaping are not clearly reported. Research into application of beam shaping for Cu-to-steel welding entails a promising prospect for further development and investigation. Furthermore, the use of advanced CFD models is a viable approach to complement experimental investigations and explore weld configurations with different beam shaping profiles that would be difficult to achieve only with experimental work. Therefore, this paper aims to study the impact of laser beam shaping on metal mixing and dynamics of the keyhole during LBW of Cu-to-steel for battery terminal-to-casing connections. Four beam shapes were tested during LBW of 300 µm Cu to 300 µm nickel-plated steel. Both experiments and CFD simulations were used to study the underlying physics. A CFD model was firstly calibrated against experiments and then deployed to explore the effect of the increasing ring-to-core diameter, as well as a tandem laser spot configuration.
2. Experimental design and model description
2.1. Experimental design
Materials used in this work are Copper SE-Cu58 2.0070 and Nickel-plated steel (commercial name: Hilumin TATA STEEL). Experiments consisted of 25 mm long welds in lap joints configuration with 300 µm Cu on top of 300 µm nickel-plated steel. Dimensions of the specimens were 65 mm × 30 mm. The laser source used was the Lumentum CORELIGHT, having 55 µm core diameter and 220 µm ring diameter, and BPP 1.4 mm·mrad and 11 mm·mrad for core and ring, respectively. The laser fiber was coupled to the Scout-200 (Laser and Control K-lab, South Korea) scanner to deliver the laser power to the specimens via 2D F-theta scanner with telecentric lenses. Fig. 2 shows the welding setup and specifications of the equipment are in Table 1. Caustic parameters were measured using PRIMES GmbH measurement system.
Fig. 2. (a) Welding setup with aluminium fixture; (b) schematical representation of the welding setup; (c) definition of weld features: top weld width, Wtop; width at the interface, Wi; weld penetration depth, Dpen.
Table 1. Specifications of the welding equipment.
Each weld seam was cut and prepared to obtain two cross sections for each experiment – cross sections were positioned at 10 mm and 15 mm away from the weld start. Three replicates were performed for each weld configuration. Sectioned samples were mounted in Bakelite resins and standard metallography procedure was performed for grinding and polishing to reveal weld profile under Nikon Eclipse LV150N optical microscope. To evaluate and characterize metal mixing with parent metals, elemental mapping of cross-sections was performed with an FEI Versa 3D dual beam scanning electron microscope using Energy Dispersive X-ray Spectroscopy (EDS mapping). Welding experiments were performed in continuous power mode without power modulation. The laser beam was focussed perpendicularly on the upper surface of the Cu sheet, and the motion of the laser was linear (no wobbling). Although the use of shielding gas tends to avoid oxidation in the process and reduce hydrogen entrapment, when using scanners to deliver the laser beam, the gas nozzle cannot be positioned in proximity of the beam. Therefore, in this work, all experiments were conducted with no shielding gas. Part-to-part gap was manually checked and set to a nominal zero. To study the impact of laser beam shaping on metal mixing and molten pool dynamics, 5 weld configurations (C1 to C5) were designed as shown in Table 2, with 4 beam shapes presented in Fig. 3. LBS#1 is single gaussian spot of 90 µm; LBS#2 super-imposes an inner core of 90 µm with an outer ring-shaped profile of 350 µm, with the ring accounting 30% of the total power. LBS#1 and LBS#2 were experimentally tested and enabled by the static beam shaping system of the Lumentum CORELIGHT source. LBS#3 follows the hollow sinh-Gaussian beam profile as defined in Liu et al. (2019), with 90 µm core and 500 µm ring, with 72% of the total power assigned to the ring. LBS#4 is a tandem beam with primary (90 µm) and secondary beam (150 µm) at a centre-to-centre distance of 300 µm, and 50% split of the power between primary and secondary beams – LBS#4 was introduced with the aim to increase the absorption rate by the pre-heating action of the secondary beam. LBS#3 and LBS#4 were only simulated since the laser beam shaping of the Lumentum CORELIGHT was only capable to work with fixed core-to-ring diameter ratio. Therefore, only a simulation-based approach (with the model pre-validated and calibrated in C1, C2 and C3) was deemed appropriate in this case to explore the effect of the increasing ring-to-core diameter and tandem laser spot configuration on material mixing.
Table 2. Process parameters used for the four selected laser beam shapes in Fig. 3.
Fig. 3. Normalized power density distribution for LBS#1, LBS#2, LBS#3 and LBS#4.
The power and speed of C1, C3, C4 and C5 were selected with an iterative process to ensure weld penetration depth, Dpen, ranging 400 – 500 µm. The choice of this penetration depth is based on the requirement that the temperature at the lower end of the steel sheet remains below 550 K. This precautionary measure aims to prevent any potential damage to the battery cell. Additionally, to minimise the effect of the weld depth on the metal mixing, a uniform depth of penetration was adopted across the different beam shapes for comparative analysis. Welding speeds were kept between 250 mm/s and 375 mm/s which is in line with the experimental work in (Perez Zapico et al., 2021). C2 is a variant of C1 and corresponds to a fully penetrated weld. Although fully penetrated welds must be avoided during LBW of battery terminals due to the risk of fire ignition, this work presents this variant for two reasons: first, to generate an additional weld configuration to validate the simulation; second, to discuss how the metal mixing behaves when transitioning from partial penetration to full penetration.
2.2. Model description
A multi-physics model was developed using the commercial CFD code FLOW-3D® (solver version: 12.0.2.01) and its module FLOW-3D® WELD (release: 7, update: 1). In order to develop a numerical model representing the essential physics during LBW of Cu-to-steel, the following assumptions were considered: (i) the liquid flow is considered Newtonian and incompressible; (ii) volumetric thermal expansion of the liquid metal due to temperature-dependent mass density is accounted; (iii) the air and vaporized metal are modelled as “void” type, with ambient temperature and pressure assigned to model the heat exchange with the metal as a natural convective flux (irradiance is neglected); (iv) the heat sinking effect of the clamping mask is neglected due to the clearance between the weld seam and the mask itself as already presented in (Chianese et al., 2022); (v) the effect of plasma plume on laser absorption is not directly modelled but is accounted in the calibration process as also proposed in previous studies by Lin et al. (2017) and Hao et al. (2021); furthermore, the laser absorption is assumed temperature dependent for Cu, constant for steel, and independent of the incidence angle. This assumption is in-line with the work presented by Huang et al. (2020), where they used the build-in ray-tracing function in FLOW-3D® WELD to predict the laser absorption in the keyhole.
2.2.1. Governing equations, boundary conditions and material properties
To reduce the computational cost of the simulations, the computational domain was divided in two zones (Fig. 4): (1) a process zone which was interested by phase change, and, (2) a thermal diffusion zone that models heat transmission in the sheets. A finer mesh size was used for cells in the process zone, and a mesh size 5 times greater than in the process zone was used for cells in the thermal diffusion zone.
Fig. 4. Top view (a) and side view (b) of a schematic representation of the computational domain and modelling approach with nested meshes (process zone and thermal diffusion zone).
Dimensions of the process zone are 2 mm × 0.8 mm× 0.775 mm. The length (2 mm) of the process zone was chosen to enable the simulation of approx. 1.8 mm weld length, which was experimentally evaluated to be sufficient for reaching the steady-state regime. The width (0.8 mm) of the process zone was selected to ensure that the molten pool was contained in it; the height of the computational domain was chosen equal to 0.8 mm so that, beside the stacked thickness of the processed sheets (0.6 mm), 0.2 mm of air (void type) are included in the computational domain. Extension of the thermal diffusion zone is calculated according to the Eq. (1), where k is the thermal conductivity, cp the specific heat at constant pressure, ρ the mass density, tend the simulation time, T the temperature, and Tamb= 20 °C the ambient temperature. The simulation time, tend, is function of the welding speed and the weld length (1.5 mm).
Four different values of the mesh size in the process zone were considered during sensitivity analysis, namely 40 µm, 20 µm, 15 µm, and 10 µm, that resulted in mesh independent solution for mesh size equal to or below 15 µm, which therefore is the selected size. This led to total number of cells approximatively equal to 528 thousand. The geometry of the thin sheets has been modelled in the computational domain, so that in-plane dimensions were parallel to X and Y axis, as shown in the top and side view in Fig. 4(a) and (b). Welding direction was parallel to X axis. The following physics have been accounted to model the welding process: continuity, fluid flow via Navier-Stokes equations, energy conservation, evaporation, keyhole formation and evolution, solidification, species conservation and tracking, surface tension with Marangoni and Laplace forces and multiple reflections. Phase change – Eq. (2) governs the evaporation phenomena which are modelled as mass transfer between the liquid phase and the void type and are proportional to the difference between the saturation pressure Psat and the partial pressure Pvap. In this equation, α is the accommodation coefficient, R is the gas constant, and T is the temperature. The saturation pressure is calculated as a function of the temperature according to the Clapeyron equation (Eq. (3)), in which the couple (Pv, Tv) represents a point on the saturation curve; γ, cv, and ΔHv are the specific heats ratio, the specific heat at constant volume, the latent heat of vaporization, respectively.
Recoil pressure – during laser welding process, intense localised heating of substrate material causes vaporization which results in recoil pressure. This pressure is proportional to the saturated vapor pressure. The relationship between the recoil pressure, Precoil, and the saturated vapor pressure, Psat, depends on the material properties and laser-to-material interaction. Eq. (4) is derived from Eq. (3) with the introduction of two coefficients, Ar and B, that will be calibrated using experimental data.
Tracking of the keyhole – surface of the keyhole is tracked by the volume of fluid (VOF) method (Daligault et al., 2022), which enables the calculation of the interface between the liquid metal and the void type, according to Eq. (5).
The interface between the cell is tracked using a scalar value f that indicates the fraction of fluid in it. A value of f=0 indicates that the cell has only void, conversely, f=1 corresponds to the case of a cell full of liquid, whereas the case of 0<f<1 indicates that the cell has both the liquid and the void type, and therefore the interface between the two falls in it. Similarly, metals involved in the welding process with fluid flow and mixing are tracked in each cell by means of a scalar value f2, which indicates the fraction of second material within the cells. Values of the generic material property ̅φ̅ in each cell is evaluated as weighted sum of the properties φ1 and φ2 of parent metals based on their mixing, as in Eq. (6).
Multiple reflections – Multiple reflections are implemented using a discrete grid cell system through the ray tracing technique. The laser beam is divided into a finite number of rays, which move in the laser beam irradiation direction. When the ray encounters the surface of the material, it is reflected according to vector Eq. (7), in which R→ is the direction of the reflected vector, I→ the direction of the incoming ray, and nˆ the normal direction of the material surface.
Laplace pressure and Marangoni effect – Recoil pressure contributes to the formation of the keyhole and mainly contributes to the velocity field in the fluid; however, surface tension-related phenomena such as Laplace pressure LP and the Marangoni force SM have great influence on the overall welding process. Laplace pressure and the Marangoni force are modelled according to (8), (9) which, σ is the surface tension, RI and RII are the principal curvature radii, and operator ∇t indicates the gradient along the tangent direction at the interface. Eq. (9) explicitly indicates the dependence of the Marangoni effect on the gradient of the surface tension, which in assumed temperature-dependent of the surface tension.
2.2.2. Boundary conditions and material properties
As shown in Fig. (4), the following boundary condition were assigned: wall in the X and Y direction (with constant ambient temperature); assigned pressure and temperature at the boundaries of the computational domain in the Z directions, with natural convective heat flux between the metallic sheets and the air. The heat source was directly imported from the power profiles defined in Fig. 3. Material properties were imported from the JMATPRO® material database. Fig. 5 shows the temperature-dependent plots.
Fig. 5. Temperature-dependent material properties defined in the model.
3. Results and discussion
3.1. Model validation
The model has been applied to simulate all the cases listed in Table 2. Model validation was conducted for the weld configurations C1, C2 and C3 by comparing the weld profile in cross sections and Fe concentration line profiles against the experimental results as shown in Fig. 6. Experimental and simulation results show that welding is done through keyhole mode. The generation of a keyhole is significantly influenced by recoil pressure. In the simulation, the recoil pressure is adjusted through the calibration of coefficients Ar and B, as indicated in Eq. 4. During the model calibration process, a value of Ar was determined to be 55,715 Pa, and the parameter B was set to 4, resulting in comparative results with those obtained in experiments. Five different mesh sizes were tested: 20 µm, 15 µm, 10 µm and 5 µm. The choice of the mesh size was driven by the need to have a minimum of 4 cells to discretise the smallest laser spot (i.e., LSB#1 has the smallest beam diameter of 90 µm among the tested beam shapes in Fig. 3). Mesh-independent solution was achieved with mesh size of 15 µm and this led to approximate a million cells in the whole computational domain.
Fig. 6. Comparison of the experimental and modelling results of the molten pool geometry and elemental maps for weld configurations C1 (a), C2 (b) and C3 (c).
The correlation was conducted looking at two cross-Section (10 mm 15 mm away from the weld start and end) – this was motivated by the need to take into account the experimental errors during the calibration and validation process.
Fig. 6 shows cross sections and elemental maps for experiments C1, C2, and C3, and corresponding simulations. Two representative cross-sections from the same weld seam are shown in each sub-figure to demonstrate the capability of the model to reproduce the geometric shape and the mixing phaenomena at different longitudinal positions along the weld seam. The fusion zones are marked in each cross section and show good correlation with predictions from simulations, as the cases with partial penetration are successfully predicted in for C1 and C3, along with full penetration in C2.
Elemental maps that were measured with EDS, and species concentration that were predicted with simulations, are reported for comparison to show capability of the model to reproduce the mixing mechanism. For each case, plots of the concentration of Fe along with line-scans are reported to quantitatively demonstrate the capability of the model to simulated diffusion of the molten metal from the bottom sheet to the upper one. They show that diffusion of Fe in Cu is well predicted in C1 and C3, as well as presence of Fe-rich clusters in the Cu near the interface between parent materials is reproduced in C2.
Good correlation between measurements and predictions of the weld geometry and metal mixing demonstrates capability of the model to simulate welding scenarios with different laser beam shapes, and weld penetration depth spanning from partial penetration to full penetration. This allows to confidently deploy the simulation model in conjunction with experiments to study the impact of laser beam shaping on metal mixing and molten pool dynamics.
3.2. Keyhole dynamics and impact on metal mixing
As keyhole instabilities have a significant impact on weld quality (Lu et al., 2015), this section highlights the impact of the laser beam shapes on the keyhole dynamics, which ultimately contributes to metal mixing. The discussion is presented by linking the laser power profile to the velocity field within the molten pool and ultimately to the metal mixing between the parent metals and the occurrence of collapse events of the keyhole.
Fig. 7 shows consecutive time frames in each weld configuration and reflects keyhole dynamic mechanisms. The keyhole’s shape and size vary, exhibiting irregularities, asymmetry and fluctuations. These shapes are directly correlated to the laser beam shape profile. The following observations are made:
Collapse events terminate in formation of pores and metal mixing. This is visible in the experimental results presented in Fig. 6(a) and (b), where relatively large pores are observed in the experimental cross-section. With a narrow beam profile (weld configuration C1, C2, C3 and C5) and high energy density, once fusion of the Cu does happen, a surplus of energy flows through the keyhole, increasing the temperature at the keyhole bottom. This generates a recoil pressure that pushes the fluid upwards. At the top surface and rear side of the keyhole, the opposing movements of the fluid, both clockwise and counter-clockwise, and driven by the Marangoni force, have an important consequence: they restrict the size of the molten pool. This restriction creates a high viscosity mushy layer that forms a barrier that limits the expansion of the molten pool. As result, closure or narrowing the top neck of the keyhole restricts the ejection of vapours out of keyhole which leads to increase in pressure within keyhole and creates a high-pressure lob. This ultimately results in pores formed to the toe of the keyhole as seen in Fig. 7(a) and (b). Although a collapse event is observed in C3 as shown Fig. 7(c), it does not necessarily create porosity in the solid front as sufficient room is available for gas vapours to escape from the bottom of the keyhole. The introduction of a pre-heat heating beam in weld configuration C5 does not produce any significant change to the keyhole dynamics as observed in Fig. 7(d). In partial penetration, narrow and deep keyhole is more unstable as slight fluctuations in fluid pressure, velocity and temperature on the rear wall of keyhole can create a collapse event. Additionally, the collapse of the keyhole in partial penetration creates a narrower fluid channel, resulting in localized increase of fluid velocity, which, in turn, affects metal mixing.
Weld configuration C4 leads to wider opening of the keyhole with greater stability as shown in Fig. 7(e). With the super-imposition of the core beam with the wider ring-shaped beam, the core beam penetrates the steel sheet, while the larger ring keeps the keyhole open at the Cu surface. This weld configuration drastically reduces the collapse events and the development of bubbles. It can be observed that the lower depth-to-width aspect ratio of the melt pool correlates to fewer number of collapse events.
Metal mixing is not only influenced by keyhole dynamics and collapse events, but there is an intricate interplay between keyhole geometry, fluid dynamics and buoyancy forces that are dependent upon density which varies with temperature in molten pool, and from top to bottom due to differences in density between Cu and steel. To test the influence of buoyancy forces, a simulation test was performed where the density of Cu and steel were artificially set to be equal. Fig. 8 shows the simulation results and confirm that buoyancy forces have an impact on the metal mixing especially at the interface between the two metals and in the Cu side of the weld. For example, the line-scan B-B in Fig. 8 shows an increase on average of the Fe vol% in the Cu side by 10%, when comparing results with same densities.
Fig. 8. Impact of buoyancy forces on the metal mixing for weld configuration C3. Sections taken at Y= 0.
Fig. 7. Consecutive time steps of the molten pool dynamics for configuration C1 (a), C2 (b), C3 (c) C5 (d) and C4 (e). The plot shows the fluid velocity (both direction and magnitude) visualized by black arrows. Cross sections taken at Y= 0.
The introduction of a ring beam (weld configuration C4 with LBS#3) in the laser welding process alters the shape of the keyhole compared to a single beam scenario (weld configuration C1 with LBS#1). In the single beam case, the keyhole walls develops predominantly in Z direction (schematically illustrated in Fig. 9(a)). The inclusion of a ring beam results in the critical change of the keyhole wall’s curvature, with a pronounced arc-like shape at the rear (Fig. 9(b)). The change of keyhole wall’s curvature plays a critical role and is explained by the complex equilibrium between the fluid pressure, the recoil pressure and the gravity load. A collapse event is associated with the non-equilibrium of the forces in the X direction. To explain this, it is first worth noting that with an idealised static molten pool (no fluid velocity) the fluid pressure would be higher at the bottom and would be governed by the hydrostatic law – with this, the pressure variation occurs linearly downwards and would be a function of the molten pool depth. Under this ideal condition, the keyhole would exhibit a stable equilibrium regime driven by the balanced effect of recoil pressure and fluid flow. With the actual molten pool, the equilibrium state is, however, perturbated by the non-linear variation of the fluid pressure due to the fast upwards motion generated by the recoil pressure itself. A near-equilibrium state is eventually achieved with the change of keyhole wall’s curvature with the resultant of the forces acting predominantly in the Z direction. The shallow angle of the keyhole wall observed at LBS#3 (θ3 < θ1) effectively decomposes the combined forces exerted by the fluid towards the Z direction, hence moving to the near-equilibrium state, with the fluid pushed downwards in Z rather than sidewise in X. It can be observed that the ring-to-core diameter and the ring-to-core power are essential to control the keyhole wall’s curvature and ultimately influence of the stability of the keyhole.
Fig. 9. Schematic representation of forces and pressures acting on the melt pool in case of welding with single laser beam (LBS#1) and ring-core configuration (LBS#3). Arrows represent forces/pressures, and the thickness is proportional of the intensity of the forces/pressures. Arrows are only shown to the rear-side of the keyhole since the physics involved there are more relevant for the dynamics of the keyhole.
3.3. Impact of beam shaping on metal mixing
Cu and steel are generally immiscible as studied by other researchers, such as Shi et al. (2013). This separation means the material solidifies as two separate phases from the liquid state. At this immiscible region a Cu-rich (α phase) and iron-rich (β phase) form FCC and BCC crystal structures, respectively. For the compositional data shown in Fig. 6, the highest amount of mixing for each of the three examples is 60%, 80% and 50% of Fe in the weld pool. When studying the Cu-Fe binary phase diagram, as performed by Chen et al. (2007), these compositions fall within the miscibility gap range. For which no IMCs are expected to form, but instead separate (α and β) phases. However, it is still clear that the formation of these separate phases still creates a mismatch in mechanical properties of the welded joint, both at the interface and enriched regions, which can lead to crack initiation, as reported by Rinne et al. (2020). For this reason, analysing the metal-mixing in dissimilar metals is an important step toward understanding and prevention of cracking mechanisms that can affect the performance of the weld. Influence of the beam shapes on the metal mixing, can be investigated by analysing velocity fields and fluid flow which are predicted with the validated model. Fig. 10(a) and (b) show that in the weld configuration C1 and C2 (corresponding to LBS#1 – single beam with circular spot and gaussian distribution) the increase in laser power leads to more steel mixing with Cu due to greater recoil pressure and to a larger melt pool with more liquid metal involved. When comparing the parameters in Fig. 11, the increased melting of the bottom steel sheet leads to a greater region of keyhole necking with collapse; this can be due to the increased laser absorption, for which steel has a greater absorptivity than the more reflective Cu (Rinne et al., 2020). The lower density of steel creates an upward buoyancy force which allows the migration of more steel into the Cu-rich region. Fig. 11(c) and (d) show weld configurations C3 and C4 respectively, with combined secondary ring-shaped and primary laser beam (LBS#2 and LBS#3, respectively). They can be compared based on similar levels of weld penetration but different width at the interface between parent metals and at the top of the weld seam. Spread of the laser power over a wider surface due to the use of a ring results in a wider weld pool compared to simulations C1 and C2, which is consistent with results found by Jabar et al. (2023). However, one difference between these two cases is that, due to different power density distributions, to achieve adequate weld penetration depth, different laser power is provided leading to different thermal fields and time that the metal stays liquid. Line-scans of the temperature profiles in the melt pool can be observed in Fig. 12, with higher peak temperature in C4, compared to simulations C1 and C2, and C5; whereas a smaller secondary ring-shaped laser beam in simulation C3 results in intermediate behaviour.
Fig. 10. Plots of metal mixing in the longitudinal and a cross sections predicted with simulations C1 (a), C2 (b), C3 (c), C4 (d) and C5 (e).
Fig. 11. (a) Temperature, (b) velocity, (c) Fe concentration and (d) actual melt pool for all the tested weld configurations C1 to C5. Cross sections taken at Y= 0.
Fig. 12. Temperature profiles for weld configurations C1 (a), C2 (b), C3 (c), C4 (d) and C5 (e). Measurements were taken at X = 1.3 mm (just behind the keyhole wall) and Z = −300 µm (interface between Cu and steel).
The higher peak temperature in C4 eventually leads to a significant thermal gradient that promotes significant upward buoyancy forces and ultimately more migration of steel towards the Cu matrix. Similarity of simulation C5 with C1 can be explained considering that the secondary laser beam pre-heats the metal without widening the keyhole. Additionally, the higher peak temperature and larger size of the melt pool in C4 lead to longer time in which the steel stays in the liquid phase with more time available to migrate toward the Cu matrix due to recoil pressure and buoyancy forces and to diffuse. For these reasons, if use of larger spot helps with keyhole stabilisation, higher laser power required to establish sound connection enhances mixing between parent metal. Therefore, selection of custom ring-to-core diameter and ring-to-core power is a decision with a trade-off between the need of stabilising the keyhole dynamics and the need to reduce the mixing. Velocity fields in Fig. 11 show also that the use of the ring-shaped secondary beam (C4), results in lower recoil pressure due to less localised laser power and vaporization. For this reason, the fluid flow and velocity of the liquid movements in considerably lower, as shown by contour plots, where regions of the molten pool in red are those in which the flow of the liquid metal is faster. The metal mixing in the molten pool of C3 weld is more homogeneous than in C1 and C2, due to the localised heat input of the ring laser beam. Rinne et al. (2020) found the addition of the ring laser produced a more homogeneous distribution of Cu and steel in the solidified structure. The lower density of the steel can also be used to explain the more even distribution of steel throughout the weld pool of C3. This is also confirmed by the EDS line-scans in Fig. 6(c) that show a significant drop of Fe into the Cu matrix compared to C1 (Fig. 6(a)). The result of metal mixing has a significant effect on the crack formation in the weld pool and heat-affected zone (HAZ). Two main types of cracking are often referred to as “hot cracking” (Rinne et al., 2020) or “liquation cracking” (Li et al., 2019). During any fusion welding process of Cu to steel the miscibility gap can be identified in the binary phase diagram of Cu-Fe (Chen et al., 2013). When both Cu and steel are melted, there is separation of the liquids during cooling, once the mixture enters the miscibility gap seen on the phase diagram the primary separation of the α and β phases occurs. The secondary separation occurs in the miscibility gap because of a lack of diffusion and a supersaturation of the α and/or β phases. The solidified weld microstructure is found inhomogeneous, consisting of the α and β phases. The difference in the thermal expansion properties of both Cu and steel can create locations of stress concentrations where cracks are often initiated, ad observed by Chen et al. (2013) and Sadeghian and Iqbal (2022). Li et al. (2019) proposed a three-stage mechanism for the formation of liquation cracks in Cu to steel laser welds. The first stage was the penetration of Cu liquid into the grain boundaries of the steel, secondly, the Cu liquid surrounds the Cu phase creating a “film” of liquid in the grain boundary. This drastically reduces the cohesive forces between the grain boundaries due to the presence of the α phase. Cracking can then be initiated in a similar manner to that detailed earlier.
4. Conclusions
A combination of multi-physics CFD modelling results and experiments have been presented to study the impact of laser beam shaping on metal mixing and molten pool dynamics during LBW of Cu-to-steel for battery terminal-to-casing connections. The multi-physics model has been validated with ex-situ EDS element mapping and weld profile’s features. The model has provided useful insights about temperature and velocity fields, mixing mechanisms and dynamics of the keyhole, all of which are difficult to access via experiments due to technological difficulties. The major findings of the work are summarized below:
Metal mixing is largely influenced by the fluid dynamics via the Marangoni, buoyancy forces and recoil pressure. With a greater laser power, recoil pressure is increased, and this leads to more weld penetration and melting of steel. Additionally, spread of the laser power results in higher width of the fusion zone. Subsequently, the buoyance forces due to the different densities of steel and Cu contribute to the upward flow of steel towards Cu, and hence impact meaningfully to the mixing. This can be clearly observed in weld configurations C1 and C2.
Due to the collapse events of the keyhole wall, porosity formation was found in welds C1, C2 and C5. Furthermore, the collapse events create a narrow fluid channel, which results in localised surges in fluid velocity, therefore, promoting metal mixing. All in all, simulations revealed that increasing depth-to-width aspect ratio is correlated to higher frequency of collapse events in the keyhole. Therefore, stabilisation of the melt pool can be achieved with tailored laser beam shapes.
The study has pointed-out that the use of larger ring beam (configuration C4) helps with keyhole stabilisation, but at the same time leads to more laser power and higher temperature that contribute to the enhancement of mixing between parent metals. This poses a trade-off in the definition of a tailored ring-to-core diameter and the ring-to-core power. Analysis of the results showed that ring-to-core diameter (350–90 µm) and 30% of ring power (weld configuration C3) resulted in more stable dynamics of the keyhole, with significant reduction of collapse events, and ultimately controlled migration of steel towards Cu. Furthermore, compared to C4 (2500 W total power), the lower thermal gradient in C3 (1530 W total power) eventually leads to a reduction in the upward buoyancy forces.
The pre-heating approach with the tandem beam (C5) only led to local fusion of Cu and no significant improvement in keyhole stability was observed.
The combination of experiments and numerical modelling provides a powerful approach to understand complex fluid flow and metal mixing processes during laser keyhole welding. This helps to study mixing behaviour along with weld pool dynamics for selection of laser welding strategies with beam shaping in case of dissimilar material welding, especially in presence of miscibility gap at higher temperature as in case of Cu and steel.
References
M.J. Brand, P.A. Schmidt, M.F. Zaeh, A. JossenWelding techniques for battery cells and resulting electrical contact resistancesJ. Energy Storage, 1 (2015), pp. 7-14
M. Buttazzoni, C. Zenz, A. Otto, R. Gómez Vázquez, G. Liedl, J.L. AriasA numerical investigation of the laser beam welding of stainless steel sheets with a gapAppl. Sci., 11 (2021), p. 2549
L. Caprio, B. Previtali, A.G. DemirEffect of in-source beam shaping and laser beam oscillation on the electromechanical properties of Ni-plated steel joints for e-vehicle battery manufacturingJ. Laser Appl., 35 (2023), Article 042030
S. Chen, J. Huang, J. Xia, H. Zhang, X. ZhaoMicrostructural characteristics of a stainless steel/copper dissimilar joint made by laser weldingMetall. Mater. Trans. A, 44 (8) (2013), pp. 3690-3696
Y.Z. Chen, F. Liu, G.C. Yang, X.Q. Xu, Y.H. ZhouRapid solidification of bulk undercooled hypoperitectic Fe–Cu alloyJ. Alloy. Compd., 427 (1–2) (2007), pp. L1-L5
G. Chianese, P. Franciosa, T. Sun, D. Ceglarek, S. PatalanoUsing photodiodes and supervised machine learning for automatic classification of weld defects in laser welding of thin foils copper-to-steel battery tabsJ. Laser Appl., 34 (2022), Article 042007
G. Chianese, S. Jabar, P. Franciosa, D. Ceglarek, S. PatalanoA multi-physics CFD study on the part-to-part gap during remote laser welding of copper-to-steel battery tab connectors with beam wobblingProcedia CIRP, 111 (2022), pp. 484-489
J. Daligault, M. Dal, C. Gorny, F. Coste, R. FabbroCombination of Eulerian and ray-tracing approaches for copper laser welding simulationJ. Laser Appl., 34 (2022), Article 042042
A. Das, D. Li, D. Williams, D. GreenwoodJoining Technologies for Automotive Battery Systems ManufacturingWorld Electr. Veh. J., 9 (2) (2018), p. 22
P. Drobniak, A. Otto, R.G. Vázquez, R.M. Arias, J.L. AriasSimulation of keyhole laser welding of stainless steel plates with a gapProcedia CIRP, 94 (2020), pp. 731-736
Y. Hao, N. Chen, H. Wang, B.E. Carlson, F. LuEffect of zinc vapor forces on spattering in partial penetration laser welding of zinc-coated steelsJ. Mater. Process. Technol., 298 (2021), Article 117282
W. Huang, H. Wang, T. Rinker, W. TanInvestigation of metal mixing in laser keyhole welding of dissimilar metalsMater. Des., 195 (2020), Article 109056
W. Huang, W. Cai, T.J. Rinker, J. Bracey, W. TanEffects of laser oscillation on metal mixing, microstructure, and mechanical property of Aluminum–Copper weldsInt. J. Mach. Tools Manuf., 188 (2023), Article 104020
M. Hummel, C. Schöler, A. Häusler, A. Gillner, R. PopraweNew approaches on laser micro welding of copper by using a laser beam source with a wavelength of 450 nmJ. Adv. Join. Process., 1 (2020), Article 100012
IEA , 2022: Global EV Outlook 2022 – Securing supplies for an electric future, 2022.
S. Jabar, A.B. Barenji, P. Franciosa, H.R. Kotadia, D. CeglarekEffects of the adjustable ring-mode laser on intermetallic formation and mechanical properties of steel to aluminium laser welded lap jointsMater. Des., 227 (2023), Article 111774
Y.B. Jeong, H.R. Jo, H.J. Park, H. Kato, K.B. KimMechanical properties and microstructural change in (Cu–Fe) immiscible metal matrix composite: Effect of Mg on secondary phase separationJ. Mater. Res. Technol., 9 (6) (2020), pp. 15989-15995
M. Kogel-HollacherThe full potential of Photonics in e-mobility: an overwievLaser Use, 97 (2020), pp. 22-23
Kogel-Hollacher, M., Nicolay, T., Reiser, J., 2022, Going Green – Laser Welding and Advanced Sensor Technology for Electric Vehicles.
N. Kumar, I. Masters, A. DasIn-depth evaluation of laser-welded similar and dissimilar material tab-to-busbar electrical interconnects for electric vehicle battery packJ. Manuf. Process., 70 (2021), pp. 78-96
J. Li, Y. Cai, F. Yan, C. Wang, Z. Zhu, C. HuPorosity and liquation cracking of dissimilar Nd:YAG laser welding of SUS304 stainless steel to T2 copperOpt. Laser Technol., 122 (2019), Article 105881
R. Lin, H. Wang, F. Lu, J. Solomon, B.E. CarlsonNumerical study of keyhole dynamics and keyhole-induced porosity formation in remote laser welding of Al alloysInt. J. Heat. Mass Transf., 108 (A) (2017), pp. 244-256
Z. Liu, X. Wang, K. HangEnhancement of trapping efficiency by utilizing a hollow sinh-Gaussian beamSci. Rep., 9 (2019), Article 10187
F. Lu, X. Li, Z. Li, X. Tang, H. CuiFormation and influence mechanism of keyhole-induced porosity in deep-penetration laser welding based on 3D transient modelingInt. J. Heat. Mass Transf., 90 (2015), pp. 1143-1152
E. Perez Zapico, A. Ascari, V. Dimatteo, A. FortunatoLaser dissimilar welding of copper and steel thin sheets for battery productionJ. Laser Appl., 33 (1) (2021), Article 012016
C. Prieto, E. Vaamonde, D. Diego-Vallejo, J. Jimenez, B. Urbach, Y. Vidne, E. ShekelDynamic laser beam shaping for laser aluminium welding in e-mobility applicationsProcedia CIRP, 94 (2020), pp. 596-600
J. Rinne, S. Nothdurft, J. Hermsdorf, S. Kaierle, L. OvermeyerAdvantages of adjustable intensity profiles for laser beam welding of steel copper dissimilar jointsProcedia CIRP, 94 (2020), pp. 661-665
J. Rinne, S. Nothdurft, J. Hermsdorf, S. Kaierle, L. OvermeyerMultivariate parametric study on laser beam welding of copper lap joints using adjustable intensity profilesProcedia CIRP, 111 (2022), pp. 415-419
A. Sadeghian, N. IqbalA review on dissimilar laser welding of steel-copper, steel-aluminum, aluminum-copper, and steel-nickel for electric vehicle battery manufacturingOpt. Laser Technol., 146 (2022), Article 107595R.P. Shi, C.P. Wang, D. Wheeler, X.J. Liu, Y. WangFormation mechanisms of self-organized core/shell and core/shell/corona microstructures in liquid droplets of immiscible alloysActa Mater., 61 (4) (2013), pp. 1229-1243
M. Sokolov, P. Franciosa, T. Sun, D. Ceglarek, V. DimatteoApplying optical coherence tomography for weld depth monitoring in remote laser welding of automotive battery tab connectorsJ. Laser Appl., 33 (1) (2021), Article 012028
J. Wagner, A. Heider, R. Ramsayer, R. Weber, F. Faure, A. Leis, N. Armon, R. Susid, O. Tsiony, E. Shekel, T. GrafInfluence of dynamic beam shaping on the geometry of the keyhole during laser beam weldingProcedia CIRP, 111 (2022), pp. 448-452
M.F.R. Zwicker, M. Moghadam, W. Zhang, C.V. NielsenJournal of Advanced Joining Processes Automotive battery pack manufacturing – a review of battery to tabJ. Adv. Join. Process., 1 (2020), Article 100017
지표 산사태로 발생한 파랑의 3차원 시뮬레이션: OpenFOAM과 FLOW-3D HYDRO 모델 비교
Ramtin Sabeti, Mohammad Heidarzadeh, Alessandro Romano, Gabriel Barajas Ojeda & Javier L. Lara
Abstract
The recent destructive landslide tsunamis, such as the 2018 Anak Krakatau event, were fresh reminders for developing validated three-dimensional numerical tools to accurately model landslide tsunamis and to predict their hazards. In this study, we perform Three-dimensional physical modelling of waves generated by subaerial solid-block landslides, and use the data to validate two numerical models: the commercial software FLOW-3D HYDRO and the open-source OpenFOAM package. These models are key representatives of the primary types of modelling tools—commercial and open-source—utilized by scientists and engineers in the field. This research is among a few studies on 3D physical and numerical models for landslide-generated waves, and it is the first time that the aforementioned two models are systematically compared. We show that the two models accurately reproduce the physical experiments and give similar performances in modelling landslide-generated waves. However, they apply different approaches, mechanisms and calibrations to deliver the tasks. It is found that the results of the two models are deviated by approximately 10% from one another. This guide helps engineers and scientists implement, calibrate, and validate these models for landslide-generated waves. The validity of this research is confined to solid-block subaerial landslides and their impact in the near-field zone.
1 Introduction and Literature Review
Subaerial landslide-generated waves represent major threats to coastal areas and have resulted in destruction and casualties in several locations worldwide (Heller et al., 2016; Paris et al., 2021). Interest in landslide-generated tsunamis has risen in the last decade due to a number of devastating events, especially after the December 2018 Anak Krakatau tsunami which left a death toll of more than 450 people (Grilli et al., 2021; Heidarzadeh et al., 2020a). Another significant subaerial landslide tsunami occurred on 16 October 1963 in Vajont dam reservoir (Northern Italy), when an impulsive landslide-generated wave overtopped the dam, killing more than 2000 people (Heller & Spinneken, 2013; Panizzo et al., 2005). The largest tsunami run-up (524 m) was recorded in Lituya Bay landslide tsunami event in 1958 where it killed five people (Fritz et al., 2009).
To achieve a better understanding of subaerial landslide tsunamis, laboratory experiments have been performed using two- and three-dimensional (2D, 3D) set-ups (Bellotti & Romano, 2017; Di Risio et al., 2009; Fritz et al., 2004; Romano et al., 2013; Sabeti & Heidarzadeh, 2022a). Results of physical models are essential to shed light on the nonlinear physical phenomena involved. Furthermore, they can be used to validate numerical models (Fritz et al., 2009; Grilli & Watts, 2005; Liu et al., 2005; Takabatake et al., 2022). However, the complementary development of numerical tools for modelling of landslide-generated waves is inevitable, as these models could be employed to accelerate understanding the nature of the processes involved and predict the detailed outcomes in specific areas (Cremonesi et al. 2011). Due to the high flexibility of numerical models and their low costs in comparison to physical models, validated numerical models can be used to replicate actual events at a fair cost and time (e.g., Cecioni et al., 2011; Grilli et al., 2017; Heidarzadeh et al., 2020b, 2022; Horrillo et al., 2013; Liu et al., 2005; Løvholt et al., 2005; Lynett & Liu, 2005).
Table 1 lists some of the existing numerical models for landslide tsunamis although the list is not exhaustive. Traditionally, Boussinesq-type models, and Shallow water equations have been used to simulate landslide tsunamis, among which are TWO-LAYER (Imamura and Imteaz,1995), LS3D (Ataie-Ashtiani & Najafi Jilani, 2007), GLOBOUSS (Løvholt et al., 2017), and BOUSSCLAW (Kim et al., 2017). Numerical models that solve Navier–Stokes equations showed good capability and reliability to simulate subaerial landslide-generated waves (Biscarini, 2010). Considering the high computational cost of solving the full version of Navier–Stokes equations, a set of methods such as RANS (Reynolds-averaged Navier–Stokes equations) are employed by some existing numerical models (Table 1), which provide an approximate averaged solution to the Navier–Stokes equations in combination with turbulent models (e.g., k–ε, k–ω). Multiphase flow models were used to simulate the complex dynamics of landslide-generated waves, including scenarios where the landslide mass is treated as granular material, as in the work by Lee and Huang (2021), or as a solid block (Abadie et al., 2010). Among the models listed in Table 1, FLOW-3D HYDRO and OpenFOAM solve Navier–Stokes equations with different approaches (e.g., solving the RANS by IHFOAM) (Paris et al., 2021; Rauter et al., 2022). They both offer a wide range of turbulent models (e.g., Large Eddy Simulation—LES, k–ε, k–ω model with Renormalization Group—RNG), and they both use the VOF (Volume of Fluid) method to track the water surface elevation. These similarities are one of the motivations of this study to compare the performance of these two models. Details of governing equations and numerical schemes are discussed in the following.
Numerical models
Approach
Developer
FLOW-3D HYDRO
This CFD package solves Navier–Stokes equations using finite-difference and finite volume approximations, along with Volume of Fluid (VOF) method for tracking the free surface
Flow Science, Inc. (https://www.flow3d.com/)
MIKE 21
This model is based on the numerical solution of 2D and 3D incompressible RANS equations subject to the assumptions of Boussinesq and hydrostatic pressure
Danish Hydraulic Institute (DHI) (https://www.mikepoweredbydhi.com/products/mike-21-3)
OpenFOAM (IHFOAM solver)
IHFOAM is a newly developed 3D numerical two-phase flow solver. Its core is based on OpenFOAM®. IHFOAM can also solve two-phase flow within porous media using RANS/VARANS equations
IHCantabria research institute (https://ihfoam.ihcantabria.com/)
NHWAVE
NHWAVE is a 3D shock-capturing non-Hydrostatic model which solves the incompressible Navier–Stokes equations in terrain and surface-following sigma coordinates
Kirby et al. (2022) (https://sites.google.com/site/gangfma/nhwave, https://github.com/JimKirby/NHWAVE)
GLOBOUSS
GloBouss is a depth-averaged model based on the standard Boussinesq equations including higher order dispersion terms, Coriolis terms, and numerical hydrostatic correction terms
Løvholt et al. (2022) (https://www.duo.uio.no/handle/10852/10184)
BOUSSCLAW
BoussClaw is a new hybrid Boussinesq type model which is an extension of the GeoClaw model. It employs a hybrid of finite volume and finite difference methods to solve Boussinesq equations
Clawpack Development Team (http://www.clawpack.org/)Kim et al. (2017)
THETIS-MUI
THETIS is a multi-fluid Navier–Stokes solver which can be considered a one-fluid model as only one velocity is defined at each point of the mesh and there is no mixing between the three considered fluids (water, air, and slide). It applies VOF method
TREFLE department of the I2M Laboratory at Bordeaux, France (https://www.i2m.u-bordeaux.fr/en)
LS3D
A 2D depth-integrated numerical model which applies a fourth-order Boussinesq approximation for an arbitrary time-variable bottom boundary
Ataie-Ashtiani and Najafi Jilani (2007)
LYNETT- Mild-Slope Equation (MSE)
MSE is a depth-integrated version of the Laplace equation operating under the assumption of inviscid flow and mildly varying bottom slopes
A simplified 3D Navier–Stokes model for two fluids (water and landslide material) using VOF for tracking of water surface
Horrillo et al. (2013)Kim et al. (2020)
(Cornell Multi-grid Coupled Tsunami Mode (COMCOT)
COMCOT adopts explicit staggered leap-frog finite difference schemes to solve Shallow Water Equations in both Spherical and Cartesian Coordinates
Liu et al. (1998); Wang and Liu (2006)
TWO-LAYER
A mathematical model for a two-layer flow along a non-horizontal bottom. Conservation of mass and momentum equations are depth integrated in each layer, and nonlinear kinematic and dynamic conditions are specified at the free surface and at the interface between fluids
Imamura and Imteaz (1995)
Table 1 Some of the existing numerical models for simulating landslide-generated waves
In this work, we apply two Computational Fluid Dynamic (CFD) frameworks, FLOW-3D HYDRO, and OpenFOAM to simulate waves generated by solid-block subaerial landslides in a 3D set-up. We calibrate and validate both numerical models using our physical experiments in a 3D wave tank and compare the performances of these models systematically. These two numerical models are selected among the existing CFD solvers because they have been reported to provide valuable insights into landslide-generated waves (Kim et al., 2020; Romano et al., 2020a, b ; Sabeti & Heidarzadeh, 2022a). As there is no study to compare the performances of these two models (FLOW-3D HYDRO and OpenFOAM) with each other in reproducing landslide-generated waves, this study is conducted to offer such a comparison, which can be helpful for model selection in future research studies or industrial projects. In the realm of tsunami generation by subaerial landslides, the solid-block approach serves as an effective representative for scenarios where the landslide mass is more cohesive and rigid, rather than granular. This methodology is particularly relevant in cases such as the 2018 Anak Krakatau or 1963 Vajont landslides, where the landslide’s nature aligns closely with the characteristics simulated by a solid-block model (Zaniboni & Tinti, 2014; Heidarzadeh et al., 2020a, 2020b).
The objectives of this research are: (i) To provide a detailed implementation and calibration for simulating solid-block subaerial landslide-generated waves using FLOW-3D HYDRO and OpenFOAM, and (ii) To compare the performance of these two numerical models based on three criteria: free surface elevation of the landslide-generated waves, capabilities of the models in simulating 3D features of the waves in the near-field, velocity fields, and velocity variations at different locations. The innovations of this study are twofold: firstly, it is a 3D study involving physical and numerical modelling and thus the data can be useful for other studies, and secondly, it compares the performance of two popular CFD models in modelling landslide-generated waves for the first time. The validated models such as those reported in this study and comparison of their performances can be useful for engineers and scientists addressing landslide tsunami hazards worldwide.
2 Data and Methods
2.1 Physical Modelling
To validate our numerical models, a series of three-dimensional physical experiments were carried out at the Hydraulic Laboratory of the Brunel University London (UK) in a 3D wave tank 2.40 m long, 2.60 m wide, and 0.60 m high (Figs. 1 and 2). To mitigate experimental errors and enhance the reliability of our results, each physical experiment was conducted three times. The reported data in the manuscript reflects the average of these three trials, assuming no anomalous outliers, thus ensuring an accurate reflection of the experimental tests. One experiment was used for validation of our numerical models. The slope angle (α) and water depth (h) were 45° and 0.246 m, respectively for this experiment. The movement of the sliding mass was recorded by a digital camera with a sampling frequency of 120 frames per second, which was used to calculate the slide impact velocity (vs). The travel distance (D), defined as the distance from the toe of the sliding mass to the water surface, was D=0.045 m. The material of the solid block used in our study was concrete with a density of 2600 kg/m3. Table 2 provides detailed information on the dimensions and kinematics of this solid block used in our physical experiments.
Figure 1. The geometrical and kinematic parameters of a subaerial landslide tsunami. Parameters are: h, water depth; aM, maximum wave amplitude; α, slope angle;vs, slide velocity; ls, length of landslide; bs, width of landslide; s, thickness of landslide; SWL, still water level; D, travel distance (the distance from the toe of the sliding mass to the water surface); L, length of the wave tank; and W, width of the wave tank and H, is the hight of the wave tank
Figure 2. a Wave tank setup of the physical experiments of this study. b Numerical simulation setup for the FLOW-3D HYDRO Model. c The numerical set-up for the OpenFOAM model. The location of the physical wave gauge (represented by numerical gauge WG-3 in the numerical simulations) is at X = 1.03 m, Y = 1.21 m, and Z = 0.046 m. d Top view showing the locations of numerical wave gauges (WG-1, WG-2, WG-3, WG-4, WG-5)
Parameter, unit
Value/type
Slide width (bs), m
0.26
Slide length (ls), m
0.20
Slide thickness (s), m
0.10
Slide volume (V), m3
2.60 × 10–3
Specific gravity, (γs)
2.60
Slide weight (ms), kg
6.86
Slide impact velocity (vs), m/s
1.84
Slide Froude number (Fr)
1.18
Material
Concrete
Table 2 Geometrical and kinematic information of the sliding mass used for physical experiments in this study
We took scale effects into account during physical experiments by considering the study by Heller et al. (2008) who proposed a criterion for avoiding scale effects. Heller et al. (2008) stated that the scale effects can be negligible as long as the Weber number (W=ρgh2/σ; where σ is surface tension coefficient) is greater than 5.0 × 103 and the Reynolds number (R=g0.5h1.5/ν; where ν is kinematic viscosity) is greater than 3.0 × 105 or water depth (h) is approximately above 0.20 m. Considering the water temperature of approximately 20 °C during our experiments, the kinematic viscosity (ν) and surface tension coefficient (σ) of water become 1.01 × 10–6 m2/s and 0.073 N/m, respectively. Therefore, the Reynolds and Weber numbers were as R= 3.8 × 105 and W= 8.1 × 105, indicating that the scale effect can be insignificant in our experiments. To record the waves, we used a twin wire wave gauge provided by HR Wallingford (https://equipit.hrwallingford.com). This wave gauge was placed at X = 1.03 m, Y = 1.21 m based on the coordinate system shown in Fig. 2a.
2.2 Numerical Simulations
The numerical simulations in this work were performed employing two CFD packages FLOW-3D HYDRO, and OpenFOAM which have been widely used in industry and academia (e.g., Bayon et al., 2016; Jasak, 2009; Rauter et al., 2021; Romano et al., 2020a, b; Yin et al., 2015).
2.2.1 Governing Equations and Turbulent Models
2.2.1.1 FLOW-3D HYDRO
The FLOW-3D HYDRO solver is based on the fundamental law of mass, momentum and energy conservation. To estimate the influence of turbulent fluctuations on the flow quantities, it is expressed by adding the diffusion terms in the following mass continuity and momentum transport equations:
quation (1) is the general mass continuity equation, where u is fluid velocity in the Cartesian coordinate directions (x), Ax is the fractional area open to flow in the x direction, VF is the fractional volume open to flow, ρ is the fluid density, R and ξ are coefficients that depend on the choice of the coordinate system. When Cartesian coordinates are used, R is set to unity and ξ is set to zero. RDIF and RSOR are the turbulent diffusion and density source terms, respectively. Uρ=Scμ∗/ρ, in which Sc is the turbulent Schmidt number, μ∗ is the dynamic viscosity, and ρ is fluid density. RSOR is applied to model mass injection through porous obstacle surfaces.
The 3D equations of motion are solved with the following Navier–Stokes equations with some additional terms:
where t is time, Gx is accelerations due to gravity, fx is viscous accelerations, and bx is the flow losses in porous media.
According to Flow Science (2022), FLOW-3D HYDRO’s turbulence models differ slightly from other formulations by generalizing the turbulence production with buoyancy forces at non-inertial accelerations and by including the influence of fractional areas/volumes of the FAVOR method (Fractional Area-Volume Obstacle Representation) method. Here we use k–ω model for turbulence modelling. The k–ω model demonstrates enhanced performance over the k-ε and Renormalization-Group (RNG) methods in simulating flows near wall boundaries. Also, for scenarios involving pressure changes that align with the flow direction, the k–ω model provides more accurate simulations, effectively capturing the effects of these pressure variations on the flow (Flow Science, 2022). The equations for turbulence kinetic energy are formulated as below based on Wilcox’s k–ω model (Flow Science, 2022):
where kT is turbulent kinetic energy, PT is the turbulent kinetic energy production, DiffKT is diffusion of turbulent kinetic energy, GT is buoyancy production, β∗=0.09 is closure coefficient, and ω is turbulent frequency.
2.2.1.2 OpenFOAM
For the simulations conducted in this study, OpenFOAM utilizes the Volume-Averaged RANS equations (VARANS) to enable the representation of flow within porous material, treated as a continuous medium. The momentum equation incorporates supplementary terms to accommodate frictional forces from the porous media. The mass and momentum conservation equations are linked to the VOF equation (Jesus et al., 2012) and are expressed as follows:
where the gravitational acceleration components are denoted bygj. The term u¯i=1Vf∫Vf0ujdV represents the volume averaged ensemble averaged velocity (or Darcy velocity) component, Vf is the fluid volume contained in the average volumeV,τ is the surface tension constant (assumed to be 1 for the water phase and 0 for the air phase), and fσi is surface tension, defined as fσi=σκ∂α∂xi, where σ (N/m) is the surface tension constant and κ (1/m) is the curvature (Brackbill et al., 1992). μeff is the effective dynamic viscosity that is defined as μeff=μ+ρνt and takes into account the dynamic molecular (μ) and the turbulent viscosity effects (ρνt). νt is eddy viscosity, which is provided by the turbulence closure model. n is the porosity, defined as the volume of voids over total volume, and P∗=1Vf∫∂Vf0P∗dS is the ensemble averaged pressure in excess of hydrostatic pressure. The coefficient A accounts for the frictional force induced by laminar Darcy-type flow, B considers the frictional force under turbulent flow conditions, and c accounts for the added mass. These coefficients (A,B, and c) are defined based on the work of Engelund (1953) and later modified by Van Gent (1995) as given below:
where D50 is the mean nominal diameter of the porous material, KC is the Keulegan–Carpenter number, a and b are empirical nondimensional coefficients (see Lara et al., 2011; Losada et al., 2016) and γ = 0.34 is a nondimensional parameter as proposed by Van Gent (1995). The k-ω Shear Stress Transport (SST) turbulence is employed to capture the effect of turbulent flow conditions (Zhang & Zhang, 2023) with the enhancement proposed by Larsen and Fuhrman (2018) for the over-production of turbulence beneath surface waves. Boundary layers are modelled with wall functions. The reader is referred to Larsen and Fuhrman (2018) for descriptions, validations, and discussions of the stabilized turbulence models.
2.2.2 FLOW-3D HYDRO Simulation Procedure
In our specific case in this study, FLOW-3D HYDRO utilizes the finite-volume method to numerically solve the equations described in the previous Sect. 2.2.1.1, ensuring a high level of accuracy in the computational modelling. The use of structured rectangular grids in FLOW-3D HYDRO offers the advantages of easier development of numerical methods, greater transparency in their relation to physical problems, and enhanced accuracy and stability of numerical solutions. (Flow Science, 2022). Curved obstacles, wall boundaries, or other geometric features are embedded in the mesh by defining the fractional face areas and fractional volumes of the cells that are open to flow (the FAVOR method). The VOF method is employed in FLOW-3D HYDRO for accurate capturing of the free-surface dynamics (Hirt and Nichols 1981). This approach then is upgraded to method of the TruVOF which is a split Lagrangian method that typically produces lower cumulative volume error than the alternative methods (Flow Science, 2022).
For numerical simulation using FLOW-3D HYDRO, the entire flow domain was 2.60 m wide, 0.60 m deep and 2.50 m long (Fig. 2b). The specific gravity (γs) for solid blocks was set to 2.60 in our model, aligning closely with the density of the actual sliding mass, which was approximately determined in our physical experiments. The fluid medium was modelled as water with a density of 1000 kg/m3 at 20 °C. A uniform grid comprising of one single mesh plane was applied with a grid size of 0.005 m. The top, front and back of the mesh areas were defined as symmetry, and the other surfaces were of wall type with no-slip conditions around the walls.
To simulate turbulent flows, k-ω model was used because of its accuracy in modelling turbulent flows (Menter 1992). Landslide movement was replicated in simulations using coupled motion objects, which implies that the movement of landslides is based on gravity and the friction between surfaces rather than a specified motion in which the model should be provided by force and torques. The time intervals of the numerical model outputs were set to 0.02 s to be consistent with the actual sampling rates of our wave gauges in the laboratory. In order to calibrate the FLOW-3D HYDRO model, the friction coefficient is set to 0.45, which is consistent with the Coulombic friction measurements in the laboratory. The Courant Number (C=UΔtΔx) is considered as the criterion for the stability of numerical simulations which gives the maximum time step (Δt) for a prespecified mesh size (Δx) and flow speed (U). The Courant number was always kept below one.
2.2.3 OpenFOAM Simulation Procedure
OpenFOAM is an open-source platform containing several C++ libraries which solves both 3D Reynolds-Averaged Navier–Stokes equations (RANS) and Volume-Averaged RANS equations (VARANS) for two-phase flows (https://www.openfoam.com/documentation/user-guide). Its implementation is based on a tensorial approach using object-oriented programming techniques and the Finite Volume Method (McDonald 1971). In order to simulate the subaerial landslide-generated waves, the IHFOAM solver based on interFoam (Higuera et al., 2013a, 2013b), and the overset mesh framework method are employed. The implementation of the overset mesh method for porous mediums in OpenFOAM is described in Romano et al. (2020a, b) for submerged rigid and impermeable landslides.
The overset mesh technique, as outlined by Romano et al. (2020a, b), uses two distinct domains: a moving domain that captures the dynamics of the rigid landslide and a static background domain to characterize the numerical wave tank. The overlapping of these domains results in a composite mesh that accurately depicts complex geometrical transformations while preserving mesh quality. A porous media with a very low permeability (n = 0.001) was used to simulate the impermeable sliding surfaces. RANS equations were solved within the porous media. The Multidimensional Universal Limiter with Explicit Solution (MULES) algorithm is employed for solving the (VOF) equation, ensuring precision in tracking fluid interfaces. Simultaneously, the PIMPLE algorithm is employed for the effective resolution of velocity–pressure coupling in the Eqs. 7 and 8. A background domain was created to reproduce the subaerial landslide waves with dimensions 2.50 m (x-direction) × 2.60 m (y-direction) × 0.6 m (z-direction) (Fig. 2c). The grid size is set to 0.005 m for the background mesh. A moving domain was applied in an area of 0.35 m (x-direction) × 0.46 m (y-direction) × 0.32 m (z-direction) with a grid spacing of 0.005 m and applying a body-fitted mesh approach, which contains the rigid and impermeable wedges. Wall condition with No-slip is defined as the boundary for the four side walls (left, right, front and back, in Fig. 1). Also, a non-slip boundary condition is specified to the bottom, whereas the top boundary is defined as open. The experimental slide movement time series is used to model the landslide motion in OpenFOAM. The applied equation is based on the analytical solution by Pelinovsky and Poplavsky (1996) which was later elaborated by Watts (1998). The motion of a sliding rigid body is governed by the following equation:
where, m represents the mass of the landslide, s is the displacement of the landslide down the slope, t is time elapsed, g stands for the acceleration due to gravity, θ is the slope angle, Cf is the Coulomb friction coefficient, Cm is the added mass coefficient, m0 denotes the mass of the water displaced by the moving landslide, A is the cross-sectional area of the landslide perpendicular to the direction of motion, ρ is the water density, and Cd is the drag coefficient.
2.2.4 Mesh Sensitivity Analysis
In order to find the most efficient mesh size, mesh sensitivity analyses were conducted for both numerical models (Fig. 3). We considered the influence of mesh density on simulated waveforms by considering three mesh sizes (Δx) of 0.0025 m, 0.005 m and 0.010 m. The results of FLOW-3D HYDRO revealed that the largest mesh deviates 9% (Fig. 3a, Δx = 0.0100 m) from two other finer meshes. Since the simulations by FLOW-3D HYDRO for the finest mesh (Δx = 0.0025 m) do not show any improvements in comparison with the 0.005 m mesh, therefore the mesh with the size of Δx = 0.0050 m is used for simulations (Fig. 3a). A similar approach was followed for mesh sensitivity of OpenFOAM mesh grids. The mesh with the grid spacing of Δx = 0.0050 m was selected for further simulations since a satisfactory independence was observed in comparison with the half size mesh (Δx = 0.0025 m). However, results showed that the mesh size with the double size of the selected mesh (Δx = 0.0100 m) was not sufficiently fine to minimize the errors (Fig. 3b).
Figure 3. a, b Sensitivity of numerical simulations to the sizes of the mesh (Δx) for FLOW-3D HYDRO, and OpenFOAM, respectively. The location of the wave gauge 3 (WG-3) is at X = 1.03 m, Y = 1.21 m, and Z = -0.55 m (see Fig. 2d)
In terms of computational cost, the time required for 2 s simulations by FLOW-3D HYDRO is approximately 4.0 h on a PC Intel® Core™ i7-8700 CPU with a frequency of 3.20 GHz equipped with a 32 GB RAM. OpenFOAM requires 20 h to run 2 s of numerical simulation on 2 processors on a PC Intel® Core™ i9-9900KF CPU with a frequency of 3.60 GHz equipped with a 364 GB RAM. Differences in computational time for simulations run with FLOW-3D HYDRO and OpenFOAM reflect the distinct characteristics of each numerical methods, and the specific hardware setups.
2.2.5 Validation
We validated both numerical models based on our laboratory experimental data (Fig. 4). The following criterion was used to assess the level of agreement between numerical simulations and laboratory observations:
where ε is the mismatch error, Obsi is the laboratory observation values, Simi is the simulation values, and the mathematical expression |X| represents the absolute value of X. The slope angle (α), water depth (h) and travel distance (D) were: α = 45°, h = 0.246 m and D = 0.045 m in both numerical models, consistent with the physical model. We find the percentage error between each simulated data point and its corresponding observed value, and subsequently average these errors to assess the overall accuracy of the simulation against the observed time series. Our results revealed that the mismatch errors between physical experiments and numerical models for the FLOW-3D HYDRO and OpenFOAM are 8% and 18%, respectively, indicating that our models reproduce the measured waveforms satisfactorily (Fig. 4). The simulated waveform by OpenFOAM shows a minor mismatch at t = 0.76 s which resulted from a droplet immediately after the slide hits the water surface in the splash zone. In term of the maximum negative amplitude, the simulated waves by OpenFOAM indicates a relatively better performance than FLOW-3D HYDRO, whereas the maximum positive amplitude (aM) simulated by FLOW-3D HYDRO is closer to the experimental value. The recorded maximum positive amplitude in physical experiment is 0.022 m, whereas it is 0.020 m for FLOW-3D HYDRO and 0.017 m for OpenFOAM simulations. In acknowledging the deviations observed, it is pertinent to highlight that while numerical models offer robust insights, the difference in meshing techniques and the distinct computational methods to resolve the governing equations in FLOW-3D HYDRO and OpenFOAM have contributed to the variance. Moreover, the intrinsic uncertainties associated with the physical experimentation process, including the precision of wave gauges and laboratory conditions, are non-negligible factors influencing the results.
Figure 4. Validation of the simulated waves (brown line for FLOW-3D HYDRO and green line for OpenFOAM) using the laboratory-measured waves (black solid diamonds). This physical experiment was conducted for wave gauge 3 (WG-3) located at X = 1.03 m, Y = 1.21 m, and Z = -0.55 m (see Fig. 2d). Here, ε shows the errors between simulations and actual physical measurements using Eq. (13)
3 Results
Following the validations of the two numerical models (FLOW-3D HYDRO and OpenFOAM), a series of simulations were performed to compare the performances of these two CFD solvers. The generation process of landslide waves, waveforms, and velocity fields are considered as the basis for comparing the performance of the two models (Figs. 5, 6, 7 and 8).
Figure 5.Comparison between the simulated waveforms by FLOW-3D HYDRO (black) and OpenFOAM (red) at four different locations in the near-field zone (WG-1,2,4 and 5). WG is the abbreviation for wave gauge. The mismatch (Δ) between the two models at each wave gauge is calculated using Eq. (14)
Figure 6. Comparison of water surface elevations produced by solid-block subaerial landslides for the two numerical models FLOW-3D-HYDRO (a–c) and OpenFOAM (e–g) at different times
Figure 7. Snapshots of the simulations at different times for FLOW-3D HYDRO (a–c) and OpenFOAM (e–g) showing velocity fields (colour maps and arrows). The colormaps indicate water particle velocity in m/s, and the lines indicate the velocities of water particles
Figure 8. Comparison of velocity variations at (WG-3) for FLOW-3D HYDRO (light blue) and OpenFOAM (brown)
3.1 Comparison of Waveforms
Five numerical wave gauges were placed in our numerical models to measure water surface oscillations in the near-field zone (Fig. 5). These gauges offer an azimuthal coverage of 60° (Fig. 2d). Figure 5 reveals that the simulated waveforms from two models (FLOW-3D HYDRO and OpenFOAM) are similar. The highest wave amplitude (aM) is recorded at WG-3 for both models, whereas the lowest amplitude is recorded at WG-5 and WG-1 which can be attributed to the longer distances of these gauges from the source region as well as their lateral offsets, resulting in higher wave energy dissipation at these gauges. The sharp peaks observed in the simulated waveforms, such as the red peak between 0.8–1.0 s in Fig. 5a from OpenFOAM, the red peak between 0.6–0.8 s in Fig. 5b also from OpenFOAM, and the black peak between 1.4–1.6 s in Fig. 5d from FLOW-3D HYDRO, are due to the models’ spatial and temporal discretization. They reflect the sensitivity of the models to capturing transient phenomena, where the chosen mesh and time-stepping intervals are key factors in the models’ ability to track rapid changes in the flow field. To quantify the deviations of the two models from one another, we apply the following equation for mismatch calculation:
where Δ is the mismatch error, Sim1 is the simulation values from FLOW-3D HYDRO, Sim2 is the simulation values from OpenFOAM, and the mathematical expression |X| implies the absolute value of X. We calculate the percentage difference for each corresponding pair of simulation results, then take the mean of these percentage differences to determine the average deviation between the two simulation time series. Using Eq. (14), we found a deviation range from 9 to 11% between the two models at various numerical gauges (Fig. 5), further confirming that the two models give similar simulation results.
3.2 Three-Dimensional Vision of Landslide Generation Process by Numerical Models
A sequence of four water surface elevation snapshots at different times is shown in Fig. 6 for both numerical modes. In both simulations, the sliding mass travels a constant distance of 0.045 m before hitting the water surface at t = 0.270 s which induces an initial change in water surface elevation (Figs. 6a and e). At t = 0.420 s, the mass is fully immersed for both simulations and an initial dipole wave is generated (Figs. 6b and f). Based on both numerical models, the maximum positive amplitude (0.020 m for FLOW-3D HYDRO, and 0.017 m for OpenFOAM) is observed at this stage (Fig. 6). The maximum propagation of landslide-simulated waves along with more droplets in the splash zone could be seen at t = 0.670 s for both models (Fig. 6c and g). The observed distinctions in water surface elevation simulations as illustrated in Fig. 6 are rooted in the unique computational methodologies intrinsic to each model. In the OpenFOAM simulations, a more diffused water surface elevation profile is evident. Such diffusion is an outcome of the simulation’s intrinsic treatment of turbulent kinetic energy dissipation, aligning with the solver’s numerical dissipation characteristics. These traits are influenced by the selected turbulence models and the numerical advection schemes, which prioritize computational stability, possibly at the expense of interface sharpness. The diffusion in the wave pattern as rendered by OpenFOAM reflects the application of a turbulence model with higher dissipative qualities, which serves to moderate the energy retained during wave propagation. This approach can provide insights into the potential overestimation of energy loss under specific simulation conditions. In contrast, the simulations from FLOW-3D HYDRO depict a more localized wave pattern, indicative of a different approach to turbulent dissipation. This coherence in wave fronts is a function of the model’s specific handling of the air–water interface and its targeted representation of the energy dynamics resulting from the landslide’s interaction with the water body. They each have specific attributes that cater to different aspects of wave simulation fidelity, thereby contributing to a more comprehensive understanding of the phenomena under study.
3.3 Wave Velocity Analysis
We show four velocity fields at different times during landslide motion in Fig. 7 and one time series of velocity (Fig. 8) for both numerical models. The velocity varies in the range of 0–1.9 m/s for both models, and the spatial distribution of water particle velocity appears to be similar in both. The models successfully reproduce the complex wavefield around the landslide generation area, which is responsible for splashing water and mixing with air around the source zone (Fig. 7). The first snapshot at t = 0.270 s (Fig. 7a and e) shows the initial contact of the sliding mass with water surface for both numerical models which generates a small elevation wave in front of the mass exhibiting a water velocity of approximately 1.2 m/s. The slide fully immerses for the first time at t = 0.420 s producing a water velocity of approximately 1.5 m/s at this time (Fig. 7b and f). The last snapshot (t = 0.670 s) shows 1.20 s after the slide hits the bottom of the wave tank. Both models show similar patterns for the propagation of the waves towards the right side of the wave tank. The differences in water surface profiles close to the slope and solid block at t = 0.67 s, observed in the FLOW-3D HYDRO and OpenFOAM simulations (Figs. 6 and 7), are due to the distinct turbulence models employed by each (RNG and k-ω SST, respectively) which handle the complex interactions of the landslide-induced waves with the structures differently. Additionally, the methods of simulating landslide movement further contribute to this discrepancy, with FLOW-3D HYDRO’s coupled motion objects possibly affecting the waves’ initiation and propagation unlike OpenFOAM’s prescribed motion from experimental data. In addition to the turbulence models, the variations in VOF methodologies between the two models also contribute to the observed discrepancies.
For the simulated time series of velocity, both models give similar patterns and close maximum velocities (Fig. 8). For both models the WG-3 located at X = 1.03 m, Y = 1.21 m, and Z = − 0.55 m (Fig. 2d) were used to record the time series. WG-3 is positioned 5 mm above the wave tank bottom, ensuring that the measurements taken reflect velocities very close to the bottom of the wave tank. The maximum velocity calculated by FLOW-3D HYDRO is 0.162 m/s while it is 0.132 m/s for OpenFOAM, implying a deviation of approximately 19% from one another. Some oscillations in velocity records are observed for both models, but these oscillations are clearer and sharper for OpenFOAM. Although it is hard to see velocity oscillations in the FLOW-3D HYDRO record, a close look may reveal some small oscillations (around t = 0.55 s and 0.9 s in Fig. 8). In fact, velocity oscillations are expected due to the variations in velocity of the sliding mass during the travel as well as due to the interferences of the initial waves with the reflected wave from the beach. In general, it appears that the velocity time series of the two models show similar patterns and similar maximum values although they have some differences in the amplitudes of the velocity oscillations. The differences between the two curves are attributed to factors such as difference in meshing between the two models, turbulence models, as well as the way that two models record the outputs.
4 Discussions
An important step for CFD modelling in academic or industrial projects is the selection of an appropriate numerical model that can deliver the task with satisfactory performance and at a reasonable computational cost. Obviously, the major drivers when choosing a CFD model are cost, capability, flexibility, and accessibility. In this sense, the existing options are of two types as follows:
Commercial models, such as FLOW-3D HYDRO, which are optimised to solve free-surface flow problems, with customer support and an intuitive Graphical User Interface (GUI) that significantly facilitates meshing, setup, simulation monitoring, visualization, and post-processing. They usually offer high-quality customer support. Although these models show high capabilities and flexibilities for numerical modelling, they are costly, and thus less accessible.
Open-source models, such as OpenFOAM, which come without a GUI but with coded tools for meshing, setup, parallel running, monitoring, post-processing, and visualization. Although these models offer no customer support, they have a big community support and online resources. Open-source models are free and widely accessible, but they may not be necessarily always flexible and capable.
OpenFOAM provides freedom for experimenting and diving through the code and formulating the problem for a user whereas FLOW-3D HYDRO comes with high-level customer supports, tutorial videos and access to an extensive set of example simulations (https://www.flow3d.com/). While FLOW-3D-HYDRO applies a semi-automatic meshing process where users only need to input the 3D model of the structure, OpenFOAM provides meshing options for simple cases, and in many advanced cases, users need to create the mesh in other software (e.g., ANSYS) (Ariza et al., 2018) and then convert it to OpenFOAM format. Auspiciously, there are numerous online resources (https://www.openfoam.com/trainings/about-trainings), and published examples for OpenFOAM (Rauter et al., 2021; Romano et al., 2020a, b; Zhang & Zhang, 2023).
The capabilities of both FLOW-3D HYDRO and OpenFOAM to simulate actual, complex landslide-generated wave events have been showcased in significant case studies. The study by Ersoy et al. (2022) applied FLOW-3D HYDRO to simulate impulse waves originating from landslides near an active fault at the Çetin Dam Reservoir, highlighting the model’s capacity for detailed, site-specific modelling. Concurrently, the work by Alexandre Paris (2021) applied OpenFOAM to model the 2017 Karrat Fjord landslide tsunami events, providing a robust validation of OpenFOAM’s utility in capturing the dynamics of real-world geophysical phenomena. Both instances exemplify the sophisticated computational approaches of these models in aiding the prediction and analysis of natural hazards from landslides.
As for limitations of this study, we acknowledge that our numerical models are validated by one real-world measured wave time series. However, it is believed that this one actual measurement was sufficient for validation of this study because it was out of the scope of this research to fully validate the FLOW-3D HYDRO and OpenFOAM models. These two models have been fully validated by more actual measurements by other researchers in the past (e.g., Sabeti & Heidarzadeh, 2022b). It is also noted that some of the comparisons made in this research were qualitative, such as the 3D wave propagation snapshots, as it was challenging to develop quantitative comparisons for snapshots. Another limitation of this study concerns the number of tests conducted here. We fixed properties such as water depth, slope angle, and travel distance throughout this study because it was out of the scope of this research to perform sensitivity analyses.
5 Conclusions
We configured, calibrated, validated and compared two numerical models, FLOW-3D HYDRO, and OpenFOAM, using physical experiments in a 3D wave tank. These validated models were used to simulate subaerial solid-block landslides in the near-field zone. Our results showed that both models are fully compatible with investigating waves generated by subaerial landslides, although they use different approaches to simulate the phenomenon. The properties of solid-block, water depth, slope angle, and travel distance were kept constant in this study as we focused on comparing the performance of the two models rather than conducting a full sensitivity analysis. The findings are as follows:
Different settings were used in the two models for modelling landslide-generated waves. In terms of turbulent flow modelling, we used the Renormalization Group (RNG) turbulence model in FLOW-3D HYDRO, and k-ω (SST) turbulence model in OpenFOAM. Regarding meshing techniques, the overset mesh method was used in OpenFOAM, whereas the structured cartesian mesh was applied in FLOW-3D HYDRO. As for simulation of landslide movement, the coupled motion objects method was used in FLOW-3D HYDRO, and the experimental slide movement time series were prescribed in OpenFOAM.
Our modelling revealed that both models successfully reproduced the physical experiments. The two models deviated 8% (FLOW-3D HYDRO) and 18% (OpenFOAM) from the physical experiments, indicating satisfactory performances. The maximum water particle velocity was approximately 1.9 m/s for both numerical models. When the simulated waveforms from the two numerical models are compared with each other, a deviation of 10% was achieved indicating that the two models perform approximately equally. Comparing the 3D snapshots of the two models showed that there are some minor differences in reproducing the details of the water splash in the near field.
Regarding computational costs, FLOW-3D HYDRO was able to complete the same simulations in 4 h as compared to nearly 20 h by OpenFOAM. However, the hardware that were used for modelling were not the same; the computer used for the OpenFOAM model was stronger than the one used for running FLOW-3D HYDRO. Therefore, it is challenging to provide a fair comparison for computational time costs.
Overall, we conclude that the two models give approximately similar performances, and they are both capable of accurately modelling landslide-generated waves. The choice of a model for research or industrial projects may depend on several factors such as availability of local knowledge of the models, computational costs, accessibility and flexibilities of the model, and the affordability of the cost of a license (either a commercial or an open-source model).
Reference
Abadie, S., Morichon, D., Grilli, S., & Glockner, S. (2010). Numerical simulation of waves generated by landslides using a multiple-fluid Navier–Stokes model. Coastal Engineering, 57(9), 779–794. https://doi.org/10.1016/j.coastaleng.2010.04.002
Ariza, C., Casado, C., Wang, R.-Q., Adams, E., & Marugán, J. (2018). Comparative evaluation of OpenFOAM® and ANSYS® Fluent for the modeling of annular reactors. Chemical Engineering & Technology, 41(7), 1473–1483. https://doi.org/10.1002/ceat.201700455
Ataie-Ashtiani, B., & Najafi Jilani, A. (2007). A higher-order Boussinesq-type model with moving bottom boundary: Applications to submarine landslide tsunami waves. Pure and Applied Geophysics, 164(6), 1019–1048. https://doi.org/10.1002/fld.1354
Bayon, A., Valero, D., García-Bartual, R., & López-Jiménez, P. A. (2016). Performance assessment of OpenFOAM and FLOW-3D in the numerical modeling of a low Reynolds number hydraulic jump. Environmental Modelling & Software, 80, 322–335. https://doi.org/10.1016/j.envsoft.2016.02.018
Bellotti, G., & Romano, A. (2017). Wavenumber-frequency analysis of landslide-generated tsunamis at a conical island. Part II: EOF and modal analysis. Coastal Engineering, 128, 84–91. https://doi.org/10.1016/j.coastaleng.2017.07.008
Biscarini, C. (2010). Computational fluid dynamics modelling of landslide generated water waves. Landslides, 7(2), 117–124. https://doi.org/10.1007/s10346-009-0194-z
Brackbill, J. U., Kothe, D. B., & Zemach, C. (1992). A continuum method for modeling surface tension. Journal of computational physics, 100(2), 335–354.
Cecioni, C., Romano, A., Bellotti, G., Di Risio, M., & De Girolamo, P. (2011). Real-time inversion of tsunamis generated by landslides. Natural Hazards & Earth System Sciences, 11(9), 2511–2520. https://doi.org/10.5194/nhess-11-2511-2011
Cremonesi, M., Frangi, A., & Perego, U. (2011). A Lagrangian finite element approaches the simulation of water-waves induced by landslides. Computers & Structures, 89(11–12), 1086–1093.
Del Jesus, M., Lara, J. L., & Losada, I. J. (2012). Three-dimensional interaction of waves and porous coastal structures: Part I: Numerical model formulation. Coastal Engineering, 64, 57–72. https://doi.org/10.1016/J.COASTALENG.2012.01.008
Di Risio, M., De Girolamo, P., Bellotti, G., Panizzo, A., Aristodemo, F., Molfetta, M. G., & Petrillo, A. F. (2009). Landslide-generated tsunamis runup at the coast of a conical island: New physical model experiments. Journal of Geophysical Research: Oceans. https://doi.org/10.1029/2008JC004858
Engelund, F., & Munch-Petersen, J. (1953). Steady flow in contracted and expanded rectangular channels. La Houille Blanche, (4), 464–481.
Ersoy, H., Oğuz Sünnetci, M., Karahan, M., & Perinçek, D. (2022). Three-dimensional simulations of impulse waves originating from concurrent landslides near an active fault using FLOW-3D software: A case study of Çetin Dam Reservoir (Southeast Turkey). Bulletin of Engineering Geology and the Environment, 81(7), 267. https://doi.org/10.1007/s10064-022-02675-8
Flow Science. (2022). FLOW-3D HYDRO version 12.0 user’s manual. Santa Fe, NM, USA. Retrieved from https://www.flow3d.com/. 6 Aug 2023.
Fritz, H. M., Hager, W. H., & Minor, H. E. (2004). Near field characteristics of landslide generated impulse waves. Journal of waterway, port, coastal, and ocean engineering, 130(6), 287–302.
Fritz, H. M., Mohammed, F., & Yoo, J. (2009). Lituya bay landslide impact generated mega-tsunami 50th anniversary. In: Tsunami science four years after the 2004 Indian Ocean Tsunami (pp. 153–175). Birkhäuser Basel, Switzerland.
Grilli, S. T., Shelby, M., Kimmoun, O., Dupont, G., Nicolsky, D., Ma, G., Kirby, J. T., & Shi, F. (2017). Modelling coastal tsunami hazard from submarine mass failures: Effect of slide rheology, experimental validation, and case studies off the US East Coast. Natural Hazards, 86(1), 353–391. https://doi.org/10.1007/s11069-016-2692-3
Grilli, S. T., & Watts, P. (2005). Tsunami generation by submarine mass failure. I: Modelling, experimental validation, and sensitivity analyses. Journal of Waterway, Port, Coastal, and Ocean Engineering, 131(6), 283–297. https://doi.org/10.1061/(ASCE)0733-950X
Grilli, S. T., Zhang, C., Kirby, J. T., Grilli, A. R., Tappin, D. R., Watt, S. F. L., et al. (2021). Modeling of the Dec. 22nd, 2018, Anak Krakatau volcano lateral collapse and tsunami based on recent field surveys: Comparison with observed tsunami impact. Marine Geology. https://doi.org/10.1016/j.margeo.2021.106566
Heidarzadeh, M., Gusman, A., Ishibe, T., Sabeti, R., & Šepić, J. (2022). Estimating the eruption-induced water displacement source of the 15 January 2022 Tonga volcanic tsunami from tsunami spectra and numerical modelling. Ocean Engineering, 261, 112165. https://doi.org/10.1016/j.oceaneng.2022.112165
Heidarzadeh, M., Ishibe, T., Sandanbata, O., Muhari, A., & Wijanarto, A. B. (2020a). Numerical modeling of the subaerial landslide source of the 22 December 2018 Anak Krakatoa volcanic tsunami, Indonesia. Ocean Engineering, 195, 106733. https://doi.org/10.1016/j.oceaneng.2019.106733
Heidarzadeh, M., Putra, P. S., Nugroho, H. S., & Rashid, D. B. Z. (2020b). Field survey of tsunami heights and runups following the 22 December 2018 Anak Krakatau volcano tsunami, Indonesia. Pure and Applied Geophysics, 177, 4577–4595. https://doi.org/10.1007/s00024-020-02587-w
Heller, V., Bruggemann, M., Spinneken, J., & Rogers, B. D. (2016). Composite modelling of subaerial landslide–tsunamis in different water body geometries and novel insight into slide and wave kinematics. Coastal Engineering, 109, 20–41. https://doi.org/10.1016/j.coastaleng.2015.12.004
Heller, V., Hager, W. H., & Minor, H. E. (2008). Scale effects in subaerial landslide generated impulse waves. Experiments in Fluids, 44(5), 691–703. https://doi.org/10.1007/s00348-007-0427-7
Heller, V., & Spinneken, J. (2013). Improved landslide-tsunami prediction: Effects of block model parameters and slide model. Journal of Geophysical Research: Oceans, 118(3), 1489–1507. https://doi.org/10.1002/jgrc.20099
Higuera, P., Lara, J. L., & Losada, I. J. (2013a). Realistic wave generation and active wave absorption for Navier–Stokes models: Application to OpenFOAM®. Coastal Engineering, 71, 102–118. https://doi.org/10.1016/j.coastaleng.2012.07.002
Higuera, P., Lara, J. L., & Losada, I. J. (2013b). Simulating coastal engineering processes with OpenFOAM®. Coastal Engineering, 71, 119–134. https://doi.org/10.1016/j.coastaleng.2012.06.002
Hirt, C.W. and Nichols, B.D., 1981. Volume of fluid (VOF) method for the dynamics of free boundaries. Journal of computational physics, 39(1), 201–225.
Horrillo, J., Wood, A., Kim, G.-B., & Parambath, A. (2013). A simplified 3-D Navier–Stokes numerical model for landslide-tsunami: Application to the Gulf of Mexico. Journal of Geophysical Research, 118(12), 6934–6950. https://doi.org/10.1002/2012JC008689
Imamura, F., & Imteaz, M. A. (1995). Long waves in two-layers: Governing equations and numerical model. Science of Tsunami Hazards, 13(1), 3–24.
Jasak, H. (2009). OpenFOAM: Open source CFD in research and industry. International Journal of Naval Architecture and Ocean Engineering, 1(2), 89–94. https://doi.org/10.2478/IJNAOE-2013-0011
Kim, G. B., Cheng, W., Sunny, R. C., Horrillo, J. J., McFall, B. C., Mohammed, F., Fritz, H. M., Beget, J., & Kowalik, Z. (2020). Three-dimensional landslide generated tsunamis: Numerical and physical model comparisons. Landslides, 17(5), 1145–1161. https://doi.org/10.1007/s10346-019-01308-2
Kim, J., Pedersen, G. K., Løvholt, F., & LeVeque, R. J. (2017). A Boussinesq type extension of the GeoClaw model-a study of wave breaking phenomena applying dispersive long wave models. Coastal Engineering, 122, 75–86. https://doi.org/10.1016/j.coastaleng.2017.01.005
Kirby, J. T., Grilli, S. T., Horrillo, J., Liu, P. L. F., Nicolsky, D., Abadie, S., Ataie-Ashtiani, B., Castro, M. J., Clous, L., Escalante, C., Fine, I., et al. (2022). Validation and inter-comparison of models for landslide tsunami generation. Ocean Modelling, 170, 101943. https://doi.org/10.1016/j.ocemod.2021.101943
Lara, J. L., Ruju, A., & Losada, I. J. (2011). Reynolds averaged Navier–Stokes modelling of long waves induced by a transient wave group on a beach. Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, 467(2129), 1215–1242.
Larsen, B. E., & Fuhrman, D. R. (2018). On the over-production of turbulence beneath surface waves in Reynolds-averaged Navier–Stokes models. Journal of Fluid Mechanics, 853, 419–460.
Lee, C. H., & Huang, Z. (2021). Multi-phase flow simulation of impulsive waves generated by a sub-aerial granular landslide on an erodible slope. Landslides, 18(3), 881–895. https://doi.org/10.1007/s10346-020-01560-z
Liu, P. L. F., Woo, S. B., & Cho, Y. S. (1998). Computer programs for tsunami propagation and inundation. Technical Report. Cornell University, Ithaca, New York.
Liu, F., Wu, T.-R., Raichlen, F., Synolakis, C. E., & Borrero, J. C. (2005). Runup and rundown generated by three-dimensional sliding masses. Journal of Fluid Mechanics, 536(1), 107–144. https://doi.org/10.1017/S0022112005004799
Losada, I. J., Lara, J. L., & del Jesus, M. (2016). Modeling the interaction of water waves with porous coastal structures. Journal of Waterway, Port, Coastal, and Ocean Engineering, 142(6), 03116003
Løvholt, F., Harbitz, C. B., & Haugen, K. (2005). A parametric study of tsunamis generated by submarine slides in the Ormen Lange/Storegga area off western Norway. In: Ormen Lange—An integrated study for safe field development in the Storegga submarine area (pp. 219–231). Elsevier. https://doi.org/10.1016/B978-0-08-044694-3.50023-8
Løvholt, F., Bondevik, S., Laberg, J. S., Kim, J., & Boylan, N. (2017). Some giant submarine landslides do not produce large tsunamis. Geophysical Research Letters, 44(16), 8463–8472
Løvholt, F. J. M. R., Griffin, J., & Salgado-Gálvez, M. A. (2022). Tsunami hazard and risk assessment on the global scale. Complexity in Tsunamis, Volcanoes, and their Hazards, 213–246
Lynett, P., & Liu, P. L. F. (2005). A numerical study of the run-up generated by three-dimensional landslides. Journal of Geophysical Research: Oceans. https://doi.org/10.1029/2004JC002443
Lynett, P. J., & Martinez, A. J. (2012). A probabilistic approach for the waves generated by a submarine landslide. Coastal Engineering Proceedings, 33, 15–15. https://doi.org/10.9753/icce.v33.currents.15
McDonald, P. W. (1971). The computation of transonic flow through two-dimensional gas turbine cascades (Vol. 79825, p. V001T01A089). American Society of Mechanical Engineers
Menter, F. R. (1992). Improved two-equation k-omega turbulence models for aerodynamic flows (No. A-92183). https://ntrs.nasa.gov/citations/19930013620
Panizzo, A., De Girolamo, P., Di Risio, M., Maistri, A., & Petaccia, A. (2005). Great landslide events in Italian artificial reservoirs. Natural Hazards and Earth System Sciences, 5(5), 733–740. https://doi.org/10.5194/nhess-5-733-2005
Paris, A. (2021). Comparison of landslide tsunami models and exploration of fields of application. Doctoral dissertation, Université de Pau et des Pays de l’Adour.
Paris, A., Heinrich, P., & Abadie, S. (2021). Landslide tsunamis: Comparison between depth-averaged and Navier–Stokes models. Coastal Engineering, 170, 104022. https://doi.org/10.1016/j.coastaleng.2021.104022
Pelinovsky, E., & Poplavsky, A. (1996). Simplified model of tsunami generation by submarine landslides. Physics and Chemistry of the Earth, 21(1–2), 13–17. https://doi.org/10.1016/S0079-1946(97)00003-7
Rauter, M., Hoße, L., Mulligan, R. P., Take, W. A., & Løvholt, F. (2021). Numerical simulation of impulse wave generation by idealized landslides with OpenFOAM. Coastal Engineering, 165, 103815. https://doi.org/10.1016/j.coastaleng.2020.103815
Rauter, M., Viroulet, S., Gylfadóttir, S. S., Fellin, W., & Løvholt, F. (2022). Granular porous landslide tsunami modelling–the 2014 Lake Askja flank collapse. Nature Communications, 13(1), 678. https://doi.org/10.1038/s41467-022-28356-2
Romano, A., Bellotti, G., & Di Risio, M. (2013). Wavenumber–frequency analysis of the landslide-generated tsunamis at a conical island. Coastal Engineering, 81, 32–43. https://doi.org/10.1016/j.coastaleng.2013.06.007
Romano, A., Lara, J., Barajas, G., Di Paolo, B., Bellotti, G., Di Risio, M., Losada, I., & De Girolamo, P. (2020a). Tsunamis generated by submerged landslides: Numerical analysis of the near-field wave characteristics. Journal of Geophysical Research: Oceans, 125(7), e2020JC016157. https://doi.org/10.1029/2020JC016157
Romano, M., Ruggiero, A., Squeglia, F., Maga, G., & Berisio, R. (2020b). A structural view of SARS-CoV-2 RNA replication machinery: RNA synthesis, proofreading and final capping. Cells, 9(5), 1267.
Sabeti, R., & Heidarzadeh, M. (2022a). Numerical simulations of water waves generated by subaerial granular and solid-block landslides: Validation, comparison, and predictive equations. Ocean Engineering, 266, 112853. https://doi.org/10.1016/j.oceaneng.2022.112853
Sabeti, R., & Heidarzadeh, M. (2022b). Numerical simulations of tsunami wave generation by submarine landslides: Validation and sensitivity analysis to landslide parameters. Journal of Waterway, Port, Coastal, and Ocean Engineering, 148(2), 05021016. https://doi.org/10.1061/(ASCE)WW.1943-5460.0000694
Takabatake, T., Han, D. C., Valdez, J. J., Inagaki, N., Mäll, M., Esteban, M., & Shibayama, T. (2022). Three-dimensional physical modeling of tsunamis generated by partially submerged landslides. Journal of Geophysical Research: Oceans, 127(1), e2021JC017826. https://doi.org/10.1029/2021JC017826
Van Gent, M. R. A. (1995). Porous flows through rubble-mound material. Journal of waterway, port, coastal, and ocean engineering, 121(3), 176–181.
Wang, X., & Liu, P. L. F. (2006). An analysis of 2004 Sumatra earthquake fault plane mechanisms and Indian Ocean tsunami. Journal of Hydraulic Research, 44, 147–154. https://doi.org/10.1080/00221686.2006.9521671
Watts, P. (1998). Wavemaker curves for tsunamis generated by underwater landslides. Journal of Waterway, Port, Coastal, and Ocean Engineering, 124(3), 127–137. https://doi.org/10.1061/(ASCE)0733-950X(1998)124:3(127)
Yin, Y., Zhang, C., Imamura, F., Harris, J. C., & Li, Z. (2015). Numerical analysis on wave generated by the Qianjiangping landslide in Three Gorges Reservoir. China. Landslides, 12(2), 355–364. https://doi.org/10.1007/s10346-015-0564-7
Zaniboni, F., & Tinti, S. (2014). Numerical simulations of the 1963 Vajont landslide, Italy: application of 1D Lagrangian modelling. Natural hazards, 70, 567–592.
Zhang, C., & Zhang, M. (2023). Numerical investigation of solitary wave attenuation and mitigation caused by vegetation using OpenFOAM. Coastal Engineering Journal, 65(2), 198–216. https://doi.org/10.1080/21664250.2022.2163844
In this study, experimental investigations were conducted on rectangular side weirs with different widths and heights. Corresponding simulations were also performed to analyze hydraulic characteristics including the water surface profile, flow velocity, and pressure. The relationship between the discharge coefficient and the Froude number, as well as the ratios of the side weir height and width to upstream water depth, was determined. A discharge formula was derived based on a dimensional analysis. The results demonstrated good agreement between simulated and experimental data, indicating the reliability of numerical simulations using FLOW-3D software (version 11.1). Notably, significant fluctuations in water surface profiles near the side weir were observed compared to those along the center line or away from the side weir in the main channel, suggesting that the entrance effect of the side weir did not propagate towards the center line of the main channel. The proposed discharge formula exhibited relative errors within 10%, thereby satisfying the flow measurement requirements for small channels and field inlets.
1. Introduction
Sharp crested weirs are used to obtain discharge in open channels by solely measuring the water head upstream of the water. Side weirs, as a kind of sharp-crested weir, are extensively used for flow measurement, flow diversion, and flow regulation in open channels. Side weirs can be placed directly in the channel direction or field inlet, without changing the original structure of the channel. Thus, side weirs have certain advantages in the promotion and application of flow measurement facilities in small channels and field inlets. The rectangular sharp-crested weir is the most commonly available, and many scholars have conducted research on it. Research on side weirs started in 1934. De Marchi studied the side weir in the rectangular channel and derived the theoretical formula based on the assumption that the specific energy of the main flow section of the rectangular channel in the side weir section was constant [1]. Ackers discussed the existing formulas for the prediction of the side weir discharge coefficient [2]. Chen concluded that the momentum theorem was more suitable for the analytical calculation of the side weir based on the experimental data [3]. Based on previous theoretical research, more and more scholars began to carry out experimental research on side weirs. Uyumaz and Muslu conducted experiments under subcritical and supercritical flow regimes and derived expressions for the side weir discharge and water surface profiles for these regimes by comparing them with experimental results [4]. Borghei et al. developed a discharge coefficient equation for rectangular side weirs in subcritical flow [5]. Ghodsian [6] and Durga and Pillai [7] developed a discharge coefficient equation of rectangular side weirs in supercritical flow. Mohamed proposed a new approach based on the video monitoring concept to measure the free surface of flow over rectangular side weirs [8]. Durga conducted experiments on rectangular side weirs of different lengths and sill heights and discussed the application of momentum and energy principles to the analysis of spatially varied flow under supercritical conditions. The results showed that the momentum principle was fitting better [7]. Omer et al. obtained sharp-crested rectangular side weirs discharge coefficients in the straight channel by using an artificial neural network model for a total of 843 experiments [9]. Emiroglu et al. studied water surface profile and surface velocity streamlines, and developed a discharge coefficient formula of the upstream Froude number, the ratios of weir length to channel width, weir length to flow depth, and weir height to flow depth [10]. Other investigators [11,12,13,14] have conducted experiments to study flow over rectangular side weirs in different flow conditions. Numerous studies have been conducted in laboratories to this day. Compared to experimental methods, the numerical simulation method has many attractive advantages. We can easily obtain a wide range of hydraulic parameters of side weirs using numerical simulation methods, without investing a lot of manpower and resources. In addition, we can conduct small changes in inlet condition, outlet condition, and geometric parameters, and study their impact on the flow characteristics of side weirs. Therefore, with the development and improvement of computational fluid dynamics, the numerical simulation method has begun to be widely applied on side weirs. Salimi et al. studied the free surface changes and the velocity field along a side weir located on a circular channel in the supercritical regime by numerical simulation [15]. Samadi et al. conducted a three-dimensional simulation on rectangular sharp-crested weirs with side contraction and without side contraction and verified the accuracy of numerical simulation compared with the experimental results [16]. Aydin investigated the effect of the sill on rectangular side weir flow by using a three-dimensional computational fluid dynamics model [17]. Azimi et al. studied the discharge coefficient of rectangular side weirs on circular channels in a supercritical flow regime using numerical simulation and experiments [18]. The discharge coefficient over the two compound side weirs (Rectangular and Semi-Circle) was modeled by using the FLOW-3D software to describe the flow characteristics in subcritical flow conditions [19]. Safarzadeh and Noroozi compared the hydraulics and 3D flow features of the ordinary rectangular and trapezoidal plan view piano key weirs (PKWs) using two-phase RANS numerical simulations [20]. Tarek et al. investigated the discharge performance, flow characteristics, and energy dissipation over PK and TL weirs under free-flow conditions using the FLOW-3D software [21]. As evident from the aforementioned, the majority of studies have primarily focused on determining the discharge coefficient, while comparatively less attention has been devoted to investigating the hydraulic characteristics of rectangular side weirs. Numerical simulations were conducted on different types of side weirs, including compound side weirs and piano key weirs, in different cross-section channels under different flow regimes. It is imperative to derive the discharge formula and investigate other crucial flow parameters such as depth, velocity, and pressure near side weirs for their effective implementation in water measurement. In this study, a combination of experimental and numerical simulation methods was employed to examine the relationship between the discharge coefficient and its influencing factors; furthermore, a dimensionless analysis was utilized to derive the discharge formula. Additionally, water surface profiles near side weirs and pressure distribution at the bottom of the side channel were analyzed to assess safety operation issues associated with installing side weirs.
2. Principle of Flow Measurement
Flow discharge over side weirs is a function of different dominant physical and geometrical quantities, which is defined as
where Q is flow discharge over the side weir, b is the side weir width, B is the channel width, P is the side weir height, v is the mean velocity, h1 is water depth upstream the side weir in the main channel, g is the gravitational acceleration, μ is the dynamic viscosity of fluid, ρ is fluid density, and i is the channel slope (Figure 1).
Figure 1. Definition sketch of parameters of rectangular side weir under subcritical flow. Note: h1 and h2 represent water depth upstream and downstream of the side weir in the main channel, respectively; y1 and y2 represent weir head upstream and downstream of the side weir in the main channel, respectively.
In experiments when the upstream weir head was over 30 mm, the effects of surface tension on discharge were found to be minor [22]. The viscosity effect was far less than the gravity effect in a turbulent flow. Hence μ and σ were excluded from the analysis [23,24]. In addition, the channel width, the channel slope, and the fluid density were all constant, so the discharge formula can be simplified as:
According to the Buckingham π theorem, the following relationship among the dimensionless parameters is established:
Selected h1 and g as basic fundamental quantities, and the remaining physical quantities were represented in terms of these fundamental quantities as follows:
In which
Based on dimensional analysis, the following equations were derived.
Namely
So the discharge formula can be simplified as:
In a sharp-crested weir, discharge over the weir is proportional to 𝐻1.51H11.5 (H1 is the upstream total head above the crest, namely H1 = y1 + v2/2 g), so Equation (6) can be transformed as follows:
Consequently, the discharge formula over rectangular side weirs is defined as follows, in which 𝑚=𝑓(𝑏ℎ1m=f(bh1,𝑃ℎ1,𝐹𝑟1)Ph1,Fr1). Parameter m represents the dimensionless discharge coefficient. Parameter Fr1 represents the Froude number at the upstream end of the side weir in the main channel.
3. Experiment Setup
The experimental setup contained a storage reservoir, a pumping station, an electromagnetic flow meter, a control valve, a stabilization pond, rectangular channels, a side weir, and a sluice gate. The layout of the experimental setup is shown in Figure 2. Water was supplied from the storage reservoir using a pump. The flow discharge was measured with an electromagnetic flow meter with precision of ±3‰. Water depth was measured with a point gauge with an accuracy of ±0.1 mm. The flow velocity was measured with a 3D Acoustic Doppler Velocimeter (Nortek Vectrino, manufactured by Nortek AS in Rud, Norway). In order to eliminate accidental and human error, multiple measurements of the water depth and flow velocity at the same point were performed and the average values were used as the actual water depth and flow velocity of the point. The main and side channels were both rectangular open channels measuring 47 cm in width and 60 cm in height. The geometrical parameters of rectangular side weirs are shown in Table 1.
Figure 2. Layout of the test system.
Table 1. The geometrical parameters of rectangular side weirs.
When water passes through a side weir, its quality point is affected not only by gravity but also by centrifugal inertia force, leading to an inclined water surface within that particular cross-section before reaching the weir. In order to examine water profiles adjacent to side weirs, cross-sectional measurements were conducted at regular intervals of 12 cm both upstream and downstream of each side weir, denoted as sections ① to ⑩, respectively. Measuring points were positioned near the side weir (referred to as “Side I”), along the center line of the main channel (referred to as “Side II”), and far away from the side weir (referred to as “Side III”) for each cross-section. The schematic diagram illustrating these measuring points is presented in Figure 3.
Figure 3. Schematic diagram of measurement points.
4. Numerical Simulation Settings
4.1. Mathematical Model
4.1.1. Governing Equations
Establishing the controlling equations is a prerequisite for solving any problem. For the flow analysis problem of water flowing over a side weir in a rectangular channel, assuming that no heat exchange occurs, the continuity equation (Equation (9)) and momentum equation (Equation (10)) can be used as the controlling equations as follows:
The continuity equation:
Momentum equation:
where: ρ is the fluid density, kg/m3; t is time, s; ui, uj are average flow velocities, u1, u2, u3 represent average flow velocity components in Cartesian coordinates x, y, and z, respectively, m/s; μ is dynamic viscosity of fluid, N·s/m2; p is the pressure, pa; Si is the body force, S1 = 0, S2 = 0, S3 = −ρg, N [24].
4.1.2. RNG k-ε Model
The water flow in the main channel is subcritical flow. When the water flows through the side weir, the flow line deviates sharply, the cross section suddenly decreases, and due to the blocking effect of the side weir, the water reflects and diffracts, resulting in strong changes in the water surface and obvious three-dimensional characteristics of the water flow [25]. Therefore the RNG k–ε model is selected. The model can better handle flows with greater streamline curvature, and its corresponding k and ε equation is, respectively, as follows:
where: k is the turbulent kinetic energy, m2/s2; μeff is the effective hydrodynamic viscous coefficient; Gk is the generation item of turbulent kinetic energy k due to gradient of the average flow velocity; C∗1εC1ε*, C2ε are empirical constants of 1.42 and 1.68, respectively; ε is turbulence dissipation rate, kg·m2/s2.
4.1.3. TruVOF Model
Because the shape of the free surface is very complex and the overall position is constantly changing, the fluid flow phenomenon with a free surface is a typical flow phenomenon that is difficult to simulate. The current methods used to simulate free surfaces mainly include elevation function method, the MAC method [26] and the VOF (Volume of Fluid) method [27]. The VOF method is a method proposed by Hirt and Nichols to deal with the complex motion of the free surface of a fluid, which can describe all the complexities of the free surface with only one function. The basic idea of the method is to define functions αw and αa, which represent the volume percentage of the calculation area occupied by water and air, respectively. In each unit cell, the sum of the volume fractions of water and air is equal to 1, i.e.,
The TruVOF calculation method can accurately track the change of free liquid level and accurately simulate the flow problems with free interface. Its equation is:
where: u_¯m is the average velocity of the mixture; t is the time; F is the volume fraction of the required fluid.
4.2. Parameter Setting and Boundary Conditions
To streamline the iterative calculation and minimize simulation time, we selected a main channel measuring 7.5 m in length and a side channel measuring 2.5 m in length for simulation. Three-dimensional geometrical models were developed using the software AutoCAD (version 2016-Simplified Chinese). The spatial domain was meshed using a constructed rectangular hexahedral mesh and each cell size was 2 cm. A volume flow rate was set in the channel inlet with an auto-adjusted fluid height. An outflow–outlet condition was positioned at the end of the side channel. A symmetry boundary condition was set in the air inlet at the top of the model, which represented that no fluid flows through the boundary. The lower Z (Zmin) and both of the side boundaries were treated as a rigid wall (W). No-slip conditions were applied at the wall boundaries. Figure 4 illustrates these boundary conditions.
Figure 4. Diagram of boundary conditions.
5. Results
5.1. Water Surface Profiles
Water surface profiles were crucial parameters for selecting water-measuring devices. Upon analyzing the consistent patterns observed in different conditions, one specific condition was chosen for further analysis. To validate the reliability of numerical simulation, measured and simulated water depths of rectangular side weirs with different widths and heights at a discharge rate of 25 L/s were extracted for comparison (Table 2 and Figure 5). The results in Table 2 and Figure 5 indicate a maximum absolute relative error value of 9.97% and all absolute relative error values within 10%, demonstrating satisfactory agreement between experimental and simulated results.
Figure 5. Comparison between measured and simulated flow depth.
P/cm
Section Position
b = 20 cm
b = 30 cm
b = 40 cm
b = 47 cm
hm/cm
hs/cm
R/%
hm/cm
hs/cm
R/%
hm/cm
hs/cm
R/%
hm/cm
hs/cm
R/%
7
④
21.49
19.4
9.73
17.74
16.9
4.74
16.07
14.51
9.71
13.79
12.50
9.35
④′
20.48
19.05
6.98
17.78
16.14
9.22
15.69
14.31
8.80
⑥
20.71
19.02
8.16
17.82
16.31
8.47
15.92
14.53
8.73
15.23
13.80
9.39
⑧′
22.00
20.22
8.09
18.27
16.74
8.37
16.59
14.96
9.83
⑧
22.37
20.17
9.83
17.73
16.80
5.25
16.27
15.08
7.31
15.36
14.36
6.51
10
④
24.15
22.6
6.42
19.96
18.84
5.61
19.03
18.58
2.36
16.83
15.85
5.82
④′
24.21
22.05
8.92
19.49
18.19
6.67
18.75
18.35
2.13
⑥
24.01
21.78
9.29
19.65
18.34
6.67
18.95
18.63
1.69
17.52
16.09
8.16
⑧′
24.88
22.4
9.97
20.65
19.21
6.97
20.12
19.29
4.13
⑧
24.03
22.96
4.45
21.16
19.34
8.60
19.71
19.43
1.42
18.39
17.36
5.60
15
④
28.85
27.56
4.47
25.86
24.09
6.84
24.05
21.89
8.98
22.73
20.80
8.49
④′
28.49
26.97
5.34
25.19
23.84
5.36
23.42
21.46
8.37
⑥
28.85
26.98
6.48
25.72
23.99
6.73
23.23
21.82
6.07
23.10
21.05
8.87
⑧′
28.96
27.30
5.73
26.38
24.19
8.30
24.18
22.27
7.90
⑧
29.18
27.96
4.18
26.57
24.54
7.64
24.57
22.33
9.12
23.20
21.10
9.05
20
④
33.29
32.34
2.85
30.63
29.02
5.26
28.49
26.87
5.69
26.99
25.81
4.37
④′
33.14
31.95
3.59
29.75
28.62
3.80
28.11
26.79
4.70
⑥
33.32
31.79
4.59
30.04
28.45
5.29
28.99
26.86
7.35
27.42
26.72
2.55
⑧′
34.02
32.39
4.79
30.69
28.95
5.67
29.59
27.25
7.91
⑧
34.62
32.84
5.14
31.44
29.29
6.84
29.51
27.31
7.46
28.21
27.00
4.29
Table 2. Comparison of measured and simulated water depths on Side I of each side weir at a discharge of 25 L/s
Due to the diversion caused by the side weir, there was a rapid variation in flow near the side weir in the main channel. In order to investigate the impact of the side weir on water flow in the main channel, water surface profiles on Side I, Side II, and Side III were plotted with a side weir width and height both set at 20 cm at a discharge rate of 25 L/s (Figure 6). As depicted in Figure 6, within a certain range of the upstream end of the main channel, water depths on Side I, Side II, and Side III were nearly equal with almost horizontal profiles. As the distance between the location of water flow and the upstream end of the weir crest decreased gradually, there was a gradual decrease in water depth on Side I along with an inclined trend in its corresponding profile; however, both Side II and Side III still maintained almost horizontal profiles. When approaching closer to the side weir area with flowing water, there was an evident reduction in water depth on Side I accompanied by a significant downward trend visible across an expanded decline range. The minimum point occurred near the upstream end of the weir crest before gradually increasing again towards downstream sections. At the crest section of the side weir, there is an upward trend observed in the water surface. The water surface tended to stabilize downstream of the main channel within a certain range from the downstream end of the weir crest. There was no significant change in the water surface profiles of Side Ⅱ and Side Ⅲ in the crest section. It can be inferred that the side weir entrance effect occurred only between Side Ⅰ and Side Ⅱ. M. Emin reported the same pattern [10].
Figure 6. Water surface profiles on Side I, Side II, and Side III with a side weir width of 20 cm and height of 15 cm at a discharge of 25 L/s.
For a more accurate study on the entrance effect of the side weir on the Water Surface Profile (WSP) for Side I; a comparative analysis conducted using different widths but the same height (15 cm) at a discharge rate of 25 L/s is presented through Figure 7, Figure 8, Figure 9 and Figure 10.
Figure 7. Water surface profile on Side Ⅰ with a side weir width of 20 cm and height of 15 cm at a discharge of 25 L/s.
Figure 8. Water surface profile on Side Ⅰ with a side weir width of 30 cm and height of 15 cm at a discharge of 25 L/s.
Figure 9. Water surface profile on Side Ⅰ with a side weir width of 40 cm and height of 15 cm at a discharge of 25 L/s.
Figure 10. Water surface profile on Side Ⅰ with a side weir width of 47 cm and height of 15 cm at a discharge of 25 L/s.
According to Figure 7, Figure 8, Figure 9 and Figure 10, the water depth upstream of the main channel started to decrease as it approached the upstream end of the weir crest and then gradually increased at the weir crest section. In other words, the water surface profile exhibited a backwater curve along the length of the weir crest. The water depth remained relatively stable downstream of the main channel within a certain range from the downstream end of the weir crest. Additionally, there was a higher water depth downstream of the main channel compared to that upstream of the main channel. Furthermore, an increase in the width of the side weir led to a gradual reduction in fluctuations on its water surface.
5.2. Velocity Distribution
The law of flow velocity distribution near the side weir is the focus of research and analysis, so the simulated and measured values of flow velocity near the side weir were compared and analyzed. Take the discharge of 25 L/s, the height of 15 cm, and the width of 30 cm of the side weir as an example to illustrate. Figure 11 shows the measured and simulated velocity distribution in the x-direction of cross-section ④. As can be seen from Figure 11, the diagrams of the measured and simulated velocity distribution were relatively consistent, and the maximum absolute relative error between the measured and simulated values at the same measurement point was 9.37%, and the average absolute relative error was 3.97%, which indicated a satisfactory agreement between the experimental and simulated results.
Figure 11. Velocity distribution in the x-direction of section ④: when the discharge is 25 L/s, the height of the side weir is 15 cm and the width of the side weir is 30 cm. (a) Measured velocity distribution; (b) Simulated velocity distribution.
From Figure 11, it can be seen that the flow velocity gradually increased from the bottom of the channel towards the water surface in the Z-direction, and the flow velocity gradually increased from Side Ⅲ to Side Ⅰ in the Y-direction. The maximum flow velocity occurred near the weir crest.
Figure 12 shows the distribution of flow velocity at different depths (z/P = 0.3, z/P = 0.8, z/P = 1.6) with a side weir width of 30 cm and height of 15 cm at a discharge of 25 L/s. The water flow line began to bend at a certain point upstream of the main channel, and the closer it was to the upstream end of the weir crest, the greater the curvature. The maximum curvature occurred at the downstream end of the weir crest. The flow patterns at the bottom, near the side weir crest, and above the side weir crest were significantly different. There was a reverse flow at the bottom of the main channel, where the forward and reverse flows intersect, resulting in a detention zone. The maximum flow velocity at the bottom layer occurred at the upstream end of the side weir crest. When the location of water flow approached the weir crest, the maximum flow velocity occurred at the upstream end of the weir crest. The maximum flow velocity on the water surface occurred at the downstream end of the weir crest. As the water depth decreased, the position of the maximum flow velocity gradually moved from the upstream end of the side weir to the downstream end of the side weir.
Figure 12. Distribution of flow velocity at different depths with a side weir width of 30 cm and height of 15 cm at a discharge of 25 L/s. (a) z/P = 0.3; (b) z/P = 0.8; (c) z/P = 1.6.
5.3. Side Channel Pressure Distribution
When water flowed through the side weir, an upstream water level was formed, resulting in a pressure zone at the junction with the side channel. This pressure zone led to increased water pressure on the floor of the side channel, which affected its stability and durability. In small channels or fields where erosion resistance is weak, excessive pressure can cause scour holes. Therefore, analyzing the pressure distribution in the side channel is necessary to select an appropriate height and width for the side weir that effectively reduces its impact on the bottom plate.
To investigate the impact of side weir width on hydraulic characteristics, pressure data was collected at a discharge rate of 25 L/s for side weirs with heights of 20 cm and widths ranging from 20 cm to 47 cm. The pressure distribution map was drawn, as shown in Figure 13.
Figure 13. Comparison of pressure distribution on the bottom plate of the side channel with different widths of side weirs when the discharge is 25 L/s and the height of side weirs is 20 cm. (a) P = 20 cm, b = 20 cm; (b) P = 20 cm, b = 30 cm; (c) P = 20 cm, b = 40 cm; (d) P = 20 cm, b = 47 cm.
As can be seen from Figure 13, the pressure at the bottom of the side channel decreased as the width of the side weir increased. This uneven distribution of water flow on the weir was caused by the sharp bending of water flow lines and the influence of centrifugal inertia force over a short period. After passing through the side weir, the water flow became symmetrically distributed with respect to the axis of the side channel, leaning towards the right bank at a certain distance. As we increased the width of the side weir, we noticed that its position gradually approached the side weir and maximum pressure decreased at this location where the water tongue formed due to flowing through it (Figure 13). For a constant height (20 cm) but varying widths (20 cm, 30 cm, 40 cm, and 47 cm), we measured maximum pressures at these positions as follows: 103,713 Pa, 103,558 Pa, 103,324 Pa, and 103,280 Pa, respectively. Consequently, increasing width reduced the impact on the side channel from water flowing through it while changing pressure distribution from concentration to dispersion in a vertical direction. In the practical application of side weirs, appropriate height should be selected based on the bottom plate’s capacity to withstand the pressure exerted by flowing water within channels.
To investigate how height affects the hydraulic characteristics of rectangular side weirs further (Figure 14), we extracted pressures on bottom plates when discharge was fixed at 25 L/s while varying heights were set as follows: 7 cm, 10 cm, 15 cm, and 20 cm, respectively.
Figure 14. Comparison of pressure distribution on the bottom plate of the side channel with different heights of side weirs when discharge is 25 L/s and the width of side weirs is 20 cm. (a) P = 7 cm, b = 20 cm; (b) P = 10 cm, b = 20 cm; (c) P = 15 cm, b = 20 cm; (d) P = 20 cm, b = 20 cm.
As shown in Figure 14, when the width of the side weir was constant, the pressure at the bottom of the side channel increased with the height of the side weir. As the height of the side weir increased, the water tongue formed by flow through the side weir gradually moved away from it in a downstream direction. In terms of vertical water flow, as the height of the side weir increased, the position of maximum pressure at which the water tongue falls shifted closer to the axis of the side channel from its right bank. Moreover, an increase in height resulted in higher maximum pressure at this falling point. For a constant width (20 cm) and varying heights (7 cm, 10 cm, 15 cm, and 20 cm), corresponding maximum pressures at this landing point were measured as 102,422 Pa, 102,700 Pa, 103,375 Pa, and 103,766 Pa, respectively. Consequently, increasing width led to a greater impact on both flow through and pressure distribution within the side channel; transforming it from scattered to concentrated along its lengthwise direction. Therefore, when applying such weirs practically one should select an appropriate width based on what pressure can be sustained by their respective channel bottom plates.
5.4. Discharge Coefficient
Based on dimensionless analysis, the influencing parameters of the discharge coefficient were obtained. To study the effect of parameters Fr1, b/h1, and P/h1, discharge coefficient values were plotted against Fr1, b/h1, and P/h1, shown in Figure 15, Figure 16 and Figure 17. The discharge coefficient decreased as parameters Fr1 and b/h1 increased. The discharge coefficient increased as parameter P/h1 increased. As Uyumaz and Muslu reported in a previous study, the variation of the discharge coefficient with respect to the Froude number showed a second-degree curve for a subcritical regime [4].
Figure 15. Variation of discharge coefficient values against Froude number.
Figure 16. Variation of discharge coefficient values against the percentage of the side weir width to the upstream flow depth over the side weir.
Figure 17. Variation of discharge coefficient values against the percentage of the side weir height to the upstream flow depth over the side weir.
Quantitative analysis between discharge coefficient values and parameters Fr1, b/h1, and P/h1 was conducted using data analysis software (IBM SPSS Statistics 19). The various coefficients obtained are shown in Table 3.
Model
Unstandardized Coefficients
Standardized Coefficients
t
Sig
B
Std. Error
Beta (β)
Constant
−1.294
0.155
−8.369
0.000
Fr1
3.430
0.286
3.401
12.013
0.000
b/h1
−0.004
0.004
−0.045
−0.944
0.348
P/h1
2.401
0.167
4.064
14.394
0.000
Table 3. Coefficient.
The value of t and Sig are the significance results of the independent variable, and the value of Sig corresponding to the value of t is less than 0.05, indicating that the independent variable has a significant impact on the dependent variable. Therefore, the values of Sig corresponding to the parameters Fr1 and P/h1 were less than 0.05, indicating that the parameters Fr1 and P/h1 have a significant impact on the discharge coefficient. On the contrary, the parameter b/h1 has less impact on the discharge coefficient. Therefore, quantitative analysis between discharge coefficient values and parameters Fr1, and P/h1 was conducted using data analysis software by removing factor b/h1. The model summary, ANOVA, and coefficient obtained are shown respectively in Table 4, Table 5 and Table 6. R and adjusted R square in Table 4 were approaching 1, which indicated the goodness of fit of the regression model was high. The value of Sig corresponding to the value of F in Table 5 was less than 0.05, which indicated that the regression equation was useful. The values of Sig corresponding to the parameters Fr1 and P/h1 in Table 6 were less than 0.05, indicating that the parameters Fr1 and P/h1 have a significant impact on the discharge coefficient.
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
0.913 a
0.833
0.829
0.03232
Table 4. Model Summary b. Note: a. Predictors:(Constant), Fr1, P/h1; b. Discharge coefficient.
Model
Sum of Squares
df
Mean Square
F
Sig
1
Regression
0.402
2
0.201
192.545
0.000 a
Residual
0.080
77
0.001
Total
0.483
79
Table 5. ANOVA b. Note: a. Predictors:(Constant), Fr1, P/h1; b. Discharge coefficient.
Model
Unstandardized Coefficients
Standardized Coefficients
t
Sig
B
Std. Error
Beta (β)
Constant
−1.326
0.151
−8.796
0.000
Fr1
3.479
0.281
3.449
12.396
0.000
P/h1
2.427
0.164
4.108
14.765
0.000
Table 6. Coefficient a. Note: a. Predictors:(Constant), Fr1, P/h1.
Based on the above analysis, the flow coefficient formula has been obtained, shown as follows:
Discharge formula were obtained by substituting Equation (15) into Equation (12), as shown in Equation (16).
where Q ∈ [0.006, 0.030], m3/s; b ∈ [0.20, 0.47], m; P ∈ [0.07, 0.20], m.
Figure 18 showed the measured discharge coefficient values with those calculated from discharge formulas in Table 3. The scatter of the data with respect to perfect line was limited to ±10%.
Figure 18. Comparison of the measured discharge coefficient values with those calculated from discharge formulas in Table 3.
6. Discussions
Determining water surface profile near the side weir in the main channel is one of the tasks of hydraulic calculation for side weirs. As the water flows through the side weir, discharge in the main channel is gradually decreasing, namely dQ/ds<0. According to the Equation (17) derived from Qimo Chen [3], it can be inferred that the value of 𝑑ℎ/𝑑𝑠 is greater than zero in subcritical flow (Fr < 1), that is, the water surface profile near the side weir in the main channel is a backwater curve. Due to the side weir entrance effect at the upstream end, water surface profiles drop slightly at the upstream end of the side weir crest, as EI-Khashab [28] and Emiroglu et al. [29] reported in previous experimental studies.
In this study, the water surface profile exhibited a backwater curve along the length of the weir crest. Therefore, during side weir application, it is crucial to ensure that downstream water levels do not exceed the highest water level of the channel.
The head on the weir is one of the important factors that flow over side weirs depends on. At the same time, the head depends on the water surface profile near the side weir in the main channel. Therefore, further research on the quantitative analysis of water surface profile needs to be conducted. Mohamed Khorchani proposed a new approach based on the video monitoring concept to measure the free surface of flow over side weirs. It points out a new direction for future research [8].
The maximum flow velocity, a key parameter in assessing the efficiency of a weir, occurs at the upstream end of the weir crest, typically near the crest. This is attributed to the convergence of the flow as it approaches the crest, resulting in a significant increase in velocity. It was found that in this study the minimum flow velocity occurred at the bottom of the main channel away from the side weir. Under such conditions, the accumulation of sediments could lead to siltation, which in turn can affect the accuracy of flow measurement through side weirs. This is because the presence of sediments can alter the flow patterns and cause errors in the measurement. Therefore, it becomes crucial to explore methods to optimize the selection of side weirs in order to minimize or eliminate the effects of sedimentation on flow measurement.
Pressure distribution plays a crucial role in ensuring structural safety for side weirs since small channels and field inlets have relatively limited pressure-bearing capacities. Therefore, it is important to select an appropriate geometrical parameter of rectangular side weirs based on their ability to withstand the pressure exerted on their bottom combined with pressure distribution data at the bottom of the side channel we have obtained in this study.
The discharge coefficient formula (Equation (15)), which incorporates Fr1 and P/h1, was derived based on dimensional analysis. However, it is worth noting that previous research has contradicted this formula by suggesting that the discharge coefficient solely depends on the Froude number. This conclusion can be observed in this study such as in Equations (18)–(23) in Table 7 of the manuscript [30,31,32,33,34,35], which clearly demonstrate the dependency of the discharge coefficient on the Froude number. In contrast, our derived discharge coefficient formula (Equation (15)) offers a more streamlined and simplified approach compared to Equation (25) [36] and Equation (29) [10]—making it easier to comprehend and apply—an advantageous feature particularly valuable in fluid dynamics where intricate calculations can be time-consuming. Furthermore, our derived discharge coefficient formula (Equation (15)) exhibits a broader application scope than that of Equation (24) [37] as shown in Table 8. Equation (26) [38] and Equation (27) [5] are specifically applicable under high flow discharge conditions. Conversely, our derived discharge coefficient formula (Equation (15)) is better suited for low-flow discharge conditions.
Table 7. Discharge coefficient formulas of rectangular side weirs presented in previous studies.
Discharge/(L·s−1)
Width of Side Weir/cm
Height of Side Weir/cm
Number of Formula
10~14
10~20
6~12
(24)
35–100
20~75
1~19
(26), (27)
6~30
20~47
7~20
(15)
Table 8. Application scope of discharge coefficient formulas.
In addition to the factors studied in the paper, factors such as the sediment content in the flow, the bottom slope, and the cross-section shape of the channel also have a certain impact on the hydraulic characteristics of the side weir. Further numerical simulation methods can be used to study the hydraulic characteristics and the influencing factors of the side weir. Water measurement facilities generally require high accuracy of water measurement, the flow of sharp-crested side weirs is complex, and the water surface fluctuates greatly. While conducting numerical simulations, experimental research on prototype channels is necessary to ensure the reliability of the results and provide reference for the body design and optimization of side weirs in small channels and field inlets.
7. Conclusions
This paper presents a comprehensive study that encompasses both experimental and numerical simulation research on rectangular side weirs of varying heights and widths within rectangular channels. A thorough analysis of the experimental and numerical simulation results has been conducted, leading to the derivation of several notable conclusions:
A comparative analysis was conducted on the measured and simulated values of water depth and flow velocity. Both of the maximum absolute relative errors were within 10%, which indicated that the numerical simulation of the side weir was feasible and effective.
The water surface profile exhibited a backwater curve along the length of the weir crest. The side weir entrance effect occurred only between Side Ⅰ and Side Ⅱ. This indicates that flow patterns and associated hydraulic forces at the weir entrance play a crucial role in determining water level distribution along the weir crest.
The maximum flow velocity of the cross-section at the upstream end of the weir crest occurred near the weir crest, while the minimum flow velocity occurred at the bottom of the main channel away from the side weir. As the water depth decreased, the position of the maximum flow velocity gradually moved from the upstream end of the side weir to the downstream end of the side weir.
When the height of the side weir remains constant, an increase in the width of the side weir leads to a decrease in pressure at the bottom of the side channel. Conversely, when the width of the side weir is kept constant, an increase in its height results in an increase in pressure at the bottom of the side channel. Therefore, during practical applications involving side weirs, it is crucial to select an appropriate weir width based on the maximum pressure that can be sustained by the channel’s bottom plate.
The discharge coefficient was found to depend on the upstream Froude number Fr1 and the percentage of the side weir height to the upstream flow depth over the side weir P/h1. The relationship between the discharge coefficient and parameters Fr1 and P/h1 was obtained using multiple regression analysis, which was of linear form and provided an easy means to estimate the discharge coefficient. The discharge formula is of high accuracy with relative errors within 10%, which met the water measurement accuracy requirements of small channels in irrigation areas.
Reference
De Marchi, G. Essay on the performance of lateral weirs. L’Energ Electr. 1934, 11, 849.
Ackers, P. A theoretical consideration of side weirs as storm water overflows. Proc. Inst. Civ. Eng. 1957, 6, 250–269.
Chen, Q.M.; Xie, P.Z.; Chen, Q.R. Experiment on hydraulic characteristics of side weir. J. Fuzhou Univ. 1979, 19, 26–29.
Uyumaz, A.; Muslu, Y. Flow over side weir in circular channels. ASCE J. Hydraul. Eng. 1985, 111, 144–160.
Borghei, M.; Jalili, M.R.; Ghodsian, M. Discharge coefficient for sharp-crested side weir in subcritical flow. ASCE J. Hydraul. Eng. 1999, 125, 1051–1056.
Ghodsian, M. Supercritical flow over rectangular side weir. Can. J. Civ. Eng. 2003, 30, 596–600.
Durga Rao, K.H.V.; Pillai, C.R.S. Study of Flow Over Side Weirs Under Supercritical Conditions. Water Resour Manag. 2008, 22, 131–143.
Khorchani, M.; Blanpain, O. Free surface measurement of flow over side weirs using the video monitoring concept. Flow Meas. Instrum. 2004, 15, 111–117.
Bilhan, O.; Emiroglu, M.E.; Kisi, O. Application of two different neural network techniques to lateral outflow over rectangular side weirs located on a straight channel. Adv. Eng. Softw. 2010, 41, 831–837.
Emiroglu, M.E.; Agaccioglu, H.; Kaya, N. Discharging capacity of rectangular side weirs in straight open channels. Flow Meas. Instrum. 2011, 22, 319–330.
Azza, N.; Al-Talib, A.N. Flow over oblique side weir. J. Damascus Univ. 2012, 28, 15–22.
Bagheri, S.; Kabiri-Samani, A.R.; Heidarpour, M. Discharge coefficient of rectangular sharp-crested side weirs part i: Traditional weir equation, Flow Measure. Instrumentation 2014, 35, 109–115.
Shariq, A.; Hussain, A.; Ansari, M.A. Lateral flow through the sharp crested side rectangular weirs in open channels. Flow Measure. Instrumentation 2018, 59, 8–17.
Li, G.D.; Shen, G.Y.; Li, S.S.; Lu, Q.N. Prediction Model of Side Weir Discharge Capacity Based on LS-SVM. J. Basic Sci. Eng. 2023, 4, 843–851.
Shabanlou, S.; Salimi, M.S. Free surface and velocity field in a circular channel along the side weir in supercritical flow conditions. Flow Meas. Instrum. 2014, 38, 108–115.
Samadi, A.; Arvanaghi, H.; Abbaspour, A. Three-Dimensional Simulation of Free Surface Flow over Rectangular Sharp crested Weirs. Int. J. Agric. Biosci. 2015, 4, 83–86.
Aydin, M.C. Investigation of a Sill Effect on Rectangular Side-Weir Flow by Using CFD. J. Irrig. Drain. Eng. 2016, 142.
Azimi, H.; Shabanlou, S.; Ebtehaj, I.; Bonakdari, H. Discharge Coefficient of Rectangular Side Weirs on Circular Channels. Int. J. Nonlinear Sci. Numer. Simul. 2016, 17, 391–399.
Khassaf, S.I.; Attiyah, A.N.; Al-Yousify, H.A. Experimental investigation of compound side weir with modeling using computational fluid dynamic. Energy Environ. 2018, 7, 169–178.
Safarzadeh, A.; Noroozi, B. 3D Hydrodynamics of Trapezoidal Piano Key Spillways. Int. J. Civ. Eng. 2017, 15, 89–101.
Selim, T.; Hamed, A.K.; Elkiki, M.; Eltarabily, M.G. Numerical investigation of flow characteristics and energy dissipation over piano key and trapezoidal labyrinth weirs under free-flow conditions. Model. Earth Syst. Environ. 2023, 10, 1253–1272.
Novak, P.; Cabelka, J. Models in Hydraulic Engineering; Pitman: London, UK, 1981.
Henderson, F.M. Open Channel Flow; Prentice-Hall: Englewood Cliffs, NJ, USA, 1966.
Wang, F.J. Computational Fluid Dynamics Analysis-Theory and Application of CFD; Tsinghua University Press: Beijing, China, 2004.
Zhu, Y.L.; Ma, X.Y.; Zhan, G.L.; Lv, J.W. Numerical simulation of flow in flat V-weir. Yellow River 2010, 32, 99–100.
Harlow, F.H.; Welch, J.E. Numberical calculation of time-dependent viscous incompressible flow of fluid with free surface. Phys. Fluids 1965, 8, 2182–2189.
Hirt, C.W.; Nichols, B.D. Volume of fluid (VOF) method for dynamics of free boundaries. Phys. Fluids 1981, 39, 201–221.
El-Khashab, A.M.M. Hydraulics of Flow over Side Weirs. Ph.D. Thesis, University of Southampton, Southampton, UK, 1975.
Emiroglu, M.E.; Kaya, N.; Agaccioglu, H. Discharge capacity of labyrinth side-weir located on a straight channel. ASCE J. Irrig. Drain. Eng. 2010, 136, 37–46.
Subramanya, K.; Awasthy, S.C. Spatially varied flow over side weirs. ASCE J. Hydraul. Div. 1972, 98, 1–10.
Nandesamoorthy, T.; Thomson, A. Discussion of spatially varied flow over side weir. ASCE J. Hydraul. Eng. 1972, 98, 2234–2235.
Yu-Tech, L. Discussion of spatially varied flow over side weir. ASCE J. Hydraul. Div. 1972, 98, 2046–2048.
Ranga Raju, K.G.; Prasad, B.; Grupta, S.K. Side weir in rectangular channel. ASCE J. Hydraul. Div. 1979, 105, 547–554.
Hager, W.H. Lateral outflow over side weirs. ASCE J. Hydraul. Eng. 1987, 113, 491–504.
Cheong, H.F. Discharge coefficient oflateral diversion from trapezoidal channel. ASCE J. Irrig. Drain. Eng. 1991, 117, 321–333.
Swamee, P.K.; Santosh, K.P.; Masoud, S.A. Side weir analysis using elementary discharge coefficient. ASCE J. Irrig. Drain. Eng. 1994, 120, 742–755.
Singh, R.; Manivannan, D.; Satyanarayana, T. Discharge coefficient of rectangular side weirs. ASCE J. Irrig. Drain. Eng. 1994, 120, 814–819.
Jalili, M.R.; Borghei, S.M. Discussion of Discharge coefficient of rectangular side weir. ASCE J. Irrig. Drain. Eng. 1996, 122, 132.
Enhanced understanding of flow structure in braided rivers is essential for river regulation, flood control, and infrastructure safety across the river. It has been revealed that the basic morphological element of braided rivers is confluence-bifurcation units. However, flow structure in these units has so far remained poorly understood with previous studies having focused mainly on single confluences/bifurcations. Here, the flow structure in a laboratory-scale confluence-bifurcation unit is numerically investigated based on the FLOW–3D® software platform. Two discharges are considered, with the central bars submerged or exposed respectively when the discharge is high or low. The results show that flow convergence and divergence in the confluence-bifurcation unit are relatively weak when the central bars are submerged. Based on comparisons with a single confluence/bifurcation, it is found that the effects of the upstream central bar on the flow structure in the confluence-bifurcation unit reign over those of the downstream central bar. Concurrently, the high-velocity zone in the confluence-bifurcation unit is less concentrated than that in a single confluence while being more concentrated than that observed in a single bifurcation. The present work unravels the flow structure in a confluence-bifurcation unit and provides a unique basis for further investigating morphodynamics in braided rivers.
1 Introduction
Confluences and bifurcations commonly exist in alluvial rivers and usually are important nodes of riverbed planform (Szupiany et al., 2012; Hackney et al., 2018). Flow convergence and divergence in these junctions result in highly three-dimensional (3D) flow characteristics, which greatly influence sediment transport, and hence riverbed evolution and channel formation (Le et al., 2019; Xie et al., 2020). Braided rivers, characterized by unstable networks of channels separated by central bars (Ashmore, 2013), have confluence-bifurcation units as their basic morphological elements (Ashmore, 1982; 1991; 2013; Federici & Paola, 2003; Jang & Shimizu, 2005). In particular, confluence-bifurcation units exhibit a distinct morphology from single confluences/bifurcations and bifurcation-confluence regions because two adjacent central bars are included. Within a confluence-bifurcation unit, two tributaries converge at the upstream bar tail and soon diverge to two anabranches again at the downstream bar head. Therefore, the flow structure in the unit may be significantly influenced by both the two central bars, and thus considerably different from that in single confluences, single bifurcations, and bifurcation-confluence regions, where the flow is affected by only one central bar. Enhanced understanding of flow structure in confluence-bifurcation units is urgently needed, which is essential for water resources management, river regulation, flood control, protection of river ecosystems and the safety of infrastructures across the rivers such as bridges, oil pipelines and communication cables (Redolfi et al., 2019; Ragno et al., 2021).
The flow dynamics, turbulent coherent structures, and turbulent characteristics in single confluences have been widely studied since the 1980s (Yuan et al., 2022). Flow dynamics at river channel confluences have been systematically and completely analyzed, which can be characterized by six major regions of flow stagnation, flow deflection, flow separation, maximum velocity, flow recovery and distinct shear layers (Best, 1987). For example, the field observation of Roy et al. (1988) and Roy and Bergeron (1990) highlighted the flow separation zones and recirculation at downstream natural confluence corners. Ashmore et al. (1992) measured the flow field in a natural confluence and found flow accelerates suddenly at the confluence junction with two separated high-velocity cores merging into one single core at the channel centre. De Serres et al. (1999) investigated the three-dimensional flow structure at a river confluence and identified the existence of the mixing layer, stagnation zones, separation zones and recovery zones. Sharifipour et al. (2015) numerically studied the flow structure in a 90° single confluence and found that the size of the separation zone decreases with the width ratio between the tributary and the main channel. Recently, three main classes of large-scale turbulent coherent structures (Duguay et al., 2022) have been presented, i.e. vertical-orientated vortices or Kelvin-Helmholtz instabilities (Rhoads & Sukhodolov, 2001; Constantinescu et al., 2011; 2016; Biron et al., 2019), channel-scale ‘back-to-back’ helical cells, (Mosley, 1976; Ashmore, 1982; Ashmore et al., 1992; Ashworth, 1996; Best, 1987; Rhoads & Kenworthy, 1995; Bradbrook et al., 1998; Lane et al., 2000), and smaller, strongly coherent streamwise-orientated vortices (Constantinescu et al., 2011; Sukhodolov & Sukhodolova, 2019; Duguay et al., 2022). However, no consensus on a universal turbulent coherent structure mode has been reached so far (Duguay et al., 2022). In addition, some studies (Ashworth, 1996; Constantinescu et al., 2011; Sukhodolov et al., 2017; Le et al., 2019; Yuan et al., 2023) have focused on turbulent characteristics, e.g. turbulent kinetic energy, turbulent dissipation rate and Reynolds stress, which can be critical parameters to further explaining the diversity of these turbulent coherent structure modes.
Investigations on the flow structure in single bifurcations have mainly focused on hydrodynamics in anabranches (Hua et al., 2009; van der Mark & Mosselman, 2013; Iwantoro et al., 2022) and around bifurcation bars (McLelland et al., 1999; Bertoldi & Tubino, 2005; 2007; Marra et al., 2014), whereas few studies have considered the effects of bifurcations on the upstream flow structure. Thomas et al. (2011) found that the velocity core upstream of the bifurcation is located near the water surface and towards the channel center in experimental investigations of a Y-shaped bifurcation. Miori et al. (2012) simulated flow in a Y-shaped bifurcation and found two circulation cells upstream of the bifurcation with flow converging at the water surface and diverging near the bed. Szupiany et al. (2012) reported velocity decreasing and back-to-back circulation cells upstream of the bifurcation junction in the field observation of a bifurcation of the Rio Parana River. These investigations provide insight into how bifurcations affect the flow patterns upstream, yet there is a need for further research on the dynamics of flow occurring immediately before the bifurcation junction.
Generally, the findings of studies on bifurcation-confluence regions are similar to those concerning single confluences and bifurcations. Hackney et al. (2018) measured the hydrodynamic characteristics in a bifurcation-confluence of the Mekong River and found the velocity cores located at the channel centre and strong secondary current occurring under low discharges. Le et al. (2019) reported a high-turbulent-kinetic-energy (high-TKE) zone located near the bed in their numerical simulation of flow in a natural bifurcation-confluence region. Moreover, a stagnation zone was found upstream of the confluence and back-to-back secondary current cells were detected at the confluence according to Xie et al. (2020) and Xu et al. (2022). Overall, these studies have further unraveled the flow patterns in river confluences and bifurcations.
Unfortunately, limited attention has been paid to the flow structure in confluence-bifurcation units. Parsons et al. (2007) investigated a large confluence-bifurcation unit in Rio Parana, Argentina, and no classical back-to-back secondary current cells were observed under a discharge of 12000 m3·s−1. To date, the differences in flow structure between confluence-bifurcation units and single confluences/bifurcations have remained far from clear. In addition, although the effects of discharge on flow structure have been investigated in several studies on single confluences/bifurcations, (Hua et al., 2009; Le et al., 2019; Luz et al., 2020; Xie et al., 2020; Xu et al., 2022), cases with fully submerged central bars were not considered, which is typical in braided rivers during floods. In-depth studies concerning these issues are urgently needed to gain better insight into the flow structure in confluence-bifurcation units of braided rivers.
This paper aims to (1) reveal the 3D flow structure in a confluence-bifurcation unit under different discharges and (2) elucidate the differences in the flow structure between confluence-bifurcation units and single confluence/bifurcation cases. Using the commercial computational fluid dynamics software FLOW-3D® (Version 11.2; https://www.flow3d.com; Flow Science, Inc.), fixed-bed simulations of a laboratory-scale confluence-bifurcation unit are conducted, and cases of a single confluence/bifurcation are also included for comparison. Two discharges are considered, with the central bars fully submerged or exposed respectively when the discharge is high or low. Based on the computational results, the 3D flow structure in the confluence-bifurcation unit conditions is analyzed from various aspects including free surface elevation, time-averaged flow velocity distribution, recirculation vortex structure, secondary current, and turbulent kinetic energy and dissipation rate. In particular, the flow structure in the confluence-bifurcation unit is compared with that in the single confluence/bifurcation cases to unravel the differences.h
2. Conceptual flume and computational cases
2.1. Conceptual flume
In this paper, a laboratory-scale conceptual flume is designed and used in numerical simulations. Figure 1(a–d) shows the morphological characteristics of the flume. To ensure that the conceptual flume reflects morphology features of natural braided channels, key parameters governing the flume morphology, e.g. unit length, width, and channel width-depth ratio, are determined according to studies on morphological characteristics of natural confluence-bifurcation units (Hundey & Ashmore, 2009; Ashworth, 1996; Orfeo et al., 2006; Parsons et al., 2007; Sambrook Smith et al., 2005; Kelly, 2006; Ashmore, 2013; Egozi & Ashmore, 2009; Redolfi et al., 2016; Ettema & Armstrong, 2019).
Figure 1. The sketch of the conceptual flume: (a) the original flume, (b) the central bar: (c) the sketch of cross-section C-C, (d) the sketch of cross-section D-D, (e) the modified part for the single confluence, (f) the modified part for the single bifurcation, (g) the position of different cross-sections. The red dashed boxes denote the regions of primary concern.
Figure 1. The sketch of the conceptual flume: (a) the original flume, (b) the central bar: (c) the sketch of cross-section C-C, (d) the sketch of cross-section D-D, (e) the modified part for the single confluence, (f) the modified part for the single bifurcation, (g) the position of different cross-sections. The red dashed boxes denote the regions of primary concern.
2.1.1. Length and width scales of the confluence-bifurcation unit
The length and width scales of the flume are first determined. The inner relation among the length LCB and average width B of a confluence-bifurcation unit and the average width Bi of a single branch was statistically studied by Hundey and Ashmore (2009), which indicates the following relations: 𝐿CB =(4∼5)𝐵 (1) 𝐵 =1.41𝐵𝑖 (2) In addition, Ashworth (1996) gave B = 2Bi in his experimental research on mid-bar formation downstream of a confluence, while the confluence-bifurcation unit of Rio Parana, Argentina has a relation of B≈1.71Bi (Orfeo et al., 2006; Parsons et al., 2007). Accordingly, the following relations are used in the present paper: 𝐿CB =4𝐵 (3) 𝐵 =1.88𝐵𝑖 (4) where LCB = 6 m, B = 1.5 m and Bi = 0.8 m.
2.1.2. Central bar morphology
The idealized plane pattern of central bars in braided rivers is a slightly fusiform leaf shape with a short upstream side and a long downstream side (Ashworth, 1996; Sambrook Smith et al., 2005; Kelly, 2006; Ashmore, 2013). To simplify the design, the bar is approximated as a combination of two different semi-ellipses (Figure 1(b)). The major axis Lb is two to ten times longer than the minor axis Bb according to the statistical data in Kelly’s study, and the regression equation is given as (Kelly, 2006): 𝐿𝑏=4.62𝐵0.96𝑏 (5) In this study, the bar width Bb is set as 0.8 m, whilst the lengths of downstream (LT1) and upstream sides (LT2) are 2 and 1.5 m, respectively (Figure 1(b)). Thus, the relation of Lb and Bb is given as: 𝐿𝑏=(𝐿𝑇1+𝐿𝑇2)=4.375𝐵𝑏 (6) The lengths of the inlet and outlet parts are determined as Lin = Lout = 8 m, which ensures negligible effects of boundary conditions without exceptional computational cost.
2.1.3. Width-depth ratio
Channel flow capacity can be significantly affected by cross-section shapes. For natural rivers, cross-section shapes can be generalized into three sorts based on the following width-depth curve (Redolfi et al., 2016): 𝐵=𝜓𝐻𝜑(7) Braided rivers usually have ψ = 5∼50 and φ>1, which indicates a rather wide and shallow cross-section. The central bar form should also be taken into account, so a parabolic cross-section shape is used here with ψ = 8 and φ>1 (Figure 1(c,d)).
2.1.4. Bed slope
In addition, natural braided rivers are usually located in mountainous areas and thus have a relatively large bed slope. According to flume experiments and field observations, the bed slope Sb is mostly in the range of 0.01∼0.02, and a few are below 0.01 (Ashworth, 1996; Egozi & Ashmore, 2009; Ashmore, 2013; Redolfi et al., 2016; Ettema & Armstrong, 2019). In this study, Sb takes 0.005.
2.1.5. Complete sketch of the conceptual flume
In summary, the flume is 29 m long, 2.4 m wide, and 0.6 m high. The plane coordinates (x-direction and y-direction) used in the calculation process are shown in Figure 1 (a). Note that the inlet corresponds to x = 0 m, and the centreline of the flume is located at y = 1.3 m. Besides, the thalweg elevation of the outlet is set as z = 0 m.
2.2. Computational cases
As stated before, the first aim of this paper is to reveal the flow structure in the confluence-bifurcation unit under different discharges. Therefore, two basic cases are set first: (1) case 1a under a low discharge (0.05 m3·s−1) with exposed central bars and (2) case 2a under a high discharge (0.30 m3·s−1) with fully submerged central bars. A total of 22 cross-sections are identified to examine the results (Figure 1(g)).
Further, cases of a single confluence/bifurcation are generated by splitting the original confluence-bifurcation unit into two parts. Part 1 only includes the upstream central bar and focuses on the flow convergence downstream of CS04 (Figure 1(e)), while Part 2 only includes the downstream central bar and focuses on the flow divergence upstream of CS19 (Figure 1(f)). Notably, the numbers of corresponding cross-sections in the original flume are reserved to facilitate comparison. The outlet section of the single confluence as well as the inlet section of the single bifurcation is extended to make the total length equivalent to the original flume (29 m). Also, two discharge conditions (0.05 and 0.30 m3·s−1), which correspond to exposed and fully submerged central bars, are considered for the single confluence/bifurcation. In total, six computational cases are conducted, as listed in Table 1. As the conceptual flume is designed to be symmetrical about the centreline, the momentum flux ratio (Mr) of the two branches should be 1 in all six cases. This is confirmed by further examining the computational results.
Case
Configuration
Qin (m3·s−1)
Zout (m)
Mr
Condition of bars
1a
CBU
0.05
0.15
1
Exposed
1b
SC
0.05
0.15
1
Exposed
1c
SB
0.05
0.15
1
Exposed
2a
CBU
0.30
0.34
1
Submerged
2b
SC
0.30
0.34
1
Submerged
2c
SB
0.30
0.34
1
Submerged
Table 1. Computational cases with inlet and outlet boundary conditions.
3. Numerical method
In this section, the 3D Large Eddy Simulation (LES) model integrated in the FLOW-3D® (Version 11.2; https://www.flow3d.com; Flow Science, Inc.) software platform is introduced, including governing equations and boundary conditions. Information on computational meshes with mesh independence test can be found in the Supplementary material.
3.1. Governing equations
The LES model was applied in the present paper to simulate flow in the laboratory-scale confluence-bifurcation unit. The LES model has been proven to be effective in simulating turbulent flow in river confluences and bifurcations (Constantinescu et al., 2011; Le et al., 2019). The basic idea of the LES model is that one should directly compute all turbulent flow structures that can be resolved by the computational meshes and only approximate those features that are too small to be resolved (Smagorinsky, 1963). Therefore, a filtering operation is applied to the original Navier-Stokes (NS) equations for incompressible fluids to distinguish the large-scale eddies and small-scale eddies (Liu et al., 2018). The filtered NS equations are then generated, which can be expressed in the form of a Cartesian tensor as (Liu, 2012):
(10) where ¯𝑢𝑖 is the resolved velocity component in the i – direction (i goes from 1 to 3, denoting the x-, y – and z-directions, respectively); t is the flow time; ρ is the density of the fluid; ¯𝑝 is the pressure; ν is the kinematic viscosity; τij is the sub-grid scale (SGS) stress; ¯𝐺𝑖 is the body acceleration. In FLOW–3D®, the full NS equations are discretized and solved using the finite-volume/finite-difference method (Bombardelli et al., 2011; Lu et al., 2023).
Due to the filtering process, the velocity can be divided into a resolved part (¯𝑢(𝑥,𝑡)) and an approximate part (𝑢′(𝑥,𝑡)) which is also known as the SGS part (Liu, 2012). To achieve model closure, the standard Smagorinsky SGS stress model is introduced here (Smagorinsky, 1963): 𝜏ij−13𝜏kk𝛿ij=−2𝜈SGS¯𝑆ij(11) where νSGS is the SGS turbulent viscosity, and ¯𝑆ij is the resolved rate-of-strain tensor for the resolved scale defined by (Smagorinsky, 1963): ¯𝑆ij=12(∂¯𝑢𝑖∂𝑥𝑗+∂¯𝑢𝑗∂𝑥𝑖)(12) In the standard Smagorinsky SGS stress model, the eddy viscosity is modelled by (Smagorinsky, 1963): 𝜈SGS=(𝐶𝑠¯𝛥)2∣¯𝑆∣,∣¯𝑆∣=√2¯𝑆ij¯𝑆ij(13) ¯𝛥=(ΔxΔyΔz)1/3(14) where Cs is the Smagorinsky constant, Δx, Δy, and Δz are mesh scales. In FLOW–3D®, Cs is between 0.1 to 0.2 (Smagorinsky, 1963). One of the key problems in simulating 3D open channel flow is the calculation of free surface. FLOW–3D® uses the Volume of Fluid (VOF) method (Hirt & Nichols, 1981) to track the change of free surface. The VOF method introduces a fluid phase fraction function f to characterize the proportion of a certain fluid in each mesh cell. In that case, the surface position can be precisely located if the mesh cell is fine enough. To monitor the change of f with time and space, the following convection equation is added:
For open channel flow, only two kinds of fluids are involved: water and air. If f is the fraction of water, the state of the fluid in each mesh cell can be defined as:
In FLOW–3D®, the interface between water and air is assumed to be shear-free, which means that the drag force on the water from the air is negligible. Moreover, in most cases, the details of the gas motion are not crucial for the heavier water motion so the computational processes will be more efficient.
3.2. Boundary conditions
Six boundary conditions need to be preset in the 3D numerical simulation process. Discharge boundary conditions are used for the inlet of the flume, where the free surface elevation is automatically calculated based on the free surface elevation boundary conditions set for the outlet. The specific information on the inlet and outlet boundary conditions for all computational cases is shown in Table 1. Moreover, because the free surface moves temporally, the free surface boundary conditions are just set as no shear stress and having a normal pressure, and the position of the free surface will be automatically adjusted over time by the VOF method in FLOW–3D®. Furthermore, the bed and two side walls are all set to be no-slip for fixed bed conditions, and a standard wall function is employed at the wall boundaries for wall treatment.
The inlet turbulent boundary conditions also need to be considered. They are set by default here. The turbulent velocity fluctuations V’ are assumed to be 10% of the mean flow velocity with the turbulent kinetic energy (TKE) (per unit mass) equaling 0.5V’2. The maximum turbulent mixing length is assumed to be 7% of the minimum computational domain scale, and the turbulent dissipation rate is evaluated automatically from the TKE.
4. Results and discussion
4.1. Flow structure in the confluence-bifurcation unit
4.1.1. Free surface elevation
Figure 2 shows the free surface elevation at five different longitudinal profiles (i.e. α = 0.2, 0.4, 0.5, 0.6, 0.8) for cases 1a and 2a. The parameter α was defined as follows:𝛼=𝑠𝐵(17) where s is the transverse distance between a certain profile and the left boundary of the flume. In general, the longitudinal change of free surface in the two cases is very similar despite different discharge levels. The free surface elevation decreases as the channel narrows from the upstream bifurcation to the front of the confluence-bifurcation unit. On the contrary, when the flow diverges again at the end of the confluence-bifurcation unit, the free surface elevation increases with channel widening. However, whether the fall or rise of free surface elevation in case 1a is much sharper than that in case 2a, especially at profiles with α = 0.2 and 0.8 (Figure 2(a)), which indicates there may be distinct flow states between the two cases. To further illustrate this finding, the Froude number Fr at different cross-sections (CS08∼CS15) is examined. In case 2a, the flow remains subcritical within the confluence-bifurcation unit. By contrast, in case 1a, a local supercritical flow is observed near the side banks of CS09 (i.e. α = 0.2 and 0.8), with Fr being about 1.2. This local supercritical flow can lead to a hydraulic drop followed by a hydraulic jump, which accounts for the sharp change of the free surface. The foregoing reveals that when central bars are exposed under relatively low discharge, supercritical flow is more likely to occur near the side banks of the confluence junction due to flow convergence.
Figure 2. Five time-averaged free surface elevation profiles in the confluence-bifurcation unit, in which α denotes the lateral position of the certain profile. Note that the black dashed line denotes the position of CS09, where Fr is about 1.2 near the side banks (α = 0.2 and 0.8) in case 1a. Z’ = z/h2, X’ = x/B, h2 is the maximum flow depth at the outlet boundary of cases 2a, 2b and 2c, h2 = 0.34 m.
Moreover, in both cases 1a and 2a, the free surface is higher at the channel centre than near the side banks, whether at the front or the end of the confluence-bifurcation unit. Thus, lateral free surface slopes from the centre to the side banks are generated. For example, the lateral free surface slopes at CS09 are 0.022 and 0.016 respectively for cases 1a and 2a. These lateral slopes can lead to lateral pressure gradient force whose direction is from the channel centreline to the side banks. Notably, the lateral surface slope in case 1a is steeper than that in case 2a, which may also result from the effect of the supercritical flow.
4.1.2. Time-averaged streamwise flow velocity
Figure 3. Time-averaged flow velocity distribution at three different slices over z-direction in the confluence-bifurcation unit: (a)∼(c) case 1a, (d)∼(f) case 2a. The flow direction is from the left to the right. StZ = Stagnation Zones, MiL = Mixing Layer. X’ = x/B, Y’ = y/B, Ui’ = Ui/Uti, Ui denotes the time-averaged streamwise flow velocity in case series i (i = 1,2), Uti denotes the cross-section-averaged streamwise flow velocity in case series i, Ut1 = 0.385 m/s, for case 2a Ut2 = 0.714 m/s.Figure 4. Time-averaged flow velocity contours at eight different cross-sections in the confluence-bifurcation unit: (a) case 1a, (b) case 2a.
Besides the shared features described above, some differences between the two cases are also identified. First, flow stagnation zones at the upstream bar tail are found exclusively in case 1a as the central bars are exposed (Figure 3 (a–c)). Second, in case 1a the mixing layer is obvious in both the lower or upper flows (Figure 3 (a–c)), while in case 2a the mixing layer can be inconspicuous in the upper flow (Figure 3 (f)). Third, in case 1a, two high-velocity cores gradually transform into one single core downstream of the confluence [Figure 4 (a), CS08∼CS11] and are divided into two cores again at the downstream bar head [Figure 4 (a), CS15]. By contrast, in case 2a, the two cores merge much more rapidly [Figure 4 (a), CS08∼CS09], and no obvious reseparation of the merged core is found at the downstream bar head (Figure 3 (d–f)). The latter two differences between cases 1a and 2a indicate that the flow convergence and divergence are relatively weak when the central bars are fully submerged. It is noticed that when the central bars are exposed, the flow in branches needs to steer around the central bar, which can cause a large angle between the two flow directions at the confluence, and thus relatively strong flow convergence and divergence may occur. By contrast, when the central bars are fully submerged, the flow behavior resembles that of a straight channel, with flow predominantly moving straight along the main axis of the central bars. Therefore, a small angle between two tributary flow forms, and thus flow convergence and divergence are relatively mild.
4.1.3. Recirculation vortex
A recirculation vortex with a vertical axis is a typical structure usually found where flow steers sharply, and is generated from flow separation (Lu et al., 2023). This vortex structure is found in the confluence-bifurcation unit in the present study, marking several significant flow separation zones. Figure 5 shows the recirculation vortex structure at the bifurcation junction of the confluence-bifurcation unit. In both cases 1a and 2a, two recirculation vortices BV1 and BV2 are found at the bifurcation junction corner. Moreover, BV1 and BV2 seem well-established near the bed but tend to transform into premature ones in the upper flow, and there is also a tendency for the cores of BV1 and BV2 to shift downstream as they transition from the lower to the upper flow (Figure 5(a–c,d–f)). This finding indicates that flow separation zones exist at the bifurcation junction corner, and the vortex structure is similar in the separation zones under low and high discharges. These flow separation zones are generated due to the inertia effect as flow suddenly diverges and steers towards the curved side banks of the channel (Xie et al., 2020). Notably, two additional vortices BV3 and BV4 are found at both sides of the downstream bar in case 1a (Figure 5(a–c)), but no such vortices exist in case 2a. This difference shows that flow separation zones at both sides of the downstream bar are hard to form when the bars are completely submerged under the high discharge.
Figure 5. Recirculation vortices at the bifurcation junction (streamline view at three different slices over z-direction): (a)∼(c) case 1a, (d)∼(f) case 2a. The red solid line marked out the position of these vortices (BV1∼BV4).
Similarly, Figure 6 shows the recirculation vortex structure at the confluence junction of the confluence-bifurcation unit. No noteworthy similarities but a key difference between the two cases are observed at this site. Two vortices CV1 and CV2 are found downstream of the confluence junction corner in case 1a (Figure 6(c)), which mark two separation zones. Conversely, no such separation zones are found in case 2a. In fact, separation zones were reported at similar sites under relatively low discharges in some previous studies (Ashmore et al., 1992, Luz et al., 2020, Sukhodolov & Sukhodolova, 2019; Xie et al., 2020). Nevertheless, the flow separation zones at the confluence corner are very restricted in the present study (Figure 6(c)). Ashmore et al. (1992) also reported that no, or very restricted flow separation zones occur downstream of natural river confluence corners, primarily because of the relatively slow change in bank orientation compared with the sharp corners of laboratory confluences where separation is pronounced (Best & Reid, 1984; Best, 1988). In the present study, the bank orientation also changes slowly, which may explain why flow separation zones are inconspicuous at the confluence corner.
Figure 6. Recirculation vortices at the confluence junction (streamline view at three different slices over z-direction): (a)∼(c) case 1a, (d)∼(f) case 2a. The red solid line marked out the position of these vortices (CV1 & CV2).
The differences in the distribution of recirculation vortices discussed above may be mainly attributed to the difference in the angle between the tributary flows under different discharges. Some previous studies have reported that the confluence/bifurcation angle can significantly influence the flow structure at confluences/bifurcations (Best & Roy, 1991; Ashmore et al., 1992; Miori et al., 2012). Although the confluence/bifurcation angle is fixed due to the determined central bar shape in the present study, the angle between two tributary flows is affected by the varying discharge. When the central bars are exposed under the low discharge, the flow is characterized by a more pronounced curvature of the streamlines, and a large angle between the two tributary flows is noted (Figure 6(b)), causing strong flow convergence and divergence. By contrast, a small angle forms as the central bars are submerged, thereby leading to relatively weak flow convergence/divergence (Figure 6(e)). Overall, the differences mentioned above can be attributed to the differences in the intensity of flow convergence and divergence under different discharges.
It should be noted that some previous studies (Constantinescu et al., 2011; Sukhodolov & Sukhodolova, 2019) presented that there is a wake mode in the mixing layer of two streams at the confluence junction. The wake mode means that in the mixing layer, multiple streamwise coherent vortices moving downstream will form, which is similar to the flow structure around a bluffing body (Constantinescu et al., 2011). However, no such structure has been found within the confluence-bifurcation unit in this study. According to the numerical simulations of Constantinescu et al. (2011), a wake mode was found at a river confluence with a concordant bed and a momentum flux ratio of about 1. The confluence has a much larger angle (∼60°) between the two streams when compared to the confluence junction of the confluence-bifurcation unit in the present study where the angle is about 25°. As flow mechanics at river confluences may include several dominant mechanisms depending on confluence morphology, momentum ratio, the angle between the tributaries and the main channel, and other factors (Constantinescu et al., 2011), the relatively small confluence angle in the present study may explain why the wake mode is absent. The possible effects of the confluence/bifurcation angle are reserved for future study. Additionally, flow separation can lead to reduced local sediment transport capacity, thus causing considerable sediment deposition under natural conditions. Hence, the bank may migrate towards the inner side of the channel at the positions of CV1, CV2, BV1, and BV2, while the bar may expand laterally at the positions of BV3 and BV4.
4.1.4. Secondary current
Secondary current is the flow perpendicular to the mainstream axis (Thorne et al., 1985) and can be categorized into two primary types based on its origin: (1) Secondary current generated by the interaction between centrifugal force and pressure gradient force; (2) Secondary current resulting from turbulence heterogeneity and anisotropy (Lane et al., 2000). There are some widely recognized definitions of secondary current strength (SCS) (Lane et al., 2000). In this paper, the secondary current cells are identified by visible vortex with a streamwise axis, and the definition of SCS proposed by Shukry (1950) is used:
where ux, uy, and uz are flow velocities in x, y, and z directions and ux represents the mainstream flow velocity.
Figure 7 presents contour plots of SCS and the secondary current structure at key cross-sections of the study area. When the central bars are exposed, at the upstream bar tail (CS08), intense transverse flow occurs with flow converging to the centreline, but no secondary current cell is formed (Figure 7(a)). This is consistent with the findings of Hackney et al. (2018). At the confluence junction (CS09), transverse flow still plays a major role in the secondary current structure, with flow converging to the centreline at the surface and diverging to side banks near the bed (Figure 7(b)). Moreover, ‘back-to-back’ helical cells, which are two vortices rotating reversely, tend to generate at CS09 with their cores located near the side banks (Figure 7(b)) (Mosley, 1976; Ashmore, 1982; Ashmore et al., 1992), yet their forms are rather premature. As the flow goes downstream, the cores of the helical cells gradually rise to the upper flow and approach towards the centreline, and the helical cells become well-established (Figure 7(c–e)). When the flow diverges again at the downstream bar head (CS15), the helical cells attenuate rapidly, and the secondary current structure is once again characterized predominantly by transverse flow (Figure 7(f)).
Figure 7. Distribution of secondary current strength and secondary current cells at six different cross-sections: (a)∼(f) case 1a, (g)∼(l) case 2a. The secondary current cells are identified by visible lateral vortices (streamline view). The zero distance of each cross-section is located on the right bank.
When the central bars are fully submerged under the high discharge, the secondary current structure at the upstream bar tail and the confluence junction exhibits a resemblance to that under the low discharge (Figure 7(g,h)). However, at CS09, two pairs of cells with different scales tend to form under the high discharge (Figure 7(h)). The large and premature helical cells are similar to those under the low discharge, whereas the small helical cells are located near side banks possibly due to wall effects. As the flow moves downstream, the large helical cells tend to diminish rapidly and merge with the small ones near both side walls (Figure 7(i–k)). Moreover, the secondary current structure is once again characterized predominantly by transverse flow at CS14 under the high discharge, which occurs earlier than that under the low discharge (Figure 7(k)). At the downstream bar head, transverse flow still takes a dominant place, while the helical cells seem to become premature with increased scale (Figure 7(l)).
In general, in both cases 1a and 2a, the lateral distribution of SCS at all cross-sections is symmetrical about the channel centreline, where SCS is relatively small. A relatively high SCS is detected at both the upstream bar tail and the downstream bar head due to the effects of centrifugal force caused by flow steering. SCS decreases rapidly from the upstream bar tail (CS08) to the entrance of the downstream bifurcation junction (CS14), followed by a sudden increase at the downstream bar head (CS15) (Figure 7 (a–e, g–k)). However, the distribution of high-SCS zones is different between the two discharges. Under the low discharge, high-SCS zones appear along the bottom near the centerline and at the free surface on both sides of the centreline. Although similar high-SCS zones are found along the bottom near the centerline under the high discharge, the high-SCS zones are not found at the free surface. Furthermore, it is noticed that more obvious high-SCS zones appear under the low discharge compared with the high discharge, especially at CS09. This may be attributed to the differences in the intensity of flow convergence and divergence under different submerging conditions of the central bars. When the central bars are exposed, flow convergence and divergence are strong and sharp flow steering occurs, thereby causing large SCS. By contrast, when the central bars are fully submerged, flow convergence and divergence are relatively weak, and thus small SCS is observed.
4.1.5. Turbulent characteristics
Turbulent characteristics reflect the performance of energy and momentum transfer activities in flow (Sukhodolov et al., 2017). Comprehensive analysis of turbulent characteristics is crucial as they greatly impact the incipient motion, settling behavior, diffusion pattern, and transport process of sediment. Here, the TKE and turbulent dissipation rate (TDR) of flow in the confluence-bifurcation unit are analyzed.
Figure 8 shows the distribution of TKE on various cross-sections in cases 1a and 2a. In the same way, Figure 10 shows the distribution of TDR. The values of TKE and TDR are nondimensionalized with mid-values of TKE = 0.005 m2·s−2 and TDR = 0.007 m3·s−2. In both cases 1a and 2a, the distributions of TKE and TDR show symmetrical patterns concerning the channel centreline. High-TKE and high-TDR zones exhibit a belt distribution near the channel bottom (McLelland et al., 1999; Ashworth, 1996; Constantinescu et al., 2011), indicating that turbulence primarily originates at the channel bottom due to the influence of bed shear stress. A sudden increase of TKE (Weber et al., 2001) and TDR occurs near the channel bottom at the confluence junction [Figure 8 and 9, CS08∼CS09] and from the entrance of the bifurcation junction (CS14) to the downstream bar head (CS15) (Figures 8 and 9).
Figure 8. Turbulent kinetic energy contours at eight different cross-sections in the confluence-bifurcation unit: (a) case 1a, (b) case 2a. TKE = turbulent kinetic energy. TKE’ = dimensionless value of TKE, with regard to a mid-value of TKE = 0.005 m2·s−2.Figure 9. Turbulent dissipation rate contours at eight different cross-sections in the confluence-bifurcation unit: (a) case 1a, (b) case 2a. TDR = turbulent dissipation rate. TDR’ = dimensionless value of TDR, with regard to a mid-value of TDR = 0.007 m3·s−2.Figure 10. Comparison of the distribution of time-averaged streamwise flow velocity along the flow depth at different cross-sections between the confluence-bifurcation unit and the single confluence. (a)∼(f) 1a vs. 1b, (g)∼(l) 2a vs. 2b.
Despite the common turbulent characteristics between cases 1a and 2a, additional high-TKE zones are found in the upper flow at the upstream bar tail (CS08), the confluence junction (CS09) and the downstream bar head (CS15) (Figure 8) when the central bars are fully submerged. The formation mechanism of these high-TKE zones near the water surface is more complicated, which may result from interactions of velocity gradient, secondary current structure and wall shear stress (Engel & Rhoads, 2017; Lu et al., 2023).
4.2. Comparison with single confluence/bifurcation cases
In this section, the results of a single confluence (cases 1b and 2b) and a single bifurcation (cases 1c and 2c) are compared with those of the confluence-bifurcation unit (cases 1a and 2a) under two discharges. Flow structure at CS08∼CS15 is mainly concerned below.
4.2.1. Comparison with single confluence cases
First, the patterns of time-averaged streamwise velocity, TKE and TDR within the single confluence (presented by contour plots in the supplementary materials) are assessed and then compared with those within the confluence-bifurcation unit (Figures 4, 8, and 9). It is found that distributions of these parameters are similar in the confluence-bifurcation unit and the single confluence from the upstream bar tail (CS08) to the entrance of the bifurcation junction (CS14), despite varying discharges. As the existence of the downstream central bar is the main difference between the single confluence and the confluence-bifurcation unit, this finding indicates that the downstream bar may have limited influence on the flow structure in the confluence-bifurcation unit. In other words, the flow structure in the confluence-bifurcation unit appears to be mainly shaped by the presence of the upstream bar, with its impact potentially reaching as far as the entrance of the bifurcation (CS14). Moreover, under the low discharge, the two high-velocity cores seem to merge later (at CS11) in the single confluence than in the confluence-bifurcation unit (at CS10), which indicates the convergence of two tributary flows may achieve a steady state faster in the confluence-bifurcation unit. To further elucidate the differences, results on the distribution of time-averaged streamwise velocity and TKE along the flow depth are discussed below.
4.2.1.1. Time-averaged streamwise velocity
Figure 10 shows the distribution of time-averaged streamwise flow velocity along the flow depth at different cross-sections. Note that α = 0.5 denotes the channel centreline and α = 0.7 denotes a position near the side banks. As only marginal differences are found at α = 0.3 and 0.7, only profiles at α = 0.7 are displayed for clarity.
Under the low discharge, no obvious difference in the distribution of time-averaged streamwise flow velocity is observed at the upstream bar tail (Figure 10(a)). At the confluence junction (Figure 10(b)), the velocities near the side banks (α = 0.7) are larger than those at the centre (α = 0.5) in both the confluence-bifurcation unit and the single confluence, which suggests that the two tributary flows have not sufficiently merged. The two tributary flows achieve convergence at CS11 in both the confluence-bifurcation unit and the single confluence (Figure 10(c)), with the velocity at the centre (α = 0.5) is larger than that near the side banks. Nevertheless, the velocities at the centre (α = 0.5) and near the side banks (α = 0.7) are closer to each other in the confluence-bifurcation unit than those in the single confluence, which represents less sufficient flow convergence in the confluence-bifurcation unit than in the single confluence. Therefore, it can be inferred that the convergence of two tributary flows may achieve a steady state faster in the confluence-bifurcation unit. After reaching the steady state, the velocity near the side banks (α = 0.7) is smaller in the single confluence than in the confluence-bifurcation unit despite close values at the centre (α = 0.5) (Figure 10(d,e)). This leads to a more pronounced disparity between velocities at the centre and near the side banks in the single confluence than that observed in the confluence-bifurcation unit. In other words, the high-velocity zone is more concentrated on the channel centreline in the single confluence, while the lateral distribution of flow velocity tends to be more uniform in the confluence-bifurcation unit. This may be attributed to the influence of the downstream central bar, which is further proved by comparing the velocity profiles at CS15 (Figure 10(e)).
As for the high discharge condition, from CS08 to CS14, the quantitative differences in velocity distribution between the confluence-bifurcation unit and the single confluence seem small. This indicates that the effect of morphology appears to be subdued when the central bars are fully submerged under the high discharge. It should be also noted that under both the low and high discharge, velocity profiles at the corresponding location exhibit the same shapes in the confluence-bifurcation unit and the single confluence, which indicates that the upstream confluence may dominate the flow structure in the confluence-bifurcation unit.
4.2.1.2. Secondary current
Figure 11 shows contour plots of SCS and the secondary current structure for single confluence cases. Compared with Figure 7, under both low and high discharge conditions, the distribution of SCS and the structure of helical cells in the confluence-bifurcation unit and the single confluence are very similar from CS08 to CS12 (Figure 7(a–d, g–j) and Figure 11(a–d, g–j)]. This indicates that the secondary current structure in the confluence-bifurcation unit exhibits certain consistent features when compared to those in the single confluence, thus proving that the effects of the upstream central bar may dominate the flow structure in the confluence-bifurcation unit. However, the secondary current structure at CS14 and CS15 is different between the confluence-bifurcation unit and the single confluence (Figure 7 and 11(e, f, k,l)). Under the low discharge, transverse flow is from the side banks to the centre and relatively high SCS is found near the side banks at CS14 in the single confluence, while the transverse flow is always from the centre to the side banks and SCS is relatively low at the corresponding sites in the confluence-bifurcation unit (Figure 11(e)). Under the high discharge, the helical cells near the side walls almost diminish in the single confluence, while they still exist in the confluence-bifurcation unit at CS14 (Figure 11(k)). Under both low and high discharges, the secondary current pattern at CS15 is similar to that at CS14 in the single confluence, while they are different in the confluence-bifurcation unit due to the existence of the downstream central bar. This comparison indicates that the existence of the downstream central bar can influence the upstream secondary current structure, nevertheless, the effects are fairly limited.
Figure 11. Secondary current at different cross-sections in the single confluence condition: (a)∼(f) case 1b, (g)∼(l) case 2b. The zero distance of each cross-section is located on the right bank.
4.2.1.3. Turbulent kinetic energy
Figure 12 shows TKE distribution along the flow depth at different cross-sections. Under the low discharge, in general, the maximum TKE tends to appear near the channel bottom in both the confluence-bifurcation unit and the single confluence. No obvious difference is observed at the upstream bar tail (CS08) (Figure 12(a)). Downstream this site (at CS09), the maximum TKE near the side banks (α = 0.7) is larger than that at the channel centre in the single confluence, while they are close to each other in the confluence-bifurcation unit (Figure 12(b)). This can also be attributed to the insufficient convergence of the two tributary flows. At CS11, flow convergence achieves a steady state in the confluence-bifurcation unit, while it remains insufficient in the single confluence. As flow convergence reaches a steady state at CS12, the maximum TKE in the single confluence exhibits a more concentrated distribution on the channel centre than that in the confluence-bifurcation unit (Figure 12(d)). This effect becomes more obvious downstream at CS14 (Figure 12(e)). The less-concentrated distribution of the maximum TKE in the confluence-bifurcation unit can be owing to the effects of the downstream central bar as well, which appears analogous to that mentioned in 4.2.1.1.
Figure 12. Comparison of the distribution of TKE along the flow depth at different cross-sections between the confluence-bifurcation unit and the single confluence. (a)∼(f) 1a vs. 1b, (g)∼(l) 2a vs. 2b.
Under the high discharge condition, two peaks of TKE appear in both the confluence-bifurcation unit and the single confluence (Figure 12(g–l)). Moreover, in both the confluence-bifurcation unit and the single confluence, from the upstream bar tail to the downstream bar head, the peak of TKE in the upper flow is larger at the channel centre (α = 0.5), while the peak of TKE in the lower flow is larger near the side banks (α = 0.7). However, the disparity between the TKE near the side banks and at the channel centre seems to be larger in the single confluence, while the TKE in the confluence-bifurcation unit takes a more uniform distribution. Even though, TKE profiles at the corresponding location exhibit highly similar shapes in the confluence-bifurcation unit and the single confluence, suggesting that the effects of channel morphology seem to be inhibited when the central bars are submerged under the high discharge.
4.2.2. Comparison with single bifurcation cases
Distributions of time-averaged streamwise velocity, TKE and TDR at corresponding cross-sections are also compared between the single bifurcation (see the Supplementary material) and the confluence-bifurcation unit (Figures 4, 8 and 9). Unlike the high similarity in flow characteristics exhibited between the confluence-bifurcation unit and the single confluence, significant differences are found between the confluence-bifurcation unit and the single bifurcation, especially at CS08∼CS14. On the one hand, the high-velocity zones are broader and asymmetrical concerning the channel centreline in the single bifurcation, with a belt-like and an approximately elliptic-like distribution respectively under the low and high discharges. By contrast, the high-velocity zone is a core that concentrates on the channel centre in the confluence-bifurcation unit. Moreover, the maximum velocity seems smaller in the single bifurcation than that in the confluence-bifurcation unit. On the other hand, the high-TKE belt near the channel bottom appears to be narrower in the single bifurcation than in the confluence-bifurcation unit, especially at CS08∼CS14 under the low discharge. Furthermore, additional high-TKE zones are found near the side walls at CS08∼CS11 in the single bifurcation, of which the scale is obviously smaller than those in the confluence-bifurcation unit. In addition, TKE at the channel centre is smaller near the free surface in the single bifurcation than that in the confluence-bifurcation unit. Nevertheless, the distributions of velocity, TKE and TDR seem to be similar in the confluence-bifurcation unit and the single bifurcation at CS15. As the existence of the upstream central bar is the main difference between the single confluence and the confluence-bifurcation unit, all the above findings indicate that the upstream central bar greatly influences the flow structure in the confluence-bifurcation unit. On the other hand, the downstream central bar may have a restricted influence on the flow structure in the confluence-bifurcation unit, whose impact may be limited to a range between the entrance of the bifurcation (CS14) and the downstream bar head (CS15). To further elucidate the differences, results on the distribution of time-averaged streamwise velocity and TKE along the flow depth are discussed below.
4.2.2.1. Time-averaged streamwise velocity
Figure 13 shows the distribution of time-averaged streamwise velocity along the flow depth at different cross-sections. Under the low discharge, distinct distribution patterns of flow velocity between the confluence-bifurcation unit and the single bifurcation are found at CS08, CS09 and CS11, which can be attributed to the effects of upstream flow convergence (Figure 13(a–c)). However, when the flow convergence reaches a steady state in the confluence-bifurcation unit (Figure 13(d–f)), the high-velocity zone is more concentrated in the confluence-bifurcation unit than in the single bifurcation due to to the significant influence of the upstream central bar on the flow structure. The velocity profiles at the downstream bar head can be a shred of evidence as well, with the maximum velocity larger at the channel centre but smaller near the side banks in the confluence-bifurcation unit than in the single bifurcation.
Figure 13. Comparison of the distribution of time-averaged streamwise flow velocity along the flow depth at different cross-sections between the confluence-bifurcation unit and the single bifurcation. (a)∼(f) 1a vs. 1c, (g)∼(l) 2a vs. 2c.
Under the high discharge, the distribution of velocity seems to exhibit limited differences between the two kinds of morphology, which indicates that the effects of channel morphology may be less noticeable when the central bars are fully submerged under the high discharge. Nevertheless, the velocity in the lower flow (below a relative depth of 0.45) shows a uniform lateral distribution in the single bifurcation, as the velocity profile at the channel centreline (α = 0.5) is in line with that near the side banks (α = 0.7) (Figure 13(g–l)). However, in the confluence-bifurcation unit, different velocity distributions in the lower flow can be observed at the channel centreline (α = 0.5) and near the side banks (α = 0.7). The foregoing results indicate that when the central bars are fully submerged, the high-velocity zones are more concentrated on the channel centreline in the confluence-bifurcation unit, while the lateral distribution of flow velocity within the single bifurcation tends to be more uniform, especially near the bifurcation junction (Figure 13(j,k)). This can also be attributed to the dominant influence of the upstream central bar over the downstream central bar.
It is also noted that the flow velocity distribution along the flow depth in the confluence-bifurcation unit is of a similar pattern despite varying discharges. As a critical point, the maximum velocity appears in the upper flow. The distribution above the critical point is approximately linear whereas it appears logarithmic below. By contrast, despite the similarity observed under the low discharge, the flow velocity distribution along the flow depth within the single bifurcation exhibits a distinct pattern under the high discharge, especially near the side banks (Figure 13(e–h)). On the one hand, the critical point in the upper flow no longer corresponds to the maximum velocity. On the other hand, the velocity distribution deviates from logarithmic below the critical point, with the maximum velocity appearing at a relative depth of 0.45. Succinctly, the distribution of streamwise velocity along the flow depth may retain the same pattern regardless of discharge levels in the confluence-bifurcation unit, while it may exhibit distinct patterns under different discharge levels in the single bifurcation.
4.2.2.2. Secondary current
Figure 14 shows contour plots of SCS and the distribution of secondary current for single bifurcation cases. In general, the value of SCS near the side banks at CS08∼CS14 (Figure 14(a–d, g–j)) in the single bifurcation seems smaller than that in the confluence-bifurcation unit (Figure 7(a–d, g–j)), especially under the low discharge. SCS distribution at CS14 is similar in the confluence-bifurcation unit and the single bifurcation under both low and high discharges. This difference in SCS distribution between the confluence-bifurcation unit and the single bifurcation indicates that the downstream bifurcation may have a restricted influence on the flow structure in the confluence-bifurcation unit. This influence is limited to a range between the entrance of the bifurcation (CS14) and the downstream bar head (CS15).
Figure 14. Secondary current at different cross-sections in the single bifurcation condition: (a)∼(f) case 1c, (g)∼(l) case 2c. The zero distance of each cross-section is located on the right bank.
In addition, the secondary current structure may also present different patterns in response to varying channel morphologies and discharge conditions. Under the low discharge condition, multiple unstable helical cells with asymmetrical distribution are formed from CS08 to CS12 in the single bifurcation (Figure 14(a–d)), while no obvious helical cells are found at CS14 and CS15 (Figure 14(d,e)). These findings are quite different from the stable and symmetrical helical cells at all cross-sections shown in the confluence-bifurcation unit (Figure 7). This difference may be attributed to the significant influence of the upstream central bar and the limited influence of the downstream central bar. Under the high discharge condition, only one pair of premature helical cells are found from CS08 to CS12 in the single bifurcation with their cores located near the side banks (Figure 14(e,f)). As the flow moves downstream, the helical cells gradually develop and become well-established (Figure 14(g,h)). These helical cells in the single bifurcation show symmetric cross-sectional distribution and a similar longitudinal development as in the confluence-bifurcation unit. However, in the confluence-bifurcation unit, two pairs of helical cells appear upstream of CS12 and CS14 and gradually fuse to one pair under the high discharge. As the ‘two-pairs’ structure in the confluence-bifurcation unit origins from the upstream confluence, the differences in the secondary current structure between the single bifurcation and the confluence-bifurcation unit under the high discharge can also be owing to the effects of the upstream central bar in excess of those of the downstream central bar.
4.2.2.3. Turbulent kinetic energy
Figure 15 shows the TKE distribution along the flow depth at different cross-sections. Under the low discharge, when the two tributary flows have not achieved sufficient convergence in the confluence-bifurcation unit, the maximum TKE is more concentrated in the single bifurcation (Figure 15(a–c)). As flow convergence achieves a steady state, more concentrated high-TKE zones appear at the channel centre within the confluence-bifurcation unit, confirming the finding that the effects of the upstream central bar reign over those of the downstream central bar in the confluence-bifurcation unit. However, things can be very complicated under the high discharge. For TKE distribution at the channel centreline, two peaks appear in the confluence-bifurcation unit with one close to the free surface and the other near the bed (Figure 15(g–l)). By contrast, only one peak near the bed is present in the single bifurcation. Therefore, a larger TKE can be found in the upper flow of the channel centreline in the confluence-bifurcation unit. For TKE distribution near the side banks, two peaks appear in both the confluence-bifurcation unit and the single bifurcation at CS09∼CS14 (Figure 15(h–l)). The upper peak is larger but the lower peak is smaller within the single bifurcation than those within the confluence-bifurcation unit. These significant discordances in TKE distribution under the high discharge further prove that the effects of the upstream bar on the flow structure in the confluence-bifurcation unit are more prominent than those of the downstream central bar.
Figure 15. Comparison of the distribution of TKE along the flow depth at different cross-sections between the confluence-bifurcation unit and the single bifurcation. (a)∼(f) 1a vs. 1c, (g)∼(l) 2a vs. 2c.
4.2.3. Further discussion of the comparisons
The above subsections have revealed significant differences in flow structure within the confluence-bifurcation unit and the single confluence and bifurcation, which directly result from the distinct channel morphologies and vary with the discharge conditions as well. These differences are summarized and further discussed below.
The distinctive morphology of a confluence-bifurcation unit plays a pivotal role in governing streamwise flow velocity distribution, secondary current structure, and turbulent kinetic energy distribution within the channel. Generally, from the upstream bar tail (CS08) to the entrance of the bifurcation (CS14), the flow structure in the confluence-bifurcation unit is highly similar to that in the single confluence, while it exhibits great differences (as shown in 4.2.2) between the confluence-bifurcation unit and the single bifurcation. This indicates that the upstream central bar greatly influences the flow structure in the confluence-bifurcation unit, with the effects spreading to the entrance of the bifurcation. At the downstream bar head (CS15), the flow structure (e.g. the transverse flow patterns) in the confluence-bifurcation unit exhibits high similarity to that in the single bifurcation. However, these similarities do not spread to upstream cross-sections, suggesting that the influence of the downstream central bar is limited at the bifurcation junction. In a word, the effects of the upstream central bar on the flow structure in the confluence-bifurcation unit are in excess of those of the downstream central bar.
However, despite the influence of channel morphology, discharge may also have some important effects on the streamwise flow velocity distribution. On the one hand, when the central bars are exposed under the low discharge, the high-velocity zone is less concentrated in the confluence-bifurcation unit than in the single confluence, while it is more concentrated in the confluence-bifurcation unit than in the single bifurcation. On the other hand, it is noticed that when the central bars are fully submerged under the high discharge, reduced differences in flow structure between the confluence-bifurcation unit and the single confluence/bifurcation are witnessed, and thus the morphology effect seems to be subdued.
4.3. Implications
The present work unravels the flow structure in a laboratory-scale confluence-bifurcation unit and takes the first step to further investigating morphodynamics in such channel morphology. Based on the comparison with a single confluence/bifurcation, the findings provide insight into the complex 3D interactions between water flow and channel morphology. The distinct flow structure in the laboratory-scale confluence-bifurcation unit may appreciably alter sediment transport and morphological evolution, of which research is underway. As the basic morphological element of braided river planform is confluence-bifurcation units, the present work should have direct implications for flow structure in natural braided rivers. This is pivotal for the sustainable management of braided rivers which deals with water and land resources planning, eco-hydrological well-being, and infrastructure safety such as cross-river bridges and oil pipelines (Redolfi et al., 2019; Ragno et al., 2021).
Notably, braided rivers worldwide (e.g. in the Himalayas, North America, and New Zealand) have undergone increased pressures and will continue to evolve due to forces of global climate change and intensified anthropogenic activities (Caruso et al., 2017; Hicks et al., 2021; Lu et al., 2022). In particular, channel aggradation caused by increased sediment supply as well as exploitation of braidplain compromise space for flood conveyance, making the rivers prone to flooding. In this sense, an enhanced understanding of the flow structure under high discharge when central bars are fully submerged is essential for mitigating flooding hazards.
5. Conclusions
This study has numerically investigated the 3D flow structure in a laboratory-scale confluence-bifurcation unit based on the LES model integrated in the FLOW–3D® software platform. Two different discharges are considered with the central bars fully submerged or exposed respectively when the discharge is high or low. Cases of a single confluence/bifurcation are included for comparison. The key findings of this paper are as follows:
Several differences are highlighted in the comparison of the flow structure in the confluence-bifurcation unit between the two discharges. When the central bars are fully submerged under the high discharge, the mixing layer of two tributary flows is less obvious, and two high-velocity cores merge more rapidly as compared with those under the low discharge. Besides, flow separation zones are found neither at the confluence corner nor on both sides of the downstream bar when the central bars are fully submerged. Moreover, SCS seems to be smaller near the side banks under the high discharge than under the low discharge. Therefore, it is suggested that flow convergence/divergence is relatively weak in the confluence-bifurcation unit when central bars are fully submerged under the high discharge.
From the upstream bar tail to the entrance of the bifurcation, the flow structure in the confluence-bifurcation unit is highly similar to that in the single confluence, while it exhibits great differences from that in the single bifurcation. Only at the downstream bar head does the flow structure in the confluence-bifurcation unit exhibit high similarity to that in the single bifurcation. Consequently, the effects of the upstream central bar on the flow structure in the confluence-bifurcation unit reign over those of the downstream central bar.
Despite the influence of channel morphology, discharge may also have significant effects on the distribution of streamwise flow velocity. On the one hand, when the central bars are exposed under the low discharge, the high-velocity zone is less concentrated in the confluence-bifurcation unit than in the single confluence, while it is more concentrated in the confluence-bifurcation unit than in the single bifurcation. On the other hand, when the central bars are fully submerged under the high discharge, reduced differences in flow structure between the confluence-bifurcation unit and the single confluence/bifurcation are witnessed, and thus the morphology effect seems to be subdued.
It is noticed that the effects of other factors (e.g. confluence and bifurcation angles, bed discordance) on the flow structure in the confluence-bifurcation unit are not discussed here. Studies on these issues are warranted and reserved for future work.
Reference
Ashmore, P. E. (1982). Laboratory modelling of gravel braided stream morphology. Earth Surface Processes and Landforms, 7(3), 201–225. https://doi.org/10.1002/esp.3290070301
Ashmore, P. E. (1991). How do gravel-bed rivers braid? Canadian Journal of Earth Sciences, 28(3), 326–341. https://doi.org/10.1139/e91-030
Ashmore, P. E. (2013). Morphology and dynamics of braided rivers. In J. Shroder, & (Editor in Chief) E. Wohl (Eds.), Treatise on geomorphology (Vol. 9, pp. 289–312). https://doi.org/10.1016/B978-0-12-374739-6.00242-6
Ashmore, P. E., Ferguson, R. I., Prestegaard, K. L., Ashworth, P. J., & Paola, C. (1992). Secondary flow in anabranch confluences of a braided, gravel-bed stream. Earth Surface Processes and Landforms, 17(3), 299–311. https://doi.org/10.1002/esp.3290170308
Ashworth, P. J. (1996). Mid channel bar growth and its relationship to local flow strength and direction. Earth Surface Processes and Landforms, 21(2), 103–123.
Bertoldi, W., & Tubino, M. (2005). Bed and bank evolution of bifurcating channels. Water Resources Research, 41(7), W07001. https://doi.org/10.1029/2004WR003333
Bertoldi, W., & Tubino, M. (2007). River bifurcations: Experimental observations on equilibrium configurations. Water Resources Research, 43(10), W10437. https://doi.org/10.1029/2007WR005907
Best, J. L. (1987). Flow dynamics at river channel confluences: Implications for sediment transport and bed morphology. In F. G. Ethridge, R. M. Flores, & M. D. Harvey (Eds.), Recent developments in fluvial sedimentology (pp. 27–35).
Best, J. L. (1988). Sediment transport and bed morphology at river channel confluences. Sedimentology, 35(3), 481–498. https://doi.org/10.1111/j.1365-3091.1988.tb00999.x
Best, J. L., & Reid, I. (1984). Separation zone at open-channel junctions. Journal of Hydraulic Engineering, 110(11), 1588–1594. https://doi.org/10.1061/(ASCE)0733-9429(1984)110:11(1588)
Best, J. L., & Roy, A. G. (1991). Mixing-layer distortion at the confluence of channels of different depth. Nature, 350(6317), 411–413. https://doi.org/10.1038/350411a0
Biron, P. M., Buffin-Bélanger, T., & Martel, N. (2019). Three-dimensional turbulent structures at a medium-sized confluence with and without an ice cover. Earth Surface Processes and Landforms, 44(15), 3042–3056. https://doi.org/10.1002/esp.4718
Bombardelli, F. A., Meireles, I., & Matos, J. (2011). Laboratory measurements and multi-block numerical simulations of the mean flow and turbulence in the nonaerated skimming flow region of steep stepped spillways. Environmental Fluid Mechanics, 11(3), 263–288. https://doi.org/10.1007/s10652-010-9188-6
Bradbrook, K. F., Biron, P. M., Lane, S. N., Richards, K. S., & Roy, A. G. (1998). Investigation of controls on secondary circulation in a simple confluence geometry using a three-dimensional numerical model. Hydrological Processes, 12(8), 1371–1396. https://doi.org/10.1002/(SICI)1099-1085(19980630)12:8<1371::AID-HYP620>3.0.CO;2-C
Caruso, B., Newton, S., King, R., & Zammit, C. (2017). Modelling climate change impacts on hydropower lake inflows and braided rivers in a mountain basin. Hydrological Sciences Journal, 62(6), 928–946. https://doi.org/10.1080/02626667.2016.1267860
Constantinescu, G., Miyawaki, S., Rhoads, B., & Sukhodolov, A. (2016). Influence of planform geometry and momentum ratio on thermal mixing at a stream confluence with a concordant bed. Environmental Fluid Mechanics, 16(4), 845–873. https://doi.org/10.1007/s10652-016-9457-0
Constantinescu, G., Miyawaki, S., Rhoads, B., Sukhodolov, A., & Kirkil, G. (2011). Structure of turbulent flow at a river confluence with momentum and velocity ratios close to 1: Insight provided by an eddy-resolving numerical simulation. Water Resources Research, 47(5), W05507. https://doi.org/10.1029/2010WR010018
De Serres, B., Roy, A. G., Biron, M. P., & Best, J. L. (1999). Three-dimensional structure of flow at a confluence of river channels with discordant beds. Geomorphology, 26(4), 313–335. https://doi.org/10.1016/S0169-555X(98)00064-6
Duguay, J., Biron, P., & Buffin-Bélanger, T. (2022). Large-scale turbulent mixing at a mesoscale confluence assessed through drone imagery and eddy-resolved modelling. Earth Surface Processes and Landforms, 47(1), 345–363. https://doi.org/10.1002/esp.5251
Egozi, R., & Ashmore, P. E. (2009). Experimental analysis of braided channel pattern response to increased discharge. Journal of Geophysical Research: Earth Surface, 114, F02012. https://doi.org/10.1029/2008JF001099
Engel, F. L., & Rhoads, B. L. (2017). Velocity profiles and the structure of turbulence at the outer bank of a compound meander bend. Geomorphology, 295, 191–201. https://doi.org/10.1016/j.geomorph.2017.06.018
Ettema, R., & Armstrong, D. L. (2019). Bedload and channel morphology along a braided, sand-bed channel: Insights from a large flume. Journal of Hydraulic Research, 57(6), 822–835. https://doi.org/10.1080/00221686.2018.1555557
Federici, B., & Paola, C. (2003). Dynamics of channel bifurcations in noncohesive sediments. Water Resources Research, 39(6), 1162. https://doi.org/10.1029/2002WR001434
Hackney, C. R., Darby, S. E., Parsons, D. R., Leyland, J., Aalto, R., Nicholas, A. P., & Best, J. L. (2018). The influence of flow discharge variations on the morphodynamics of a diffluence-confluence unit on a large river. Earth Surface Processes and Landforms, 43(2), 349–362. https://doi.org/10.1002/esp.4204
Hicks, D. M., Baynes, E. R. C., Measures, R., Stecca, G., Tunnicliffe, J., & Fredrich, H. (2021). Morphodynamic research challenges for braided river environments: Lessons from the iconic case of New Zealand. Earth Surface Processes and Landforms, 46(1), 188–204. https://doi.org/10.1002/esp.5014
Hirt, C. W., & Nichols, B. D. (1981). Volume of fluid (VOF) method for the dynamics of free boundaries. Journal of Computational Physics, 39(1), 201–225. https://doi.org/10.1016/0021-9991(81)90145-5
Hua, Z. L., Gu, L., & Chu, K. J. (2009). Experiments of three-dimensional flow structure in braided rivers. Journal of Hydrodynamics, 21(2), 228–237. https://doi.org/10.1016/S1001-6058(08)60140-7
Hundey, E. J., & Ashmore, P. E. (2009). Length scale of braided river morphology. Water Resources Research, 45(8), W08409. https://doi.org/10.1029/2008WR007521
Iwantoro, A. P., van der Vegt, M., & Kleinhans, M. G. (2022). Stability and asymmetry of tide-influenced river bifurcations. Journal of Geophysical Research: Earth Surface, 127(6), e2021JF006282. https://doi.org/10.1029/2021JF006282
Jang, C. L., & Shimizu, Y. (2005). Numerical simulation of relatively wide, shallow channels with erodible banks. Journal of Hydraulic Engineering, 131, 565–575.
Kelly, S. (2006). Scaling and hierarchy in braided rivers and their deposits: Examples and implications for reservoir modelling. In G. H. Smith, J. L. Best, C. S. Bristow, & G. E. Petts (Eds.), Braided rivers: Process, deposits, ecology and management (pp. 75–106).
Lane, S. N., Bradbrook, K. F., Richards, K. S., Biron, P. M., & Roy, A. G. (2000). Secondary circulation cells in river channel confluences: Measurement artefacts or coherent flow structures? Hydrological Processes, 14(11-12), 2047–2071. https://doi.org/10.1002/1099-1085(20000815/30)14:11/12<2047::AID-HYP54>3.0.CO;2-4
Le, T. B., Khosronejad, A., Sotiropoulos, F., Bartelt, N., Woldeamlak, S., & Dewall, P. (2019). Large-eddy simulation of the Mississippi River under base-flow condition: Hydrodynamics of a natural diffluence-confluence region. Journal of Hydraulic Research, 57(6), 836–851. https://doi.org/10.1080/00221686.2018.1534282
Liu, C. B., Li, J., Bu, W. Y., Ma, W. X., Shen, G., & Yuan, Z. (2018). Large eddy simulation for improvement of performance estimation and turbulent flow analysis in a hydrodynamic torque converter. Engineering Applications of Computational Fluid Mechanics, 12(1), 635–651. https://doi.org/10.1080/19942060.2018.1489896
Liu, Z. (2012). Investigation of flow characteristics around square cylinder with inflow turbulence. Engineering Applications of Computational Fluid Mechanics, 6(3), 426–446. https://doi.org/10.1080/19942060.2012.11015433
Lu, G. W., Liu, J. X., Cao, Z. X., Li, Y. W., Lei, X. T., & Li, Y. (2023). A computational study of 3D flow structure in two consecutive bends subject to the influence of tributary inflow in the middle Yangtze River. Engineering Applications of Computational Fluid Mechanics, 17(1), 2183901. https://doi.org/10.1080/19942060.2023.2183901
Lu, H. Y., Li, Z. W., Hu, X. Y., Chen, B., & You, Y. C. (2022). Morphodynamic processes in a large gravel–bed braided channel in response to runof change: A case study in the Source Region of Yangtze River. Arabian Journal of Geosciences, 15(5), 377. https://doi.org/10.1007/s12517-022-09641-y
Luz, L. D., Szupiany, R. N., Parolin, M., Silva, A., & Stevaux, J. C. (2020). Obtuse-angle vs. confluent sharp meander bends: Insights from the Paraguay-Cuiabá confluence in the tropical Pantanal wetlands, Brazil. Geomorphology, 348, 106907. https://doi.org/10.1016/j.geomorph.2019.106907
Marra, W. A., Parsons, D. R., Kleinhans, M. G., Keevil, G. M., & Thomas, R. E. (2014). Near-bed and surface flow division patterns in experimental river bifurcations. Water Resources Research, 50(2), 1506–1530. https://doi.org/10.1002/2013WR014215
McLelland, S. J., Ashworth, P. J., Best, J. L., Roden, J., & Klaassen, G. J. (1999). Flow structure and transport of sand-grade suspended sediment around an evolving braid bar, Jamuna River, Bangladesh. Fluvial Sedimentology VI, 28, 43–57. https://doi.org/10.1002/9781444304213.ch4
Miori, S., Hardy, R. J., & Lane, S. N. (2012). Topographic forcing of flow partition and flow structures at river bifurcations. Earth Surface Processes and Landforms, 37(6), 666–679. https://doi.org/10.1002/esp.3204
Mosley, M. P. (1976). An experimental study of channel confluences. The Journal of Geology, 84(5), 535–562. https://doi.org/10.1086/628230
Orfeo, O., Parsons, D. R., Best, J. L., Lane, S. N., Hardy, R. J., Kostaschuk, R., Szupiany, R. N., & Amsler, M. L. (2006). Morphology and flow structures in a large confluence-diffluence: Rio Parana, Argentina. In R. M. L. Ferreira, C. T. L. Alves, G. A. B. Leal, & A. H. Cardoso (Eds.), River Flow 2006 (pp. 1277–1282).
Parsons, D. R., Best, J. L., Lane, S. N., Orfeo, O., Hardy, R. J., & Kostaschuk, R. (2007). Form roughness and the absence of secondary flow in a large confluence–diffluence, Rio Paraná, Argentina. Earth Surface Processes and Landforms, 32(1), 155–162. https://doi.org/10.1002/esp.1457
Ragno, N., Redolfi, M., & Tubino, M. (2021). Coupled morphodynamics of river bifurcations and confluences. Water Resources Research, 57(1), e2020WR028515. https://doi.org/10.1029/2020WR028515
Redolfi, M., Tubino, M., Bertoldi, W., & Brasington, J. (2016). Analysis of reach-scale elevation distribution in braided rivers: Definition of a new morphologic indicator and estimation of mean quantities. Water Resources Research, 52(8), 5951–5970. https://doi.org/10.1002/2015WR017918
Redolfi, M., Zolezzi, G., & Tubino, M. (2019). Free and forced morphodynamics of river bifurcations. Earth Surface Processes and Landforms, 44(4), 973–987. https://doi.org/10.1002/esp.4561
Rhoads, B. L., & Kenworthy, S. T. (1995). Flow structure at an asymmetrical stream confluence. Geomorphology, 11(4), 273–293. https://doi.org/10.1016/0169-555X(94)00069-4
Rhoads, B. L., & Sukhodolov, A. N. (2001). Field investigation of three-dimensional flow structure at stream confluences: 1. Thermal mixing and time-averaged velocities. Water Resources Research, 37(9), 2393–2410. https://doi.org/10.1029/2001WR000316
Roy, A. G., & Bergeron, N. (1990). Flow and particle paths at a natural river confluence with coarse bed material. Geomorphology, 3(2), 99–112. https://doi.org/10.1016/0169-555X(90)90039-S
Roy, A. G., Roy, R., & Bergeron, N. (1988). Hydraulic geometry and changes in flow velocity at a river confluence with coarse bed material. Earth Surface Processes and Landforms, 13(7), 583–598. https://doi.org/10.1002/esp.3290130704
Sambrook Smith, G. H., Ashworth, P. J., Best, J. L., Woodward, J., & Simpson, C. J. (2005). The morphology and facies of sandy braided rivers: Some considerations of scale invariance. In M. D. Blum, S. B. Marriott, & S. F. Leclair (Eds.), Fluvial sedimentology VII. International association of sedimentologists. Special Publication No. 35 (pp. 145–158). Blackwell.
Sharifipour, M., Bonakdari, H., Zaji, A. H., & Shamshirband, S. (2015). Numerical investigation of flow field and flowmeter accuracy in open-channel junctions. Engineering Applications of Computational Fluid Mechanics, 9(1), 280–290. https://doi.org/10.1080/19942060.2015.1008963
Shukry, A. (1950). Flow around bends in an open flume. Transactions of the American Society of Civil Engineers, 115(1), 751–778. https://doi.org/10.1061/TACEAT.0006426
Smagorinsky, J. (1963). General circulation experiments with the primitive equations. Monthly Weather Review, 91(3), 99–164. https://doi.org/10.1175/1520-0493(1963)091<0099:GCEWTP>2.3.CO;2
Sukhodolov, A. N., Krick, J., Sukhodolova, T. A., Cheng, Z. Y., Rhoads, B. L., & Constantinescu, G. S. (2017). Turbulent flow structure at a discordant river confluence: Asymmetric jet dynamics with implications for channel morphology. Journal of Geophysical Research: Earth Surface, 122(6), 1278–1293. https://doi.org/10.1002/2016JF004126
Sukhodolov, A. N., & Sukhodolova, T. A. (2019). Dynamics of flow at concordant gravel bed river confluences: Effects of junction angle and momentum flux ratio. Journal of Geophysical Research: Earth Surface, 124(2), 588–615. https://doi.org/10.1029/2018JF004648
Szupiany, R. N., Amsler, M. L., Hernandez, J., Parsons, D. R., Best, J. L., Fornari, E., & Trento, A. (2012). Flow fields, bed shear stresses, and suspended bed sediment dynamics in bifurcations of a large river. Water Resources Research, 48(11), W11515. https://doi.org/10.1029/2011WR011677.
Thomas, R. E., Parsons, D. R., Sandbach, S. D., Keevil, G. M., Marra, W. A., Hardy, R. J., Best, J. L., Lane, S. N., & Ross, J. A. (2011). An experimental study of discharge partitioning and flow structure at symmetrical bifurcations. Earth Surface Processes and Landforms, 36(15), 2069–2082. https://doi.org/10.1002/esp.2231
Thorne, C. R., Zevenbergen, L. W., Pitlick, J. C., Rais, S., Bradley, J. B., & Julien, P. Y. (1985). Direct measurements of secondary currents in a meandering sand-bed river. Nature, 315, 746–747. https://doi.org/10.1038/315746a0.
van der Mark, C. F., & Mosselman, E. (2013). Effects of helical flow in one-dimensional modelling of sediment distribution at river bifurcations. Earth Surface Processes and Landforms, 38(5), 502–511. https://doi.org/10.1002/esp.3335
Wang, X. G., Yan, Z. M., & Guo, W. D. (2007). Three-dimensional simulation for effects of bed discordance on flow dynamics at Y-shaped open channel confluences. Journal of Hydrodynamics, 19(5), 587–593. https://doi.org/10.1016/S1001-6058(07)60157-7
Weber, L. J., Schumate, E. D., & Mawer, N. (2001). Experiments on flow at a 90° open-channel junction. Journal of Hydraulic Engineering, 127(5), 340–350. https://doi.org/10.1061/(ASCE)0733-9429(2001)127:5(340)
Xie, Q. C., Yang, J., & Lundström, T. S. (2020). Flow and sediment behaviours and morpho-dynamics of a diffluence−Confluence unit. River Research and Applications, 36(8), 1515–1528. https://doi.org/10.1002/rra.3697
Xu, L., Yuan, S. Y., Tang, H. W., Qiu, J. J., Whittaker, C., & Gualtieri, C. (2022). Mixing dynamics at the large confluence between the Yangtze River and Poyang Lake. Water Resources Research, 58(11), e2022WR032195. https://doi.org/10.1029/2022WR032195
Yuan, S. Y., Xu, L., Tang, H. W., Xiao, Y., & Gualtieri, C. (2022). The dynamics of river confluences and their effects on the ecology of aquatic environment: A review. Journal of Hydrodynamics, 34(1), 1–14. https://doi.org/10.1007/s42241-022-0001-z
Yuan, S. Y., Yan, G. H., Tang, H. W., Xiao, Y., Rahimi, H., Aye, M. N., & Gualtieri, C. (2023). Effects of tributary floodplain on confluence hydrodynamics. Journal of Hydraulic Research, 61(4), 552–572. https://doi.org/10.1080/00221686.2023.2231413
자유 표면 흐름은 가정과 사무실 환경 모두에서 사용되는 소비자 제품의 설계 및 제조에서 일반적입니다. 예를 들어, 병 채우기는 매일 대규모로 이루어지는 프로세스입니다. 생산 속도를 극대화하면서 낭비를 최소화하도록 이러한 프로세스를 설계하면 시간이 지남에 따라 상당한 비용 절감으로 이어질 수 있습니다. FLOW-3D는 또한 스프레이 노즐을 설계하고 다공성 재료 및 기타 소비재 구성 요소의 흡수 기능을 모델링하는 데 사용할 수 있습니다. FLOW-3D 의 공기 유입, 다공성 매체 및 표면 장력을 포함한 고급 다중 물리 모델을 사용하면 소비자 제품 설계를 정확하게 시뮬레이션하고 최적화하는 것이 쉽습니다.
충전재
유입된 공기는 생산 라인에서 용기를 채울 때 액체의 부피를 늘릴 수 있습니다. 아래 왼쪽 이미지는 높이가 약 20cm인 병을 1.2초 동안 채우는 것을 보여줍니다. 색상 음영은 액체에 있는 공기의 부피 분율을 나타냅니다. 병에서 혼합 시간이 짧고 혼합 정도가 높기 때문에 공기가 표면으로 올라가 빠져나갈 시간이 없었습니다. 그러나 오른쪽 이미지에서 볼 수 있듯이 약 1.7초의 추가 시간이 지나면 공기가 표면으로 올라가면서 발생하는 액체 부피 감소가 명확하게 보입니다. FLOW-3D 의 드리프트 플럭스 모델을 사용하면 액체에 있는 기포와 같은 구성 요소를 분리하여 분리할 수 있습니다.
이 기사에서는 FLOW-3D를 사용하여 새로운 타이드 병 디자인의 충전을 모델링하는 방법을 설명하며, Procter and Gamble Company의 기술 섹션 책임자인 John McKibben이 기고했습니다 .
지금 오전 9시인데 긴급 이메일을 받았다고 상상해보세요.
방금 새로운 Tide® 병 디자인 중 하나가 손잡이에 채워지고 충전 장비에 문제가 생길 수 있다는 것을 깨달았습니다. 우리는 프로토타입 병이 없으며 몇 주 동안 없을 것입니다. 디자이너와 소비자는 디자인의 모습을 좋아하지만, 채우는 방식이 생산 시설에 쇼스토퍼가 될 수 있습니다.
이런 상황이 제게 주어졌을 때, 저는 3D 지오메트리(그림 1)의 스테레오 리소그래피(.stl) 파일을 요청하여 응답을 시작했고, 제가 무엇을 할 수 있는지 알아보고자 했습니다. 저는 FLOW-3D가 .stl 파일을 사용하여 지오메트리를 입력하고 충전을 위한 자유 표면 문제를 해결할 수 있을 것이라는 것을 알고 있었습니다. 저는 이것이 잠재적인 문제에 대한 좋은 정성적 이해를 제공할 것으로 기대했지만, 이 애플리케이션에 얼마나 정확할지에 대해 약간 불확실했습니다.
시뮬레이션 설정 및 실행
오후 1시경에 저는 지오메트리 파일, 유량, 유체 특성을 받았습니다. 몇 시간 이내에 시뮬레이션이 실행되어 예비 결과가 나왔습니다. 저는 제 고객을 초대하여 결과를 잠깐 살펴보게 했고 그는 “사장의 상사”를 데려와서 살펴보게 했습니다. 그래서 저녁 5시경에 예비 결과를 살펴보고 원래 우려했던 것이 문제가 아니라는 것을 확인했습니다.
하지만 결과는 몇 가지 다른 의문을 제기했습니다. 손잡이에 채우면 유입 유체 제트가 많이 깨졌습니다. 이렇게 하면 유입 공기와 거품의 양이 늘어날 것이라는 걸 알았습니다(결국 세탁 세제를 채우고 있으니까요). FLOW-3D 공기 유입 모델을 테스트하기로 했습니다. 이 모델은 원래 난류 제트용으로 개발되었고, 이 층류 문제를 살펴보면 얼마나 잘 수행될지 확신할 수 없었습니다.
그림 2: 채워진 결과그림 3: 실험 비교
그림 2는 공기 유입 모델이 있는 경우와 없는 경우 병 충전 모델의 결과를 보여줍니다. 유입 공기가 포함되면 충전 레벨이 상당히 증가한다는 점에 유의하십시오. 유입 공기가 병 상단에서 유체를 강제로 밀어내지는 않지만 공기 유입 정확도를 확인해야 할 만큼 충분히 가깝습니다. 그림 3은 공기 유입 레벨을 몇 주 후에 실행한 실험 이미지와 비교합니다(시제품 병이 출시된 후). 제트 분리 및 충전 레벨의 질적 일치는 우수하며 시뮬레이션이 병 설계를 선별하기에 충분히 정확하다는 것을 확인했습니다.
홍조
변기가 어떻게 작동하는지 궁금한 적이 있나요? 사실 꽤 복잡합니다. 손잡이를 밀면 물이 변기 그릇을 채우기 시작합니다. 변기 그릇의 유체 수위가 트랩 상단(변기 그릇 뒤) 위로 올라가면 웨어 유형의 흐름이 시작됩니다. 흐름이 충분히 빠르면 변기 그릇에 거품이 형성되어 사이펀이 생성됩니다. 그 지점에서 사이펀이 변기 그릇에서 물을 끌어내고 변기가 물을 흘립니다. 많은 지역에서 물 절약은 중요한 문제이며, 저유량 변기는 가정과 상업용 모두에 필요합니다. 하지만 변기가 첫 번째 시도에서 제 역할을 하지 못하면 물 절약 목표는 달성되지 않습니다. FLOW-3D를 사용하면 다양한 설계를 모델링하여 최적의 결과를 얻을 수 있습니다.
식품 가공
식품 가공 산업은 복잡한 유체, 일반적으로 비뉴턴 유체, 슬러리, 고체와 유체의 혼합물을 관리하여 분배 장비를 최적으로 설계하고 제조하기 위한 다양한 요구 사항이 있습니다. 이는 상업용 장비의 일관성과 내구성 및 품질에 필수적입니다. 또한 포장 디자인의 혁신을 통해 한 제품을 다른 제품과 명확히 구별할 수 있습니다. 예를 들어, 꿀, 케첩 또는 크리머를 깨끗하고 정확하게 분배하는 것은 소비자가 매장에서 내리는 선택일 수 있습니다. 운송 및 보관 요구 사항에는 더 나은 모양 엔지니어링과 더 많은 용기 재료 선택이 필요합니다. 1.5리터 물병이나 세탁 세제를 움직이거나 떨어뜨리는 동안의 유체 하중은 상류 설계의 중요한 부분이 될 수 있습니다.
꿀, 옥수수 시럽, 치약과 같은 점성 유체는 일반적으로 고체 표면에 닿으면 코일을 형성하는 경향이 있습니다. 이 효과는 관찰하기에 흥미롭고 재미있지만, 공기가 제품에 끌려들어 포장이 어려워질 수 있는 포장 공정에서는 환영받지 못할 수 있습니다. 코일링이 발생하는 조건은 유체의 점도, 유체가 떨어지는 거리, 유체의 속도에 따라 달라집니다. FLOW-3D는 다양한 물리적 공정 매개변수를 연구하여 효율적인 공정을 설계하는 데 도움이 되는 정확한 도구를 제공합니다.
혼입
지난 수십 년 동안 컴퓨터화된 측정 및 시뮬레이션 기술의 발전으로 인해 혼합에 대한 이해가 크게 진전되었습니다. 유동 모델링 기술의 지속적인 발전 덕분에 혼합 장비의 유동 의존적 프로세스에 대한 자세한 통찰력을 CFD 소프트웨어를 사용하여 쉽게 시뮬레이션하고 이해할 수 있습니다. 오늘날 블렌딩에서 고체 현탁액, 재킷 반응기의 열 전달에서 발효에 이르기까지 광범위한 응용 분야가 FLOW-3D 의 혼합 기술을 사용하여 모델링됩니다. FLOW-3D 시뮬레이션은 임펠러의 모든 구성과 모든 용기 형상의 혼합 조건에서 블렌딩 시간, 순환 및 전력 수와 같은 주요 혼합 매개변수를 평가하는 데 도움이 될 수 있습니다. 이러한 시뮬레이션은 실험적 방법을 사용하여 보완합니다. 이러한 장비의 유동 의존적 프로세스를 예측하고 이해하기 위해 CFD 소프트웨어를 사용하면 제품 품질을 향상시키고 많은 제품의 비용과 출시 시간을 모두 줄일 수 있습니다.
비뉴턴 유체
혈액, 케첩, 치약, 샴푸, 페인트, 로션과 같은 비뉴턴 유체는 다양한 점도를 가진 복잡한 유동학을 가지고 있습니다. FLOW-3D 는 변형 및/또는 온도에 따라 달라지는 비뉴턴 점도를 가진 이러한 유체를 모델링합니다. 전단 및 온도에 따른 점도는 Carreau, 거듭제곱 법칙 함수 또는 단순히 표 형식의 입력을 통해 설명됩니다. 일부 폴리머, 세라믹 및 반고체 금속의 특징인 시간 종속 또는 틱소트로피 거동도 시뮬레이션할 수 있습니다.
핸드 로션 펌프는 종종 여러 가지 설계 문제와 관련이 있습니다. 펌프가 공기 공극을 가두지 않고 효과적으로 작동하고 로션의 연속적인 흐름을 생성하는 것이 중요합니다. 좋은 설계는 노력이 덜 필요하고 이상적으로는 로션을 원하는 곳으로 향하게 합니다. FLOW-3D 의 이동 객체 모델은 노즐이 아래로 눌리는 것을 시뮬레이션하여 저장소의 로션을 가압하는 데 사용됩니다. 로션의 압력과 로션을 추출하는 데 필요한 힘을 연구할 수 있습니다. 여러 설계 변수는 동일한 고정 구조 메시 내에서 쉽게 분석할 수 있습니다.
다공성 재료
다공성 매체에서 유체의 이동에 대한 수치 모델링은 어려울 수 있지만 FLOW-3D 에는 다공성 재료와 관련된 문제를 해결하는 데 유용한 기능이 많이 포함되어 있습니다. FAVOR™ 기술에는 사용자가 연속적인 다공성 매체를 표현할 수 있도록 하는 데 필요한 다공성 변수가 포함되어 있습니다. FLOW-3D를 사용하면 사용자가 포화 및 불포화 흐름 조건을 모두 시뮬레이션할 수 있습니다. 거듭제곱 법칙 관계를 사용하면 불포화 흐름 조건에서 모세관 압력 과 포화 사이의 비선형 관계를 모델링 할 수 있습니다. 별도의 충전 및 배수 곡선을 사용하여 히스테리시스 현상을 모델링할 수 있습니다. 서로 직접 접촉하는 경우에도 서로 다른 다공성, 투과성 및 습윤성 속성을 서로 다른 장애물에 할당할 수 있습니다. 투과성은 흐름 방향에 따라 지정할 수 있으므로 사용자가 다공성 매체의 이방성 동작을 모델링할 수 있습니다. 유체와 다공성 매체 간의 열 전달을 고려할 수 있습니다.
분무
소용돌이 분무 노즐은 화학 세정제, 의약품 및 연료에서 액체를 분사하는 일반적인 방법입니다. 액체를 성공적으로 분무하려면 일반적으로 노즐로 침투하는 공기 코어를 형성해야 합니다. CFD는 최적의 분무 콘에 대한 기하학, 소용돌이 속도 및 유체 특성의 영향을 탐색하는 효과적인 방법입니다.
이 예에서 2차원 축대칭 소용돌이 흐름이 시뮬레이션되었습니다. 대칭 축을 따라 공기 코어가 노즐의 전체 길이를 거의 관통했습니다. 왼쪽 플롯은 평면에서 속도 분포를 나타내는 벡터가 있는 압력 분포입니다. 오른쪽 플롯은 속도의 소용돌이 구성 요소로 채색되어 있으며 빨간색은 더 높은 값을 나타냅니다.
분무 콘의 규모와 입자 크기가 너무 광범위하기 때문에 분무의 완전한 분무를 직접 계산하는 것은 불가능합니다. 또한 분무는 외부 교란, 노즐의 미세한 결함 및 기타 영향과 밀접하게 관련된 혼란스러운 프로세스입니다. 그러나 노즐을 떠날 때 분무 콘의 특성(예: 벽 두께, 콘 각도, 축 및 방위 속도)을 예측할 수 있다면 이러한 유형의 흐름 장치를 최적화하는 데 큰 도움이 됩니다.
소용돌이 분무 노즐의 FLOW-3D 시뮬레이션
Products
자유 표면 흐름은 가정과 사무실 환경 모두에서 사용되는 소비자 제품의 설계 및 제조에서 일반적입니다.
예를 들어, 병 채우기는 매일 대규모로 진행되는 프로세스입니다. 생산 속도를 최대화하면서 낭비를 최소화하도록 이러한 프로세스를 설계하면 시간이 지남에 따라 상당한 비용 절감으로 이어질 수 있습니다. FLOW-3D는 또한 스프레이 노즐을 설계하고 다공성 재료 및 기타 소비재 구성 요소의 흡수 기능을 모델링하는 데 사용할 수 있습니다.
공기 혼입, 다공성 매질 및 표면 장력을 포함한 FLOW-3D의 고급 다중 물리 모델을 사용하면 소비자 제품 설계를 정확하게 시뮬레이션하고 최적화 할 수 있습니다.
September 2023 DOI:10.30955/gnc2023.00436 Conference: 18th International Conference on Environmental Science and Technology CEST2023, 30 August to 2 September 2023, Athens, ...
Numerical Study of the Local Scouring Process and Influencing Factors of Semi-Exposed Submarine Cables by Qishun Li,Yanpeng Hao *,Peng Zhang,Haotian Tan,Wanxing Tian,Linhao ...
Analyses of Cryogenic Propellant Tank Pressurization based upon Experiments and Numerical SimulationsCarina Ludwig? and Michael Dreyer***DLR - German Aerospace Center, ...
Coupled thermodynamic-fluid-dynamic solution for a liquid-hydrogen tank G. D. Grayson Published Online:23 May 2012 https://doi.org/10.2514/3.26706 Read Now Tools Share Introduction ...
자유 표면 흐름은 가정과 사무실 환경 모두에서 사용되는 소비자 제품의 설계 및 제조에서 일반적입니다. 예를 들어, 병 채우기는 매일 대규모로 이루어지는 프로세스입니다. 생산 속도를 극대화하면서 낭비를 최소화하도록 이러한 프로세스를 설계하면 시간이 지남에 따라 상당한 비용 절감으로 이어질 수 있습니다. FLOW-3D는 또한 스프레이 노즐을 설계하고 다공성 재료 및 기타 소비재 구성 요소의 흡수 기능을 모델링하는 데 사용할 수 있습니다. FLOW-3D 의 공기 유입, 다공성 매체 및 표면 장력을 포함한 고급 다중 물리 모델을 사용하면 소비자 제품 설계를 정확하게 시뮬레이션하고 최적화하는 것이 쉽습니다.
충전재
유입된 공기는 생산 라인에서 용기를 채울 때 액체의 부피를 늘릴 수 있습니다. 아래 왼쪽 이미지는 높이가 약 20cm인 병을 1.2초 동안 채우는 것을 보여줍니다. 색상 음영은 액체에 있는 공기의 부피 분율을 나타냅니다. 병에서 혼합 시간이 짧고 혼합 정도가 높기 때문에 공기가 표면으로 올라가 빠져나갈 시간이 없었습니다. 그러나 오른쪽 이미지에서 볼 수 있듯이 약 1.7초의 추가 시간이 지나면 공기가 표면으로 올라가면서 발생하는 액체 부피 감소가 명확하게 보입니다. FLOW-3D 의 드리프트 플럭스 모델을 사용하면 액체에 있는 기포와 같은 구성 요소를 분리하여 분리할 수 있습니다.
이 기사에서는 FLOW-3D를 사용하여 새로운 타이드 병 디자인의 충전을 모델링하는 방법을 설명하며, Procter and Gamble Company의 기술 섹션 책임자인 John McKibben이 기고했습니다 .
지금 오전 9시인데 긴급 이메일을 받았다고 상상해보세요.
방금 새로운 Tide® 병 디자인 중 하나가 손잡이에 채워지고 충전 장비에 문제가 생길 수 있다는 것을 깨달았습니다. 우리는 프로토타입 병이 없으며 몇 주 동안 없을 것입니다. 디자이너와 소비자는 디자인의 모습을 좋아하지만, 채우는 방식이 생산 시설에 쇼스토퍼가 될 수 있습니다.
이런 상황이 제게 주어졌을 때, 저는 3D 지오메트리(그림 1)의 스테레오 리소그래피(.stl) 파일을 요청하여 응답을 시작했고, 제가 무엇을 할 수 있는지 알아보고자 했습니다. 저는 FLOW-3D가 .stl 파일을 사용하여 지오메트리를 입력하고 충전을 위한 자유 표면 문제를 해결할 수 있을 것이라는 것을 알고 있었습니다. 저는 이것이 잠재적인 문제에 대한 좋은 정성적 이해를 제공할 것으로 기대했지만, 이 애플리케이션에 얼마나 정확할지에 대해 약간 불확실했습니다.
시뮬레이션 설정 및 실행
오후 1시경에 저는 지오메트리 파일, 유량, 유체 특성을 받았습니다. 몇 시간 이내에 시뮬레이션이 실행되어 예비 결과가 나왔습니다. 저는 제 고객을 초대하여 결과를 잠깐 살펴보게 했고 그는 “사장의 상사”를 데려와서 살펴보게 했습니다. 그래서 저녁 5시경에 예비 결과를 살펴보고 원래 우려했던 것이 문제가 아니라는 것을 확인했습니다.
하지만 결과는 몇 가지 다른 의문을 제기했습니다. 손잡이에 채우면 유입 유체 제트가 많이 깨졌습니다. 이렇게 하면 유입 공기와 거품의 양이 늘어날 것이라는 걸 알았습니다(결국 세탁 세제를 채우고 있으니까요). FLOW-3D 공기 유입 모델을 테스트하기로 했습니다. 이 모델은 원래 난류 제트용으로 개발되었고, 이 층류 문제를 살펴보면 얼마나 잘 수행될지 확신할 수 없었습니다.
그림 2: 채워진 결과그림 3: 실험 비교
그림 2는 공기 유입 모델이 있는 경우와 없는 경우 병 충전 모델의 결과를 보여줍니다. 유입 공기가 포함되면 충전 레벨이 상당히 증가한다는 점에 유의하십시오. 유입 공기가 병 상단에서 유체를 강제로 밀어내지는 않지만 공기 유입 정확도를 확인해야 할 만큼 충분히 가깝습니다. 그림 3은 공기 유입 레벨을 몇 주 후에 실행한 실험 이미지와 비교합니다(시제품 병이 출시된 후). 제트 분리 및 충전 레벨의 질적 일치는 우수하며 시뮬레이션이 병 설계를 선별하기에 충분히 정확하다는 것을 확인했습니다.
홍조
변기가 어떻게 작동하는지 궁금한 적이 있나요? 사실 꽤 복잡합니다. 손잡이를 밀면 물이 변기 그릇을 채우기 시작합니다. 변기 그릇의 유체 수위가 트랩 상단(변기 그릇 뒤) 위로 올라가면 웨어 유형의 흐름이 시작됩니다. 흐름이 충분히 빠르면 변기 그릇에 거품이 형성되어 사이펀이 생성됩니다. 그 지점에서 사이펀이 변기 그릇에서 물을 끌어내고 변기가 물을 흘립니다. 많은 지역에서 물 절약은 중요한 문제이며, 저유량 변기는 가정과 상업용 모두에 필요합니다. 하지만 변기가 첫 번째 시도에서 제 역할을 하지 못하면 물 절약 목표는 달성되지 않습니다. FLOW-3D를 사용하면 다양한 설계를 모델링하여 최적의 결과를 얻을 수 있습니다.
식품 가공
식품 가공 산업은 복잡한 유체, 일반적으로 비뉴턴 유체, 슬러리, 고체와 유체의 혼합물을 관리하여 분배 장비를 최적으로 설계하고 제조하기 위한 다양한 요구 사항이 있습니다. 이는 상업용 장비의 일관성과 내구성 및 품질에 필수적입니다. 또한 포장 디자인의 혁신을 통해 한 제품을 다른 제품과 명확히 구별할 수 있습니다. 예를 들어, 꿀, 케첩 또는 크리머를 깨끗하고 정확하게 분배하는 것은 소비자가 매장에서 내리는 선택일 수 있습니다. 운송 및 보관 요구 사항에는 더 나은 모양 엔지니어링과 더 많은 용기 재료 선택이 필요합니다. 1.5리터 물병이나 세탁 세제를 움직이거나 떨어뜨리는 동안의 유체 하중은 상류 설계의 중요한 부분이 될 수 있습니다.
꿀, 옥수수 시럽, 치약과 같은 점성 유체는 일반적으로 고체 표면에 닿으면 코일을 형성하는 경향이 있습니다. 이 효과는 관찰하기에 흥미롭고 재미있지만, 공기가 제품에 끌려들어 포장이 어려워질 수 있는 포장 공정에서는 환영받지 못할 수 있습니다. 코일링이 발생하는 조건은 유체의 점도, 유체가 떨어지는 거리, 유체의 속도에 따라 달라집니다. FLOW-3D는 다양한 물리적 공정 매개변수를 연구하여 효율적인 공정을 설계하는 데 도움이 되는 정확한 도구를 제공합니다.
혼입
지난 수십 년 동안 컴퓨터화된 측정 및 시뮬레이션 기술의 발전으로 인해 혼합에 대한 이해가 크게 진전되었습니다. 유동 모델링 기술의 지속적인 발전 덕분에 혼합 장비의 유동 의존적 프로세스에 대한 자세한 통찰력을 CFD 소프트웨어를 사용하여 쉽게 시뮬레이션하고 이해할 수 있습니다. 오늘날 블렌딩에서 고체 현탁액, 재킷 반응기의 열 전달에서 발효에 이르기까지 광범위한 응용 분야가 FLOW-3D 의 혼합 기술을 사용하여 모델링됩니다. FLOW-3D 시뮬레이션은 임펠러의 모든 구성과 모든 용기 형상의 혼합 조건에서 블렌딩 시간, 순환 및 전력 수와 같은 주요 혼합 매개변수를 평가하는 데 도움이 될 수 있습니다. 이러한 시뮬레이션은 실험적 방법을 사용하여 보완합니다. 이러한 장비의 유동 의존적 프로세스를 예측하고 이해하기 위해 CFD 소프트웨어를 사용하면 제품 품질을 향상시키고 많은 제품의 비용과 출시 시간을 모두 줄일 수 있습니다.
비뉴턴 유체
혈액, 케첩, 치약, 샴푸, 페인트, 로션과 같은 비뉴턴 유체는 다양한 점도를 가진 복잡한 유동학을 가지고 있습니다. FLOW-3D 는 변형 및/또는 온도에 따라 달라지는 비뉴턴 점도를 가진 이러한 유체를 모델링합니다. 전단 및 온도에 따른 점도는 Carreau, 거듭제곱 법칙 함수 또는 단순히 표 형식의 입력을 통해 설명됩니다. 일부 폴리머, 세라믹 및 반고체 금속의 특징인 시간 종속 또는 틱소트로피 거동도 시뮬레이션할 수 있습니다.
핸드 로션 펌프는 종종 여러 가지 설계 문제와 관련이 있습니다. 펌프가 공기 공극을 가두지 않고 효과적으로 작동하고 로션의 연속적인 흐름을 생성하는 것이 중요합니다. 좋은 설계는 노력이 덜 필요하고 이상적으로는 로션을 원하는 곳으로 향하게 합니다. FLOW-3D 의 이동 객체 모델은 노즐이 아래로 눌리는 것을 시뮬레이션하여 저장소의 로션을 가압하는 데 사용됩니다. 로션의 압력과 로션을 추출하는 데 필요한 힘을 연구할 수 있습니다. 여러 설계 변수는 동일한 고정 구조 메시 내에서 쉽게 분석할 수 있습니다.
다공성 재료
다공성 매체에서 유체의 이동에 대한 수치 모델링은 어려울 수 있지만 FLOW-3D 에는 다공성 재료와 관련된 문제를 해결하는 데 유용한 기능이 많이 포함되어 있습니다. FAVOR™ 기술에는 사용자가 연속적인 다공성 매체를 표현할 수 있도록 하는 데 필요한 다공성 변수가 포함되어 있습니다. FLOW-3D를 사용하면 사용자가 포화 및 불포화 흐름 조건을 모두 시뮬레이션할 수 있습니다. 거듭제곱 법칙 관계를 사용하면 불포화 흐름 조건에서 모세관 압력 과 포화 사이의 비선형 관계를 모델링 할 수 있습니다. 별도의 충전 및 배수 곡선을 사용하여 히스테리시스 현상을 모델링할 수 있습니다. 서로 직접 접촉하는 경우에도 서로 다른 다공성, 투과성 및 습윤성 속성을 서로 다른 장애물에 할당할 수 있습니다. 투과성은 흐름 방향에 따라 지정할 수 있으므로 사용자가 다공성 매체의 이방성 동작을 모델링할 수 있습니다. 유체와 다공성 매체 간의 열 전달을 고려할 수 있습니다.
분무
소용돌이 분무 노즐은 화학 세정제, 의약품 및 연료에서 액체를 분사하는 일반적인 방법입니다. 액체를 성공적으로 분무하려면 일반적으로 노즐로 침투하는 공기 코어를 형성해야 합니다. CFD는 최적의 분무 콘에 대한 기하학, 소용돌이 속도 및 유체 특성의 영향을 탐색하는 효과적인 방법입니다.
이 예에서 2차원 축대칭 소용돌이 흐름이 시뮬레이션되었습니다. 대칭 축을 따라 공기 코어가 노즐의 전체 길이를 거의 관통했습니다. 왼쪽 플롯은 평면에서 속도 분포를 나타내는 벡터가 있는 압력 분포입니다. 오른쪽 플롯은 속도의 소용돌이 구성 요소로 채색되어 있으며 빨간색은 더 높은 값을 나타냅니다.
분무 콘의 규모와 입자 크기가 너무 광범위하기 때문에 분무의 완전한 분무를 직접 계산하는 것은 불가능합니다. 또한 분무는 외부 교란, 노즐의 미세한 결함 및 기타 영향과 밀접하게 관련된 혼란스러운 프로세스입니다. 그러나 노즐을 떠날 때 분무 콘의 특성(예: 벽 두께, 콘 각도, 축 및 방위 속도)을 예측할 수 있다면 이러한 유형의 흐름 장치를 최적화하는 데 큰 도움이 됩니다.
소용돌이 분무 노즐의 FLOW-3D 시뮬레이션
Products
자유 표면 흐름은 가정과 사무실 환경 모두에서 사용되는 소비자 제품의 설계 및 제조에서 일반적입니다.
예를 들어, 병 채우기는 매일 대규모로 진행되는 프로세스입니다. 생산 속도를 최대화하면서 낭비를 최소화하도록 이러한 프로세스를 설계하면 시간이 지남에 따라 상당한 비용 절감으로 이어질 수 있습니다. FLOW-3D는 또한 스프레이 노즐을 설계하고 다공성 재료 및 기타 소비재 구성 요소의 흡수 기능을 모델링하는 데 사용할 수 있습니다.
공기 혼입, 다공성 매질 및 표면 장력을 포함한 FLOW-3D의 고급 다중 물리 모델을 사용하면 소비자 제품 설계를 정확하게 시뮬레이션하고 최적화 할 수 있습니다.
September 2023 DOI:10.30955/gnc2023.00436 Conference: 18th International Conference on Environmental Science and Technology CEST2023, 30 August to 2 September 2023, Athens, ...
Numerical Study of the Local Scouring Process and Influencing Factors of Semi-Exposed Submarine Cables by Qishun Li,Yanpeng Hao *,Peng Zhang,Haotian Tan,Wanxing Tian,Linhao ...
Analyses of Cryogenic Propellant Tank Pressurization based upon Experiments and Numerical SimulationsCarina Ludwig? and Michael Dreyer***DLR - German Aerospace Center, ...
Coupled thermodynamic-fluid-dynamic solution for a liquid-hydrogen tank G. D. Grayson Published Online:23 May 2012 https://doi.org/10.2514/3.26706 Read Now Tools Share Introduction ...
인텔 : 모델명이 ‘2’로 시작하고 ‘V’로 끝나는 코어 울트라 시리즈 2(Core Ultra Series 2). 예를 들면 인텔 코어 울트라 5 226V(시리즈2)가 있다.
AMD : 라이젠 AI 300 시리즈. 예시로 AMD 라이젠 AI 7 프로 360.
퀄컴 : 스냅드래곤 X 시리즈의 플러스(Plus) 또는 엘리트(Elite) 제품
이 세 가지 프로세서는 성능과 배터리 수명 면에서 애플 맥북의 M 시리즈와 경쟁하도록 설계됐다. 그러나 노트북을 선택할 때는 프로세서뿐 아니라 다양한 요소를 함께 고려해야 한다.
인텔 프로세서
인텔의 최신 프로세서는 다음 세 가지 범주로 나뉜다.
인텔 코어 울트라(Intel Core Ultra) : 프리미엄 칩으로, AI 전용 프로세서를 탑재했다(예 : 인텔 코어 울트라 7 155U).
인텔 코어(Intel Core) : 주류 노트북에 사용되는 칩으로, 코어 울트라보다 한 단계 아래다(예 : 인텔 코어 7 150U).
인텔 프로세서(Intel Processor) : 과거 펜티엄과 셀러론 브랜드를 대체하는 저가형 PC 칩이다(예 : 인텔 프로세서 N200).
인텔은 프로세서를 성능 등급에 따라 ‘3’, ‘5’, ‘7’, ‘9’로 세분화했다. 숫자가 높을수록 더 많은 코어를 가지고 있다는 의미이며, 이미지 처리 및 비디오 작업 속도가 향상된다. 코어 5와 코어 울트라 5 칩은 웹 브라우징 및 오피스 작업에 적합하다.
Intel
모델명 뒤에 붙는 접미사도 중요하다. 이 글자는 프로세서가 어떻게 최적화되었는지를 나타낸다. 긴 접미사 목록 중에 알아두어야 할 주요 단어는 ‘U’와 ‘H’다. U는 배터리 수명을, H는 성능을 강조한다. 코어 울트라 5 226V의 ‘V’는 코어 울트라 제품 라인에만 적용되는 접미사다.
구형 모델은 12세대 코어 i5 1235U처럼 이름에 ‘i’와 세대 번호가 포함되어 있다. 14세대에 이르러 인텔은 모든 것을 재설정하고 이제 ‘시리즈 1’부터 세기 시작했다(예 : 코어 울트라 155U). 즉, 최신 인텔 칩의 모델명은 구형 모델보다 짧다. 가격이 적당한 경우라면 구형 모델도 여전히 고려해 볼만하다.
AMD 프로세서
AMD는 인텔만큼 브랜딩 개편에 적극적이지는 않다. 애플 및 퀄컴과 경쟁하는 AI 300 시리즈 칩 외에 나머지 프로세서는 2023년 도입된 더 길고 혼란스러운 명명 체계를 따르고 있다.
AMD
예시로 AMD 라이젠 5 8640HS를 살펴본다.
첫 번째 숫자 ‘8’은 세대를 의미하며, 2024년에 출시된 칩을 나타낸다(7735HS는 2023년 제품).
‘5’는 성능 등급을 나타내며, 인텔과 마찬가지로 숫자가 높을수록 성능이 좋다는 의미다. 인텔 코어 5와 코어 7 체계와 유사하게 홀수로 계산된다.
마지막 글자는 프로세서의 최적화 방식이다. ‘U’는 배터리 수명, ‘H’는 성능을 우선시한다.
이 명명 체계를 따르는 칩은 AMD의 구형 젠 4(Zen 4) 아키텍처를 기반으로 하지만, 최신 AI 300 시리즈는 젠 5 아키텍처를 사용한다. AMD가 프로세서 라인 대부분을 최신 아키텍처로 전환함에 따라 이에 맞는 새로운 브랜드가 등장할 것으로 예상된다.
퀄컴 프로세서
퀄컴은 올해 초 전력 효율성에 중점을 두고 PC CPU 경쟁에 합류했다. 퀄컴의 스냅드래곤 X 칩은 휴대폰, 태블릿, 애플의 M 시리즈 프로세서에서 볼 수 있는 것과 동일한 Arm 기반 아키텍처를 사용하며, 우수한 PC 성능과 긴 배터리 수명을 제공한다. 무엇보다 퀄컴의 직관적인 브랜드 전략이 신선하게 다가온다.
스냅드래곤 X 엘리트(Snapdragon X Elite) : 최고급 모델
스냅드래곤 X 플러스(Snapdragon X Plus) : 그보다 한 단계 낮은 모델
마이크로소프트 서피스 노트북에 탑재된 스냅드래곤 X 플러스를 사용해 본 경험에 따르면, 충분한 성능과 하루 종일 지속되는 배터리 수명을 제공했다.
다만, Arm 기반 프로세서가 모든 윈도우 소프트웨어와 호환되는 것은 아니다. 스냅드래곤 PC에서 Arm이 아닌 앱을 실행하는 마이크로소프트의 에뮬레이션 엔진에서도 호환성 문제가 발생할 수 있다. 에뮬레이션 개선과 Arm 버전의 소프트웨어를 출시하는 개발자가 늘어나면서 상황이 점점 개선되고 있지만, 인텔과 AMD 노트북에서는 겪지 않아도 될 골칫거리가 여전히 남아 있다.
CPU 시장의 긍정적인 변화
복잡한 이름을 살펴보는 것이 혼란스러울 수 있고 AI에 대한 강조가 다소 과장된 면이 있지만, PC 프로세서 분야에서 3가지 업체가 경쟁하는 덕분에 상황은 개선되고 있다. 지난 4년간 애플은 전력 효율성 측면에서 독보적인 성과를 보여줬다. 그러나 인텔, AMD, 퀄컴이 새로운 프로세서를 내놓으며 애플의 수준에 도달하고 있다.
물론 복잡한 브랜드와 명명 체계는 단점이지만, 이런 경쟁 덕분에 더 나은 성능과 배터리 수명을 갖춘 제품이 등장하고 있다. 사용자에게 긍정적인 변화다. dl-itworldkorea@foundryco.com
아래 과거 자료도 선택에 큰 도움이 됩니다.
2023년 01월 11일
본 자료는 IT WORLD에서 인용한 자료입니다.
일반적으로 수치해석을 주 업무로 사용하는 경우 노트북을 사용하는 경우는 그리 많지 않습니다. 그 이유는 CPU 성능을 100%로 사용하는 해석 프로그램의 특성상 발열과 부품의 성능 측면에서 데스크탑이나 HPC의 성능을 따라 가기는 어렵기 때문입니다.
그럼에도 불구하고, 이동 편의성이나 발표, Demo 등의 업무 필요성이 자주 있는 경우, 또는 계산 시간이 짧은 경량 해석을 주로 하는 경우, 노트북이 주는 이점이 크기 때문에 수치해석용 노트북을 고려하기도 합니다.
보통 수치해석용 컴퓨터를 검토하는 경우 CPU의 Core수나 클럭, 메모리, 그래픽카드 등을 신중하게 검토하게 되는데 모든 것이 예산과 직결되어 있기 때문입니다. 따라서 해석용 컴퓨터 구매 시 어떤 것을 선정 우선순위에 두는지에 따라 사양이 달라지게 됩니다.
해석용으로 노트북을 고려하는 경우, 보통 CPU의 클럭은 비교적 선택 기준이 명확합니다. 메모리 또한 용량에 따라 가격이 정해지기 때문에 이것도 비교적 명확합니다. 나머지 가격에 가장 큰 영향을 주는 것이 그래픽카드인데, 이는 그래픽 카드의 경우 일반적인 게임용이나 포토샵으로 일반적인 이미지 처리 작업을 수행하는 그래픽카드와 3차원 CAD/CAE에 사용되는 업무용 그래픽 카드는 명확하게 분리되어 있고, 이는 가격 측면에서 매우 차이가 많이 납니다.
통상 게임용 그래픽카드는 수치해석의 경우 POST 작업시 문제가 발생하는 경우가 종종 발생하기 때문에 일반적으로 선택 우선 순위에서 충분한 확인을 한 후 구입하는 것이 좋습니다.
FLOW-3D는 OpenGL 드라이버가 만족스럽게 수행되는 최신 그래픽 카드가 적합합니다. 최소한 OpenGL 3.0을 지원하는 것이 좋습니다. FlowSight는 DirectX 11 이상을 지원하는 그래픽 카드에서 가장 잘 작동합니다. 권장 옵션은 NVIDIA의 Quadro K 시리즈와 AMD의 Fire Pro W 시리즈입니다.
특히 엔비디아 쿼드로(NVIDIA Quadro)는 엔비디아가 개발한 전문가 용도(워크스테이션)의 그래픽 카드입니다. 일반적으로 지포스 그래픽 카드가 게이밍에 초점이 맞춰져 있지만, 쿼드로는 다양한 산업 분야의 전문가가 필요로 하는 영역에 광범위한 용도로 사용되고 있습니다. 주로 산업계의 그래픽 디자인 분야, 영상 콘텐츠 제작 분야, 엔지니어링 설계 분야, 과학 분야, 의료 분석 분야 등의 전문가 작업용으로 사용되고 있습니다. 따라서 일반적인 소비자를 대상으로 하는 지포스 그래픽 카드와는 다르계 산업계에 포커스 되어 있으며 가격이 매우 비싸서 도입시 예산을 고려해야 합니다.
MSI가 새로운 노트북 CPU 벤치마크, 그리고 그 CPU가 내장돼 있는 신제품 노트북 제품군을 모두 CES 2023에서 공개했다. CES에서 인텔은 노트북용 13세대 코어 칩, 코드명 랩터 레이크와 핵심 제품인 코어 i9-13980HX를 발표했다.
ⓒ PCWorld
새로운 노트북용 13세대 코어 칩이 게임 플레이에서 12% 더 빠르다는 정도의 약간의 정보는 이미 알려져 있다. 사용자가 기다리는 것은 실제 CPU가 탑재된 노트북에서의 성능이지만 보통 벤치마크는 제품 출시가 임박해서야 공개되는 것이 보통이다. 올해는 다르다.
CES 2023에서 MSI는 인텔 최고급 제품군인 코어 i9-13980HX 프로세서가 탑재된 타이탄 GT77 HX과 레이더 GE78 HX를 공개했다. 이례적으로 여기에 더해 PCI 익스프레서 5 SSD의 실제 성능을 측정하는 크리스털디스크마크, 모바일 프로세서 실행 속도를 측정하는 시네벤치 벤치마크 점수도 함께 제공했다. 다음 영상의 결과부터 말하자면 인텔 최신 프로세서를 큰 폭으로 따돌릴 만한 수치다.
MSI는 레이더 GE78 HX 외에도 레이더 GE68 HX 그리고 게이밍 노트북 같지 않은 외관의 스텔스 16 스튜디오, 스텔스 14, 사이보그 14 등 2023년에 출시될 다른 노트북도 전시했다. 오래된 PC 애호가라면 MSI 노트북 전면을 장식한 화려한 복고풍의 라이트 브라이트(Lite Brite) LED를 반가워할지도 모른다. 바닥면 섀시가 투명한 플라스틱 소재로 MSI 로고가 새겨져 있는 제품도 있다. 상세한 가격, 출시일, 사양 등은 추후 공개 예정이다. editor@itworld.co.kr
고성능 노트북을 구매할 때는 코어 i7과 코어 i9 사이에서 선택의 갈림길에 서게 된다. 코어 i7 CPU도 강력하지만 코어 i9는 최고의 성능을 위해 만들어진 CPU이며 보통 그에 상응하는 높은 가격대로 판매된다.
CPU에 초점을 둔다면 관건은 성능이다. 성능을 좌우하는 두 가지 주요소는 CPU의 동작 클록 속도(MHz), 그리고 탑재된 연산 코어의 수다. 그러나 노트북에서 한 가지 중요한 제약 요소는 냉각이다. 냉각이 제대로 되지 않으면 고성능도 쓸모가 없다. 가장 적합한 노트북 CPU를 결정하는 데 도움이 되도록 인텔의 지난 3개 세대 CPU의 코어 i7과 i9에 대한 정보를 모았다. 최신 세대부터 시작해 역순으로 살펴보자.
11세대: 코어 i9 vs. 코어 i7
인텔의 11세대 타이거 레이크(Tiger Lake) H는 한 가지 큰 이정표를 달성했다. 인텔이 2015년부터 H급 CPU에 사용해 온 14nm 공정을 마침내 최신 10nm 슈퍼핀(SuperFin) 공정으로 바꾼 것이다. 오랫동안 기다려온 변화다.
인텔이 자랑할 만한 10nm 고성능 칩을 내놓자 타이거 레이크 H를 장착한 노트북도 속속 발표됐다. 얇고 가볍고 예상외로 가격도 저렴한 에이서 프레데터 트라이톤(Acer Predator Triton) 300 SE를 포함해 일부는 벌써 매장에 출시됐다. 모든 타이거 레이크 H 칩이 8코어 CPU라는 점도 달라진 부분이다. 이전 세대의 경우 같은 제품군 내에서 코어 수에 차이를 둬 성능 기대치를 구분했다.
클록 차이도 크지 않다. 코어 i7-11800H의 최대 클록은 4.6GHz, 코어 i9-11980HK는 5GHz로, 클록 속도 증가폭은 약 8.6% 차이다. 나쁘지 않은 수치지만 둘 다 8코어 CPU임을 고려하면 대부분의 사용자에게 코어 i9는 큰 매력은 없다.
다만 코어 i9에 유리한 부분을 하나 더 꼽자면 코어 i9-11980HK가 65W의 열설계전력(TDP)을 옵션으로 제공한다는 점이다. 높은 TDP는 최상위 코어 i9에만 제공되는데, 이는 전력 및 냉각 요구사항을 충족하는 노트북에서는 코어 i7 버전보다 더 높은 지속 클록 속도를 제공할 수 있음을 의미한다.
대신 이런 노트북은 두껍고 크기도 클 가능성이 높다. 따라서 두 개의 얇은 랩톱 중에서(하나는 코어 i9, 하나는 코어 i7) 고민하는 사람에겐 열 및 전력 측면의 여유분은 두께와 크기를 희생할 만큼의 가치는 없을 것이다.
*11세대의 승자: 대부분의 사용자에게 코어 i7
10세대: 코어 i9 vs. 코어 i7
인텔은 10세대 코멧 레이크(Comet Lake) H 제품군에서 14nm를 고수했다. 그 대신 코어 i9 CPU 외에 코어 i7에도 8코어 CPU를 도입, 사용자가 비싼 최상위 CPU를 사지 않고도 더 뛰어난 성능을 누릴 수 있게 했다.
11세대 노트북이 나오기 시작했지만 10세대 CPU 제품 중에서도 아직 괜찮은 제품이 많다. 예를 들어 MSI GE76 게이밍 노트북은 빠른 CPU와 고성능 155W GPU를 탑재했고, 전면 모서리에는 RGB 라이트가 달려 있다.
11세대 칩과 마찬가지로 코어와 클록 속도의 차이가 크지 않으므로 대부분의 사용자에게 코어 i7과 코어 i9 간의 차이는 미미하다. 코어 i9-10980HK의 최대 부스트 클록은 5.3GHz, 코어 i7-10870H는 5GHz로, 두 칩의 차이는 약 6%다. PC를 최대 한계까지 사용해야 하는 경우가 아니라면 더 비싼 비용을 들여 10세대 코어 i9를 구매할 이유가 없다.
*10세대 승자: 대부분의 사용자에게 코어 i7
9세대: 코어 i9 대 코어 i7
인텔은 9세대 커피 레이크 리프레시(Coffee Lake Refresh) 노트북 H급 CPU에서 14nm 공정을 계속 유지했다. 코어 i9는 더 높은 클록 속도(최대 5GHz)를 제공하며 8개의 CPU 코어를 탑재했다. 물론 이 칩은 2년 전에 출시됐지만 인텔이 설계를 도운 XPG 제니아(Xenia) 15 등 아직 괜찮은 게이밍 노트북이 있다. 얇고 가볍고 빠르며 엔비디아 RTX GPU를 내장했다.
8코어 4.8GHz 코어 i9-9880HK와 4.6GHz 6코어 코어 i7-9850의 클록 속도 차이는 약 4%로, 실제 사용 시 유의미한 차이로 이어지는 경우는 극소수다. 두 CPU 모두 기업용 노트북에 많이 사용됐다. 대부분의 소비자용 노트북에는 8코어 5GHz 코어 i9-9880HK와 6코어 4.5GHz 코어 i7-9750H가 탑재됐다. 이 두 CPU의 클록 차이는 약 11%로, 이 정도면 유의미한 차이지만 마찬가지로 대부분의 경우 실제로 체감하기는 어렵다.
그러나 코어 수의 차이는 멀티 스레드 애플리케이션에서 큰 체감 효과로 이어지는 경우가 많다. 3D 모델링 테스트인 씨네벤치(Cinebench) R20에서 코어 i9-9980HK를 탑재한 구형 XPS 15의 점수는 코어 i7-9750H를 탑재한 게이밍 노트북보다 42% 더 높았다. 8코어 코어 i9의 발열을 심화하는 무거운 부하에서는 성능 차이가 약 7%로 줄어들었다. 여기에는 노트북의 설계가 큰 영향을 미칠 것이다. 어쨌든 일부 상황에서는 8코어가 6코어보다 유리하다.
또한 수치해석의 경우 결과를 분석하는 작업중의 많은 부분이 POST 작업으로 그래픽처리가 필요하다. 따라서 아래 영상편집을 위한 노트북에 대한 자료도 선택에 도움이 될것으로 보인다.
영상 편집을 위한 최고의 노트북 9선
Brad Chacos, Ashley Biancuzzo, Sam Singleton | PCWorld
영상을 편집하다 보면 컴퓨터의 여러 리소스를 집약적으로 사용하기 마련이다. 그래서 영상 편집은 대부분 데스크톱 PC에서 하는 경우가 많지만, 노트북에서 영상을 편집하려 한다면 PC만큼 강력한 사양이 뒷받침되어야 한다.
ⓒ Gordon Mah Ung / IDG
영상 편집용 노트북을 구매할 때 가장 비싼 제품을 선택할 필요는 없다. 사용 환경에 맞게 프로세서, 디스플레이의 품질, 포트 종류 등을 다양하게 고려해야 한다. 다음은 영상 편집에 최적화된 노트북 제품이다. 추천 제품을 확인한 후 영상 편집용 노트북을 테스트하는 팁도 참고하자.
1. 영상 편집용 최고의 노트북, 델 XPS 17(2022)
ⓒ IDG
장점 • 가격 대비 강력한 기능 • 밝고 풍부한 색채의 대형 디스플레이 • 썬더볼트 4 포트 4개 제공 • 긴 배터리 수명 • 시중에서 가장 빠른 GPU인 RTX 3060
단점 • 무겁고 두꺼움 • 평범한 키보드 • USB-A, HDMI, 이더넷 미지원
델 XPS 17(2022)이야말로 콘텐츠 제작에 최적화된 노트북이다. 인텔 12세대 코어 i7-12700H 프로세서 및 엔비디아 지포스 RTX 3060는 편집을 위한 뛰어난 성능을 제공한다. 1TB SSD도 함께 지원되기에 데이터를 옮길 때도 편하다.
XPS 17은 SD카드 리더, 여러 썬더볼트 4 포트, 3840×2400 해상도의 17인치 터치스크린 패널, 16:10 화면 비율과 같은 영상 편집자에게 필요한 기능을 포함한다. 무게도 2.5kg 대로 비교적 가볍다. 배터리 지속 시간은 한번 충전 시 11시간인데, 이전 XPS 17 버전보다 1시간 이상 늘어난 수치다.
2. 영상 편집에 최적화된 스크린, 델 XPS 15 9520
ⓒ IDG
장점 • 뛰어난 OLED 디스플레이 • 견고하고 멋진 섀시(Chassis) • 강력한 오디오 • 넓은 키보드 및 터치패드
단점 • 다소 부족한 화면 크기 • 실망스러운 배터리 수명 • 시대에 뒤떨어진 웹캠 • 제한된 포트
델 XPS 15 9520은 놀라운 OLED 디스플레이를 갖추고 있으며, 최신 인텔 코어 i7-12700H CPU 및 지포스 RTX 3050 Ti 그래픽이 탑재되어 있다. 컨텐츠 제작 및 영상 편집용으로 가장 선호하는 제품이다. 시스템도 좋지만 투박하면서 금속 소재로 이루어진 외관이 특히 매력적이다.
15인치 노트북이지만 매일 갖고 다니기에 다소 무거운 것은 단점이다. XPS 17 모델에서 제공되는 포트도 일부 없다. 그러나 멋진 OLED 디스플레이가 단연 돋보이며, 3456X2160 해상도, 16:10 화면 비율, 그리고 매우 선명하고 정확한 색상을 갖추고 있어 좋다.
3. 최고의 듀얼 모니터 지원, 에이수스 젠북 프로 14 듀오 올레드
ⓒ IDG
장점 • 놀라운 기본 디스플레이와 보기 쉬운 보조 디스플레이 • 탁월한 I/O 옵션 및 무선 연결 • 콘텐츠 제작에 알맞은 CPU 및 GPU 성능
단점 • 생산성 노트북 치고는 부족한 배터리 수명 • 작고 어색하게 배치된 트랙패드 • 닿기 어려운 포트 위치
에이수스 젠북 프로 14 듀오(Asus Zenbook Pro 14 Duo OLED)는 일반적이지 않은 노트북이다. 일단 사양은 코어 i7 프로세서, 지포스 RTX 3050 그래픽, 16GB DDR5 메모리, 빠른 1TB NVMe SSD를 포함해 상당한 성능을 자랑한다. 또한 초광도의 547니트로 빛을 발하는 한편 DCI-P3 색영역의 100%를 커버하는 14.5인치 4K 터치 OLED 패널을 갖추고 있다. 사실상 콘텐츠 제작자를 위해 만들어진 제품이라 볼 수 있다.
가장 흥미로운 부분은 키보드 바로 위에 위치한 12.7인치 2880×864 스크린이다. 윈도우에서는 해당 모니터를 보조 모니터로 간주하며, 사용자는 번들로 제공된 에이수스 소프트웨어를 사용해 트랙패드로 사용하거나 어도비 앱을 위한 터치 제어 패널을 표시할 수 있다. 어떤 작업이든 유용하게 써먹을 수 있다.
젠북 프로 14 듀오 올레드는 기본적으로 휴대용이자 중간급 워크스테이션이다. 단, 배터리 수명은 평균 수준이기 때문에 중요한 작업 수행이 필요한 경우, 반드시 충전 케이블을 가지고 다녀야 한다. 그럼에도 불구하고 젠북 프로 14 듀오 올레드는 3D 렌더링 및 인코딩과 같은 작업에서 탁월한 성능을 보여 콘텐츠 제작자들에게 맞춤화 된 컴퓨터이다. 듀얼 스크린은 역대 최고의 기능이다.
4. 영상 편집하기 좋은 포터블 노트북, 레이저 블레이드 14(2021)
ⓒ IDG
장점 • AAA 게임에서 뛰어난 성능 • 훌륭한 QHD 패널 • 유난히 적은 소음
단점 • 700g으로 무거운 AC 어댑터 • 비싼 가격 • 썬더볼트 4 미지원
휴대성이 핵심 고려 사항이라면, 레이저 블레이드 14(Razer Blade 14) (2021)를 선택해 보자. 노트북 두께는 1.5cm, 무게는 1.7kg에 불과해 비슷한 수준의 노트북보다 훨씬 가볍다. 사양은 AMD의 8-코어 라이젠 9 5900HX CPU, 엔비디아의 8GB 지포스 RTX 3080, 1TB NVMe SSD, 16GB 메모리를 탑재하고 있어 사양도 매우 좋다.
그러나 휴대성을 대가로 몇 가지 이점을 포기해야 할 수 있다. 일단 14인치 IPS 등급 스크린은 공장에서 보정된 상태로 제공되지만, 최대 해상도는 2560×1440다. 또 풀 DCI-P3 색영역을 지원하지만 4K 영상 편집은 불가능하다. 거기에 레이저 블레이드 14는 SD 카드 슬롯도 없다. 다만 편집 및 렌더링을 위한 강력한 성능을 갖추고 있고 가방에 쉽게 넣을 수 있는 제품인 것은 분명하다.
5. 배터리 수명이 긴 노트북, 델 인스피론 16
ⓒ Dell
장점 • 넉넉한 16인치 16:10 디스플레이 • 긴 배터리 수명 • 경쟁력 있는 애플리케이션 성능 • 편안한 키보드 및 거대한 터치패드 • 쿼드 스피커(Quad speakers)
단점 • GPU 업그레이드 어려움 • 512GB SSD 초과 불가 • 태블릿 모드에서는 어색하게 느껴질 수 있는 큰 스크린
긴 배터리 수명을 가장 최우선으로 고려한다면, 델 인스피론 16(Dell Inspiron 16)을 살펴보자. 콘텐츠 제작 작업을 하며테스트해보니, 인스피론 16은 한 번 충전으로 16.5시간 동안 이용할 수 있다. 외부에서 작업을 마음껏 편집할 수 있는 시간이다. 그러나 무거운 배터리로 인해 무게가 2.1 kg에 달하므로 갖고 다니기에 적합한 제품은 아니다.
가격은 저렴한 편이나 몇 가지 단점이 있다. 일단 인텔 코어 i7-1260P CPU, 인텔 아이리스 Xe 그래픽, 16GB 램, 512GB SSD 스토리지를 탑재하고 있다. 이 정도 사양으로 영상 편집 프로젝트 대부분을 작업할 수 있으나, 스토리지 용량이 부족하기 때문에 영상 파일을 저장할 경우 외장 드라이브가 필요하다. 그러나 델 인스피론 16이 진정으로 빛을 발하는 부분은 단연 배터리 수명이다. 또한 강력한 쿼드 스피커 시스템도 사용해 보면 만족할 것이다. 포트의 경우, USB 타입-C 2개, USB-A 3.2 Gen 1 1개, HDMI 1개, SD 카드 리더 1개, 3.5mm 오디오 잭 1개가 제공된다.
6. 게이밍과 영상 편집 모두에 적합한 노트북, MSI GE76 레이더
ⓒ MSI
장점 • 뛰어난 성능을 발휘하는 12세대 코어 i9-12900HK • 팬 소음을 크게 줄이는 AI 성능 모드 • 1080p 웹캠과 훌륭한 마이크 및 오디오로 우수한 화상 회의 경험 제공
단점 • 동일한 유형의 세 번째 버전 • 어수선한 UI • 비싼 가격
사양이 제일 좋은 제품을 찾고 있을 경우, 크고 무거운 게이밍 노트북을 선택해 보자. MSI GE76 레이더(Raider)는 강력한 14-코어 인텔 코어 i9-12900HK 칩, 175와트의 엔비디아 RTX 3080 Ti가 탑재됐고, 충분한 내부 냉각 성능 덕분에 UL의 프로시온(Procyon) 벤치마크의 어도비 프리미어 테스트에서 다른 노트북보다 훨씬 뛰어난 성능을 보였다. MSI GE76 레이더는 심지어 고속 카드 전송을 위해 PCle 버스에 연결된 SD 익스프레스(SD Express) 카드 리더도 갖추고 있다.
동일한 제품의 작년 모델은 게이머 중심의 360Hz 1080p 디스플레이를 지원한다. 영상 편집 과정에서는 그닥 이상적이지 않은 사양이다. 그러나 2022년의 12UHS 고급 버전은 4K, 120Hz 패널을 추가했는데, 이 패널은 콘텐츠 생성에 맞춰 튜닝 되지는 않았으나 17.3인치의 넓은 스크린 크기이기에 영상 편집자에게 꽤 유용하다.
7. 가성비 좋은 노트북, HP 엔비 14t-eb000(2021)
ⓒ IDG
장점 • 높은 가격 대비 우수한 성능 • 환상적인 배터리 수명 • 성능 조절이 감지되지 않을 정도의 저소음 팬 • 썬더볼트 4 지원
단점 • 약간 특이한 키보드 레이아웃 • 비효율적인 웹캠의 시그니처 기능
가장 빠른 영상 편집 및 렌더링을 원할 경우 하드웨어에 더 많은 비용을 들여야 하지만, 예산이 넉넉하지 않을 때가 있다. 이때 HP 엔비(Envy) 14 14t-eb000) (2021)를 이용해보면 좋다. 가격은 상대적으로 저렴한 편이고 견고한 기본 컨텐츠 제작에 유용하다.
엔트리 레벨의 지포스 GTX 1650 Ti GPU 및 코어 i5-1135G7 프로세서는 그 자체로 업계 최고 제품은 아니다. 하지만 일반적인 편집 작업을 충분히 수행할 수 있는 사양이다. 분명 가성비 좋은 제품이다. 14인치 1900×1200 디스플레이는 16:10 화면 비율로 생산성을 향상하고, 공장 색 보정과 DCI-P3는 지원하지 않지만 100% sRGB 지원을 제공한다. 그뿐만 아니라, HP 엔비 14의 경우 중요한 SD 카드 및 썬더볼트 포트가 포함되며, 놀라울 정도로 조용하게 실행된다.
8. 컨텐츠 제작에 알맞은 또다른 게이밍 노트북, 에이수스 ROG 제피러스 S17
ⓒ
장점 • 뛰어난 CPU 및 GPU 성능 • 강력하고 혁신적인 디자인 • 편안한 맞춤형 키보드
단점 • 약간의 압력이 필요한 트랙패드 • 상당히 높은 가격
에이수스 ROG 제피러스(Zephyrus) S17은 영상 편집자의 궁극적인 꿈이다. 이 노트북은 초고속 GPU 및 CPU 성능과 함께 120Hz 화면 재생률을 갖춘 놀라운 17.3인치 4K 디스플레이를 탑재하고 있다. 견고한 전면 금속 섀시, 6개의 스피커 사운드 시스템 및 맞춤형 키보드는 프리미엄급 경험을 더욱 향상한다. 거기다 SD 카드 슬롯 및 풍부한 썬더볼트 포트가 포함되어 있어 더욱 좋다. 그러나 이를 위해 상당한 비용을 지불해야 한다. 예산이 넉넉하고 최상의 제품을 원한다면 제피루스 S17을 선택하면 된다.
9. 강력한 휴대성을 가진 노트북, XPG 제니아 15 KC
ⓒ XPG
장점 • 가벼운 무게 • 조용함 • 상대적으로 빠른 속도
단점 • 중간 수준 이하의 RGB • 평범한 오디오 성능 • 느린 SD 카드 리더
사양이 좋은 노트북의 경우, 대부분 부피가 크고 무거워서 종종 2.2kg 또는 2.7kg를 넘기도 한다. XPG 제니아 15 KC(XPG Xenia 15 KC)만은 예외다. XPG 제니아 15 KC의 무게는 1.8kg가 조금 넘는 수준으로, 타제품에 비해 상당히 가볍다. 또한 소음도 별로 없다. 원래 게이밍 노트북 자체가 소음이 크기에 비교해보면 큰 장점이 될 수 있다. 1440p 디스플레이와 상대적으로 느린 SD 카드 리더 성능으로 인해 일부 콘텐츠 제작자들이 구매를 주저할 수 있으나, 조용하고 휴대하기 좋은 제품을 찾고 있다면 제니아 15 KC가 좋은 선택지다.
영상 편집 노트북 구매 시 고려 사항
영상 편집 노트북 구매 시 고려해야 할 가장 중요한 사항은 CPU 및 GPU다. 하드웨어가 빨라질수록 편집 속도도 빨라진다. 필자는 UL 프로시온 영상 편집 테스트(UL Procyon Video Editing Test)를 통해 속도를 테스트해보았다. 이 벤치마크는 2개의 서로 다른 영상 프로젝트를 가져와 색상 그레이딩 및 전환과 같은 시각적 효과를 적용한 다음, 1080p와 4K 모두에서 H.264, H.265를 사용해 내보내는 작업을 어도비 프리미어가 수행하도록 한다.
ⓒ Gordon Mah Ung / IDG
성능은 인텔의 11세대 프로세서를 실행하는 크고 무거운 노트북에서 가장 높았고, AMD의 비피 라이젠 9(beefy Ryzen 9) 프로세서를 탑재한 노트북이 바로 뒤를 이었다. 10세대 인텔 칩은 여전히 상당한 점수를 기록하고 있다. 위의 차트에는 없으나 새로운 인텔 12세대 노트북은 더 빨리 실행된다. 최고 성능의 노트북은 모두 최신 인텔 CPU 및 엔비디아의 RTX 30 시리즈 GPU를 결합했는데, 두 기업 모두 어도비 성능 최적화에 많은 시간 및 리소스를 투자했기 때문에 놀라운 일은 아니다.
GPU는 어도비 프리미어 프로에서 CPU보다 더 중요하지만, 매우 빠르게 수확체감 지점에 다다른다. 최고급 RTX 3080 그래픽을 사용하는 노트북은 RTX 3060 그래픽을 사용하는 노트북보다 영상 편집 속도가 더 빠르나, 속도 차이가 크지는 않다. 델 XPS 17 9710의 점수를 살펴보면, 지포스 RTX 3060 노트북 GPU는 MSI GE76 레이더의 가장 빠른 RTX 3080보다 14% 더 느릴 수 있다. 특히 GE76 레이더가 델 노트북에 비해 얼마나 더 크고 두꺼운지를 고려할 때 수치가 크지는 않다.
일반적으로 그래픽과 영상 편집을 위해 적어도 RTX 3060을 갖추는 것을 권장한다. 그러나 영상 편집은 워크플로에 크게 의존한다. 특정 작업 및 도구는 CPU 집약적이거나 프리미어보다 GPU에 더 의존할 수 있다. 이 경우 원하는 요소의 우선순위를 조정하길 바란다. 앞서 언급한 목록은 기본적으로 여러 요소를 종합적으로 고려해서 만든 내용이다.
인텔 및 엔비디아는 각각 퀵 싱크(Quick Sync) 및 쿠다(CUDA)와 같은 도구를 구축하는 데 수년을 보냈고, 이로 인해 많은 영상 편집 앱의 속도는 크게 향상될 수 있다. AMD 하드웨어는 영상 편집에 적합하나 특히 워크플로가 공급업체별 소프트웨어 최적화에 의존하는 경우, 특별한 이유가 없는 한 인텔 및 엔비디아를 사용하는 것을 추천한다.
영상 촬영 ⓒ Gordon Mah Ung/IDG
그러나 내부 기능만 신경 써서는 안된다. PC월드의 영상 디렉터인 아담 패트릭 머레이는 “영상 편집에 이상적인 노트북에는 카메라로 촬영 중 영상 파일을 저장하는 SD 카드 리더가 포함되어 있다”라고 강조한다. 또한 머레이는 영상 편집에 이상적인 게임용 노트북에서 흔히 볼 수 있는 초고속 1080p 패널보다 4k, 60Hz 패널을 갖춘 노트북을 선택할 것을 추천한다.
4K 영상을 잘 편집하려면 4K 패널이 필요하며, 초고속 화면 재생률은 게임에서처럼 영상 편집에는 아무런 의미가 없다. 예를 들어, 개인 유튜브 채널용으로 일상적인 영상만 만드는 경우 색상 정확도가 중요하지 않을 수 있다. 그러나 색상 정확도가 중요할 경우, 델타 E < 2 색상 정확도와 더불어 DCI-P3 색 영역 지원은 필수적이다.
게임용 노트북은 사양이 좋지만 콘텐츠 제작용으로는 조금 부족해 보일 수 있다. 게임용과 콘텐츠 제작용으로 함께 쓰는 노트북을 원한다면, 게임용으로 노트북 한 대를 구매하고, 색상을 정확히 파악하기 위한 모니터를 추가로 구매하는 것도 방법이다. editor@itworld.co.kr
코어가 많은 그래픽카드의 경우 가격이 상상 이상으로 높습니다. 빠르면 빠를수록 좋겠지만 어디까지나 예산에 맞춰 구매를 해야 하는 현실을 감안할 수 밖에 없는 것 같습니다.
한가지 유의할 점은 엔비디아의 GTX 게이밍 하드웨어는 모델에 따라 다르기는 하지만, 볼륨 렌더링의 속도가 느리거나 오동작 등 몇 가지 제한 사항이 있습니다. 일반적으로 노트북에 내장된 통합 그래픽 카드보다는 개별 그래픽 카드를 강력하게 추천합니다. 최소한 그래픽 메모리는 512MB 이상이어야 하고 1GB이상을 권장합니다.
2021-12-15 현재 그래픽카드의 성능 순위는 위와 다음과 같습니다. 출처: https://www.videocardbenchmark.net/high_end_gpus.html
주요 Notebook
출시된 모든 그래픽 카드가 노트북용으로 장착되어 출시되지는 않기 때문에, 현재 오픈마켓 검색서비스를 제공하는 네이버에서 Lenovo Quadro 그래픽카드를 사용하는 노트북을 검색하면 아래와 같습니다. 검색 시점에 따라 상위 그래픽카드를 장착한 노트북의 대략적인 가격을 볼 수 있을 것입니다.
<검색 방법> 네이버 쇼핑 검색 키워드 : 컴퓨터 제조사 + 그래픽카드 모델 + NoteBook 형태로 검색 Lenovo quadro notebook or HP quadro notebook 또는 Lenovo firepro notebookorHP firepronotebook
( 2021-12-15기준)
대부분 검색 시점에 따라 최신 CPU와 최신 그래픽카드를 선택하여 검색을 하면 예산에 적당한 노트북을 자신에게 맞는 최상의 노트북을 어렵지 않게 선택할 수 있습니다.
PCI-Express(또는 PCI-E) 표준을 사용하는 최근 출시된 AMD 비디오 카드(예: AMD RX 6950 XT)와 nVidia 그래픽 카드(예: nVidia GeForce RTX 3090)는 하이엔드 비디오 카드 차트에서 흔히 볼 수 있습니다.
PassMark – G3D Mark High End Videocards / Price
FLOW-3D POST 성능과 밀접한 그래픽카드의 이해
FLOW Science, inc의 최첨단 POST Processor인 FLOW-3D POST를 최대한 활용하려면 좋은 하드웨어가 있어야 합니다. 이 블로그에서 소프트웨어 엔지니어링의 GUI 개발자/관리자인 Stephen Sanchez는 이러한 하드웨어 권장 사항에 따라 최적의 FLOW-3D POST 경험을 얻을 수 있는 방법에 대해 정보를 제공 합니다.
고품질 그래픽 하드웨어
최소 3GB의 VRAM 이 있는 그래픽 카드로 시작하는 것이 좋습니다 . 이것은 많은 볼륨 렌더링을 수행할 경우 특히 중요합니다. 볼륨 렌더링은 FLOW-3D POST 의 고급 기능으로 iso-surface가 아닌 유체 도메인 전체에서 변수의 세부 사항을 시각화합니다. 이 기능은 매우 통찰력 있지만 후 처리 중에 효과적으로 사용하려면 좋은 하드웨어가 필요합니다.
다음으로 Intel 통합 그래픽을 기본 그래픽 하드웨어로 사용해서는 안됩니다. 인텔 통합 그래픽은 전용 그래픽 하드웨어가 있는 랩톱에서도 대부분의 랩톱에서 일반적입니다(자세한 내용은 아래 참조).
대부분의 FLOW-3D POST 기능은 이 구성에서 작동하지 않으므로 Intel 통합 그래픽을 지원하지 않습니다.
FLOW-3D POST 는 NVIDIA 그래픽 카드 와 함께 사용할 때 가장 잘 수행됩니다. FLOW-3D POST 가 잘 작동하는 것으로 확인되었으므로 Maxwell 아키텍처 제품군 이상의 NVIDIA 그래픽 하드웨어를 적극 권장 합니다.
NVIDIA Quadro 카드는 가장 안정적인 것으로 입증되었습니다. 고급 AMD 카드도 작동해야 하지만 NVIDIA 하드웨어 및 드라이버만큼 안정적이지 않다는 사실을 발견 했으므로 항상 AMD보다 NVIDIA를 권장합니다.
노트북의 듀얼 그래픽 카드 – 간단하지만 숨겨진 솔루션
이제 많은 노트북에 NVIDIA 그래픽 카드와 Intel 통합 그래픽 간에 전환 할 수 있는 기능이 있습니다. NVIDIA 카드로 FLOW-3D POST 가 실행되고 있는지 확인하는 것이 중요합니다 . NVIDIA 제어판을 통해 NVIDIA 카드로 노트북을 강제로 실행할 수 있습니다.
비디오 드라이버 업데이트
비디오 드라이버가 업데이트 되었는지 확인하는 것이 좋습니다. FLOW-3D POST 에서 비디오 드라이버를 업데이트하여 쉽게 해결할 수 있는 아티팩트 및 디스플레이 문제에 대한 보고가 있었습니다 . 비디오 드라이버를 최신 상태로 유지하는 것은 이러한 문제를 방지하는 좋은 방법입니다.
RAM, RAM, RAM!
메모리가 충분하지 않으면 시뮬레이션 후 처리가 불가능할뿐만 아니라 메모리 요구 사항을 인식하는 것이 중요합니다. 최대 10 배의 성능 저하로 이어질 수 있습니다! FLOW-3D POST 에 필요한 RAM 양은 여러 요소, 특히 시뮬레이션 크기에 따라 다릅니다. 사용자에게 최대한의 유연성을 제공하기 위해 메시의 셀 수에 따라 다음과 같은 RAM 권장 사항이 있습니다.
초대형 (2 억 개 이상의 셀) : 최소 128GB
대용량 (6 천 ~ 1 억 5 천만 셀) : 64-128GB
중간 (3 천만 ~ 6 천만 셀) : 32-64GB
소형 (3,000 만 셀 이하) : 최소 32GB
FLOW-3D POST 는 메모리 집약적 일 수 있습니다. 실행할 시뮬레이션 크기에 대한 대략적인 아이디어가 있는 경우, 이 지침을 가능한 한 잘 따르는 것이 좋습니다. 즉, 유연성을 극대화하고 가장 원활한 FLOW-3D POST 경험을 보장하기 위해 문제 크기에 관계없이 가능한 한 많은 RAM을 확보하는 것이 좋습니다.
그래픽 카드를 업그레이드 교체 설치하는 방법
그래픽 카드를 업그레이드하는 것은 성능 향상을 위한 좋은 방법이다. 그래픽 카드 업그레이드를 통해 시각적으로 고사양을 요구하는 POST 작업을 쉽게 소화할 수 있는 컴퓨터로 진화할 수 있다.
업그레이드를 위한 그래픽 카드 구매시 고려 사항, 기존 PC에 적합한가?
원하는 그래픽 카드를 결정하는 것은 복잡하고 미묘한 문제다. AMD와 엔비디아는 200달러 미만에서부터 최대 1,500달러에 이르는 지포스(GeForce) RTX 3090에 이르기까지 거의 모든 예산에 대한 선택지를 제공하기 때문이다.
카드의 소음, 발열, 전력 소비 등과 같은 사항을 고려할 수 있겠지만, 일반적으로는 비용 대비 가장 큰 효과를 제공하는 그래픽 카드를 원한다.
컴퓨터가 새 그래픽 카드를 지원하는 적절한 하드웨어인지 확인한다.
사용자가 겪는 가장 일반적인 문제는 부적절한 파워 서플라이(power supply)다. 충분한 전력을 공급할 수 없거나 사용 가능한 PCI-E 전원 커넥터가 충분하지 않을 수 있다. 필자의 경험상 파워 서플라이는 적어도 제조업체에서 권장하는 파워 서플라이의 요구 사항을 충족해야 한다. 예를 들어, 350W를 소비하는 지포스 GTX 3090을 구입했다면 8핀 전원 커넥터 한 쌍과 함께 엔비디아에서 제안한 최소 750W의 전력 공급 장치를 갖춰야 한다.
현재 파워 서플라이가 얼마나 많은 전력을 제공하는지 알아보려면 PC 본체를 열고 모든 파워 서플라이에 기본 정보가 나열된 표준 식별 스티커를 확인하면 된다. 또한 사용 가능한 6핀 및 8핀 PCI-E 커넥터의 수를 확인할 수 있다.
ⓒ Thomas Ryan 파워서플라이
마지막으로 본체 내부에 새 그래픽 카드를 넣을 충분한 공간이 있는지 확인한다. 일부 고급 그래픽 카드는 길이가 상당히 길어 30Cm 이상일 수 있으며, 확장 슬롯이 2개 또는 3개가 될 수 있다. 해당 그래픽 카드의 실제 크기는 제조업체 웹사이트에서 찾을 수 있다.
여기까지 해결했다면 이제 본격적으로 설치 작업에 착수한다.
생각보다 간단한 그래픽 카드 설치 작업
그래픽 카드 설치에는 새 그래픽 카드, 컴퓨터, 그리고 십자 드라이버 3가지만 있으면 된다. 설치하기 전 PC를 끄고 전원 플러그를 뽑는다.
기존 GPU를 제거해야 하는 경우가 아니면, 먼저 프로세서의 방열판에 가장 가까운 긴 PCI-E x16 슬롯을 찾아야 한다. 이 슬롯은 메인보드의 첫 번째 또는 두 번째 확장 슬롯이다.
이 슬롯에 접근을 차단하는 느슨한 전선이 없는지 확인한다. 기존 그래픽 카드를 교체하는 경우, 연결된 케이블을 모두 분리하고, PC 본체 후면 내부에 고정 브래킷에서 나사를 제거한 다음, 카드를 제거한다. 대부분의 메인보드에는 그래픽 카드를 제자리에 고정하는 PCI-E 슬롯 끝에 작은 플라스틱 걸쇠(latch)가 있다. 이 걸쇠를 눌러 이전 그래픽 카드의 잠금을 해제하고 분리한다.
ⓒ Thomas Ryan PCI-E x16 슬롯에 설치
이제 새 그래픽 카드를 개방형 PCI-E x16 슬롯에 설치할 수 있다. 카드를 슬롯에 완전히 삽입한 다음, PCI-E 슬롯 끝에 있는 플라스틱 걸쇠를 눌러 제자리에 고정한다. 그런 다음 나사를 사용해 그래픽 카드의 금속 고정 브래킷을 PC 본체에 고정한다. 덮개 브래킷 또는 이전 그래픽 카드를 고정했던 나사를 재사용할 수 있다.
ⓒ Thomas Ryan 그래픽 카드에는 추가 전원 커넥터 연결
대부분의 게임용 그래픽 카드에는 추가 전원 커넥터가 필요하다. 추가 전원이 필요한 경우, 해당 PCI-E 전원 케이블을 연결했는지 확인한다. 전원이 제대로 공급되지 않으면 그래픽 카드가 제대로 작동하지 않는다. 이 PCI-E 전원 케이블을 연결하지 않으면 PC 자체가 부팅되지 않을 수 있다.
그래픽 카드를 고정하고 난 후, 전원을 켠 상태에서 본체 측면 패널을 제자리로 밀어넣고 디스플레이 케이블을 새 그래픽 카드에 연결해 작업을 완료한다. 이제 컴퓨터를 켠다.
이제 그래픽 카드의 소프트웨어를 업그레이드할 단계가 왔다.
새 그래픽 카드가 이전 카드와 동일한 브랜드일 경우에는 절차가 간단하다. 제조업체의 웹사이트로 이동해 운영체제에 맞는 최신 드라이버 패키지를 다운로드한다. 그래픽 드라이버는 일반적으로 약 500MB로, 상당히 크다. 인터넷 연결 속도에 따라 다운로드하는 데 시간이 걸릴 수도 있다. 드라이버를 설치하고 컴퓨터를 다시 시작하면 이제 새 그래픽 카드가 제공하는 부드럽고 매끄러운 프레임 속도를 즐길 수 있다.
그래픽 카드 제조업체가 바뀐 경우(인털에서 AMD로, 혹은 AMD에서 인텔로), 새 그래픽 카드용 드라이버를 설치하기 전에 이전 그래픽 드라이버를 제거하고 컴퓨터를 다시 시작해야 한다. 이전 드라이버를 제거하지 않으면 새 드라이버와 충돌할 수 있다.
editor@itworld.co.kr 기사 일부 발췌 인용
그래픽 카드 GPU 온도 확인하는 방법
그래픽 카드 온도 확인은 아주 쉽다. 윈도우에서 바로 온도를 확인할 수 있는 내장 도구도 추가됐다. 또한, 무료 GPU 모니터링 도구가 많이 있고 그중 대다수가 온도를 측정해준다. 조금 더 자세히 알아보자.
ⓒ MARK HACHMAN / IDG 그래픽카드 온도 확인
마이크로소프트가 윈도우 10 2020년 5월 업데이트에서 GPU 온도 모니터링 툴을 작업 관리자에 추가했다. 무려 24년이나 걸렸다.
Ctrl+Shift+Esc를 열어 작업 관리자 대화창을 열거나 Ctrl+Alt+Delete에서 ‘작업 관리자’를 선택하거나 윈도우 시작 메뉴 아이콘을 오른쪽 클릭해서 ‘작업 관리자’를 선택한다. 여기에서 ‘성능’ 탭으로 들어가면 왼쪽에 GPU를 확인할 수 있을 것이다. 윈도우 10 2020년 5월 업데이트 혹은 그 이후 버전의 윈도우가 설치되어 있을 때만 사용할 수 있는 기능이다.
하지만 이 기능은 매우 단순하다. 시간 흐름에 따른 온도 변화를 추적하지 않고, 현재의 온도만을 보여준다. 그리고 업무를 하거나 오버클럭 조정 중에 작업 관리자를 여는 것도 귀찮을 수 있다. 마침내 윈도우에 GPU 온도를 확인할 수 있는 기능이 들어간 것은 환영하지만, 뒤이어 설명할 서드파티 도구가 훨씬 더 나은 GPU 온도 확인 옵션을 제공한다.
AMD 라데온 그래픽 카드 사용자가 라데온 세팅(Radeon Setting) 앱을 최신 버전으로 유지하고 있다면 방법은 쉽다. 2017년 AMD는 시각 설정을 변경할 수 있는 라데온 오버레이(Radeon Overlay)를 출시했다. 여기에도 GPU 온도와 다른 중요한 정보를 확인할 수 있는 성능 모니터 기능이 있다.
프로그램을 활성화하려면 Alt+R 키를 눌러 라데온 오버레이를 불러온다. 성능 모니터링 섹션에서 원하는 탭을 선택한다. Ctrl+Shift + 0을 눌러서 성능 모니터링 도구 설정을 단독으로 불러올 수 있다.
라데온 세팅 앱에서 오버클럭 도구인 와트맨(Wattman)으로 이동해 GPU 온도를 확인할 수 있다. 윈도우 바탕 화면을 우클릭하고, 라데온 설정을 선택한 후 게이밍(Gaming) > 글로벌 세팅(Global Setting) > 글로벌 와트맨(Global Wattman) 항목으로 이동한다. 도구를 사용해 지나친 오버클럭으로 그래픽 카드를 날려버리지 않겠다고 서약한 후에는 와트맨에 액세스하고 GPU 온도, 그리고 그래프 형태로 된 핵심적 통계 수치를 볼 수 있다. 여기까지가 전부다.
라데온 사용자가 아닌 사람도 많을 것이다. 스팀의 하드웨어 설문 조사는 전체 응답자 PC 중 75%가 엔비디아 지포스 그래픽 카드를 탑재했다는 결과를 발표했다. 그리고 지포스 익스피리언스 소프트웨어는 GPU 온도 확인 기능을 제공하지 않아서 서드파티 소프트웨어의 손을 빌려야 한다.
그래픽 카드 제조 업체는 보통 GPU 오버 클럭을 위한 특수한 소프트웨어를 제공한다. 이 도구에는 라데온 오버레이처럼 가장 중요한 측정을 실행할 때 OSD(On-Screen Display)를 지속하는 옵션 등이 있다. 여러 종류 중에서 가장 추천하는 것은 다재다능함을 갖춘 MSI의 애프터버너(Afterburner) 도구다. 이 제품은 오랫동안 인기를 얻었는데 엔비디아 지포스, AMD 라데온 그래픽 카드 두 제품 모두에서 잘 작동하고, 반길 만한 다른 기능도 더했다.
이제 그래픽 카드를 모니터링하는 소프트웨어를 갖췄다. 하지만 화면을 채우는 숫자는 맥락이 없이는 아무것도 아니다. 그래픽 카드 온도는 어디까지 괜찮은 것일까?
쉬운 대답은 없다. 제품마다 다르다. 이럴 때는 구글이 친구가 된다. 대다수 칩은 섭씨 90도 중반에도 작동하고, 게이밍 노트북에서도 90도까지 온도가 올라가는 경우가 흔히 있다. 그러나 일반 데스크톱 PC 온도가 90도 이상으로 올라간다면 구조 신호나 다름없다. 공기 흐름이 원활한 GPU 1대 시스템에서는 80도 이상 올라가면 위험하다. 팬이 여러 개 달린 커스텀 그래픽 카드는 무거운 워크로드 하에서도 60~70도가 적당하고, 수냉쿨러가 달린 GPU라면 온도가 더 낮아야 할 것이다.
그래픽 카드가 최근 5년 안에 생산된 제품이고 90도 이상으로 뜨거워진다면, 또는 최근 몇 주간 온도가 급격히 상승했다면 다음의 냉각 방법을 고려해보자.
그래픽 카드 온도 낮추는 법
그래픽 카드 온도가 높아졌을 때 하드웨어 업그레이드에 돈을 들이지 않고 개선하지 않기란 어렵다. 그러나 돈을 쏟아붓기 전에 정말 그래야 하는지 필요성을 점검해 보자. 다시 한번 강조하지만 그래픽 카드는 뜨거운 온도를 버틸 수 있도록 설계되어 있다. PC가 무거운 게임이나 영상 편집 중에 강제 종료되는 경우가 아니라면 아마도 걱정할 필요가 없을 것이다.
우선, 시스템의 케이블을 깨끗하게 정리해 GPU 주변의 공기가 원활하게 순환되는지 확인하라. 케이블이 깔끔하게 정리됐다면 케이스에 팬을 추가하는 것도 고려한다. 모든 PC는 최적의 성능을 위해 공기를 빨아들이고 내보내는 팬이 여럿 달려 있는데, POST PC라면 팬은 더 많아야 한다. 저렴한 팬은 10달러부터 구입할 수 있고, RGB 조명이 붙은 화려한 제품은 조금 더 가격이 높다.
마지막으로, GPU와 히트싱크의 써멀 페이스트가 오래되어 말라 있다면 효율이 떨어질 수 있다. 특히 오래된 그래픽 카드라면 더더욱 그렇다. 그리고 아주 드문 경우지만 품질이 좋지 않은 써멀 페이스트가 발라져서 출시되는 경우도 있다. 다른 방법이 모두 효과가 없다면 써멀 페이스트를 다시 바르는 것을 시도해보자. 그러나 과정이 매우 어려울 수 있고 카드마다 조금씩 다르고, 잘못 손댈 경우 사용자 보증 기한의 보호를 받을 수 없게 된다.
온도를 확실하게 낮추려면 수랭 쿨러를 위한 쿨링 시스템을 고려한다. 대다수 사용자에게는 지나친 모험이지만 대부분 수냉쿨러는 발열과 노이즈 감소 효과가 확실하고 공기 냉각에 있어 병목 현상도 없다.
“업무 효율 향상의 기본” 멀티 모니터 구축 가이드
듀얼 모니터를 사용하면 업무 생산성이 높아진다는 연구 결과가 있지만, 모니터가 많을수록 생산성이 높아지는지 여부에 대해서는 아직 이렇다 할 근거는 없다. 그러나 업무 생산성을 생각하지 않더라도 모니터를 여러 대(3대~6대까지) 사용하는 것은 멋진 일이며, 많은 화면을 봐야 하는 엔지니어는 정말 필요할지도 모른다.
모니터를 세로로 세워두면 긴 문서를 볼 때 스크롤을 적게 해도 된다는 장점이 있다. 멀티 디스플레이 환경을 구축하기 위해 고려해야 할 모든 것들을 살펴보겠다.
멀티 모니터 구축 가이드(www.itworld.co.kr)
1단계 : 그래픽 카드 확인하기
보조 모니터를 구입하기 전에 컴퓨터가 물리적으로 이 모든 모니터들을 감당할 수 있을지 점검해 봐야 한다. 가장 쉬운 방법은 PC의 뒷면을 보고, 그래픽 포트(DVI, HDMI, 디스플레이포트, VGA 등)가 몇 개나 있는지 확인하는 것이다.
별도의 그래픽 카드가 없다면 포트를 2개밖에 발견하지 못할 것이다. 그래픽이 통합된 대부분의 마더보드는 모니터 2개 밖에 설치하지 못한다. 별도의 그래픽 카드가 있다면, 마더보드의 포트를 제외하고 최소 3개의 포트를 발견할 수 있을 것이다.
팁 : 마더보드와 별도 그래픽 카드의 포트를 모두 이용해서 멀티 모니터를 설치할 수도 있지만, 이 경우 성능 저하와 모니터끼리의 속도 차이가 발생할 것이다. 그래도 이렇게 하고 싶다면, PC의 BIOS에서 Configuration > Video > Integrated graphics 로 진입한 다음, ‘always enable’로 설정한다.
그러나 별도의 그래픽 카드에 3개 이상의 포트가 있다고 해서 이것을 모두 동시에 사용할 수 있다는 의미는 아니다. 예를 들어서 구형 엔비디아 카드는 포트가 2개 이상이어도 하나의 카드에 모니터를 2개 이상 연결할 수 없다. 자신의 그래픽 카드가 멀티 모니터를 지원하는지 판단하는 가장 좋은 방법은 그래픽 카드 모델명을 찾아서 원하는 모니터 개수와 함께 검색하는 것이다. 예를 들어, ‘엔비디아 GTX 1660 모니터 4대’라고 검색하면 된다.
EVGA 지포스 RTX 2060 KO 같은 현대적인 그래픽카드는 여러 디스플레이를 동시에 연결할 수 있다. ⓒ BRAD CHACOS/IDG
그래픽 카드가 원하는 만큼 충분히 모니터를 지원할 수 있으면 좋지만, 그렇지 않다면 추가 그래픽 카드를 구입해야 한다. 그래픽 카드를 추가로 구입하기 전 타워 안에 충분한 공간(PCI 슬롯)이 있는지, 전원 공급은 충분한지 확인해야 한다.
멀티 모니터용으로만 그래픽 카드를 구입한다면 최신 그래픽 카드 중에서도 저렴한 옵션을 선택하는 것이 좋다.
아니면 멀티 스트리밍이 지원되는 디스플레이포트를 탑재한 신형 모니터를 사용하는 방법도 있다. 그래픽 카드의 디스플레이포트 1.2에 연결하고, 디스플레이포트 케이블을 사용해 다음 모니터로 연결하는 것이다. 모니터의 크기나 해상도가 같지 않아도 된다. 뷰소닉(ViewSonic)의 VP2468이 이런 제품 중 하나다. 아마존에서 약 210달러에 판매되는 이 24인치 모니터는 디스플레이포트 아웃 외에도 프리미엄 IPS 스크린, 아주 얇은 베젤 등 멀티 모니터 설정에 이상적인 특징을 제공한다.
2단계 : 모니터 선택하기
그래픽 카드에 대해서 파악했다면 이제 추가 모니터를 구입할 차례다. 사용자에 따라서 기존에 사용하고 있는 모니터, 책상 크기, 추가 모니터 용도 등에 따라서 완벽한 모니터가 달라질 것이다.
필자의 경우, 이미 24인치 모니터 2대를 가지고 있었기 때문에 중앙에 설치할 더 큰 모니터가 필요해서 27인치 모니터를 선택했다. 게임을 하지 않기 때문에 모니터 크기 차이는 상관없었다. 하지만 사용자에 따라서 멀티 모니터로 POST를 하거나 동영상을 보기 위해서는 이러한 구성보다 같은 모니터를 연결하는 것이 더 좋을 것이다.
모니터를 구입하기 전에 PC와 모니터의 포트 호환성을 설펴야 한다. DVI-HDMI 혹은 디스플레이포트-DVI 등 전환해주는 케이블을 이용할 수도 있지만 다소 귀찮다. 그러나 PC나 모니터에 VGA 포트가 있다면, 교체를 권한다. VGA는 아날로그 커넥터이기 때문에 선명도가 떨어진다.
3단계 : PC설정
모니터를 구입하고 나면 PC에 연결하고 PC의 전원을 켠다. 이것으로 모니터 설치가 끝났다. 하지만 완전히 끝난 것은 아니다.
윈도우가 멀티 모니터 환경에서 잘 동작하게 만들어야 하는데, 윈도우 7이나 윈도우 8 사용자라면 바탕화면에서 오른쪽 클릭하고 ‘화면 해상도’를 선택한다. 윈도우 10 사용자라면 ‘디스플레이 설정’을 클릭한다. 그러면 디스플레이를 정렬할 수 있는 창이 나타난다.
ⓒ ITWorld 디스플레이 설정
여기서 모니터들이 모두 탐지되는지 확인할 수 있다. ‘식별’을 클릭하면 각 디스플레이에 큰 숫자가 나타난다. 주 모니터(작업 표시줄과 시작 버튼이 나타나는 모니터)로 사용할 모니터에 1번이 나타나야 하는데, 원하는 것을 선택한 다음 아래 여러 디스플레이 설정에서 ‘이 디스플레이를 주 모니터로 만들기’를 클릭한다. 그 다음 ‘다중 디스플레이’ 드롭다운 메뉴에서 복제할 것인지 확장할 것인지를 선택하면 되는데, 대부분의 경우 ‘디스플레이 확장’이 적합하다.
GPU 제어판에서도 다중 모니터를 설정할 수 있다. 바탕화면에서 오른쪽 클릭을 하고 엔비디아, AMD, 인텔 등 그래픽 제조사의 제어판 메뉴를 열어 윈도우와 유사한 방식으로 디스플레이를 설정할 수 있다.
멀티 디스플레이를 구축할 경우에는 같은 모델을 이용하는 것이 해상도나 선명도, 색보정 등의 문제가 발생하지 않아 ‘끊김 없는’ 경험을 할 수 있다.
본 자료는 Computer에 대한 전문적인 지식보다는 수치해석을 주 목적으로 FLOW-3D 를 이용하기 위한 해석용 컴퓨터를 선택할 때 도움을 주기 위한 자료입니다.
흔히 고성능 컴퓨터는 표준 데스크톱 컴퓨터와 어떻게 다른지 궁금하게 생각하는데, 보통 HPC(high performance computing)는 더욱 강력한 프로세서, 더 큰 메모리, 뛰어난 그래픽 성능을 갖추고 있어 일반적인 표준 데스크톱보다 훨씬 빠르게 여러 가지 복잡한 작업을 동시에 처리할 수 있습니다. 따라서 시중에서 판매하는 고사양의 컴퓨터에 CPU, Memory, Graphic Card 등을 보완하거나 고사양으로 만들어진 컴퓨터를 구매하게 되면, 단일 노드의 HPC(high performance computing)와 유사한 성능을 확보할 수 있습니다.
하지만 개인이 여러 컴퓨터를 대상으로 테스트를 수행하기 어렵기 때문에, 전문가들이 테스트를 수행하여 공개하는 보고서를 참조하여 도움을 얻는 것이 효율적입니다.
아래 전문 성능비교 테스트 보고서가 시스템 선택에 도움이 될 것으로 생각합니다. 참고로, 당사는 기사를 제공하는 기관과 전혀 관련이 없음을 알려드립니다.
In this blog, Flow Science’s IT Manager Matthew Taylor breaks down the different hardware components and suggests some ideal configurations for getting the most out of your FLOW-3D products.
개요
본 자료는 Flow Science의 IT 매니저 Matthew Taylor가 작성한 자료를 기반으로 STI C&D에서 일부 자료를 보완한 자료입니다. 본 자료를 통해 FLOW-3D 사용자는 최상의 해석용 컴퓨터를 선택할 때 도움을 받을 수 있을 것으로 기대합니다.
수치해석을 하는 엔지니어들은 사용하는 컴퓨터의 성능에 무척 민감합니다. 그 이유는 수치해석을 하기 위해 여러 준비단계와 분석 시간들이 필요하지만 당연히 압도적으로 시간을 소모하는 것이 계산 시간이기 때문일 것입니다.
따라서 수치해석용 컴퓨터의 선정을 위해서 단위 시간당 시스템이 처리하는 작업의 수나 처리량, 응답시간, 평균 대기 시간 등의 요소를 복합적으로 검토하여 결정하게 됩니다.
또한 수치해석에 적합한 성능을 가진 컴퓨터를 선별하는 방법으로 CPU 계산 처리속도인 Flops/sec 성능도 중요하지만 수치해석을 수행할 때 방대한 계산 결과를 디스크에 저장하고, 해석결과를 분석할 때는 그래픽 성능도 크게 좌우하기 때문에 SSD 디스크와 그래픽카드에도 관심을 가져야 합니다.
FLOW SCIENCE, INC. 에서는 일반적인 FLOW-3D를 지원하는 최소 컴퓨터 사양과 O/S 플랫폼 가이드를 제시하지만, 도입 담당자의 경우, 최상의 조건에서 해석 업무를 수행해야 하기 때문에 가능하면 최고의 성능을 제공하는 해석용 장비 도입이 필요합니다. 이 자료는 2022년 현재 FLOW-3D 제품을 효과적으로 사용하기 위한 하드웨어 선택에 대해 사전에 검토되어야 할 내용들에 대해 자세히 설명합니다. 그리고 실행 중인 시뮬레이션 유형에 따라 다양한 구성에 대한 몇 가지 아이디어를 제공합니다.
CPU 최신 뉴스
2024년 04월 01일 기준
이미지 출처 : https://www.cpubenchmark.net/high_end_cpus.html
CPU의 선택
CPU는 전반적인 성능에 큰 영향을 미치며, 대부분의 경우 컴퓨터의 가장 중요한 구성 요소입니다. 그러나 데스크탑 프로세서를 구입할 때가 되면 Intel 과 AMD의 모델 번호와 사양을 이해하는 것이 어려워 보일 것입니다. 그리고, CPU 성능을 평가하는 방법에 의해 가장 좋은 CPU를 고른다고 해도 보드와, 메모리, 주변 Chip 등 여러가지 조건에 의해 성능이 달라질 수 있기 때문에 성능평가 결과를 기준으로 시스템을 구입할 경우, 단일 CPU나 부품으로 순위가 정해진 자료보다는 시스템 전체를 대상으로 평가한 순위표를 보고 선정하는 지혜가 필요합니다.
PassMark – CPU Mark
High End CPUs
Updated 31st of March 2024
수치해석을 수행하는 CPU의 경우 예산에 따라 Core가 많지 않은 CPU를 구매해야 하는 경우도 있을 수 있습니다. 보통 Core가 많다고 해석 속도가 선형으로 증가하지는 않으며, 해석 케이스에 따라 적정 Core수가 있습니다. 이 경우 예산에 맞는 성능 대비 최상의 코어 수가 있을 수 있기 때문에 Single thread Performance 도 매우 중요합니다. 아래 성능 도표를 참조하여 예산에 맞는 최적 CPU를 찾는데 도움을 받을 수 있습니다.
CPU 성능 분석 방법
부동소수점 계산을 하는 수치해석과 밀접한 Computer의 연산 성능 벤치마크 방법은 대표적으로 널리 사용되는 아래와 같은 방법이 있습니다.
FLOW-3D의 CFD 솔버 성능은 CPU의 부동 소수점 성능에 전적으로 좌우되기 때문에 계산 집약적인 프로그램입니다. 현재 출시된 사용 가능한 모든 CPU를 벤치마킹할 수는 없지만 상대적인 성능을 합리적으로 비교할 수는 있습니다.
특히, 수치해석 분야에서 주어진 CPU에 대해 FLOW-3D 성능을 추정하거나 여러 CPU 옵션 간의 성능을 비교하기 위한 최상의 옵션은 Standard Performance Evaluation Corporation의 SPEC CPU2017 벤치마크(현재까지 개발된 가장 최신 평가기준임)이며, 특히 SPECspeed 2017 Floating Point 결과가 CFD Solver 성능을 매우 잘 예측합니다.
이는 유료 벤치마크이므로 제공된 결과는 모든 CPU 테스트 결과를 제공하지 않습니다. 보통 제조사가 ASUS, Dell, Lenovo, HP, Huawei 정도의 제품에 대해 RAM이 많은 멀티 소켓 Intel Xeon 기계와 같은 값비싼 구성으로 된 장비 결과들을 제공합니다.
CPU 비교를 위한 또 다른 옵션은 Passmark Software의 CPU 벤치마크입니다. PerformanceTest 제품군은 유료 소프트웨어이지만 무료 평가판을 사용할 수 있습니다. 대부분의 CPU는 저렴한 옵션을 포함하여 나열됩니다. 부동 소수점 성능은 전체 벤치마크의 한 측면에 불과하지만 다양한 워크로드에서 전반적인 성능을 제대로 테스트합니다.
예산을 결정하고 해당 예산에 해당하는 CPU를 선택한 후에는 벤치마크를 사용하여 가격에 가장 적합한 성능을 결정할 수 있습니다.
다른 컴퓨터 시스템에서 컴퓨팅 계산에 대한 집약적인 워크로드를 비교하는데 사용할 수 있는 성능 측정을 제공하도록 설계된 SPEC CPU 2017에는 SPECspeed 2017 정수, SPECspeed 2017 부동 소수점, SPECrate 2017 정수 및 SPECrate 2017 부동 소수점의 4 가지 제품군으로 구성된 43 개의 벤치 마크가 포함되어 있습니다. SPEC CPU 2017에는 에너지 소비 측정을 위한 선택적 메트릭도 포함되어 있습니다.
<SPEC CPU 벤치마크 보고서>
벤치마크 결과보고서는 제조사별, 모델별로 테스트한 결과를 아래 사이트에 가면 볼 수 있습니다.
Designed to provide performance measurements that can be used to compare compute-intensive workloads on different computer systems, SPEC CPU 2017 contains 43 benchmarks organized into four suites: SPECspeed 2017 Integer, SPECspeed 2017 Floating Point, SPECrate 2017 Integer, and SPECrate 2017 Floating Point. SPEC CPU 2017 also includes an optional metric for measuring energy consumption.
일반적으로 클럭 속도가 높은 칩은 CPU 코어를 더 적게 포함합니다. FLOW-3D는 병렬화가 잘되어 있지만, 디스크 쓰기와 같이 일부 작업은 기본적으로 단일 스레드 방식으로 수행됩니다. 따라서 데이터 출력이 빈번하거나 큰 시뮬레이션은 종종 더 많은 코어가 아닌, 더 높은 클럭 속도를 활용합니다. 마찬가지로 코어 및 소켓의 다중 스레딩은 오버헤드를 발생시키므로 작은 문제의 해석일 경우 사용되는 코어 수를 제한하면 성능이 향상될 수 있습니다.
CPU 아키텍처
CPU 아키텍처는 중요합니다. 최신 CPU는 일반적으로 사이클당 더 많은 기능을 제공합니다. 즉, 현재 세대의 CPU는 일반적으로 동일한 클럭 속도에서 이전 CPU보다 성능이 우수합니다. 또한 전력 효율이 높아져 와트당 성능이 향상될 수 있습니다. Flow Science에는 구형 멀티 소켓 12, 16, 24 코어 Xeon보다 성능이 뛰어난 최근 세대 10~12 Core i9 CPU 시스템을 보유하고 있습니다.
오버클럭
해석용 장비에서는 CPU를 오버클럭 하지 않는 것이 좋습니다. 하드웨어를 다년간의 투자라고 생각한다면, 오버클럭화는 발열을 증가시켜 수명을 단축시킵니다. CPU에 따라 안정성도 저하될 수 있습니다. CPU를 오버클럭 할 때는 세심한 열 관리가 권장됩니다.
하이퍼스레딩
<이미지출처:https://gameabout.com/krum3/4586040>
하이퍼스레딩은 물리적으로 1개의 CPU를 가상으로 2개의 CPU처럼 작동하게 하는 기술로 파이프라인의 단계수가 많고 각 단계의 길이가 짧을때 유리합니다. 다만 수치해석 처럼 모든 코어의 CPU를 100% 사용중인 장시간 수행 시뮬레이션은 일반적으로 Hyper Threading이 비활성화 된 상태에서 더 잘 수행됩니다. FLOW-3D는 100% CPU 사용률이 일반적이므로 새 하드웨어를 구성할 때 Hyper Threading을 비활성화하는 것이 좋습니다. 설정은 시스템의 BIOS 설정에서 수행합니다.
몇 가지 워크로드의 경우에는 Hyper Threading을 사용하여 약간 더 나은 성능을 보이는 경우가 있습니다. 따라서, 최상의 런타임을 위해서는 두 가지 구성중에서 어느 구성이 더 적합한지 시뮬레이션 유형을 테스트하는 것이 좋습니다.
스케일링
여러 코어를 사용할 때 성능은 선형적이지 않습니다. 예를 들어 12 코어 CPU에서 24 코어 CPU로 업그레이드해도 시뮬레이션 런타임이 절반으로 줄어들지 않습니다. 시뮬레이션 유형에 따라 16~32개 이상의 CPU 코어를 선택할 때는 FLOW-3D 및 FLOW-3D CAST의 HPC 버전을 사용하거나 FLOW-3D CLOUD로 이동하는 것을 고려하여야 합니다.
AMD Ryzen 또는 Epyc CPU
AMD는 일부 CPU로 벤치마크 차트를 석권하고 있으며 그 가격은 매우 경쟁력이 있습니다. FLOW SCIENCE, INC. 에서는 소수의 AMD CPU로 FLOW-3D를 테스트했습니다. 현재 Epyc CPU는 이상적이지 않고 Ryzen은 성능이 상당히 우수합니다. 발열은 여전히 신중하게 다뤄져야 할 문제입니다.
FLOW-3D는 OpenGL 드라이버가 만족스럽게 수행되는 최신 그래픽 카드가 필요합니다. 최소한 OpenGL 3.0을 지원하는 것이 좋습니다. 권장 옵션은 엔비디아의 쿼드로 K 시리즈와 AMD의 파이어 프로 W 시리즈입니다.
특히 엔비디아 쿼드로(NVIDIA Quadro)는 엔비디아가 개발한 전문가 용도(워크스테이션)의 그래픽 카드입니다. 일반적으로 지포스 그래픽 카드가 게이밍에 초점이 맞춰져 있지만, 쿼드로는 다양한 산업 분야의 전문가가 필요로 하는 영역에 광범위한 용도로 사용되고 있습니다. 주로 산업계의 그래픽 디자인 분야, 영상 콘텐츠 제작 분야, 엔지니어링 설계 분야, 과학 분야, 의료 분석 분야 등의 전문가 작업용으로 사용되고 있습니다. 따라서 일반적인 소비자를 대상으로 하는 지포스 그래픽 카드와는 다르계 산업계에 포커스 되어 있으며 가격이 매우 비싸서 도입시 예산을 고려해야 합니다.
유의할 점은 엔비디아의 GTX 게이밍 하드웨어는 볼륨 렌더링의 속도가 느리거나 오동작 등 몇 가지 제한 사항이 있습니다. 일반적으로 노트북에 내장된 통합 그래픽 카드보다는 개별 그래픽 카드를 강력하게 추천합니다. 최소한 그래픽 메모리는 512MB 이상을 권장합니다.
Flow Science는 nVidia 드라이버 버전이 341.05 이상인 nVidia Quadro K, M 또는 P 시리즈 그래픽 하드웨어를 권장합니다. 이 카드와 드라이버 조합을 사용하면 원격 데스크톱 연결이 완전한 3D 가속 기능을 갖춘 기본 하드웨어에서 자동으로 실행됩니다.
원격 데스크톱 세션에 연결할 때 nVidia Quadro 그래픽 카드가 설치되어 있지 않으면 Windows는 소프트웨어 렌더링을 사용합니다. FLOW-3D 가 소프트웨어 렌더링을 사용하고 있는지 확인하려면 FLOW-3D 도움말 메뉴에서 정보를 선택하십시오. GDI Generic을 소프트웨어 렌더링으로 사용하는 경우 GL_RENDERER 항목에 표시됩니다.
하드웨어 렌더링을 활성화하는 몇 가지 옵션이 있습니다. 쉬운 방법 중 하나는 실제 콘솔에서 FLOW-3D를 시작한 다음 원격 데스크톱 세션을 연결하는 것입니다. Nice Software DCV 와 같은 일부 VNC 소프트웨어는 기본적으로 하드웨어 렌더링을 사용합니다.
RAM 고려 사항
프로세서 코어당 최소 4GB의 RAM은 FLOW-3D의 좋은 출발입니다. POST Processor를 사용하여 후처리 작업을 할 경우 충분한 양의 RAM을 사용하는 것이 좋습니다.
현재 주력제품인 DDR4보다 2배 빠른 DDR5가 곧 출시된다는 소식도 있습니다.
일반적으로 FLOW-3D를 이용하여 해석을 할 경우 격자(Mesh)수에 따라 소요되는 적정 메모리 크기는 아래와 같습니다.페이지 보기
초대형 (2억개 이상의 셀) : 최소 128GB
대형 (60 ~ 1억 5천만 셀) : 64 ~ 128GB
중간 (30-60백만 셀) : 32-64GB
작음 (3 천만 셀 이하) : 최소 32GB
HDD 고려 사항
수치해석은 해석결과 파일의 데이터 양이 매우 크기 때문에 읽고 쓰는데, 속도면에서 매우 빠른 SSD를 적용하면 성능면에서 큰 도움이 됩니다. 다만 SSD 가격이 비싸서 가성비 측면을 고려하여 적정수준에서 결정이 필요합니다.
CPU와 저장장치 간 데이터가 오고 가는 통로가 그림과 같이 3가지 방식이 있습니다. 이를 인터페이스라 부르며 SSD는 흔히 PCI-Express 와 SATA 통로를 이용합니다.
흔히 말하는 NVMe는 PCI-Express3.0 지원 SSD의 경우 SSD에 최적화된 NVMe (NonVolatile Memory Express) 전송 프로토콜을 사용합니다. 주의할 점은 MVMe중에서 SATA3 방식도 있기 때문에 잘 구별하여 구입하시기 바랍니다.
그리고 SSD를 선택할 경우에도 SSD 종류 중에서 PCI Express 타입은 매우 빠르고 가격이 고가였지만 최근에는 많이 저렴해졌습니다. 따라서 예산 범위내에서 NVMe SSD등 가장 효과적인 선택을 하는 것이 좋습니다. ( 참고 :해석용 컴퓨터 SSD 고르기 참조 )
기존의 물리적인 하드 디스크의 경우, 디스크에 기록된 데이터를 읽기 위해서는 데이터를 읽어내는 헤드(바늘)가 물리적으로 데이터가 기록된 위치까지 이동해야 하므로 이동에 일정한 시간이 소요됩니다. (이러한 시간을 지연시간, 혹은 레이턴시 등으로 부름) 따라서 하드 디스크의 경우 데이터를 읽기 위한 요청이 주어진 뒤에 데이터를 실제로 읽기까지 일정한 시간이 소요되는데, 이 시간을 일정한 한계(약 10ms)이하로 줄이는 것이 불가능에 가까우며, 데이터가 플래터에 실제 기록된 위치에 따라서 이러한 데이터에의 접근시간 역시 차이가 나게 됩니다.
하지만 HDD의 최대 강점은 가격대비 용량입니다. 현재 상용화되어 판매하는 대용량 HDD는 12TB ~ 15TB가 공급되고 있으며, 이는 데이터 저장이나 백업용으로 가장 좋은 선택이 됩니다. 결론적으로 데이터를 직접 읽고 쓰는 드라이브는 SSD를 사용하고 보관하는 용도의 드라이브는 기존의 HDD를 사용하는 방법이 효과적인 선택이 될 수 있습니다.
상기 벤치마크 테스트는 테스트 조건에 따라 그 성능 곡선이 달라질 수 있기 때문에 조건을 확인할 필요가 있습니다. 예를 들어 Windows7, windows8, windows10 , windows11 모두에서 테스트한 결과를 평균한 점수와 자신이 사용할 컴퓨터 O/S에서 테스트한 결과는 다를 수 있습니다. 상기 결과에 대한 테스트 환경에 대한 내용은 아래 사이트를 참고하시기 바랍니다.
하부 테일워터를 위한 USBR 및 쐐기형 배플 블록 분지를 사용한 유압 점프의 수치적 조사
Muhammad Waqas Zaffar; Ishtiaq Hassan; Zulfiqar Ali; Kaleem Sarwar; Muhammad Hassan; Muhammad Taimoor Mustafa; Faizan Ahmed Waris
Abstract
The stilling basin of the Taunsa barrage is a modified form of the United States Bureau of Reclamation (USBR) Type-III basin, which consists of baffle and friction blocks. Studies revealed uprooting of baffle blocks due to their vertical face. Additionally, the literature highlighted issues of rectangular face baffle blocks: less drag, smaller wake area, and flow reattachment. In contrast, the use of wedge-shaped baffle blocks (WSBBs) is limited downstream of open-channel flows. Therefore, this study developed numerical models to investigate the effects of USBR and WSBB basins on the hydraulic jump (HJ) downstream of the Taunsa barrage under lower tailwater conditions. Surface profiles in WSBB and modified USBR basins showed agreement with previous studies, for which the coefficient of determination (R2) reached 0.980 and 0.970, respectively. The HJ efficiencies reached 57.9 and 58.6% in WSBB and modified USBR basins, respectively. The results of sequent depths, roller length, and velocity profiles in the WSBB basin were found more promising than the modified USBR basin, which further confirmed the suitability of the WSBB basin for barrages. Furthermore, WSBB improved flow behaviors in the basin, which showed no fluid reattachment on the sides of WSBB, increased wake regions, and decreased turbulent kinetic energies.
Barrages in Pakistan were built about 50–100 years ago and play an important role in the economy. However, as time passed, the stability of these barrages was compromised due to hydraulic and structural deficiencies (Zaidi et al. 2011). Similarly, Taunsa Barrage Punjab, on the mighty river Indus, is one of the major hydraulic structures, which was constructed about 65 years ago. The barrage was designed for a design discharge capacity of 28,313 m3/s, and its stilling basin was a modified form of the United States Bureau of Reclamation (USBR) Type-III basin that consisted of USBR impact friction and baffle blocks. These arrangements dissipate excessive kinetic energy, enhance turbulence, kill rollers, and stabilize the hydraulic jump (HJ) even in case of less tailwater depth. The barrage consists of 64 bays, and the total width of the barrage between the abatements is 1,324.60 m. Of total width, 1,176.5 m is a clear waterway (Zaffar & Hassan 2023a). Figure 1 shows the typical cross-section of the Taunsa barrage.
Figure 1
Typical cross-section of the Taunsa barrage.
Soon after the barrage operation in 1958, multiple problems occurred on the barrage downstream, such as uprooting of the impact baffle blocks due to their vertical face, damage to the basin’s floor, lowering of tailwater levels, and bed retrogression (Zulfiqar & Kaleem 2015). During 1959–1962, repair works were carried out to cater to these issues, but the problems remained persistent. To resolve these issues, the Punjab Government constituted a committee of experts in 1966 and 1973, but no specific measures were taken, and the issues continued to aggravate (Zaidi et al. (2004). Additionally, these traditional impact blocks also face flow reattachment on the sides that decreases the drag force (Frizell & Svoboda 2012). On the contrary, after investigating the wedge-shaped baffle blocks (WSBBs) downstream of pipe outlets, research scholars (Pillai et al. 1989; Verma & Goel 2003; Verma et al. 2004; Goel 2007; Goel 2008; Tiwari et al. 2010) reported that these blocks increased the energy dissipation and created more eddies and wake regions on either side. These studies further mentioned that upon the use of WSBBs, the overall length of stilling basins was also reduced from 15 to 25%.
Energy dissipation is the most common issue faced in the design of hydraulic structures. The kinetic energy typically comes from the upstream of the dams (El Baradei et al. 2022), spillways (Sutopo et al. 2022), chute, sluice gates, and weirs, which are further induced by the HJ and its turbulent structure (Elsaeed et al. 2016). In the HJ, flow suddenly changes from supercritical to subcritical conditions which dissipate the energy of the upstream flow, thereby saving the hydraulic structures from damage. The HJ occurs when Froude Number (Fr) falls below unity, which is the ratio of inertia to gravitational forces that can be calculated by the following equation (Bayon-Barrachina & Lopez-Jimenez 2015; Bayon-Barrachina et al. 2015).
(1)
where v, h, and g are stream-wise velocities, flow depth, and acceleration due to gravity, respectively. Hager & Sinniger (1985) investigated characteristics of the HJ for abrupt changes in horizontal bed and proposed the following equation to compute the efficiency of HJs.
(2)
where Fr1 is the Froude number in the supercritical flow before the HJ.
Bakhmeteff & Matzke (1936) developed HJ similarity models and proposed dimensionless Equation (3) for the free surface profile of HJs.
(3)
where is the water depth at x (hi) and the variable X is the dimensionless longitudinal coordinates (x), as shown in the following dimensionless equations, respectively.
(4)
(5)
where D is the gate opening and h1 and h2 are the water depths in supercritical and subcritical regions, respectively. X1 and X2 are the functions of variable X, and their values can be calculated at the toe of the HJ and the end of the roller region, respectively. The components of Equations (4) and (5) are shown in Figure 2.
Figure 2
Schematic diagram for the dimensionless free surface profiles of HJs.
Habibzadeh et al. (2012) conducted experiments to investigate the role of baffle blocks for submerged HJs and energy dissipation downstream of low-head hydraulic structures. Chachereau & Chanson (2011) and Wang & Chanson (2015) investigated free surface profiles and turbulent fluctuation within the HJ for a wide range of initial Froude number (Fr1). The velocity profiles showed a wall jet-like profile, and turbulence intensities were high due to the fluctuations at the free surface. Maleki & Fiorotto (2021) developed a semi-empirical method to investigate HJs on a rough bed. The results showed that the characteristic length scale was linearly changing with Fr1. Macián-Pérez et al. (2020b) carried out experiments on the USBR-II stilling basin to investigate the characteristics of HJs. The results of sequent depths and HJ efficiency agreed with the experimental studies, which reached 99.2 and 97%, respectively. The results of the dimensionless free surface profile also agreed with the previous studies for which the value of the coefficient of determination (R2) reached 0.979. Murzyn & Chanson (2009) conducted experiments for a wide range of Fr1 up to 8.3 to investigate bubbly and turbulence structure within the HJ. The results showed that void fraction (Cmax) and bubble frequency (Fmax) were found in the developing region, and vertical interfacial velocity agreed with the wall jet-like profile. Qasim et al. (2022) conducted experiments on bed discordance downstream of different weirs. The results indicated that as the bed discordance increased, the dimensionless flow depth decreased downstream of the discordance, which increased the Froude number. The results further showed that as the configuration of bed discordance was changed, the free surface profiles were also changed, which affected the flow depths and velocity profiles. Bhosekar et al. (2014) conducted experiments to investigate the characteristics of discharge downstream of orifice spillways. The results showed that free surface profiles were not elliptical due to the flat curve near the gate opening. The results further indicated that the flat curve developed a negative pressure on roof profiles, which reduced the discharge capacity of the spillway.
To stabilize the HJ in the stilling basins, different shapes of baffle blocks are employed, i.e., baffle blocks (Habibzadeh et al. 2012), friction blocks (Chaudary & Sarwar 2014), end sill (Mansour et al. 2004) and vertical sill (Alikhani et al. 2010), splitter blocks (Verma & Goel 2003), curved (Eloubaidy et al. 1999), T-shaped and triangular (Tiwari & Goel 2016), and WSBB (Pillai et al. 1989; Goel 2007; Goel 2008). These arrangements control the HJs in case of fewer tailwater depths (Peterka 1984) and minimize the erosion downstream of structures (Zaffar et al. 2023). Sayyadi et al. (2022) investigated HJ characteristics for negative steps in the stilling basin. The results showed that the negative step increased the energy dissipation up to 11%. Pillai et al. (1989) compared three different stilling basins for the Fr1 up to 4.5. The results showed that the stilling basin with the WSBB reduced the scour and overall length of the basin. Goel (2008), Goel (2007), and Tiwari et al. (2010) conducted experiments to investigate HJ characteristics downstream of square and circular pipe outlets using WSBBs. The results showed that as compared to the impact USBR-VI basin, the WSBB basin spread the fluid efficiently in the lateral direction and reduced the basin length up to 50%.
In the former section, the experimental studies on HJs, velocity distribution, free surface profiles, and turbulent kinetic energy (TKE) are discussed, which could be assisted by Computational Fluid Dynamics (CFD) models (Ghaderi et al. 2020). Furthermore, over these hydraulic structures, the flow is very complex and associated with secondary currents, which characterized it as highly turbulent in all directions. Hence, using laboratory and field experiments, it is hard to accurately measure the free surface profile, velocities, secondary currents, and TKE over these hydraulic structures (Jothiprakash et al. 2015). Furthermore, physical experiments and on-site measurements are usually expensive and time-consuming. In contrast, the improvements in computational speed, storage, and turbulence modeling have made CFD a viable complementary investigation tool for hydraulic modeling (Ghaderi et al. 2021). Consequently, the use of numerical modeling tools such as Open Foam (Bayon-Barrachina & Lopez-Jimenez 2015), ANSYS Fluent (Aydogdu et al. 2022), and FLOW-3D (Hirt & Sicilian 1985) has become prevalent to get hydraulic characteristics of grade-control structures. Such modeling tools are helpful, especially when the basic fundamental equations are unable to provide desired outputs, like in the case of multifaceted geometries (Herrera-Granados & Kostecki 2016). So far, many researchers have employed numerical models in the hydraulic investigations of HJs and energy dissipation, but only a few of the latest studies are highlighted here. FLOW-3D numerical models were employed to investigate the HJ (Zaffar & Hassan 2023a) and baffle blocks (Zaffar & Hassan 2023b) for different stilling basins of the Taunsa barrage. These studies focused on velocity distribution, TKE, free surface profiles, energy loss energy, and the effects of baffle blocks on the HJ characteristics. Macián-Pérez et al. (2020a) carried out a numerical investigation on a high Reynolds of 210,000 to study the HJ characteristics. Upon comparison, the FLOW-3D model showed 93% accuracy in the roller length of HJs. The results also indicated 94.2 and 94.3% accuracy for sequent depths and HJ efficiency, respectively. Nikmehr & Aminpour (2020) examined the HJ characteristics on rough beds using FLOW-3D, and compared results with the experiments. The results indicated that roughness height and its distance affected the HJ length. Gadge et al. (2018) conducted a numerical study to investigate the impact of roof profiles on the discharge capacity of orifice spillways and validated the models with experimental results. The study revealed that in addition to the pond level and height of orifice (d), the bottom and roof profiles also affected the discharge coefficient (Cd).
From the literature review, it is found that only a few studies are conducted on the flow characteristics downstream of the Taunsa barrage (Zaidi et al. 2004, 2011; Chaudhry 2010). These studies were carried out in laboratory flume and investigated the effects of tailwater on the location of HJs. However, the studies were lacking in providing the data for other essential hydraulic parameters, i.e., velocity distribution, free surface profiles, TKE, and relative energy loss in the stilling basin. On the contrary, the literature has revealed many experimental and numerical studies on different shapes of baffle blocks downstream of open-channel flow, but the use of WSBB downstream of river diversion barrage is found limited. In the previous studies (Pillai et al. 1989; Verma & Goel 2003; Verma et al. 2004; Goel 2008, 2007; Tiwari et al. 2010), these blocks have only been tested downstream of pipe outlet basins for the initial Froude number of 4.5. Therefore, in the present study, FLOW-3D numerical models are developed to investigate the effects of presently available USBR baffle blocks in the stilling basin of the Tuansa barrage. Due to the uprooting problems of these blocks, the study also investigates the suitability of WSBBs downstream of the studied barrage and draws a comparison between the results of modified USBR and WSBB basins. In this study, based on results from the literature, WSBB with a vertex angle of 150° and cutback angle of 90° is applied for Fr1 up to 6.64. The main objective of this study is to investigate HJs and flow behavior with USBR baffle blocks and WSBB downstream of an investigated barrage at 44 m3/s discharge. At 44 m3/s discharge, the numerical models are operated at the minimum tailwater level of 129.10 m, and investigated free surface profiles, sequent depths, roller lengths, HJ efficiency, velocity profile, and TKE in the two different stilling basins.
MATERIALS AND METHODS
Existing and proposed stilling basins, appurtenances
The present numerical models are developed downstream of the Taunsa barrage, Pakistan. The stilling basin of the barrage includes USBR impact friction and baffle blocks (Zaffar & Hassan 2023b; Zaffar et al. 2023). In the basin, floor level and weir crest are fixed at 126.79 and 130.44 m, respectively. The slopes of upstream and downstream weir glacis are maintained at 1:3 and 1:4 (H:V), respectively. In both the studied basins, the blocks are installed 14.63 m away from the centerline of the crest and are placed in a staggered position. The overall length and height of the USBR blocks is 1.37 m as shown in Figures 3(a) and 3(b). Additionally, between the two staggered rows of baffle blocks, a 1.37-m distance is maintained, while the top width of all the USBR blocks is 0.46 m, which is angled at 45° from the rear side. On the other hand, in the WSBB basin, WSBB is placed at the locations of impact USBR baffle blocks. Furthermore, in both the studied basins, two staggered rows of friction blocks are also placed at the basin’s end about 28.95 m away from the weir’s crest. These friction blocks are 1.37 m long, 1.22 m wide, and 1.37 m high. The top surface of these blocks is identical to their bottom. The overall length, width, and height of the WSBB are kept at 1.37 m, and a detailed geometry of the investigated WSBB can be seen in Figures 3(c) and 3(d). Currently, for the investigated WSBB, a vertex angle of 150° and a cutback angle of 90° are employed.
Figure 3
Baffle block geometry for the basins: (a) top view of USBR baffle blocks, (b) isometric view representing front of USBR baffle blocks, (c) top view of WSBBs and (d) isometric view representing front of WSBBs.
Numerical model implementation
Environmental flows are governed by the laws of physics and represented by Navier–Stokes Equations (NSEs), which are inherently nonlinear, time-dependent, and contain three-dimensional partial deferential schemes (Viti et al. 2018). These partial differential equations explain the procedures of continuity, momentum, heat, and mass transfer. For one- and two-dimensional models, these equations can be solved analytically, while for the solution of three-dimensional models, CFD models are employed to discretize the NSEs. In these models, flow equations, i.e., NSEs and continuity equations, are discretized in each cell. Generally, these models start with a mesh, which further contains multiple interconnected cells in the employed mesh blocks. These meshes subdivide the physical space into small volumes, which are associated with several nodes. The values of unknown parameters are stored on these nodes, such as velocity, temperature, and pressure. Different numerical techniques are available to discretize the NSEs, i.e., Direct Numerical Simulation (DNS) (Jothiprakash et al. 2015), Large Eddy Simulation (LES) (Ghosal & Moin 1995), and Reynold Averaged Navier–Stokes (RANS) Equation (Kamath et al. 2019). However, as compared to DNS and LES models, due to less computation cost and simulation time, the RANS model is frequently used in river and hydraulic investigations. Using the RANS model, two additional variables are generated, for which turbulence closure models are usually employed (Carvalho et al. 2008). These models find closure by averaging the Reynolds stress terms in NSEs and append additional variables for turbulent viscosity and transport equations.
Presently, FLOW-3D models are developed to investigate the effects of different shapes of baffle blocks on HJ downstream of the river diversion barrage. The models employ RANS equations to solve algorithms and equations of incompressible fluid in each computational cell. To further address the additional terms, i.e., Reynolds stresses and turbulent viscosity, the Renormalization group (RNG K–ɛ) method is applied. For the discretization of RANS and other algorithms, at present, the Volume of Fluid (VOF) method (finite volume method (FVM)) is employed, while the equations of the controlled volume are formulated with area and volume porosity functions. This formulation is called the ‘Fractional Area/Volume Obstacle Representation’ (FAVOR) method (Hirt & Sicilian 1985). The proceeding section describes the equations used for the present models.
Assuming the flow is steady and neglecting the fluctuation of specific weight ( = 0), Equations (6) and (7) are used for the turbulent flow (Carvalho et al. 2008).
(6)
(7)
Following the above equations, it is apparent that these relationships consist of three momentum and one continuity equation, in which there are 10 unknowns (p, u, v, w, and six Reynolds stress components). In the present study, the flow is considered incompressible, which implies the following equation to solve the flow domain (Viti et al. 2018).
(8)
In Equations (6)–(8) u, v, and w, are velocity components in x, y, and z directions, respectively. and p are total pressure and fluid density while the terms are known as the Reynolds stresses. Ax, Ay, and Az are flow areas while R,, and RSOR are the model’s coefficient, flow generic property, and mass source term, respectively.
Turbulence modeling and free surface tracking
Six turbulence models are available in FLOW-3D, which employs numerous equations to solve the closure problems. Among various models, the two-equation turbulence models such as standard K–ɛ (Bradshaw 1997, RNG K–ɛ (Yakhot et al. 1991), and K–ω (Wilcox 2008) are widely used in hydraulic investigations.
The standard K–ɛ and RNG K–ɛ models solve transport equations for TKE and its dissipation. The formulation of both models is identical, but the former derives model coefficients empirically. However, RNG K–ɛ applies a statistical approach to derive the transport equations explicitly, which has shown better capability in low-turbulence and high-shear regions (Macián-Pérez et al. 2020a, 2020b) than the standard K–ɛ model. In contrast, the K–ω model was implemented for stream-wise pressure gradient and near-wall boundaries (Wilcox 2008). The model replaces turbulent dissipation rate with turbulent frequency. In the K–ω model, the values of TKE and turbulent frequency are specified at the inlet boundaries. The above-mentioned turbulence models were investigated by Macián-Pérez et al. (2020b) for flow behaviors in the USBR-II stilling basin and the study indicated that the RNG K–ɛ model showed better accuracy. Also, RNG K–ɛ (Macián-Pérez et al. 2020a) showed more promising results for a grid convergence index (GCI), free surface, roller lengths, and HJ efficiency. Additionally, the RNG K–ɛ model predicted good results in flow rate, free surface profiles, and velocity profiles. Based on bibliographical results, this study also employed the RNG K–ɛ model, for which the following transport equations (Equations (9) and (10)) are utilized for TKE (K) and its dissipation (ɛ), respectively (Macián-Pérez et al. 2020a).
(9)
(10)
where is the coordinate in the x direction; is the dynamic viscosity; is the turbulent dynamic viscosity; K is the turbulent kinetic energy; is the turbulent dissipation; is the fluid density, and is the production of TKE. Finally, the terms, , , and are model parameters whose values are given in Yakhot et al. (1991).
For free surface modeling, the VOF method is employed. VOF applies an additional variable, which is called fraction of fluid (F), in which F represents the proportion of fluid. To compute F in the domain, the following equation (Hirt & Sicilian 1985) is used:
(11)
In FLOW-3D, the fluid fraction (F) in each cell is usually presented by three possibilities:
(A)F = 0, cell is empty.
(B)F = 1, a cell is fully occupied by fluid.
(C)0 < F < 1, cell represents the surface between the two fluids.
One fluid (water) with a free surface is considered in the present models, for which FLOW-3D automatically selects the free surface method from the availableVOF advection scheme. For the free surface tracking, 0.5 value is assigned in each computation cell.
Pressure velocity coupling
One of the major issues in solving the NSEs is pressure–velocity coupling, and for that, a network of algorithms (SIMPLE (Patankar & Spalding 1972) and PISO (ISSA 1985)) has been developed. These above-mentioned algorithms use under- and over-relaxing factors for pressure correction in the continuity and momentum equations, which contain large memory. Additionally, due to the relaxation factors, sometimes the solution becomes unstable and does not find convergence. On the contrary, FLOW-3D employs the Generalized Minimum Residual Method (GMRES) (Joubert 1994) because it possesses good convergence, high speed, and uses less memory. Additionally, GMRES does not apply any relaxation factor and possesses an additional algorithm, ‘Generalized Minimum Residual Solver (GCG),’ to treat the viscous terms.
The solid geometry of the barrage bay was designed in AutoCAD and converted into a stereo lithography file. Before importing the stereo lithography file into FLOW-3D, it was tested in Netfabb-basic software to remove any holes, facets, and boundary edges. Figures 4(a) and 4(b) show stereo lithography files used for the studied models.
Figure 4
Geometry details of the studied stilling basins: (a) modified USBR baffle block basin and (b) WSBB basin.
Meshing and boundary condition
The structured rectangular hexahedral mesh was employed to resolve the geometry and flow domain. To resolve the flow domain, a coarse mesh block was initiated upstream of the barrage (Xmin = 10 m) which ended at the upstream side of the gate (Xmax = 32.90 m). However, the fine mesh block was started from Xmin = 32.90 m which was extended up to the basin’s end (Xmax = 71 m). In total, 60 m of the model domain was simulated, out of which 22.90 m comprised upstream while the remaining included downstream side. For discharge measurement, mesh sensitivity analysis was performed. The blocks with coarse meshes were initially employed to calculate the volume flow rate (Q). The total number of mesh cells used in the coarse mesh block was 1,108,705. Out of 1,108,705 mesh cells, 481,000 cells were employed for the upstream block, while 627,705 mesh cells were utilized for the downstream mesh block. On the contrary, upon the use of fine mesh blocks, a total of 2,991,820 mesh cells were employed, out of which 481,000 cells were contained in the upstream mesh block while 2,510,820 cells were employed on the downstream side. Figure 5 shows mesh grids employed for flow and solid domains.
Figure 5
Meshing setup of the modeling domain.
It is essential to mention that in both meshing scenarios, fine mesh blocks were used on the downstream side of the bay because the focus of the present investigation was made around the baffle blocks and in the HJ regions. The details of mesh cell size and mesh quality indicators for the various mesh blocks are provided in Tables 1 and 2, respectively. Notably, except for discharge analysis, the results of other hydraulic parameters are produced from the fine meshing.
Table 1
Details of mesh blocks and cell sizes
Mesh block
Number of cells
Maximum adjacent ratio
Maximum aspect ratio
Block-1
X = 196; Y = 65; Z = 37
X
Y
Z
X–Y
Y–Z
Z–X
1.0
1.0
1.0
1.999
1.0
1.993
Block-2
X = 261; Y = 130; Z = 74
1.0
1.0
1.0
1.0
1.0
1.0
Table 2
Meshing quality indicators for various mesh blocks
Scenarios
Mesh block-1 (cell characteristics)
Mesh block-2 (cell characteristics)
Coarse meshing
Δx (m)
Δy (m)
Δz (m)
Δx (m)
Δy (m)
Δz (m)
0.142
0.284
0.284
0.142
0.284
0.284
Fine meshing
Δx (m)
Δy (m)
Δz (m)
Δx (m)
Δy (m)
Δz (m)
0.142
0.284
0.284
0.142
0.142
0.142
A vertical gate of 18.5-m width, 0.53-m length, and 6.10-m height was mounted upstream of the weir crest. Pond levels of 135.93 and 136.24 m were maintained for free and orifice flows, respectively. Table 3 shows the conditions used for models’ operation.
Table 3
Free and gated flow conditions for operation of the simulations (Zaffar & Hassan 2023b)
Discharge through barrage (m3/s)
Single bay discharge (m3/s)
Pond level (m)
Tailwater levels for jump formation (m)
Gate opening (m)
Turbulence model
28,313
444
135.93
133.80
Free flow
RNG K–ɛ
2,831
44
136.24
129.10
0.280
RNG K–ɛ
For the first mesh block, the upstream and downstream boundaries were set as pressure (P), while for the second block upstream boundary was set as symmetry (S). The lateral sides were set as rigid boundaries (W), and no-slip conditions were expressed as zero tangential and normal velocity (u=v=w = 0), where u, v, and w are the velocities in x, y, and z directions, respectively. These boundaries indicate a wall law velocity profile, which further expresses that the average velocity of turbulent flows is proportional to the logarithm of the distance from that point to the fluid boundary. For all variables (except pressure (P) (which was set to zero), upper boundaries (Zmax) were set as atmospheric pressure to allow water to null von Neumann. For both mesh blocks, the lower boundaries (Zmin) were set as walls.
For the present models, the stability and convergence at each iteration were checked by Courant number (Ghaderi et al. 2020), which affected the time steps from 0.06 to 0.0023 and 0.015 to 0.0025 for free and gated flow, respectively. It is worth mentioning here that for the free flow analysis of higher discharge such as 444 m3/s, the steady state solution can only be achieved by mass-averaged fluid kinetic energy (MAFKE) and volume flow rate (VFR) at the inlet and outlet boundaries. Therefore, the time at which the MAFKE and VFR reach the steady state is assigned as the simulation time (Ts) of models. Presently, VFRs at the inlet and outlet boundaries are considered as the stability and convergence indicators. Based on the criterion mentioned above, the present free and gated models achieved hydraulic stability at Ts = 60 s while the actual time (Ta) of models ranged between 30 and 48 h. However, to accommodate free surface fluctuations, the models were run for Ts = 80 s.
Models’ verification and validation
Analysis of design discharge
For performance assessment of the numerical models, He/Hd = 0.998 (Johnson & Savage 2006; Gadge et al. 2019; Zaffar & Hassan 2023a, 2023b) was implemented for free flow analysis, where He and Hd are effective and designed heads, respectively. This was the design discharge of the Taunsa barrage, for which the models were operated on the pond and tailwater levels of 135.93 and 133.8 m, respectively, as provided in Table 3.
Gated flow modeling
Computational discharge is of paramount importance in hydraulic modeling, and the following discharge formula is used for gated flow operations (Gadge et al. 2018). To model 44 m3/s of discharge, gate opening and designed head for orifice are set at D = 0.280 m and Hd = 136.24 m, respectively. Figure 6 illustrates the typical cross-section for gated flow operations.
(12)
where Q (m3/s) is discharged through the orifice opening, A (m2) is the area of the orifice, g (m/s2) is the acceleration due to gravity and hc is the centerline head (hc = Hd–D/2). The values of coefficients of discharge (Cd) used and simulated were 0.816 and 0.819, respectively, and were found well within the range of Cd values calculated by Bhosekar et al. (2014).
Figure 6
Cross-section showing gated flow through the orifice.
RESULTS AND DISCUSSION
Discharges and flow evolution
Figure 7 shows the time instant of flow evolution for designed discharge. At the start of the simulation, due to the inlet velocity, the free surface was found to be changed. However, when it reached a steady state, a fully developed flow on the downstream glacis was achieved, as shown in Figure 7. A stable free jump was observed at Ts = 60 s for free and gated flow, and the free surface on the downstream side was found to be stable with little fluctuation. The accuracy of FLOW-3D simulations was checked by comparing those discharges with designed values. At Ts = 80 s, the modified USBR baffle block basin underestimated the discharge, which reached 440.48 m3/s and displayed a 0.80% error. Similarly, at Ts = 80 s, for the WSBB basin, the model produced 440.17 m3/s discharge, in which the maximum error reached 0.893%. However, upon the use of coarse meshing, the errors in the computed discharge were increased, which reached −5 and −4% in modified USBR and WSBB basins, respectively. The free flow analysis of modified USBR and WSBB models showed acceptable validation with the designed flow, which allowed us to run the models for orifice discharge.
Figure 7
Flow evolution at the designed flow: USBR basin (a–c) and WSBB basin (d–f).
For the gated flow, on using fine meshing with similar meshing and boundary conditions, the modified USBR basin produced 44.14 m3/s of discharge, for which the maximum error reached 0.32%. However, in the WSBB basin, the numerical model underestimated the flow, which displayed only a 1.14% error. The evolutionary process of gated flows is shown in Figure 8. Based on the validation results of free and gated flows, further analysis was performed on the characteristics of HJs in the two different basins, and their results were compared with the relevant literature.
Figure 8
Evolutionary process of orifice discharge: USBR basin (a–c) and WSBB basin (d–f).
Free surface profiles
Due to the limited results of investigated hydraulic parameters on the studied barrage, the models’ results are compared with the previous relevant experimental and numerical studies. For such comparison, the models require some similarity in boundary and initial conditions, as obtained from Bayon-Barrachina & Lopez-Jimenez (2015) and Wang & Chanson (2015). Similar to the studies by Bayon-Barrachina & Lopez-Jimenez (2015) and Wang & Chanson (2015), in the present gated models, the upstream and downstream initial conditions are set to fluid elevation, i.e., pond and tailwater level, with hydrostatic pressure boundaries, while upstream boundaries are set to atmospheric pressure. Additionally, the sides and bottom are set to wall boundaries as described in Bayon-Barrachina & Lopez-Jimenez (2015) and Wang & Chanson (2015). However, the present models differ from the basin appurtenances as the compared studies have investigated HJ and other parameters on the horizontal flat beds. Furthermore, Bayon-Barrachina & Lopez-Jimenez (2015) and Wang & Chanson (2015) have investigated hydraulic parameters such as sequent depths, roller lengths, free surface profiles, energy dissipation, and TKE for Fr1 of 6.10 and 3.8 < Fr1 < 8.5, respectively. Similar to the above-mentioned study, the present modified USBR and WSBB basins are investigated for the Fr1 of 6.5 and 6.64, respectively. Hence, to confirm the results of free surface profiles, the studies of Bayon-Barrachina & Lopez-Jimenez (2015) and Wang & Chanson (2015) are utilized, for which the relevant discussion is made in the proceeding paragraphs.
Using Equation (3) of Bakhmeteff & Matzke (1936), the free surface profiles of HJs were obtained by the VOF method. Figure 9 compares the results of free surface profiles with Bayon-Barrachina & Lopez-Jimenez (2015) and Wang & Chanson (2015). The results of the present model agreed well with Bayon-Barrachina & Lopez-Jimenez (2015) (coefficient of determination (R2) = 0.992), for which the value of R2 in WSBB and USBR basins reached 0.980 and 0.970, respectively. Similarly, after comparing free surface profiles with Wang & Chanson (2015), the results of the present model were found to be more promising. However, as compared to the USBR basin, the results of free surface profiles in the WSBB basin showed more agreement with the compared studies, as can be seen in Figure 9.
Figure 9
Comparison of dimensionless free surface profiles of HJs with the literature.
To further assess the performance and gain deeper insight into the models’ efficiency, residual plots are drawn for the investigated hydraulic parameters. These errors referred to the difference between the observed (literature) and predicted data, which monitored the regression quality (Hassanpour et al. 2021). At a 5% level of significance, a homo-scedasticity analysis measured the residual errors, and the results of predicted residual errors were compared with the previous study.
Figure 10 compares the residual errors of free surface profiles of present models with Wang & Chanson (2015) and Bayon-Barrachina & Lopez-Jimenez (2015). Notably, the solid horizontal line in Figure 10 is the agreement line. From Figure 8(b), the maximum residual errors of free surface profiles in the experimental study of Wang and Chanson ranged between −0.2 and 0.13, while from the agreement line, the maximum negative and positive residual errors in the numerical study of Bayon-Barrachina & Lopez-Jimenez (2015) reached −0.12 to 0.20. In comparison to the previous studies, at the start of the regression line, the residual errors in the present models indicated a random scattered pattern below the agreement line, which further showed that the residual errors in the present and compared models were not normally distributed. After X = 0.3, above the agreement line, the distribution pattern of residual errors was also not normal. The stilling basin with USBR baffle blocks indicated the maximum negative and positive residual values of −0.160 to 0.10, respectively, while −0.184 to 0.124 values of residuals were noticed in the WSBB basin. Furthermore, from Figure 10, it is evident from residual analysis that the free profiles of HJs within modified USBR and WSBB basins have followed the trend of previous studies and the residuals of the present model are found less, which indicates reasonable accuracy of the models. Overall, the regression analysis of free surface profiles both for the present and compared studies revealed a curvilinear pattern that showed a heteroscedasticity residual.
Figure 10
Residual error diagram of free surface profiles of HJs with the literature.
Sequent depth ratio
In 1840, Belanger developed a famous equation for the sequent depth of HJs in smooth rectangular channels, which is widely used by numerous researchers to validate the results. Similarly, in the laboratory experimentation for different upstream and downstream tailwater water levels, Hager & Bremen (1989) developed a relationship of sequent depth (y2/y1) against a wide range of initial Froude numbers (Fr1). For the present models, 6.5 and 6.64 values of Fr1 are obtained in WSBB and USBR basins, respectively, and their relationship with sequent depths is developed as shown in Figure 11(a). The results of y2/y1 against Fr1 are compared with the experiments of other authors (Belanger 1841; Hager & Bremen 1989; Kucukali & Chanson 2008) and with the numerical study of Bayon-Barrachina & Lopez-Jimenez (2015).
Figure 11
(a) Comparison of sequent depth ratio with previous studies and (b) comparison of residual errors of sequent depths with the literature.
The sequent depths obtained from WSBB and modified USBR basins were 8.96 and 8.68, respectively, which were found to agree with the experimental results of other studies (Hager & Bremen 1989 and Belanger 1841), as shown in Figure 11(a). The results were also compared with the experiments of Kucukali & Chanson (2008), in which the value of Fr1 was 6.9. However, the Fr1 values obtained from present numerical models were 6.5 and 6.64 within WSBB and USBR basins, respectively. The comparison indicated that the present models overestimated the sequent depths, for which the errors reached 8.6 and 5.7% in WSBB and modified USBR basins, respectively. Furthermore, upon comparison with Bayon-Barrachina & Lopez-Jimenez (2015), results showed that present models underestimated the sequent depths, for which the maximum errors reached −13.2 and −9.8% errors in WSBB and USBR basins, respectively.
Figure 11(b) shows the comparison of residual errors of sequent depths with the previous studies. The maximum positive and negative residual values of 0.250 and −0.222 were found in WSBB and USBR baffle block stilling basins, respectively. The pattern of residual errors indicated an equal variance along the agreement line, which showed normal destitution of residual errors, thereby a homoscedastic pattern was noticed. The residual errors of sequent depths in both the tested basins were found within the ranges of Bayon-Barrachina & Lopez-Jimenez (2015) and Kucukali & Chanson (2008).
Roller length
Figure 12 compares the roller length (Lr/d1) of two different stilling basins with the previous studies, where Lr is the roller length of HJs and d1 is the initial flow depth before the HJ. Following Figure 10, the results showed that both the stilling basins produced almost similar roller lengths. Furthermore, in comparison to the previous studies, the relationship between roller lengths (Lr/d1) and Fr1 was found close to Kucukali and Chanson’s experiments (2008). However, the comparison with other studies indicated that the present models underestimated the roller lengths. The reason for reduced roller lengths was the effects of the basin’s appurtenance, which controlled the HJ lengths, as shown in Figures 13(a) and 13(b) (encircled regions). However, as compared to the modified USBR basin, the roller length in the WSBB basin was found to be less.
Figure 12
Comparison of roller lengths of HJ and initial Froude number with previous studies.
Figure 13
Roller lengths and energy dissipators: (a) WSBB basin and (b) modified USBR basin.
Figure 14 compares the residual errors for the roller lengths of present basins with the previous studies. The analysis showed a random distribution of residual errors and indicated homo-scedasticity of residuals. The results of residual errors for both the tested basins showed a good agreement with the compared studies and remained within the range of their residual errors. However, as compared to the USBR basin, the residual errors in the WSBB basin were found less, which reached −1.38.
Figure 14
Comparison of residual errors for the roller lengths with the previous studies.
HJ efficiency
The efficiency of the HJ is the ratio of energy loss to the upstream hydraulic head. Flow depth (hi), velocity (vi), and acceleration due to gravity (g) are the variables of HJ efficiency. The following equation was used to measure the efficiency of the HJ (Bayon-Barrachina & Lopez-Jimenez 2015).
(13)
where H1 and H2 are the specific energy heads upstream and downstream of HJs, respectively.
The results of numerical models showed 57.9 and 58.6% efficiencies in WSBB and modified USBR basins, respectively. The efficiencies for both the basins were also computed by Equation (2) and the results indicated 61.2 and 61.9% efficiencies for WSBB and modified USBR basins, respectively. The comparison further revealed that the present model underestimated the efficiencies, which reached the maximum errors of 5.41 and 5.45% in WSBB and modified USBR basins, respectively.
Figure 15(a) compares the efficiencies of present models with the previous studies. Upon comparing with Wu & Rajaratnam (1996), the results showed that the present model underestimated efficiencies for which the errors reached 6.6 and 5.54% in WSBB and modified USBR basins, respectively. Similarly, after comparing with Kucukali & Chanson (2008), the present model also showed a reduction in the HJ efficiencies for which the maximum errors reached 5.07 and 3.99% in WSBB and modified USBR basins, respectively. However, after comparing with Bayon-Barrachina & Lopez-Jimenez (2015), the results showed good agreement and indicated only 0.34 and 1.37% errors in WSBB and modified USBR basins, respectively. Overall, based on the bibliographic comparison, the overall accuracy of the present models for energy dissipation reached 93%.
Figure 15
(a) Comparison of HJ efficiency and (b) comparison of residual errors for the hydraulic jump efficiency.
Figure 15(b) indicates the residual errors of for modified USBR and WSBB basins and compares the errors with the previous experimental and numerical studies. Upon comparison with the modified USBR basin and with the literature, the HJ efficiency in the WSBB basin showed a close agreement with the zero residual line, for which the maximum error reached 0.001. On the other hand, the residual error in the modified USBR basin reached 0.003. Figure 14 also showed that the maximum residual error in the HJ efficiencies was found to be less than that was observed in previous studies, which remained within the limits of the compared studies (Wu & Rajaratnam 1996; Kucukali & Chanson 2008; Bayon-Barrachina & Lopez-Jimenez 2015).
Velocity distribution
Velocity distribution was measured at different flow depths to obtain vertical velocity profiles in two different stilling basins. Figure 16 shows the typical profiles in HJs, where (δ) is the y value at which the maximum velocity (Umax) occurs, while (b) is the length scale where u = 0.5Umax and ∂u/∂y < 0 (Ead & Rajaratnam 2002; Nasrabadi et al. 2012). The results indicated that in both basins, the velocity profiles showed a wall jet-like structure (Ead & Rajaratnam 2002). The results further showed that as the distance from the HJ-initiating locations was increased the maximum velocity decreased, thereby boundary growth layers were also decreased.
Figures 17 and 18 show that due to the supercritical velocity, a contracted jet was impinging near the beds of basins, and velocity decreased in upper fluid regions. The sections A-A and B-B in the basins indicated reverse flow and eddies in the HJs. The results of the upper fluid region of the HJ indicated typical backward velocity profiles as described by other authors (Ead & Rajaratnam 2002; Nasrabadi et al. 2012). Over time, these reverse fluid circulations were found to be stabilized and showed stagnation zones (Yamini et al. 2022). The analysis further showed a recirculation region within the HJ, and the maximum backward velocity profiles were found in the developed regions. The results also showed that after the jump termination, the negative velocity profiles converted into forward velocity profiles, as can be seen in sections C-C of Figures 17 and 18. Additionally, the results showed that after the WSBB and USBR baffle blocks, the velocity near the bed decreased and became positive at the free surface.
Figure 17
2D illustration of vertical velocity profiles in the USBR basin.
Figure 18
2D illustration of vertical velocity profiles in the WSBB basin.
Figure 19 shows the vertical velocity profile in the HJs at five different horizontal sections. The dimensionless plots between (y/b) and (U/Umax) illustrated that the velocity profiles followed the wall jet-like structure and were found to be agreed with Ead & Rajaratnam (2002), where y was the flow depth, b was the length scale, and Umax was the maximum velocity in the vertical section. The results also showed that in the HJ regions, both stilling basins produced identical structures of forward velocity profiles as can be seen in Figure 19(a) and 19(b).
From Figure 19, results showed that as the distance from the HJ toe increased, the vertical distance of Umax and inner layer thickness also increased. The analysis further indicated that as the distance from the initial location of HJs increased, the position of Umax was increased, which leveled off after the HJ, as can be seen in sections (C-C) of Figures 17 and 18. In both stilling basins, at X = 2 m, from the HJ initial location, the forward velocity profiles were found well agreed with the profile of Ead & Rajaratnam (2002) and the values of R2 reached 0.937 and 0.887 for WSBB and modified USBR basins, respectively, as shown in Figures 18(a) and 18(b), respectively. However, at X = 5.4 m, as compared to velocity (U/Umax = 0.36) in the modified USBR basin, the results showed less forward velocity (U/Umax = 0.21) in the WSBB basin at the upper fluid region.
Figure 20(a) shows the residual errors of velocity profiles in the WSBB basin. At x = 2 m, x = 3.2 m, and x = 4.3 m, the residual errors were found close to the agreement line. At the above-mentioned locations in the WSBB basin, the residual errors were also found less than Ead & Rajaratnam (2002). At x = 5.4 m in the WSBB basin, the maximum positive and negative residual errors of velocity profiles were 0.179 and −0.371, respectively, which were less than those that were found in the modified USBR basin. Figure 20(b) compares residual errors of velocity profiles in the modified USBR basin with the literature. It was evident from the residual diagrams that as the stream-wise distance from the jump-initiating location was increased, the maximum positive and negative errors also increased. The maximum positive and negative residual errors in the USBR basin were found at x = 5.4 m, which reached 1.139 and −1.352, respectively, and showed deviation from Ead & Rajaratnam (2002).
Figure 20
Comparison of residual errors with previous studies: (a) WSBB and (b) modified USBR basins.
Turbulent kinetic energy
TKE is the averaged velocity value in x, y, and z directions and describes energy dissipation at two different flow sections. The root mean square values of the velocity fluctuations are used to compute TKE. By considering the successive velocity values, the root mean square velocity (Urms) can be computed by the following equation (Gray et al. 2005).
(14)
where u1, u2, and u3 are successive velocities in the flow direction. Now, TKE can be calculated by the following equation.
(15)
where urms, vrms, and wrms are the root mean square velocities in x, y and z directions, respectively. Figure 21 shows the depth-wise (X–Y) TKE in a modified USBR basin. The results showed that the maximum TKE was found within and foreside of HJs while after the HJ, TKE continued to decline up to the end of the basin. Near the foreside of basins, flows were found to be strongly turbulent, dissipating most of the TKEs. Figure 19 shows the distribution of TKEs at seven vertical sections (X–Y), i.e., at Z = 0 m, 0.47 m, 0.93 m, 1.39 m, 1.85 m, 2.51 m, and free surface. At Z = 0 m, the maximum TKE in the USBR basin was found in the HJ region, which reached 0.30 J/kg, as shown in Figure 21(a). It is observed that behind the baffle blocks, TKEs were reduced due to eddies and fluid circulations, which dissipated the TKEs. Due to the impact of supercritical flows, a small number of eddies and fluid circulations were also noticed in front of the baffle blocks. The results showed that at the basin’s floor, TKEs traveled up to X = 21 m from the HJ-initiating location. Figures 21(b)–21(d) show that as the vertical distance from the basin’s floor was increased, the TKEs also increased, while their magnitude in the longitudinal direction was found to be reduced. The maximum TKEs at Z = 0.93 m, 1.39 m, and 1.85 m were 4.2, 4.5, and 3.6 J/kg, respectively. The results further indicated that as compared to the floor level, the TKEs from the central fluid depth to the free surface were found to be increased and their distribution in horizontal and lateral directions also increased. In the modified USBR basin, the maximum TKEs were noted at the toe of the HJ, which gradually reduced as the flow moved downstream, reaching 0.1 J/kg at the basin end, as shown in Figure 21(g).
Figure 21
Depth-wise distribution of TKEs in the USBR stilling basin at (a) Z = 0 m (floor level), (b) Z = 0.47 m, (c) Z = 0.93 m, (d) Z= 1.39 m, (e) 1.85 m, (f) Z= 2.51 m, and (g) free surface.
In the WSBB basin, at Z = 0 m (floor level), maximum TKEs reached 0.20 J/kg as shown in Figure 22(a). The results showed that as compared to USBR baffle blocks, the WSBBs were spreading the flow more efficiently in the lateral direction. Due to the spreading of fluid in the lateral direction, the results indicated that in the WSBB basin, the TKEs declined earlier in the basin and less energy was reached at the basin’s end. In the WSBB basin, only the TKEs in central fluid depths traveled downstream, which ended at X = 13 m from the toe of HJs. It is worth mentioning here that as compared to the modified USBR basin, the TKEs in the WSBB basin declined earlier, which indicated 8 m less distance than the USBR basin.
Figure 22
Depth-wise distribution of TKEs in the WSBB stilling basin at (a) Z= 0 m (floor level), (b) Z= 0.47 m, (c) Z= 0.93 m, (d) Z= 1.39 m, (e) 1.85 m, (f) Z= 2.51 m, and (g) free surface.
The results further showed that upon the use of WSBBs, no flow reattachment was witnessed on either side of the baffle blocks. Due to reduced reattachment, more wake areas were generated on the side of the WSBB basin, and the results showed agreement with the statement of other authors (Verma & Goel 2003; Verma et al. 2004; Goel & Verma 2006). At Z = 0.47 m, 0.93 m, and 1.39 m, the maximum TKEs were noticed in the HJ region, which reached 4.3, 4.6, and 3.5 J/kg, respectively, as shown in Figure 22(b)–22(d), respectively. In the WSBB basin, after the HJ, the baffle blocks declined the TKEs due to the development of sharp discontinuities in the flow. After Z = 1.39 m, the value of TKEs up to the free surface gradually reduced, as shown in Figure 22(e) and 22(f). Figure 22(g) shows 2D illustrations of TKEs on the free surface, and the results indicate that as compared to the modified USBR basin, the magnitude of TKEs was lower and traveled less distance in the WSBB basin.
CONCLUSIONS
This study developed numerical models on the rigid bed to investigate the effects of USBR and WSBB baffle blocks on the HJ downstream of the river diversion barrage using FLOW-3D. VOF and RNG K–ɛ models were employed to track the free surface and turbulence, respectively. For the proposed new basin (WSBB basin), WSBB with a vertex angle of 150° and cutback of 90° is employed in the baffle block region, while the friction block region remained unchanged. The performance of the two different basins is assessed by HJs and other hydraulic parameters such as free surface profile, sequent depths, roller lengths, HJ efficiency, velocity profile, and TKE. Furthermore, the results of the present modified USBR Type-III and WSBB basins are compared with the relevant literature, for which regression analysis is performed and residual error diagrams are plotted. However, the present models are limited to the single discharge of 44 m3/s and employ only one turbulence model, i.e., RNG K–ɛ. Additionally, the present models were designed for a single bay of the barrage.
Upon use of fine meshing, in comparison to the designed discharge, the present models showed 0.80 and 0.90% of errors in modified USBR Type-III and WSBB basins, respectively. Similarly, for the gated flow, the results indicated 0.32 and 1.14% errors in the modified USBR Type-III and WSBB basins, respectively.
After employing regression analysis, the results of free surface profiles showed agreement with the previous studies for which R2 reached 0.980 and 0.970 in WSBB and modified USBR basins, respectively. From the results, it can be believed that as compared to the modified USBR Type-III, the newly proposed WSBB basin produced a better free surface profile of HJs.
Due to the inclusion of the baffle blocks in the studied basins, the roller lengths of HJs were contained efficiently, and thereby, as compared to the literature, lesser roller lengths were observed in the modified USBR Type-III and WSBB basins.
The overall efficiency of HJs in modified USBR and WSBB basins reached 58.60 and 57.90%, respectively, which showed good agreement with the literature. Based on the results of the efficiency of HJs, the accuracy of the present models reached 93%.
In the hydraulic regions, the results of dimensionless velocity profiles indicated a wall jet-like structure, which agreed well with the literature. In addition, as compared to the modified USBR Type-III basin, the velocity profiles in the WSBB basin were found to be more promising, for which R2 reached 0.937. Additionally, after the HJ, as compared to the USBR Type-III basin, the forward velocity (U/Umax) in the WSBB basin was found to be less. Conclusively, it can be said that in comparison to the modified USBR Type-III basin, at the lower discharges, the WSBB basin decays the velocities more efficiently.
The results of TKEs indicated that the flow was strongly turbulent near the foreside of the HJs, and the maximum TKEs were noted in the central fluid depths. In the WSBB basin, no fluid reattachment was observed on either side of the baffle blocks, and the results further indicated that as compared to the modified USBR Type-III basin, fewer TKEs were found at the end of the WSBB basin.
Based on the models’ results, the study confirms the suitability of WSBB downstream of the barrage for lower tailwater conditions. From the results, it is believed that FLOW-3D is a very effective and efficient tool for the hydraulic investigation of flow behavior downstream of the barrage. However, in Pakistan, the use of such modeling tools is found very limited, therefore, the study results will help hydraulic and civil engineers to assess different energy dissipation arrangements within the stilling basins and will provide suitable alternative solutions. The present study was limited to the fixed geometry of the WSBB, therefore, it is suggested to investigate HJ and flow characteristics with other vertex and cutback angles. In addition, it is also recommended to study the hydraulics of WSBB downstream of barrages by employing multiple bays of barrage and other turbulence models.
REFERENCES
Alikhani A., Behrozi-Rad R. & Fathi-Moghadam M. 2010 Hydraulic jump in stilling basin with vertical end sill. Int. J. Phys. Sci. 5, 25–29.
Aydogdu M., Gul E. & Dursun O. F. 2022 Experimentally verified numerical investigation of the sill hydraulics for abruptly expanding stilling basin. Arabian J. Sci. Eng. https://doi.org/10.1007/s13369-022-07089-6.
Bakhmeteff B. A. & Matzke A. E. 1936 The hydraulic jump in terms of dynamic similarity. Trans. ASCE 100, 630–680.
Bayon-Barrachina A. & Lopez-Jimenez P. A. 2015 Numerical analysis of hydraulic jumps using OpenFOAM. J. Hydroinf. 17, 662–678. https://doi.org/10.2166/hydro.2015.041.
Bayon-Barrachina A., Valles-Moran F. J., Lopes-Jiménez P. A., Bayn A., Valles-Morn F. J. & Lopes-Jimenez P. A. 2015 Numerical analysis and validation of South Valencia sewage collection system. In E-proceedings 36th IAHR World Congr, 28 June – 3 July, 2015, Hague, Netherlands Numer. 17, pp. 1–11.
Belanger, 1841. Bélanger, J. B. ‘Notes sur l’Hydraulique.’ Ecole Royale des Ponts et Chaussées, Paris, France, session 1842 (1841): 223. Bhosekar V. V., Patnaik S., Gadge P. P. & Gupta I. D. 2014 Discharge characteristics of orifice spillway. Int. J. Dam. Eng. XXIV, 5–18.
Bradshaw P. 1997 Understanding and prediction of turbulent flow – 1996. Int. J. Heat Fluid Flow 18, 45–54. https://doi.org/10.1016/S0142-727X(96)00134-8.
Carvalho R. F., Lemos C. M. & Ramos C. M. 2008 Numerical computation of the flow in hydraulic jump stilling basins. J. Hydraul. Res. 46, 739–752. https://doi.org/10.1080/00221686.2008.9521919.
Chachereau Y. & Chanson H. 2011 Free-surface fluctuations and turbulence in hydraulic jumps. Exp. Therm. Fluid Sci. 35, 896–909. https://doi.org/10.1016/j.expthermflusci.2011.01.009.
Chaudary Z. A. & Sarwar M. K. 2014 Rehabilitated Taunsa Barrage : prospects and concerns. Sci. Technol. Dev. 33, 127–131.
Chaudhry Z. A. 2010 Surface flow hydraulics of Taunsa Barrage : before and after rehabilitation. Pak. J. Sci. 62, 116–119.
Ead S. A. & Rajaratnam N. 2002 Hydraulic jumps on corrugated beds. J. Hydraul. Eng. 128, 656–663. https://doi.org/10.1061/(asce)0733-9429(2002)128:7(656).
El Baradei S. A., Abodonya A., Hazem N., Ahmed Z., El Sharawy M., Abdelghaly M. & Nabil H. 2022 Ethiopian dam optimum hydraulic operating conditions to reduce unfavorable impacts on downstream countries. Civ. Eng. J. 8, 1906–1919. https://doi.org/10.28991/CEJ-2022-08-09-011.
Eloubaidy A., Al-Baidhani J. & Ghazali A. 1999 Dissipation of hydraulic energy by curved baffle blocks. Pertanika J. Sci. Technol. 7, 69–77.
Elsaeed G., Ali A., Abdelmageed N. & Ibrahim A. 2016 Effect of end step shape in the performance of stilling basins downstream radial gates. J. Sci. Res. Rep. 9, 1–9. https://doi.org/10.9734/jsrr/2016/21452.
Frizell K. & Svoboda C. 2012 Performance of Type III Stilling Basins-Stepped Spillway Studies. US Bur. Reclam, Denver, CO, USA.
Gadge P. P., Jothiprakash V. & Bhosekar V. V. 2018 Hydraulic investigation and design of roof profile of an orifice spillway using experimental and numerical models. J. Appl. Water Eng. Res. 6, 85–94. https://doi.org/10.1080/23249676.2016.1214627.
Gadge P. P., Jothiprakash V. & Bhosekar V. V. 2019 Hydraulic design considerations for orifice spillways. ISH J. Hydraul. Eng. 5010, 1–7. https://doi.org/10.1080/09715010.2018.1423579.
Ghaderi A., Daneshfaraz R., Dasineh M. & Di Francesco S. 2020 Energy dissipation and hydraulics of flow over trapezoidal-triangular labyrinth weirs. Water (Switzerland) 12. https://doi.org/10.3390/w12071992
Ghaderi A., Dasineh M. & Aristodemo F. 2021 Numerical simulations of the flow field of a submerged hydraulic jump over triangular macroroughnesses. Water (Switzerland) 13, 1–24.
Ghosal S. & Moin P. 1995 The basic equations for the large eddy simulation of turbulent flows in complex geometry. J. Comput. Phys. 118, 24–37. https://doi.org/10.1006/JCPH.1995.1077.
Goel A. 2007 Experimental study on stilling basins for square outlets. In 3rd WSEAS Int. Conf. Appl. Theor. Mech., pp. 157–162.
Goel A. 2008 Design of stilling basin for circular pipe outlets. Can. J. Civ. Eng. 35, 1365–1374. https://doi.org/10.1139/L08-085.
Goel A. & Verma D. V. S. 2006 Alternate designs of stilling basins for pipe outlets. Irrig. Drain. Syst. 20, 139–150. https://doi.org/10.1007/s10795-006-7901-x.
Gray T. E., Alexander J. & Leeder M. R. 2005 Quantifying velocity and turbulence structure in depositing sustained turbidity currents across breaks in slope. Sedimentology 52, 467–488. https://doi.org/10.1111/j.1365-3091.2005.00705.x.
Habibzadeh A., Loewen M. R. & Rajaratnam N. 2012 Performance of baffle blocks in submerged hydraulic jumps. J. Hydraul. Eng. 138, 902–908. https://doi.org/10.1061/(asce)hy.1943-7900.0000587.
Hager W. H. & Bremen R. 1989 Classical hydraulic jump: sequent depths. J. Hydraul. Res. 27, 565–585. https://doi.org/10.1080/00221688909499111.
Hager W. H. & Sinniger R. 1985 Flow characteristics of the hydraulic jump in a stilling basin with an abrupt bottom rise. J. Hydraul. Res. 23, 101–113. https://doi.org/10.1080/00221688509499359.
Hassanpour N., Dalir A. H. & Bayon A. 2021 Pressure fluctuations in the spatial hydraulic jump in stilling basins with different expansion ratio. Water (Switzerland) 13 (1), 1–15.
Herrera-Granados O. & Kostecki S. W. 2016 Numerical and physical modeling of water flow over the ogee weir of the new Niedów barrage. J. Hydrol. Hydromech. 64, 67–74. https://doi.org/10.1515/johh-2016-0013.
Hirt C. M. & Sicilian J. M. 1985 A porosity technique for the definition of obstacles in rectangular cell meshes. In: International Conference on Numerical Ship Hydrodynamics, 4th., pp. 1–19.
Issa R. I. 1985 Solution of the implicitly discretized fluid flow equations by operator-splitting. J. Comput. Phys. 62, 40–65. https://doi.org/10.1080/10407782.2016.1173467.
Johnson M. C. & Savage B. M. 2006 Physical and numerical comparison of flow over ogee spillway in the presence of tailwater. J. Hydraul. Eng. 132, 1353–1357. https://doi.org/10.1061/(asce)0733-9429(2006)132:12(1353).
Jothiprakash V., Bhosekar V. V. & Deolalikar P. B. 2015 Flow characteristics of orifice spillway aerator : numerical model studies. ISH J. Hydraul. Eng. 5010, 1–15. https://doi.org/10.1080/09715010.2015.1007093.
Joubert W. 1994 A robust GMRES-based adaptive polynomial preconditioning algorithm for nonsymmetric linear systems. SIAM J. Sci. Comput. 15, 427–439. https://doi.org/10.1137/0915029.
Kamath A., Fleit G. & Bihs H. 2019 Investigation of free surface turbulence damping in RANS simulations for complex free surface flows. Water (Switzerland) 3, 456. https://doi.org/10.3390/w11030456.
Kucukali S. & Chanson H. 2008 Turbulence measurements in the bubbly flow region of hydraulic jumps. Exp. Therm. Fluid Sci. 33, 41–53. https://doi.org/10.1016/j.expthermflusci.2008.06.012.
Macián-Pérez J. F., Bayón A., García-Bartual R., Amparo López-Jiménez P. & Vallés-Morán F. J. 2020a Characterization of structural properties in high reynolds hydraulic jump based on CFD and physical modeling approaches. J. Hydraul. Eng. 146, 04020079. https://doi.org/10.1061/(asce)hy.1943-7900.0001820.
Macián-Pérez J. F., García-Bartual R., Huber B., Bayon A. & Vallés-Morán F. J. 2020b Analysis of the flow in a typified USBR II stilling basin through a numerical and physical modeling approach. Water (Switzerland) 12, 6–20. https://doi.org/10.3390/w12010227.
Maleki S. & Fiorotto V. 2021 Hydraulic jump stilling basin design over rough beds. J. Hydraul. Eng. 147, 04020087. https://doi.org/10.1061/(asce)hy.1943-7900.0001824.
Mansour B. G. S., Nashed N. F. & Mansour S. G. S. 2004 Model study to optimise the hydraulic performance of the New Naga Hammadi Barrage stilling basin. Bridg. Gap Meet. World’s Water Environ. Resour. Challenges – Proc. World Water Environ. Resour. Congr. 2001 111, 1–9. https://doi.org/10.1061/40569(2001)461. Murzyn F. & Chanson H. 2009 Experimental investigation of bubbly flow and turbulence in hydraulic jumps. Environ. Fluid Mech. 9, 143–159. https://doi.org/10.1007/s10652-008-9077-4.
Nasrabadi M., Omid M. H. & Farhoudi J. 2012 Submerged hydraulic jump with sediment-laden flow. Int. J. Sediment Res. 27, 100–111. https://doi.org/10.1016/S1001-6279(12)60019-5.
Nikmehr S. & Aminpour Y. 2020 Numerical simulation of hydraulic jump over rough beds. Period. Polytech. Civ. Eng. 64, 396–407. https://doi.org/10.3311/PPci.15292.
Patankar S. V. & Spalding D. B. 1972 A calculation procedure for heat, mass and momentum transfer in three-dimensional parabolic flows. Int. J. Heat Mass Transf. 15, 1787–1806. https://doi.org/10.1016/0017-9310(72)90054-3.
Peterka A. J. 1984 Hydraulic design of stilling basins and energy dissipators. Water Resour. Tech. Publ. – US Dep. Inter. 240, 1–240.
Pillai N. N., Goel A. & Dubey A. K. 1989 Hydraulic jump type stilling basin for low froude numbers. J. Hydraul. Eng. 115, 989–994. https://doi.org/10.1061/(asce)0733-9429(1989)115:7(989).
Qasim R. M., Mohammed A. A. & Abdulhussein I. A. 2022 An investigating of the impact of Bed flume discordance on the weir-gate hydraulic structure. HighTechnol. Innovation J. 3, 341–355. https://doi.org/10.28991/HIJ-2022-03-03-09.
Sayyadi K., Heidarpour M. & Ghadampour Z. 2022 Effect of bed roughness and negative step on characteristics of hydraulic jump in rectangular stilling basin. Shock Vib. 2022. https://doi.org/https://doi.org/10.1155/2022/1722065.
Sutopo Y., Utomo K. S. & Tinov N. 2022 The effects of spillway width on outflow discharge and flow elevation for the Probable Maximum Flood (PMF). Civ. Eng. J. 8, 723–733. https://doi.org/10.28991/CEJ-2022-08-04-08.
Tiwari H. L. & Goel A. 2016 Effect of impact wall on energy dissipation in stilling basin. KSCE J. Civ. Eng. 20, 463–467. https://doi.org/10.1007/s12205-015-0292-5.
Tiwari H. L., Gahlot V. K. & Goel A. 2010 Stilling basins below outlet works – an overview. Int. J. Eng. Sci. 2, 6380–6385.
Verma D. V. S. & Goel A. 2003 Development of efficient stilling basins for pipe outlets. J. Irrig. Drain. Eng. 129, 194–200. https://doi.org/10.1061/(asce)0733-9437(2003)129:3(194).
Verma D. V. S., Goel A. & Rai V. 2004 New stilling basins designs for deep rectangular outlets. IJE Trans. A Basics 17, 1–10.
Viti N., Valero D. & Gualtieri C. 2018 Numerical simulation of hydraulic jumps. Part 2: recent results and future outlook. Water (Switzerland) 11, 1–18. https://doi.org/10.3390/w11010028.
Wang H. & Chanson H. 2015 Experimental study of turbulent fluctuations in hydraulic jumps. J. Hydraul. Eng. 141, 04015010. https://doi.org/10.1061/(asce)hy.1943-7900.0001010.
Wilcox D. C. 2008 Formulation of the k-ω turbulence model revisited. AIAA J. 46, 2823–2838. https://doi.org/10.2514/1.36541.
Wu S. & Rajaratnam N. 1996 Transition from hydraulic jump to open channel flow. J. Hydraul. Eng 122 (9), 526–528.
Yakhot V., Thangam S., Gatski T. B., Orszag S. A. & Speziale C. G. 1991 Development of turbulence models for shear flows by a double expansion technique. Phys. Fluids A 4, 1510–1520.
Yamini O. A., Movahedi A., Mousavi S. H., Kavianpour M. R. & Kyriakopoulos G. L. 2022 Hydraulic performance of seawater intake system using CFD modeling. J. Mar. Sci. Eng. 10. https://doi.org/10.3390/jmse10070988.
Zaffar M. W. & Hassan I. 2023a Hydraulic investigation of stilling basins of the barrage before and after remodelling using FLOW-3D. Water Supply 23, 796–820. https://doi.org/10.2166/ws.2023.032.
Zaffar M. W. & Hassan I. 2023b Numerical investigation of hydraulic jump for different stilling basins using FLOW-3D. AQUA – Water Infrastructure, Ecosystems and Society 72 (7), 1320–1343. jws2023290. https://doi.org/10.2166/aqua.2023.290.
Zaffar M. W., Hassan I., Latif U., Jahan S. & Ullah Z. 2023 Numerical investigation of scour downstream of diversion barrage for different stilling basins at flood discharge. Sustainability 15, 11032. https://doi.org/https://doi.org/10.3390/su151411032.
Zaidi S. M. A., Khan M. A. & Rehman S. U. 2004 Plan. Des. Taunsa Barrage Rehabil. Proj. Pakistan Eng. Congr. Lahore. 71st Annu. Sess. Proceedings, Pap.687, pp. 228–286. Zaidi S. M. A., Amin M. & Ahmadani M. A. 2011 Perform. Eval. Taunsa barrage Emerg. Rehabil. Mod. Proj. Pakistan Eng. Congr. 71st Annu. Sess. Proceedings, Pap. pp. 650–682. Zulfiqar C. & Kaleem S. M. 2015 Launching/Disappearance of stone a`pron, block floor downstream of the Taunsa Barrage and unprecedent drift of the river towards Kot Addu Town. Sci. Technol. Dev. 34, 60–65. https://doi.org/10.3923/std.2015.60.65.
In practical use, there is shrinkage in the width direction in existing overflow water film. This study introduces the ” flow deflectors ” to solve the problem of discontinuous water film. According to the principle of the experimental devices, the model was established by FLOW-3D software and the corresponding boundary conditions were set up, the minimum size of the grid was 0.25mm × 0.25mm × 0.25mm. The film model was investigated under ten working conditions with three factors. Through the analysis of the distribution curves of velocities and thickness, impacts of unit width flux, tank width and the distance between the flow deflectors on the velocity and thickness of the liquid film were obtained. The results indicated that the water film was able to keep continuous and airtight with flow deflectors. The variation law of water film with unit width flux was obtained and a correlation formula was proposed according to Nusselt correlation. And, it was found that the tank width has little influence on the water film itself. The “dry zones” were also founded on the film when the distance between flow deflectors increased.
References
Gong H, Jiang J, Qiu J, Jiang R and Zhang W 2016 Theoretical and Experimental Study on the Water Curtain Shape and Mechanical Relationship of the Smoke Control System Advances in Engineering Research 500-7 Gong H, Jiang J, Zhang W and Guan Y 2017 Theory and Simulation of the Relationship between Smoke-proof Water Curtain’s Tank Structure and Flow Non-uniformity DEStech Transactions on Environment, Energy and Earth Science iceepe 257-63 ZHENG Li, QIAN Zhongdong, CAO Zhixian and TAN Guangming 2009 Comparison of computed result for dam over-topping flow with two 3D models Engineering Journal of Wuhan University 06 758-63 Hartley DE and Murgatroyd W 1964 Criteria for the break-up of thin liquid layers flowing isothermally over solid surfaces International Journal of Heat and Mass Transfer 9 1003-15 QIN Wei, LIU Jianhua, WENG Zemin and JIANG Zhangyan 1997 Characteristics of Interfacial Surface Wave of Free-Falling Liquid Film JOURNAL OF JIMEI NAVIGATION INSTITUTE 02 3-8 Ye X, Yan W, Jiang Z and Li C 2002 Hydrodynamics of Free-Falling Turbulent Wavy Films and Implications for Enhanced Heat Transfer Heat Transfer Engineering 1 48-60
Deniz Velioglu Sogut ,Erdinc Sogut ,Ali Farhadzadeh,Tian-Jian Hsu
Abstract
The present study evaluates the performance of two numerical approaches in estimating non-equilibrium scour patterns around a non-slender square structure subjected to a transient wave, by comparing numerical findings with experimental data. This study also investigates the impact of the structure’s positioning on bed evolution, analyzing configurations where the structure is either attached to the sidewall or positioned at the centerline of the wave flume. The first numerical method treats sediment particles as a distinct continuum phase, directly solving the continuity and momentum equations for both sediment and fluid phases. The second method estimates sediment transport using the quadratic law of bottom shear stress, yielding robust predictions of bed evolution through meticulous calibration and validation. The findings reveal that both methods underestimate vortex-induced near-bed vertical velocities. Deposits formed along vortex trajectories are overestimated by the first method, while the second method satisfactorily predicts the bed evolution beneath these paths. Scour holes caused by wave impingement tend to backfill as the flow intensity diminishes. The second method cannot sufficiently capture this backfilling, whereas the first method adequately reflects the phenomenon. Overall, this study highlights significant variations in the predictive capabilities of both methods in regard to the evolution of non-equilibrium scour at low Keulegan–Carpenter numbers.
Keywords
Keulegan-Carpenter number, Solitary wave, non slender, wave-structure interaction, FLOW-3D, WedWaveFoam
Omega-Luitex법을 이용한 수력점프 발생시 러프 베드의 와류 진화 예측 및 영향 분석
Cong Trieu Tran, Cong Ty Trinh
Abstract
The dissipation of energy downstream of hydropower projects is a significant issue. The hydraulic jump is exciting and widely applied in practice to dissipate energy. Many hydraulic jump characteristics have been studied, such as length of jump Lj and sequent flow depth y2. However, understanding the evolution of the vortex structure in the hydraulic jump shows a significant challenge. This study uses the RNG k-e turbulence model to simulate hydraulic jumps on the rough bed. The Omega-Liutex method is compared with Q-criterion for capturing vortex structure in the hydraulic jump. The formation, development, and shedding of the vortex structure at the rough bed in the hydraulic jumper are analyzed. The vortex forms and rapidly reduces strength on the rough bed, resulting in fast dissipation of energy. At the rough block rows 2nd and 3rd, the vortex forms a vortex rope that moves downstream and then breaks. The vortex-shedding region represents a significant energy attenuation of the flow. Therefore, the rough bed dissipates kinetic energy well. Adding reliability to the vortex determined by the Liutex method, the vorticity transport equation is used to compare the vorticity distribution with the Liutex distribution. The results show a further comprehension of the hydraulic jump phenomenon and its energy dissipation.
[1] Viti, N., Valero, D., & Gualtieri, C. (2019). Numerical Simulation of Hydraulic Jumps. Part 2: Recent Results and Future Outlook. Water, 11(1), 28. https://doi.org/10.3390/w11010028
[2] Peterka, A. J. (1978.) Hydraulic Design of Stilling Basins and Energy Dissipators. Department of the Interior, Bureau of Reclamation.
[3] Bejestan, M. S. & Neisi, K. (2009). A new roughened bed hydraulic jump stilling basin. Asian journal of applied sciences, 2(5), 436-445. https://doi.org/10.3923/ajaps.2009.436.445
[4] Tokyay, N. D. (2005). Effect of channel bed corrugations on hydraulic jumps. Impacts of Global Climate Change, 1-9. https://doi.org/10.1061/40792(173)408
[5] Nikmehr, S. & Aminpour, Y. (2020). Numerical Simulation of Hydraulic Jump over Rough Beds. Periodica Polytechnica Civil Engineering, 64(2), 396-407. https://doi.org/10.3311/PPci.15292
[6] Hunt, J. C., Wray, A. A., & Moin, P. (1988). Eddies, streams, and convergence zones in turbulent flows. Studying turbulence using numerical simulation databases. 2. Proceedings of the 1988 summer program.
[7] Gao, Y. & Liu, C. (2018). Rortex and comparison with eigenvalue-based vortex identification criteria. Physics of Fluids, 30(8), 085107. https://doi.org/10.1063/1.5040112
[8] Liu, C., Gao, Y., Tian, S., & Dong, X. (2018). Rortex – A new vortex vector definition and vorticity tensor and vector decompositions. Physics of Fluids, 30(3), 035103. https://doi.org/10.1063/1.5023001
[9] Liu, C. et al. (2019). Third generation of vortex identification methods: Omega and Liutex/Rortex based systems. Journal of Hydrodynamics, 31(2), 205-223. https://doi.org/10.1007/s42241-019-0022-4
[10] Liu, C., Wang, Y., Yang, Y. et al (2016). New omega vortex identification method. Science China Physics, Mechanics & Astronomy, (8), 56-64. https://doi.org/10.1007/s11433-016-0022-6
[11] Tran, C. T. & Pham, D. C. (2022). Application of Liutex and Entropy Production to Analyze the Influence of Vortex Rope in the Francis-99 Turbine Draft Tube. Tehnički vjesnik, 29(4), 1177-1183. https://doi.org/10.17559/TV-20210821070801
[12] Dong, X., Gao, Y., & Liu, C. (2019). New normalized Rortex/vortex identification method. Physics of Fluids, 31(1), 011701. https://doi.org/10.1063/1.5066016
[13] Wang, L., Zheng, Z., Cai, W. et al. (2019). Extension Omega and Omega-Liutex methods applied to identify vortex structures in viscoelastic turbulent flow. Journal of Hydrodynamics, 31(5), 911-921. https://doi.org/10.1007/s42241-019-0045-x
[14] Xu, H., Cai, X., & Liu, C. (2019). Liutex (vortex) core definition and automatic identification for turbulence vortex structures. Journal of Hydrodynamics, 31(5), 857-863. https://doi.org/10.1007/s42241-019-0066-5
[15] Tran, C. T. et al. (2020). Prediction of the precessing vortex core in the Francis-99 draft tube under off-design conditions by using Liutex/Rortex method. Journal of Hydrodynamics, 32, 623-628. https://doi.org/10.1007/s42241-020-0031-3
[16] Liu, C. et al. (2019). A Liutex based definition of vortex axis line. arXiv preprint arXiv:1904.10094. https://doi.org/10.48550/arXiv.1904.10094
[17] Samadi-Boroujeni, H. et al. (2013). Effect of triangular corrugated beds on the hydraulic jump characteristics. Canadian Journal of Civil Engineering, 40(9), 841-847. https://doi.org/10.1139/cjce-2012-0019
[18] Ghaderi, A. et al. (2020). Characteristics of free and submerged hydraulic jumps over different macroroughnesses. Journal of Hydroinformatics, 22(6), 1554-1572. https://doi.org/10.2166/hydro.2020.298
[19] Wu, Z. et al. (2021). Analysis of the influence of transverse groove structure on the flow of a flat-plate surface based on Liutex parameters. Engineering Applications of Computational Fluid Mechanics, 15(1), 1282-1297. https://doi.org/10.1080/19942060.2021.1968955
[20] Ji, B., et al. (2014). Numerical simulation of threedimensional cavitation shedding dynamics with special emphasis on cavitation – vortex interaction. Ocean Engineering, 87, 64-77. https://doi.org/10.1016/j.oceaneng.2014.05.005
[21] Tran, C., Bin, J., & Long, X. (2019). Simulation and Analysis of Cavitating Flow in the Draft Tube of the Francis Turbine with Splitter Blades at Off-Design Condition. Tehnicki vjesnik – Technical Gazette, 26(6). https://doi.org/10.17559/TV-20190316042929
Mohammad Nazari-Sharabian, Aliasghar Nazari-Sharabian, Moses Karakouzian, Mehrdad Karami
Abstract
Scour is defined as the erosive action of flowing water, as well as the excavating and carrying away materials from beds and banks of streams, and from the vicinity of bridge foundations, which is one of the main causes of river bridge failures. In the present study, implementing a numerical approach, and using the FLOW-3D model that works based on the finite volume method (FVM), the applicability of using sacrificial piles in different configurations in front of a bridge pier as countermeasures against scouring is investigated. In this regard, the numerical model was calibrated based on an experimental study on scouring around an unprotected circular river bridge pier. In simulations, the bridge pier and sacrificial piles were circular, and the riverbed was sandy. In all scenarios, the flow rate was constant and equal to 45 L/s. Furthermore, one to five sacrificial piles were placed in front of the pier in different locations for each scenario. Implementation of the sacrificial piles proved to be effective in substantially reducing the scour depths. The results showed that although scouring occurred in the entire area around the pier, the maximum and minimum scour depths were observed on the sides (using three sacrificial piles located upstream, at three and five times the pier diameter) and in the back (using five sacrificial piles located upstream, at four, six, and eight times the pier diameter) of the pier. Moreover, among scenarios where single piles were installed in front of the pier, installing them at a distance of five times the pier diameter was more effective in reducing scour depths. For other scenarios, in which three piles and five piles were installed, distances of six and four times the pier diameter for the three piles scenario, and four, six, and eight times the pier diameter for the five piles scenario were most effective.
Keywords
Scouring; River Bridges; Sacrificial Piles; Finite Volume Method (FVM); FLOW-3D.
References
Karakouzian, Chavez, Hayes, and Nazari-Sharabian. “Bulbous Pier: An Alternative to Bridge Pier Extensions as a Countermeasure Against Bridge Deck Splashing.” Fluids 4, no. 3 (July 24, 2019): 140. doi:10.3390/fluids4030140.
Karami, Mehrdad, Abdorreza Kabiri-Samani, Mohammad Nazari-Sharabian, and Moses Karakouzian. “Investigating the Effects of Transient Flow in Concrete-Lined Pressure Tunnels, and Developing a New Analytical Formula for Pressure Wave Velocity.” Tunnelling and Underground Space Technology 91 (September 2019): 102992. doi:10.1016/j.tust.2019.102992.
Karakouzian, Moses, Mohammad Nazari-Sharabian, and Mehrdad Karami. “Effect of Overburden Height on Hydraulic Fracturing of Concrete-Lined Pressure Tunnels Excavated in Intact Rock: A Numerical Study.” Fluids 4, no. 2 (June 19, 2019): 112. doi:10.3390/fluids4020112.
Chiew, Yee-Meng. “Scour protection at bridge piers.” Journal of Hydraulic Engineering 118, no. 9 (1992): 1260-1269. doi:10.1061/(ASCE)0733-9429(1992)118:9(1260).
Shen, Hsieh Wen, Verne R. Schneider, and Susumu Karaki. “Local scour around bridge piers.” Journal of the Hydraulics Division (1969): 1919-1940.
Richardson, E.V., and Davis, S.R. “Evaluating Scour at Bridges”. Hydraulic Engineering Circular. (2001), 18 (HEC-18), Report no. FHWA NHI 01–001, U.S. Department of Transportation, Federal Highway Administration, Washington, DC, USA.
Elsaeed, Gamal, Hossam Elsersawy, and Mohammad Ibrahim. “Scour Evaluation for the Nile River Bends on Rosetta Branch.” Advances in Research 5, no. 2 (January 10, 2015): 1–15. doi:10.9734/air/2015/17380.
Chang, Wen-Yi, Jihn-Sung Lai, and Chin-Lien Yen. “Evolution of scour depth at circular bridge piers.” Journal of Hydraulic Engineering 130, no. 9 (2004): 905-913. doi:10.1061/(ASCE)0733-9429(2004)130:9(905).
Unger, Jens, and Willi H. Hager. “Riprap failure at circular bridge piers.” Journal of Hydraulic Engineering 132, no. 4 (2006): 354-362. doi:10.1061/(ASCE)0733-9429(2006)132:4(354).
Abdeldayem, Ahmed W., Gamal H. Elsaeed, and Ahmed A. Ghareeb. “The effect of pile group arrangements on local scour using numerical models.” Advances in Natural and Applied Sciences 5, no. 2 (2011): 141-146.
Sheppard, D. M., B. Melville, and H. Demir. “Evaluation of Existing Equations for Local Scour at Bridge Piers.” Journal of Hydraulic Engineering 140, no. 1 (January 2014): 14–23. doi:10.1061/(asce)hy.1943-7900.0000800.
Melville, Bruce W., and Anna C. Hadfield. “Use of sacrificial piles as pier scour countermeasures.” Journal of Hydraulic Engineering 125, no. 11 (1999): 1221-1224. doi:10.1061/(ASCE)0733-9429(1999)125:11(1221).
Yao, Weidong, Hongwei An, Scott Draper, Liang Cheng, and John M. Harris. “Experimental Investigation of Local Scour Around Submerged Piles in Steady Current.” Coastal Engineering 142 (December 2018): 27–41. doi:10.1016/j.coastaleng.2018.08.015.
Link, Oscar, Marcelo García, Alonso Pizarro, Hernán Alcayaga, and Sebastián Palma. “Local Scour and Sediment Deposition at Bridge Piers During Floods.” Journal of Hydraulic Engineering 146, no. 3 (March 2020): 04020003. doi:10.1061/(asce)hy.1943-7900.0001696.
Khan, Mujahid, Mohammad Tufail, Muhammad Fahad, Hazi Muhammad Azmathullah, Muhammad Sagheer Aslam, Fayaz Ahmad Khan, and Asif Khan. “Experimental analysis of bridge pier scour pattern.” Journal of Engineering and Applied Sciences 36, no. 1 (2017): 1-12.
Yang, Yifan, Bruce W. Melville, D. M. Sheppard, and Asaad Y. Shamseldin. “Clear-Water Local Scour at Skewed Complex Bridge Piers.” Journal of Hydraulic Engineering 144, no. 6 (June 2018): 04018019. doi:10.1061/(asce)hy.1943-7900.0001458.
Moussa, Yasser Abdallah Mohamed, Tarek Hemdan Nasr-Allah, and Amera Abd-Elhasseb. “Studying the Effect of Partial Blockage on Multi-Vents Bridge Pier Scour Experimentally and Numerically.” Ain Shams Engineering Journal 9, no. 4 (December 2018): 1439–1450. doi:10.1016/j.asej.2016.09.010.
Guan, Dawei, Yee-Meng Chiew, Maoxing Wei, and Shih-Chun Hsieh. “Characterization of Horseshoe Vortex in a Developing Scour Hole at a Cylindrical Bridge Pier.” International Journal of Sediment Research 34, no. 2 (April 2019): 118–124. doi:10.1016/j.ijsrc.2018.07.001.
Dougherty, E.M. “CFD Analysis of Bridge Pier Geometry on Local Scour Potential” (2019). LSU Master’s Theses. 5031.
Vijayasree, B. A., T. I. Eldho, B. S. Mazumder, and N. Ahmad. “Influence of Bridge Pier Shape on Flow Field and Scour Geometry.” International Journal of River Basin Management 17, no. 1 (November 10, 2017): 109–129. doi:10.1080/15715124.2017.1394315.
Farooq, Rashid, and Abdul Razzaq Ghumman. “Impact Assessment of Pier Shape and Modifications on Scouring Around Bridge Pier.” Water 11, no. 9 (August 23, 2019): 1761. doi:10.3390/w11091761.
Link, Oscar, Cristian Castillo, Alonso Pizarro, Alejandro Rojas, Bernd Ettmer, Cristián Escauriaza, and Salvatore Manfreda. “A Model of Bridge Pier Scour During Flood Waves.” Journal of Hydraulic Research 55, no. 3 (November 18, 2016): 310–323. doi:10.1080/00221686.2016.1252802.
Karakouzian, Moses, Mehrdad Karami, Mohammad Nazari-Sharabian, and Sajjad Ahmad. “Flow-Induced Stresses and Displacements in Jointed Concrete Pipes Installed by Pipe Jacking Method.” Fluids 4, no. 1 (February 21, 2019): 34. doi:10.3390/fluids4010034.
Flow Science, Inc. FLOW-3D User’s Manual, Flow Science (2018).
Brethour, J. Modeling Sediment Scour. Flow Science, Santa Fe, NM. (2003).
Brethour, James, and Jeff Burnham. “Modeling sediment erosion and deposition with the FLOW-3D sedimentation & scour model.” Flow Science Technical Note, FSI-10-TN85 (2010): 1-22.
Balouchi, M., and Chamani, M.R. “Investigating the Effect of using a Collar around a Bridge Pier, on the Shape of the Scour Hole”. Proceedings of the First International Conference on Dams and Hydropower (2012) (In Persian).
Bayon, Arnau, Daniel Valero, Rafael García-Bartual, Francisco José Vallés-Morán, and P. Amparo López-Jiménez. “Performance Assessment of OpenFOAM and FLOW-3D in the Numerical Modeling of a Low Reynolds Number Hydraulic Jump.” Environmental Modelling & Software 80 (June 2016): 322–335. doi:10.1016/j.envsoft.2016.02.018.
Aminoroayaie Yamini, O., S. Hooman Mousavi, M. R. Kavianpour, and Azin Movahedi. “Numerical Modeling of Sediment Scouring Phenomenon Around the Offshore Wind Turbine Pile in Marine Environment.” Environmental Earth Sciences 77, no. 23 (November 24, 2018). doi:10.1007/s12665-018-7967-4.
Nazari-Sharabian, Mohammad, Masoud Taheriyoun, Sajjad Ahmad, Moses Karakouzian, and Azadeh Ahmadi. “Water Quality Modeling of Mahabad Dam Watershed–Reservoir System under Climate Change Conditions, Using SWAT and System Dynamics.” Water 11, no. 2 (February 24, 2019): 394. doi:10.3390/w11020394.