Lab-on-a-Chip 시스템의 혈류 역학에 대한 검토: 엔지니어링 관점

Review on Blood Flow Dynamics in Lab-on-a-Chip Systems: An Engineering Perspective

  • Bin-Jie Lai
  • Li-Tao Zhu
  • Zhe Chen*
  • Bo Ouyang*
  • , and 
  • Zheng-Hong Luo*

Abstract

다양한 수송 메커니즘 하에서, “LOC(lab-on-a-chip)” 시스템에서 유동 전단 속도 조건과 밀접한 관련이 있는 혈류 역학은 다양한 수송 현상을 초래하는 것으로 밝혀졌습니다.

본 연구는 적혈구의 동적 혈액 점도 및 탄성 거동과 같은 점탄성 특성의 역할을 통해 LOC 시스템의 혈류 패턴을 조사합니다. 모세관 및 전기삼투압의 주요 매개변수를 통해 LOC 시스템의 혈액 수송 현상에 대한 연구는 실험적, 이론적 및 수많은 수치적 접근 방식을 통해 제공됩니다.

전기 삼투압 점탄성 흐름에 의해 유발되는 교란은 특히 향후 연구 기회를 위해 혈액 및 기타 점탄성 유체를 취급하는 LOC 장치의 혼합 및 분리 기능 향상에 논의되고 적용됩니다. 또한, 본 연구는 보다 정확하고 단순화된 혈류 모델에 대한 요구와 전기역학 효과 하에서 점탄성 유체 흐름에 대한 수치 연구에 대한 강조와 같은 LOC 시스템 하에서 혈류 역학의 수치 모델링의 문제를 식별합니다.

전기역학 현상을 연구하는 동안 제타 전위 조건에 대한 보다 실용적인 가정도 강조됩니다. 본 연구는 모세관 및 전기삼투압에 의해 구동되는 미세유체 시스템의 혈류 역학에 대한 포괄적이고 학제적인 관점을 제공하는 것을 목표로 한다.

KEYWORDS: 

1. Introduction

1.1. Microfluidic Flow in Lab-on-a-Chip (LOC) Systems

Over the past several decades, the ability to control and utilize fluid flow patterns at microscales has gained considerable interest across a myriad of scientific and engineering disciplines, leading to growing interest in scientific research of microfluidics. 

(1) Microfluidics, an interdisciplinary field that straddles physics, engineering, and biotechnology, is dedicated to the behavior, precise control, and manipulation of fluids geometrically constrained to a small, typically submillimeter, scale. 

(2) The engineering community has increasingly focused on microfluidics, exploring different driving forces to enhance working fluid transport, with the aim of accurately and efficiently describing, controlling, designing, and applying microfluidic flow principles and transport phenomena, particularly for miniaturized applications. 

(3) This attention has chiefly been fueled by the potential to revolutionize diagnostic and therapeutic techniques in the biomedical and pharmaceutical sectorsUnder various driving forces in microfluidic flows, intriguing transport phenomena have bolstered confidence in sustainable and efficient applications in fields such as pharmaceutical, biochemical, and environmental science. The “lab-on-a-chip” (LOC) system harnesses microfluidic flow to enable fluid processing and the execution of laboratory tasks on a chip-sized scale. LOC systems have played a vital role in the miniaturization of laboratory operations such as mixing, chemical reaction, separation, flow control, and detection on small devices, where a wide variety of fluids is adapted. Biological fluid flow like blood and other viscoelastic fluids are notably studied among the many working fluids commonly utilized by LOC systems, owing to the optimization in small fluid sample volumed, rapid response times, precise control, and easy manipulation of flow patterns offered by the system under various driving forces. 

(4)The driving forces in blood flow can be categorized as passive or active transport mechanisms and, in some cases, both. Under various transport mechanisms, the unique design of microchannels enables different functionalities in driving, mixing, separating, and diagnosing blood and drug delivery in the blood. 

(5) Understanding and manipulating these driving forces are crucial for optimizing the performance of a LOC system. Such knowledge presents the opportunity to achieve higher efficiency and reliability in addressing cellular level challenges in medical diagnostics, forensic studies, cancer detection, and other fundamental research areas, for applications of point-of-care (POC) devices. 

(6)

1.2. Engineering Approach of Microfluidic Transport Phenomena in LOC Systems

Different transport mechanisms exhibit unique properties at submillimeter length scales in microfluidic devices, leading to significant transport phenomena that differ from those of macroscale flows. An in-depth understanding of these unique transport phenomena under microfluidic systems is often required in fluidic mechanics to fully harness the potential functionality of a LOC system to obtain systematically designed and precisely controlled transport of microfluids under their respective driving force. Fluid mechanics is considered a vital component in chemical engineering, enabling the analysis of fluid behaviors in various unit designs, ranging from large-scale reactors to separation units. Transport phenomena in fluid mechanics provide a conceptual framework for analytically and descriptively explaining why and how experimental results and physiological phenomena occur. The Navier–Stokes (N–S) equation, along with other governing equations, is often adapted to accurately describe fluid dynamics by accounting for pressure, surface properties, velocity, and temperature variations over space and time. In addition, limiting factors and nonidealities for these governing equations should be considered to impose corrections for empirical consistency before physical models are assembled for more accurate controls and efficiency. Microfluidic flow systems often deviate from ideal conditions, requiring adjustments to the standard governing equations. These deviations could arise from factors such as viscous effects, surface interactions, and non-Newtonian fluid properties from different microfluid types and geometrical layouts of microchannels. Addressing these nonidealities supports the refining of theoretical models and prediction accuracy for microfluidic flow behaviors.

The analytical calculation of coupled nonlinear governing equations, which describes the material and energy balances of systems under ideal conditions, often requires considerable computational efforts. However, advancements in computation capabilities, cost reduction, and improved accuracy have made numerical simulations using different numerical and modeling methods a powerful tool for effectively solving these complex coupled equations and modeling various transport phenomena. Computational fluid dynamics (CFD) is a numerical technique used to investigate the spatial and temporal distribution of various flow parameters. It serves as a critical approach to provide insights and reasoning for decision-making regarding the optimal designs involving fluid dynamics, even prior to complex physical model prototyping and experimental procedures. The integration of experimental data, theoretical analysis, and reliable numerical simulations from CFD enables systematic variation of analytical parameters through quantitative analysis, where adjustment to delivery of blood flow and other working fluids in LOC systems can be achieved.

Numerical methods such as the Finite-Difference Method (FDM), Finite-Element-Method (FEM), and Finite-Volume Method (FVM) are heavily employed in CFD and offer diverse approaches to achieve discretization of Eulerian flow equations through filling a mesh of the flow domain. A more in-depth review of numerical methods in CFD and its application for blood flow simulation is provided in Section 2.2.2.

1.3. Scope of the Review

In this Review, we explore and characterize the blood flow phenomena within the LOC systems, utilizing both physiological and engineering modeling approaches. Similar approaches will be taken to discuss capillary-driven flow and electric-osmotic flow (EOF) under electrokinetic phenomena as a passive and active transport scheme, respectively, for blood transport in LOC systems. Such an analysis aims to bridge the gap between physical (experimental) and engineering (analytical) perspectives in studying and manipulating blood flow delivery by different driving forces in LOC systems. Moreover, the Review hopes to benefit the interests of not only blood flow control in LOC devices but also the transport of viscoelastic fluids, which are less studied in the literature compared to that of Newtonian fluids, in LOC systems.

Section 2 examines the complex interplay between viscoelastic properties of blood and blood flow patterns under shear flow in LOC systems, while engineering numerical modeling approaches for blood flow are presented for assistance. Sections 3 and 4 look into the theoretical principles, numerical governing equations, and modeling methodologies for capillary driven flow and EOF in LOC systems as well as their impact on blood flow dynamics through the quantification of key parameters of the two driving forces. Section 5 concludes the characterized blood flow transport processes in LOC systems under these two forces. Additionally, prospective areas of research in improving the functionality of LOC devices employing blood and other viscoelastic fluids and potentially justifying mechanisms underlying microfluidic flow patterns outside of LOC systems are presented. Finally, the challenges encountered in the numerical studies of blood flow under LOC systems are acknowledged, paving the way for further research.

2. Blood Flow Phenomena

ARTICLE SECTIONS

Jump To


2.1. Physiological Blood Flow Behavior

Blood, an essential physiological fluid in the human body, serves the vital role of transporting oxygen and nutrients throughout the body. Additionally, blood is responsible for suspending various blood cells including erythrocytes (red blood cells or RBCs), leukocytes (white blood cells), and thrombocytes (blood platelets) in a plasma medium.Among the cells mentioned above, red blood cells (RBCs) comprise approximately 40–45% of the volume of healthy blood. 

(7) An RBC possesses an inherent elastic property with a biconcave shape of an average diameter of 8 μm and a thickness of 2 μm. This biconcave shape maximizes the surface-to-volume ratio, allowing RBCs to endure significant distortion while maintaining their functionality. 

(8,9) Additionally, the biconcave shape optimizes gas exchange, facilitating efficient uptake of oxygen due to the increased surface area. The inherent elasticity of RBCs allows them to undergo substantial distortion from their original biconcave shape and exhibits high flexibility, particularly in narrow channels.RBC deformability enables the cell to deform from a biconcave shape to a parachute-like configuration, despite minor differences in RBC shape dynamics under shear flow between initial cell locations. As shown in Figure 1(a), RBCs initiating with different resting shapes and orientations displaying display a similar deformation pattern 

(10) in terms of its shape. Shear flow induces an inward bending of the cell at the rear position of the rim to the final bending position, 

(11) resulting in an alignment toward the same position of the flow direction.

Figure 1. Images of varying deformation of RBCs and different dynamic blood flow behaviors. (a) The deforming shape behavior of RBCs at four different initiating positions under the same experimental conditions of a flow from left to right, (10) (b) RBC aggregation, (13) (c) CFL region. (18) Reproduced with permission from ref (10). Copyright 2011 Elsevier. Reproduced with permission from ref (13). Copyright 2022 The Authors, under the terms of the Creative Commons (CC BY 4.0) License https://creativecommons.org/licenses/by/4.0/. Reproduced with permission from ref (18). Copyright 2019 Elsevier.

The flexible property of RBCs enables them to navigate through narrow capillaries and traverse a complex network of blood vessels. The deformability of RBCs depends on various factors, including the channel geometry, RBC concentration, and the elastic properties of the RBC membrane. 

(12) Both flexibility and deformability are vital in the process of oxygen exchange among blood and tissues throughout the body, allowing cells to flow in vessels even smaller than the original cell size prior to deforming.As RBCs serve as major components in blood, their collective dynamics also hugely affect blood rheology. RBCs exhibit an aggregation phenomenon due to cell to cell interactions, such as adhesion forces, among populated cells, inducing unique blood flow patterns and rheological behaviors in microfluidic systems. For blood flow in large vessels between a diameter of 1 and 3 cm, where shear rates are not high, a constant viscosity and Newtonian behavior for blood can be assumed. However, under low shear rate conditions (0.1 s

–1) in smaller vessels such as the arteries and venules, which are within a diameter of 0.2 mm to 1 cm, blood exhibits non-Newtonian properties, such as shear-thinning viscosity and viscoelasticity due to RBC aggregation and deformability. The nonlinear viscoelastic property of blood gives rise to a complex relationship between viscosity and shear rate, primarily influenced by the highly elastic behavior of RBCs. A wide range of research on the transient behavior of the RBC shape and aggregation characteristics under varied flow circumstances has been conducted, aiming to obtain a better understanding of the interaction between blood flow shear forces from confined flows.

For a better understanding of the unique blood flow structures and rheological behaviors in microfluidic systems, some blood flow patterns are introduced in the following section.

2.1.1. RBC Aggregation

RBC aggregation is a vital phenomenon to be considered when designing LOC devices due to its impact on the viscosity of the bulk flow. Under conditions of low shear rate, such as in stagnant or low flow rate regions, RBCs tend to aggregate, forming structures known as rouleaux, resembling stacks of coins as shown in Figure 1(b). 

(13) The aggregation of RBCs increases the viscosity at the aggregated region, 

(14) hence slowing down the overall blood flow. However, when exposed to high shear rates, RBC aggregates disaggregate. As shear rates continue to increase, RBCs tend to deform, elongating and aligning themselves with the direction of the flow. 

(15) Such a dynamic shift in behavior from the cells in response to the shear rate forms the basis of the viscoelastic properties observed in whole blood. In essence, the viscosity of the blood varies according to the shear rate conditions, which are related to the velocity gradient of the system. It is significant to take the intricate relationship between shear rate conditions and the change of blood viscosity due to RBC aggregation into account since various flow driving conditions may induce varied effects on the degree of aggregation.

2.1.2. Fåhræus-Lindqvist Effect

The Fåhræus–Lindqvist (FL) effect describes the gradual decrease in the apparent viscosity of blood as the channel diameter decreases. 

(16) This effect is attributed to the migration of RBCs toward the central region in the microchannel, where the flow rate is higher, due to the presence of higher pressure and asymmetric distribution of shear forces. This migration of RBCs, typically observed at blood vessels less than 0.3 mm, toward the higher flow rate region contributes to the change in blood viscosity, which becomes dependent on the channel size. Simultaneously, the increase of the RBC concentration in the central region of the microchannel results in the formation of a less viscous region close to the microchannel wall. This region called the Cell-Free Layer (CFL), is primarily composed of plasma. 

(17) The combination of the FL effect and the following CFL formation provides a unique phenomenon that is often utilized in passive and active plasma separation mechanisms, involving branched and constriction channels for various applications in plasma separation using microfluidic systems.

2.1.3. Cell-Free Layer Formation

In microfluidic blood flow, RBCs form aggregates at the microchannel core and result in a region that is mostly devoid of RBCs near the microchannel walls, as shown in Figure 1(c). 

(18) The region is known as the cell-free layer (CFL). The CFL region is often known to possess a lower viscosity compared to other regions within the blood flow due to the lower viscosity value of plasma when compared to that of the aggregated RBCs. Therefore, a thicker CFL region composed of plasma correlates to a reduced apparent whole blood viscosity. 

(19) A thicker CFL region is often established following the RBC aggregation at the microchannel core under conditions of decreasing the tube diameter. Apart from the dependence on the RBC concentration in the microchannel core, the CFL thickness is also affected by the volume concentration of RBCs, or hematocrit, in whole blood, as well as the deformability of RBCs. Given the influence CFL thickness has on blood flow rheological parameters such as blood flow rate, which is strongly dependent on whole blood viscosity, investigating CFL thickness under shear flow is crucial for LOC systems accounting for blood flow.

2.1.4. Plasma Skimming in Bifurcation Networks

The uneven arrangement of RBCs in bifurcating microchannels, commonly termed skimming bifurcation, arises from the axial migration of RBCs within flowing streams. This uneven distribution contributes to variations in viscosity across differing sizes of bifurcating channels but offers a stabilizing effect. Notably, higher flow rates in microchannels are associated with increased hematocrit levels, resulting in higher viscosity compared with those with lower flow rates. Parametric investigations on bifurcation angle, 

(20) thickness of the CFL, 

(21) and RBC dynamics, including aggregation and deformation, 

(22) may alter the varying viscosity of blood and its flow behavior within microchannels.

2.2. Modeling on Blood Flow Dynamics

2.2.1. Blood Properties and Mathematical Models of Blood Rheology

Under different shear rate conditions in blood flow, the elastic characteristics and dynamic changes of the RBC induce a complex velocity and stress relationship, resulting in the incompatibility of blood flow characterization through standard presumptions of constant viscosity used for Newtonian fluid flow. Blood flow is categorized as a viscoelastic non-Newtonian fluid flow where constitutive equations governing this type of flow take into consideration the nonlinear viscometric properties of blood. To mathematically characterize the evolving blood viscosity and the relationship between the elasticity of RBC and the shear blood flow, respectively, across space and time of the system, a stress tensor (τ) defined by constitutive models is often coupled in the Navier–Stokes equation to account for the collective impact of the constant dynamic viscosity (η) and the elasticity from RBCs on blood flow.The dynamic viscosity of blood is heavily dependent on the shear stress applied to the cell and various parameters from the blood such as hematocrit value, plasma viscosity, mechanical properties of the RBC membrane, and red blood cell aggregation rate. The apparent blood viscosity is considered convenient for the characterization of the relationship between the evolving blood viscosity and shear rate, which can be defined by Casson’s law, as shown in eq 1.

𝜇=𝜏0𝛾˙+2𝜂𝜏0𝛾˙⎯⎯⎯⎯⎯⎯⎯√+𝜂�=�0�˙+2��0�˙+�

(1)where τ

0 is the yield stress–stress required to initiate blood flow motion, η is the Casson rheological constant, and γ̇ is the shear rate. The value of Casson’s law parameters under blood with normal hematocrit level can be defined as τ

0 = 0.0056 Pa and η = 0.0035 Pa·s. 

(23) With the known property of blood and Casson’s law parameters, an approximation can be made to the dynamic viscosity under various flow condition domains. The Power Law model is often employed to characterize the dynamic viscosity in relation to the shear rate, since precise solutions exist for specific geometries and flow circumstances, acting as a fundamental standard for definition. The Carreau and Carreau–Yasuda models can be advantageous over the Power Law model due to their ability to evaluate the dynamic viscosity at low to zero shear rate conditions. However, none of the above-mentioned models consider the memory or other elastic behavior of blood and its RBCs. Some other commonly used mathematical models and their constants for the non-Newtonian viscosity property characterization of blood are listed in Table 1 below. 

(24−26)Table 1. Comparison of Various Non-Newtonian Models for Blood Viscosity 

(24−26)

ModelNon-Newtonian ViscosityParameters
Power Law(2)n = 0.61, k = 0.42
Carreau(3)μ0 = 0.056 Pa·s, μ = 0.00345 Pa·s, λ = 3.1736 s, m = 2.406, a = 0.254
Walburn–Schneck(4)C1 = 0.000797 Pa·s, C2 = 0.0608 Pa·s, C3 = 0.00499, C4 = 14.585 g–1, TPMA = 25 g/L
Carreau–Yasuda(5)μ0 = 0.056 Pa·s, μ = 0.00345 Pa·s, λ = 1.902 s, n = 0.22, a = 1.25
Quemada(6)μp = 0.0012 Pa·s, k = 2.07, k0 = 4.33, γ̇c = 1.88 s–1

The blood rheology is commonly known to be influenced by two key physiological factors, namely, the hematocrit value (H

t) and the fibrinogen concentration (c

f), with an average value of 42% and 0.252 gd·L

–1, respectively. Particularly in low shear conditions, the presence of varying fibrinogen concentrations affects the tendency for aggregation and rouleaux formation, while the occurrence of aggregation is contingent upon specific levels of hematocrit. 

(27) The study from Apostolidis et al. 

(28) modifies the Casson model through emphasizing its reliance on hematocrit and fibrinogen concentration parameter values, owing to the extensive knowledge of the two physiological blood parameters.The viscoelastic response of blood is heavily dependent on the elasticity of the RBC, which is defined by the relationship between the deformation and stress relaxation from RBCs under a specific location of shear flow as a function of the velocity field. The stress tensor is usually characterized by constitutive equations such as the Upper-Convected Maxwell Model 

(29) and the Oldroyd-B model 

(30) to track the molecule effects under shear from different driving forces. The prominent non-Newtonian features, such as shear thinning and yield stress, have played a vital role in the characterization of blood rheology, particularly with respect to the evaluation of yield stress under low shear conditions. The nature of stress measurement in blood, typically on the order of 1 mPa, is challenging due to its low magnitude. The occurrence of the CFL complicates the measurement further due to the significant decrease in apparent viscosity near the wall over time and a consequential disparity in viscosity compared to the bulk region.In addition to shear thinning viscosity and yield stress, the formation of aggregation (rouleaux) from RBCs under low shear rates also contributes to the viscoelasticity under transient flow 

(31) and thixotropy 

(32) of whole blood. Given the difficulty in evaluating viscoelastic behavior of blood under low strain magnitudes and limitations in generalized Newtonian models, the utilization of viscoelastic models is advocated to encompass elasticity and delineate non-shear components within the stress tensor. Extending from the Oldroyd-B model, Anand et al. 

(33) developed a viscoelastic model framework for adapting elasticity within blood samples and predicting non-shear stress components. However, to also address the thixotropic effects, the model developed by Horner et al. 

(34) serves as a more comprehensive approach than the viscoelastic model from Anand et al. Thixotropy 

(32) typically occurs from the structural change of the rouleaux, where low shear rate conditions induce rouleaux formation. Correspondingly, elasticity increases, while elasticity is more representative of the isolated RBCs, under high shear rate conditions. The model of Horner et al. 

(34) considers the contribution of rouleaux to shear stress, taking into account factors such as the characteristic time for Brownian aggregation, shear-induced aggregation, and shear-induced breakage. Subsequent advancements in the model from Horner et al. often revolve around refining the three aforementioned key terms for a more substantial characterization of rouleaux dynamics. Notably, this has led to the recently developed mHAWB model 

(35) and other model iterations to enhance the accuracy of elastic and viscoelastic contributions to blood rheology, including the recently improved model suggested by Armstrong et al. 

(36)

2.2.2. Numerical Methods (FDM, FEM, FVM)

Numerical simulation has become increasingly more significant in analyzing the geometry, boundary layers of flow, and nonlinearity of hyperbolic viscoelastic flow constitutive equations. CFD is a powerful and efficient tool utilizing numerical methods to solve the governing hydrodynamic equations, such as the Navier–Stokes (N–S) equation, continuity equation, and energy conservation equation, for qualitative evaluation of fluid motion dynamics under different parameters. CFD overcomes the challenge of analytically solving nonlinear forms of differential equations by employing numerical methods such as the Finite-Difference Method (FDM), Finite-Element Method (FEM), and Finite-Volume Method (FVM) to discretize and solve the partial differential equations (PDEs), allowing for qualitative reproduction of transport phenomena and experimental observations. Different numerical methods are chosen to cope with various transport systems for optimization of the accuracy of the result and control of error during the discretization process.FDM is a straightforward approach to discretizing PDEs, replacing the continuum representation of equations with a set of finite-difference equations, which is typically applied to structured grids for efficient implementation in CFD programs. 

(37) However, FDM is often limited to simple geometries such as rectangular or block-shaped geometries and struggles with curved boundaries. In contrast, FEM divides the fluid domain into small finite grids or elements, approximating PDEs through a local description of physics. 

(38) All elements contribute to a large, sparse matrix solver. However, FEM may not always provide accurate results for systems involving significant deformation and aggregation of particles like RBCs due to large distortion of grids. 

(39) FVM evaluates PDEs following the conservation laws and discretizes the selected flow domain into small but finite size control volumes, with each grid at the center of a finite volume. 

(40) The divergence theorem allows the conversion of volume integrals of PDEs with divergence terms into surface integrals of surface fluxes across cell boundaries. Due to its conservation property, FVM offers efficient outcomes when dealing with PDEs that embody mass, momentum, and energy conservation principles. Furthermore, widely accessible software packages like the OpenFOAM toolbox 

(41) include a viscoelastic solver, making it an attractive option for viscoelastic fluid flow modeling. 

(42)

2.2.3. Modeling Methods of Blood Flow Dynamics

The complexity in the blood flow simulation arises from deformability and aggregation that RBCs exhibit during their interaction with neighboring cells under different shear rate conditions induced by blood flow. Numerical models coupled with simulation programs have been applied as a groundbreaking method to predict such unique rheological behavior exhibited by RBCs and whole blood. The conventional approach of a single-phase flow simulation is often applied to blood flow simulations within large vessels possessing a moderate shear rate. However, such a method assumes the properties of plasma, RBCs and other cellular components to be evenly distributed as average density and viscosity in blood, resulting in the inability to simulate the mechanical dynamics, such as RBC aggregation under high-shear flow field, inherent in RBCs. To accurately describe the asymmetric distribution of RBC and blood flow, multiphase flow simulation, where numerical simulations of blood flows are often modeled as two immiscible phases, RBCs and blood plasma, is proposed. A common assumption is that RBCs exhibit non-Newtonian behavior while the plasma is treated as a continuous Newtonian phase.Numerous multiphase numerical models have been proposed to simulate the influence of RBCs on blood flow dynamics by different assumptions. In large-scale simulations (above the millimeter range), continuum-based methods are wildly used due to their lower computational demands. 

(43) Eulerian multiphase flow simulations offer the solution of a set of conservation equations for each separate phase and couple the phases through common pressure and interphase exchange coefficients. Xu et al. 

(44) utilized the combined finite-discrete element method (FDEM) to replicate the dynamic behavior and distortion of RBCs subjected to fluidic forces, utilizing the Johnson–Kendall–Roberts model 

(45) to define the adhesive forces of cell-to-cell interactions. The iterative direct-forcing immersed boundary method (IBM) is commonly employed in simulations of the fluid–cell interface of blood. This method effectively captures the intricacies of the thin and flexible RBC membranes within various external flow fields. 

(46) The study by Xu et al. 

(44) also adopts this approach to bridge the fluid dynamics and RBC deformation through IBM. Yoon and You utilized the Maxwell model to define the viscosity of the RBC membrane. 

(47) It was discovered that the Maxwell model could represent the stress relaxation and unloading processes of the cell. Furthermore, the reduced flexibility of an RBC under particular situations such as infection is specified, which was unattainable by the Kelvin–Voigt model 

(48) when compared to the Maxwell model in the literature. The Yeoh hyperplastic material model was also adapted to predict the nonlinear elasticity property of RBCs with FEM employed to discretize the RBC membrane using shell-type elements. Gracka et al. 

(49) developed a numerical CFD model with a finite-volume parallel solver for multiphase blood flow simulation, where an updated Maxwell viscoelasticity model and a Discrete Phase Model are adopted. In the study, the adapted IBM, based on unstructured grids, simulates the flow behavior and shape change of the RBCs through fluid-structure coupling. It was found that the hybrid Euler–Lagrange (E–L) approach 

(50) for the development of the multiphase model offered better results in the simulated CFL region in the microchannels.To study the dynamics of individual behaviors of RBCs and the consequent non-Newtonian blood flow, cell-shape-resolved computational models are often adapted. The use of the boundary integral method has become prevalent in minimizing computational expenses, particularly in the exclusive determination of fluid velocity on the surfaces of RBCs, incorporating the option of employing IBM or particle-based techniques. The cell-shaped-resolved method has enabled an examination of cell to cell interactions within complex ambient or pulsatile flow conditions 

(51) surrounding RBC membranes. Recently, Rydquist et al. 

(52) have looked to integrate statistical information from macroscale simulations to obtain a comprehensive overview of RBC behavior within the immediate proximity of the flow through introduction of respective models characterizing membrane shape definition, tension, bending stresses of RBC membranes.At a macroscopic scale, continuum models have conventionally been adapted for assessing blood flow dynamics through the application of elasticity theory and fluid dynamics. However, particle-based methods are known for their simplicity and adaptability in modeling complex multiscale fluid structures. Meshless methods, such as the boundary element method (BEM), smoothed particle hydrodynamics (SPH), and dissipative particle dynamics (DPD), are often used in particle-based characterization of RBCs and the surrounding fluid. By representing the fluid as discrete particles, meshless methods provide insights into the status and movement of the multiphase fluid. These methods allow for the investigation of cellular structures and microscopic interactions that affect blood rheology. Non-confronting mesh methods like IBM can also be used to couple a fluid solver such as FEM, FVM, or the Lattice Boltzmann Method (LBM) through membrane representation of RBCs. In comparison to conventional CFD methods, LBM has been viewed as a favorable numerical approach for solving the N–S equations and the simulation of multiphase flows. LBM exhibits the notable advantage of being amenable to high-performance parallel computing environments due to its inherently local dynamics. In contrast to DPD and SPH where RBC membranes are modeled as physically interconnected particles, LBM employs the IBM to account for the deformation dynamics of RBCs 

(53,54) under shear flows in complex channel geometries. 

(54,55) However, it is essential to acknowledge that the utilization of LBM in simulating RBC flows often entails a significant computational overhead, being a primary challenge in this context. Krüger et al. 

(56) proposed utilizing LBM as a fluid solver, IBM to couple the fluid and FEM to compute the response of membranes to deformation under immersed fluids. This approach decouples the fluid and membranes but necessitates significant computational effort due to the requirements of both meshes and particles.Despite the accuracy of current blood flow models, simulating complex conditions remains challenging because of the high computational load and cost. Balachandran Nair et al. 

(57) suggested a reduced order model of RBC under the framework of DEM, where the RBC is represented by overlapping constituent rigid spheres. The Morse potential force is adapted to account for the RBC aggregation exhibited by cell to cell interactions among RBCs at different distances. Based upon the IBM, the reduced-order RBC model is adapted to simulate blood flow transport for validation under both single and multiple RBCs with a resolved CFD-DEM solver. 

(58) In the resolved CFD-DEM model, particle sizes are larger than the grid size for a more accurate computation of the surrounding flow field. A continuous forcing approach is taken to describe the momentum source of the governing equation prior to discretization, which is different from a Direct Forcing Method (DFM). 

(59) As no body-conforming moving mesh is required, the continuous forcing approach offers lower complexity and reduced cost when compared to the DFM. Piquet et al. 

(60) highlighted the high complexity of the DFM due to its reliance on calculating an additional immersed boundary flux for the velocity field to ensure its divergence-free condition.The fluid–structure interaction (FSI) method has been advocated to connect the dynamic interplay of RBC membranes and fluid plasma within blood flow such as the coupling of continuum–particle interactions. However, such methodology is generally adapted for anatomical configurations such as arteries 

(61,62) and capillaries, 

(63) where both the structural components and the fluid domain undergo substantial deformation due to the moving boundaries. Due to the scope of the Review being blood flow simulation within microchannels of LOC devices without deformable boundaries, the Review of the FSI method will not be further carried out.In general, three numerical methods are broadly used: mesh-based, particle-based, and hybrid mesh–particle techniques, based on the spatial scale and the fundamental numerical approach, mesh-based methods tend to neglect the effects of individual particles, assuming a continuum and being efficient in terms of time and cost. However, the particle-based approach highlights more of the microscopic and mesoscopic level, where the influence of individual RBCs is considered. A review from Freund et al. 

(64) addressed the three numerical methodologies and their respective modeling approaches of RBC dynamics. Given the complex mechanics and the diverse levels of study concerning numerical simulations of blood and cellular flow, a broad spectrum of numerical methods for blood has been subjected to extensive review. 

(64−70) Ye at al. 

(65) offered an extensive review of the application of the DPD, SPH, and LBM for numerical simulations of RBC, while Rathnayaka et al. 

(67) conducted a review of the particle-based numerical modeling for liquid marbles through drawing parallels to the transport of RBCs in microchannels. A comparative analysis between conventional CFD methods and particle-based approaches for cellular and blood flow dynamic simulation can be found under the review by Arabghahestani et al. 

(66) Literature by Li et al. 

(68) and Beris et al. 

(69) offer an overview of both continuum-based models at micro/macroscales and multiscale particle-based models encompassing various length and temporal dimensions. Furthermore, these reviews deliberate upon the potential of coupling continuum-particle methods for blood plasma and RBC modeling. Arciero et al. 

(70) investigated various modeling approaches encompassing cellular interactions, such as cell to cell or plasma interactions and the individual cellular phases. A concise overview of the reviews is provided in Table 2 for reference.

Table 2. List of Reviews for Numerical Approaches Employed in Blood Flow Simulation

ReferenceNumerical methods
Li et al. (2013) (68)Continuum-based modeling (BIM), particle-based modeling (LBM, LB-FE, SPH, DPD)
Freund (2014) (64)RBC dynamic modeling (continuum-based modeling, complementary discrete microstructure modeling), blood flow dynamic modeling (FDM, IBM, LBM, particle-mesh methods, coupled boundary integral and mesh-based methods, DPD)
Ye et al. (2016) (65)DPD, SPH, LBM, coupled IBM-Smoothed DPD
Arciero et al. (2017) (70)LBM, IBM, DPD, conventional CFD Methods (FDM, FVM, FEM)
Arabghahestani et al. (2019) (66)Particle-based methods (LBM, DPD, direct simulation Monte Carlo, molecular dynamics), SPH, conventional CFD methods (FDM, FVM, FEM)
Beris et al. (2021) (69)DPD, smoothed DPD, IBM, LBM, BIM
Rathnayaka (2022) (67)SPH, CG, LBM

3. Capillary Driven Blood Flow in LOC Systems

ARTICLE SECTIONS

Jump To


3.1. Capillary Driven Flow Phenomena

Capillary driven (CD) flow is a pivotal mechanism in passive microfluidic flow systems 

(9) such as the blood circulation system and LOC systems. 

(71) CD flow is essentially the movement of a liquid to flow against drag forces, where the capillary effect exerts a force on the liquid at the borders, causing a liquid–air meniscus to flow despite gravity or other drag forces. A capillary pressure drops across the liquid–air interface with surface tension in the capillary radius and contact angle. The capillary effect depends heavily on the interaction between the different properties of surface materials. Different values of contact angles can be manipulated and obtained under varying levels of surface wettability treatments to manipulate the surface properties, resulting in different CD blood delivery rates for medical diagnostic device microchannels. CD flow techniques are appealing for many LOC devices, because they require no external energy. However, due to the passive property of liquid propulsion by capillary forces and the long-term instability of surface treatments on channel walls, the adaptability of CD flow in geometrically complex LOC devices may be limited.

3.2. Theoretical and Numerical Modeling of Capillary Driven Blood Flow

3.2.1. Theoretical Basis and Assumptions of Microfluidic Flow

The study of transport phenomena regarding either blood flow driven by capillary forces or externally applied forces under microfluid systems all demands a comprehensive recognition of the significant differences in flow dynamics between microscale and macroscale. The fundamental assumptions and principles behind fluid transport at the microscale are discussed in this section. Such a comprehension will lay the groundwork for the following analysis of the theoretical basis of capillary forces and their role in blood transport in LOC systems.

At the macroscale, fluid dynamics are often strongly influenced by gravity due to considerable fluid mass. However, the high surface to volume ratio at the microscale shifts the balance toward surface forces (e.g., surface tension and viscous forces), much larger than the inertial force. This difference gives rise to transport phenomena unique to microscale fluid transport, such as the prevalence of laminar flow due to a very low Reynolds number (generally lower than 1). Moreover, the fluid in a microfluidic system is often assumed to be incompressible due to the small flow velocity, indicating constant fluid density in both space and time.Microfluidic flow behaviors are governed by the fundamental principles of mass and momentum conservation, which are encapsulated in the continuity equation and the Navier–Stokes (N–S) equation. The continuity equation describes the conservation of mass, while the N–S equation captures the spatial and temporal variations in velocity, pressure, and other physical parameters. Under the assumption of the negligible influence of gravity in microfluidic systems, the continuity equation and the Eulerian representation of the incompressible N–S equation can be expressed as follows:

∇·𝐮⇀=0∇·�⇀=0

(7)

−∇𝑝+𝜇∇2𝐮⇀+∇·𝝉⇀−𝐅⇀=0−∇�+�∇2�⇀+∇·�⇀−�⇀=0

(8)Here, p is the pressure, u is the fluid viscosity, 

𝝉⇀�⇀ represents the stress tensor, and F is the body force exerted by external forces if present.

3.2.2. Theoretical Basis and Modeling of Capillary Force in LOC Systems

The capillary force is often the major driving force to manipulate and transport blood without an externally applied force in LOC systems. Forces induced by the capillary effect impact the free surface of fluids and are represented not directly in the Navier–Stokes equations but through the pressure boundary conditions of the pressure term p. For hydrophilic surfaces, the liquid generally induces a contact angle between 0° and 30°, encouraging the spread and attraction of fluid under a positive cos θ condition. For this condition, the pressure drop becomes positive and generates a spontaneous flow forward. A hydrophobic solid surface repels the fluid, inducing minimal contact. Generally, hydrophobic solids exhibit a contact angle larger than 90°, inducing a negative value of cos θ. Such a value will result in a negative pressure drop and a flow in the opposite direction. The induced contact angle is often utilized to measure the wall exposure of various surface treatments on channel walls where different wettability gradients and surface tension effects for CD flows are established. Contact angles between different interfaces are obtainable through standard values or experimental methods for reference. 

(72)For the characterization of the induced force by the capillary effect, the Young–Laplace (Y–L) equation 

(73) is widely employed. In the equation, the capillary is considered a pressure boundary condition between the two interphases. Through the Y–L equation, the capillary pressure force can be determined, and subsequently, the continuity and momentum balance equations can be solved to obtain the blood filling rate. Kim et al. 

(74) studied the effects of concentration and exposure time of a nonionic surfactant, Silwet L-77, on the performance of a polydimethylsiloxane (PDMS) microchannel in terms of plasma and blood self-separation. The study characterized the capillary pressure force by incorporating the Y–L equation and further evaluated the effects of the changing contact angle due to different levels of applied channel wall surface treatments. The expression of the Y–L equation utilized by Kim et al. 

(74) is as follows:

𝑃=−𝜎(cos𝜃b+cos𝜃tℎ+cos𝜃l+cos𝜃r𝑤)�=−�(cos⁡�b+cos⁡�tℎ+cos⁡�l+cos⁡�r�)

(9)where σ is the surface tension of the liquid and θ

bθ

tθ

l, and θ

r are the contact angle values between the liquid and the bottom, top, left, and right walls, respectively. A numerical simulation through Coventor software is performed to evaluate the dynamic changes in the filling rate within the microchannel. The simulation results for the blood filling rate in the microchannel are expressed at a specific time stamp, shown in Figure 2. The results portray an increasing instantaneous filling rate of blood in the microchannel following the decrease in contact angle induced by a higher concentration of the nonionic surfactant treated to the microchannel wall.

Figure 2. Numerical simulation of filling rate of capillary driven blood flow under various contact angle conditions at a specific timestamp. (74) Reproduced with permission from ref (74). Copyright 2010 Elsevier.

When in contact with hydrophilic or hydrophobic surfaces, blood forms a meniscus with a contact angle due to surface tension. The Lucas–Washburn (L–W) equation 

(75) is one of the pioneering theoretical definitions for the position of the meniscus over time. In addition, the L–W equation provides the possibility for research to obtain the velocity of the blood formed meniscus through the derivation of the meniscus position. The L–W equation 

(75) can be shown below:

𝐿(𝑡)=𝑅𝜎cos(𝜃)𝑡2𝜇⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯√�(�)=��⁡cos(�)�2�

(10)Here L(t) represents the distance of the liquid driven by the capillary forces. However, the generalized L–W equation solely assumes the constant physical properties from a Newtonian fluid rather than considering the non-Newtonian fluid behavior of blood. Cito et al. 

(76) constructed an enhanced version of the L–W equation incorporating the power law to consider the RBC aggregation and the FL effect. The non-Newtonian fluid apparent viscosity under the Power Law model is defined as

𝜇=𝑘·(𝛾˙)𝑛−1�=�·(�˙)�−1

(11)where γ̇ is the strain rate tensor defined as 

𝛾˙=12𝛾˙𝑖𝑗𝛾˙𝑗𝑖⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯√�˙=12�˙���˙��. The stress tensor term τ is computed as τ = μγ̇

ij. The updated L–W equation by Cito 

(76) is expressed as

𝐿(𝑡)=𝑅[(𝑛+13𝑛+1)(𝜎cos(𝜃)𝑅𝑘)1/𝑛𝑡]𝑛/𝑛+1�(�)=�[(�+13�+1)(�⁡cos(�)��)1/��]�/�+1

(12)where k is the flow consistency index and n is the power law index, respectively. The power law index, from the Power Law model, characterizes the extent of the non-Newtonian behavior of blood. Both the consistency and power law index rely on blood properties such as hematocrit, the appearance of the FL effect, the formation of RBC aggregates, etc. The updated L–W equation computes the location and velocity of blood flow caused by capillary forces at specified time points within the LOC devices, taking into account the effects of blood flow characteristics such as RBC aggregation and the FL effect on dynamic blood viscosity.Apart from the blood flow behaviors triggered by inherent blood properties, unique flow conditions driven by capillary forces that are portrayed under different microchannel geometries also hold crucial implications for CD blood delivery. Berthier et al. 

(77) studied the spontaneous Concus–Finn condition, the condition to initiate the spontaneous capillary flow within a V-groove microchannel, as shown in Figure 3(a) both experimentally and numerically. Through experimental studies, the spontaneous Concus–Finn filament development of capillary driven blood flow is observed, as shown in Figure 3(b), while the dynamic development of blood flow is numerically simulated through CFD simulation.

Figure 3. (a) Sketch of the cross-section of Berthier’s V-groove microchannel, (b) experimental view of blood in the V-groove microchannel, (78) (c) illustration of the dynamic change of the extension of filament from FLOW 3D under capillary flow at three increasing time intervals. (78) Reproduced with permission from ref (78). Copyright 2014 Elsevier.

Berthier et al. 

(77) characterized the contact angle needed for the initiation of the capillary driving force at a zero-inlet pressure, through the half-angle (α) of the V-groove geometry layout, and its relation to the Concus–Finn filament as shown below:

𝜃<𝜋2−𝛼sin𝛼1+2(ℎ2/𝑤)sin𝛼<cos𝜃{�<�2−�sin⁡�1+2(ℎ2/�)⁡sin⁡�<cos⁡�

(13)Three possible regimes were concluded based on the contact angle value for the initiation of flow and development of Concus–Finn filament:

𝜃>𝜃1𝜃1>𝜃>𝜃0𝜃0no SCFSCF without a Concus−Finn filamentSCF without a Concus−Finn filament{�>�1no SCF�1>�>�0SCF without a Concus−Finn filament�0SCF without a Concus−Finn filament

(14)Under Newton’s Law, the force balance with low Reynolds and Capillary numbers results in the neglect of inertial terms. The force balance between the capillary forces and the viscous force induced by the channel wall is proposed to derive the analytical fluid velocity. This relation between the two forces offers insights into the average flow velocity and the penetration distance function dependent on time. The apparent blood viscosity is defined by Berthier et al. 

(78) through Casson’s law, 

(23) given in eq 1. The research used the FLOW-3D program from Flow Science Inc. software, which solves transient, free-surface problems using the FDM in multiple dimensions. The Volume of Fluid (VOF) method 

(79) is utilized to locate and track the dynamic extension of filament throughout the advancing interface within the channel ahead of the main flow at three progressing time stamps, as depicted in Figure 3(c).

4. Electro-osmotic Flow (EOF) in LOC Systems

ARTICLE SECTIONS

Jump To


The utilization of external forces, such as electric fields, has significantly broadened the possibility of manipulating microfluidic flow in LOC systems. 

(80) Externally applied electric field forces induce a fluid flow from the movement of ions in fluid terms as the “electro-osmotic flow” (EOF).Unique transport phenomena, such as enhanced flow velocity and flow instability, induced by non-Newtonian fluids, particularly viscoelastic fluids, under EOF, have sparked considerable interest in microfluidic devices with simple or complicated geometries within channels. 

(81) However, compared to the study of Newtonian fluids and even other electro-osmotic viscoelastic fluid flows, the literature focusing on the theoretical and numerical modeling of electro-osmotic blood flow is limited due to the complexity of blood properties. Consequently, to obtain a more comprehensive understanding of the complex blood flow behavior under EOF, theoretical and numerical studies of the transport phenomena in the EOF section will be based on the studies of different viscoelastic fluids under EOF rather than that of blood specifically. Despite this limitation, we believe these studies offer valuable insights that can help understand the complex behavior of blood flow under EOF.

4.1. EOF Phenomena

Electro-osmotic flow occurs at the interface between the microchannel wall and bulk phase solution. When in contact with the bulk phase, solution ions are absorbed or dissociated at the solid–liquid interface, resulting in the formation of a charge layer, as shown in Figure 4. This charged channel surface wall interacts with both negative and positive ions in the bulk sample, causing repulsion and attraction forces to create a thin layer of immobilized counterions, known as the Stern layer. The induced electric potential from the wall gradually decreases with an increase in the distance from the wall. The Stern layer potential, commonly termed the zeta potential, controls the intensity of the electrostatic interactions between mobile counterions and, consequently, the drag force from the applied electric field. Next to the Stern layer is the diffuse mobile layer, mainly composed of a mobile counterion. These two layers constitute the “electrical double layer” (EDL), the thickness of which is directly proportional to the ionic strength (concentration) of the bulk fluid. The relationship between the two parameters is characterized by a Debye length (λ

D), expressed as

𝜆𝐷=𝜖𝑘B𝑇2(𝑍𝑒)2𝑐0⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯√��=��B�2(��)2�0

(15)where ϵ is the permittivity of the electrolyte solution, k

B is the Boltzmann constant, T is the electron temperature, Z is the integer valence number, e is the elementary charge, and c

0 is the ionic density.

Figure 4. Schematic diagram of an electro-osmotic flow in a microchannel with negative surface charge. (82) Reproduced with permission from ref (82). Copyright 2012 Woodhead Publishing.

When an electric field is applied perpendicular to the EDL, viscous drag is generated due to the movement of excess ions in the EDL. Electro-osmotic forces can be attributed to the externally applied electric potential (ϕ) and the zeta potential, the system wall induced potential by charged walls (ψ). As illustrated in Figure 4, the majority of ions in the bulk phase have a uniform velocity profile, except for a shear rate condition confined within an extremely thin Stern layer. Therefore, EOF displays a unique characteristic of a “near flat” or plug flow velocity profile, different from the parabolic flow typically induced by pressure-driven microfluidic flow (Hagen–Poiseuille flow). The plug-shaped velocity profile of the EOF possesses a high shear rate above the Stern layer.Overall, the EOF velocity magnitude is typically proportional to the Debye Length (λ

D), zeta potential, and magnitude of the externally applied electric field, while a more viscous liquid reduces the EOF velocity.

4.2. Modeling on Electro-osmotic Viscoelastic Fluid Flow

4.2.1. Theoretical Basis of EOF Mechanisms

The EOF of an incompressible viscoelastic fluid is commonly governed by the continuity and incompressible N–S equations, as shown in eqs 7 and 8, where the stress tensor and the electrostatic force term are coupled. The electro-osmotic body force term F, representing the body force exerted by the externally applied electric force, is defined as 

𝐹⇀=𝑝𝐸𝐸⇀�⇀=���⇀, where ρ

E and 

𝐸⇀�⇀ are the net electric charge density and the applied external electric field, respectively.Numerous models are established to theoretically study the externally applied electric potential and the system wall induced potential by charged walls. The following Laplace equation, expressed as eq 16, is generally adapted and solved to calculate the externally applied potential (ϕ).

∇2𝜙=0∇2�=0

(16)Ion diffusion under applied electric fields, together with mass transport resulting from convection and diffusion, transports ionic solutions in bulk flow under electrokinetic processes. The Nernst–Planck equation can describe these transport methods, including convection, diffusion, and electro-diffusion. Therefore, the Nernst–Planck equation is used to determine the distribution of the ions within the electrolyte. The electric potential induced by the charged channel walls follows the Poisson–Nernst–Plank (PNP) equation, which can be written as eq 17.

∇·[𝐷𝑖∇𝑛𝑖−𝑢⇀𝑛𝑖+𝑛𝑖𝐷𝑖𝑧𝑖𝑒𝑘𝑏𝑇∇(𝜙+𝜓)]=0∇·[��∇��−�⇀��+����������∇(�+�)]=0

(17)where D

in

i, and z

i are the diffusion coefficient, ionic concentration, and ionic valence of the ionic species I, respectively. However, due to the high nonlinearity and numerical stiffness introduced by different lengths and time scales from the PNP equations, the Poisson–Boltzmann (PB) model is often considered the major simplified method of the PNP equation to characterize the potential distribution of the EDL region in microchannels. In the PB model, it is assumed that the ionic species in the fluid follow the Boltzmann distribution. This model is typically valid for steady-state problems where charge transport can be considered negligible, the EDLs do not overlap with each other, and the intrinsic potentials are low. It provides a simplified representation of the potential distribution in the EDL region. The PB equation governing the EDL electric potential distribution is described as

∇2𝜓=(2𝑒𝑧𝑛0𝜀𝜀0)sinh(𝑧𝑒𝜓𝑘b𝑇)∇2�=(2���0��0)⁡sinh(����b�)

(18)where n

0 is the ion bulk concentration, z is the ionic valence, and ε

0 is the electric permittivity in the vacuum. Under low electric potential conditions, an even further simplified model to illustrate the EOF phenomena is the Debye–Hückel (DH) model. The DH model is derived by obtaining a charge density term by expanding the exponential term of the Boltzmann equation in a Taylor series.

4.2.2. EOF Modeling for Viscoelastic Fluids

Many studies through numerical modeling were performed to obtain a deeper understanding of the effect exhibited by externally applied electric fields on viscoelastic flow in microchannels under various geometrical designs. Bello et al. 

(83) found that methylcellulose solution, a non-Newtonian polymer solution, resulted in stronger electro-osmotic mobility in experiments when compared to the predictions by the Helmholtz–Smoluchowski equation, which is commonly used to define the velocity of EOF of a Newtonian fluid. Being one of the pioneers to identify the discrepancies between the EOF of Newtonian and non-Newtonian fluids, Bello et al. attributed such discrepancies to the presence of a very high shear rate in the EDL, resulting in a change in the orientation of the polymer molecules. Park and Lee 

(84) utilized the FVM to solve the PB equation for the characterization of the electric field induced force. In the study, the concept of fractional calculus for the Oldroyd-B model was adapted to illustrate the elastic and memory effects of viscoelastic fluids in a straight microchannel They observed that fluid elasticity and increased ratio of viscoelastic fluid contribution to overall fluid viscosity had a significant impact on the volumetric flow rate and sensitivity of velocity to electric field strength compared to Newtonian fluids. Afonso et al. 

(85) derived an analytical expression for EOF of viscoelastic fluid between parallel plates using the DH model to account for a zeta potential condition below 25 mV. The study established the understanding of the electro-osmotic viscoelastic fluid flow under low zeta potential conditions. Apart from the electrokinetic forces, pressure forces can also be coupled with EOF to generate a unique fluid flow behavior within the microchannel. Sousa et al. 

(86) analytically studied the flow of a standard viscoelastic solution by combining the pressure gradient force with an externally applied electric force. It was found that, at a near wall skimming layer and the outer layer away from the wall, macromolecules migrating away from surface walls in viscoelastic fluids are observed. In the study, the Phan-Thien Tanner (PTT) constitutive model is utilized to characterize the viscoelastic properties of the solution. The approach is found to be valid when the EDL is much thinner than the skimming layer under an enhanced flow rate. Zhao and Yang 

(87) solved the PB equation and Carreau model for the characterization of the EOF mechanism and non-Newtonian fluid respectively through the FEM. The numerical results depict that, different from the EOF of Newtonian fluids, non-Newtonian fluids led to an increase of electro-osmotic mobility for shear thinning fluids but the opposite for shear thickening fluids.Like other fluid transport driving forces, EOF within unique geometrical layouts also portrays unique transport phenomena. Pimenta and Alves 

(88) utilized the FVM to perform numerical simulations of the EOF of viscoelastic fluids considering the PB equation and the Oldroyd-B model, in a cross-slot and flow-focusing microdevices. It was found that electroelastic instabilities are formed due to the development of large stresses inside the EDL with streamlined curvature at geometry corners. Bezerra et al. 

(89) used the FDM to numerically analyze the vortex formation and flow instability from an electro-osmotic non-Newtonian fluid flow in a microchannel with a nozzle geometry and parallel wall geometry setting. The PNP equation is utilized to characterize the charge motion in the EOF and the PTT model for non-Newtonian flow characterization. A constriction geometry is commonly utilized in blood flow adapted in LOC systems due to the change in blood flow behavior under narrow dimensions in a microchannel. Ji et al. 

(90) recently studied the EOF of viscoelastic fluid in a constriction microchannel connected by two relatively big reservoirs on both ends (as seen in Figure 5) filled with the polyacrylamide polymer solution, a viscoelastic fluid, and an incompressible monovalent binary electrolyte solution KCl.

Figure 5. Schematic diagram of a negatively charged constriction microchannel connected to two reservoirs at both ends. An electro-osmotic flow is induced in the system by the induced potential difference between the anode and cathode. (90) Reproduced with permission from ref (90). Copyright 2021 The Authors, under the terms of the Creative Commons (CC BY 4.0) License https://creativecommons.org/licenses/by/4.0/.

In studying the EOF of viscoelastic fluids, the Oldroyd-B model is often utilized to characterize the polymeric stress tensor and the deformation rate of the fluid. The Oldroyd-B model is expressed as follows:

𝜏=𝜂p𝜆(𝐜−𝐈)�=�p�(�−�)

(19)where η

p, λ, c, and I represent the polymer dynamic viscosity, polymer relaxation time, symmetric conformation tensor of the polymer molecules, and the identity matrix, respectively.A log-conformation tensor approach is taken to prevent convergence difficulty induced by the viscoelastic properties. The conformation tensor (c) in the polymeric stress tensor term is redefined by a new tensor (Θ) based on the natural logarithm of the c. The new tensor is defined as

Θ=ln(𝐜)=𝐑ln(𝚲)𝐑Θ=ln(�)=�⁡ln(�)�

(20)in which Λ is the diagonal matrix and R is the orthogonal matrix.Under the new conformation tensor, the induced EOF of a viscoelastic fluid is governed by the continuity and N–S equations adapting the Oldroyd-B model, which is expressed as

∂𝚯∂𝑡+𝐮·∇𝚯=𝛀Θ−ΘΩ+2𝐁+1𝜆(eΘ−𝐈)∂�∂�+�·∇�=�Θ−ΘΩ+2�+1�(eΘ−�)

(21)where Ω and B represent the anti-symmetric matrix and the symmetric traceless matrix of the decomposition of the velocity gradient tensor ∇u, respectively. The conformation tensor can be recovered by c = exp(Θ). The PB model and Laplace equation are utilized to characterize the charged channel wall induced potential and the externally applied potential.The governing equations are numerically solved through the FVM by RheoTool, 

(42) an open-source viscoelastic EOF solver on the OpenFOAM platform. A SIMPLEC (Semi-Implicit Method for Pressure Linked Equations-Consistent) algorithm was applied to solve the velocity-pressure coupling. The pressure field and velocity field were computed by the PCG (Preconditioned Conjugate Gradient) solver and the PBiCG (Preconditioned Biconjugate Gradient) solver, respectively.Ranging magnitudes of an applied electric field or fluid concentration induce both different streamlines and velocity magnitudes at various locations and times of the microchannel. In the study performed by Ji et al., 

(90) notable fluctuation of streamlines and vortex formation is formed at the upper stream entrance of the constriction as shown in Figure 6(a) and (b), respectively, due to the increase of electrokinetic effect, which is seen as a result of the increase in polymeric stress (τ

xx). 

(90) The contraction geometry enhances the EOF velocity within the constriction channel under high E

app condition (600 V/cm). Such phenomena can be attributed to the dependence of electro-osmotic viscoelastic fluid flow on the system wall surface and bulk fluid properties. 

(91)

Figure 6. Schematic diagram of vortex formation and streamlines of EOF depicting flow instability at (a) 1.71 s and (b) 1.75 s. Spatial distribution of the elastic normal stress at (c) high Eapp condition. Streamline of an electro-osmotic flow under Eapp of 600 V/cm (90) for (d) non-Newtonian and (e) Newtonian fluid through a constriction geometry. Reproduced with permission from ref (90). Copyright 2021 The Authors, under the terms of the Creative Commons (CC BY 4.0) License https://creativecommons.org/licenses/by/4.0/.

As elastic normal stress exceeds the local shear stress, flow instability and vortex formation occur. The induced elastic stress under EOF not only enhances the instability of the flow but often generates an irregular secondary flow leading to strong disturbance. 

(92) It is also vital to consider the effect of the constriction layout of microchannels on the alteration of the field strength within the system. The contraction geometry enhances a larger electric field strength compared with other locations of the channel outside the constriction region, resulting in a higher velocity gradient and stronger extension on the polymer within the viscoelastic solution. Following the high shear flow condition, a higher magnitude of stretch for polymer molecules in viscoelastic fluids exhibits larger elastic stresses and enhancement of vortex formation at the region. 

(93)As shown in Figure 6(c), significant elastic normal stress occurs at the inlet of the constriction microchannel. Such occurrence of a polymeric flow can be attributed to the dominating elongational flow, giving rise to high deformation of the polymers within the viscoelastic fluid flow, resulting in higher elastic stress from the polymers. Such phenomena at the entrance result in the difference in velocity streamline as circled in Figure 6(d) compared to that of the Newtonian fluid at the constriction entrance in Figure 6(e). 

(90) The difference between the Newtonian and polymer solution at the exit, as circled in Figure 6(d) and (e), can be attributed to the extrudate swell effect of polymers 

(94) within the viscoelastic fluid flow. The extrudate swell effect illustrates that, as polymers emerge from the constriction exit, they tend to contract in the flow direction and grow in the normal direction, resulting in an extrudate diameter greater than the channel size. The deformation of polymers within the polymeric flow at both the entrance and exit of the contraction channel facilitates the change in shear stress conditions of the flow, leading to the alteration in streamlines of flows for each region.

4.3. EOF Applications in LOC Systems

4.3.1. Mixing in LOC Systems

Rather than relying on the micromixing controlled by molecular diffusion under low Reynolds number conditions, active mixers actively leverage convective instability and vortex formation induced by electro-osmotic flows from alternating current (AC) or direct current (DC) electric fields. Such adaptation is recognized as significant breakthroughs for promotion of fluid mixing in chemical and biological applications such as drug delivery, medical diagnostics, chemical synthesis, and so on. 

(95)Many researchers proposed novel designs of electro-osmosis micromixers coupled with numerical simulations in conjunction with experimental findings to increase their understanding of the role of flow instability and vortex formation in the mixing process under electrokinetic phenomena. Matsubara and Narumi 

(96) numerically modeled the mixing process in a microchannel with four electrodes on each side of the microchannel wall, which generated a disruption through unstable electro-osmotic vortices. It was found that particle mixing was sensitive to both the convection effect induced by the main and secondary vortex within the micromixer and the change in oscillation frequency caused by the supplied AC voltage when the Reynolds number was varied. Qaderi et al. 

(97) adapted the PNP equation to numerically study the effect of the geometry and zeta potential configuration of the microchannel on the mixing process with a combined electro-osmotic pressure driven flow. It was reported that the application of heterogeneous zeta potential configuration enhances the mixing efficiency by around 23% while the height of the hurdles increases the mixing efficiency at most 48.1%. Cho et al. 

(98) utilized the PB model and Laplace equation to numerically simulate the electro-osmotic non-Newtonian fluid mixing process within a wavy and block layout of microchannel walls. The Power Law model is adapted to describe the fluid rheological characteristic. It was found that shear-thinning fluids possess a higher volumetric flow rate, which could result in poorer mixing efficiency compared to that of Newtonian fluids. Numerous studies have revealed that flow instability and vortex generation, in particular secondary vortices produced by barriers or greater magnitudes of heterogeneous zeta potential distribution, enhance mixing by increasing bulk flow velocity and reducing flow distance.To better understand the mechanism of disturbance formed in the system due to externally applied forces, known as electrokinetic instability, literature often utilize the Rayleigh (Ra) number, 

(1) as described below:

𝑅𝑎𝑣=𝑢ev𝑢eo=(𝛾−1𝛾+1)2𝑊𝛿2𝐸el2𝐻2𝜁𝛿Ra�=�ev�eo=(�−1�+1)2��2�el2�2��

(22)where γ is the conductivity ratio of the two streams and can be written as 

𝛾=𝜎el,H𝜎el,L�=�el,H�el,L. The Ra number characterizes the ratio between electroviscous and electro-osmotic flow. A high Ra

v value often results in good mixing. It is evident that fluid properties such as the conductivity (σ) of the two streams play a key role in the formation of disturbances to enhance mixing in microsystems. At the same time, electrokinetic parameters like the zeta potential (ζ) in the Ra number is critical in the characterization of electro-osmotic velocity and a slip boundary condition at the microchannel wall.To understand the mixing result along the channel, the concentration field can be defined and simulated under the assumption of steady state conditions and constant diffusion coefficient for each of the working fluid within the system through the convection–diffusion equation as below:

∂𝑐𝒊∂𝑡+∇⇀(𝑐𝑖𝑢⇀−𝐷𝑖∇⇀𝑐𝒊)=0∂��∂�+∇⇀(���⇀−��∇⇀��)=0

(23)where c

i is the species concentration of species i and D

i is the diffusion coefficient of the corresponding species.The standard deviation of concentration (σ

sd) can be adapted to evaluate the mixing quality of the system. 

(97) The standard deviation for concentration at a specific portion of the channel may be calculated using the equation below:

𝜎sd=∫10(𝐶∗(𝑦∗)−𝐶m)2d𝑦∗∫10d𝑦∗⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯�sd=∫01(�*(�*)−�m)2d�*∫01d�*

(24)where C*(y*) and C

m are the non-dimensional concentration profile and the mean concentration at the portion, respectively. C* is the non-dimensional concentration and can be calculated as 

𝐶∗=𝐶𝐶ref�*=��ref, where C

ref is the reference concentration defined as the bulk solution concentration. The mean concentration profile can be calculated as 

𝐶m=∫10(𝐶∗(𝑦∗)d𝑦∗∫10d𝑦∗�m=∫01(�*(�*)d�*∫01d�*. With the standard deviation of concentration, the mixing efficiency 

(97) can then be calculated as below:

𝜀𝑥=1−𝜎sd𝜎sd,0��=1−�sd�sd,0

(25)where σ

sd,0 is the standard derivation of the case of no mixing. The value of the mixing efficiency is typically utilized in conjunction with the simulated flow field and concentration field to explore the effect of geometrical and electrokinetic parameters on the optimization of the mixing results.

5. Summary

ARTICLE SECTIONS

Jump To


5.1. Conclusion

Viscoelastic fluids such as blood flow in LOC systems are an essential topic to proceed with diagnostic analysis and research through microdevices in the biomedical and pharmaceutical industries. The complex blood flow behavior is tightly controlled by the viscoelastic characteristics of blood such as the dynamic viscosity and the elastic property of RBCs under various shear rate conditions. Furthermore, the flow behaviors under varied driving forces promote an array of microfluidic transport phenomena that are critical to the management of blood flow and other adapted viscoelastic fluids in LOC systems. This review addressed the blood flow phenomena, the complicated interplay between shear rate and blood flow behaviors, and their numerical modeling under LOC systems through the lens of the viscoelasticity characteristic. Furthermore, a theoretical understanding of capillary forces and externally applied electric forces leads to an in-depth investigation of the relationship between blood flow patterns and the key parameters of the two driving forces, the latter of which is introduced through the lens of viscoelastic fluids, coupling numerical modeling to improve the knowledge of blood flow manipulation in LOC systems. The flow disturbances triggered by the EOF of viscoelastic fluids and their impact on blood flow patterns have been deeply investigated due to their important role and applications in LOC devices. Continuous advancements of various numerical modeling methods with experimental findings through more efficient and less computationally heavy methods have served as an encouraging sign of establishing more accurate illustrations of the mechanisms for multiphase blood and other viscoelastic fluid flow transport phenomena driven by various forces. Such progress is fundamental for the manipulation of unique transport phenomena, such as the generated disturbances, to optimize functionalities offered by microdevices in LOC systems.

The following section will provide further insights into the employment of studied blood transport phenomena to improve the functionality of micro devices adapting LOC technology. A discussion of the novel roles that external driving forces play in microfluidic flow behaviors is also provided. Limitations in the computational modeling of blood flow and electrokinetic phenomena in LOC systems will also be emphasized, which may provide valuable insights for future research endeavors. These discussions aim to provide guidance and opportunities for new paths in the ongoing development of LOC devices that adapt blood flow.

5.2. Future Directions

5.2.1. Electro-osmosis Mixing in LOC Systems

Despite substantial research, mixing results through flow instability and vortex formation phenomena induced by electro-osmotic mixing still deviate from the effective mixing results offered by chaotic mixing results such as those seen in turbulent flows. However, recent discoveries of a mixing phenomenon that is generally observed under turbulent flows are found within electro-osmosis micromixers under low Reynolds number conditions. Zhao 

(99) experimentally discovered a rapid mixing process in an AC applied micromixer, where the power spectrum of concentration under an applied voltage of 20 V

p-p induces a −5/3 slope within a frequency range. This value of the slope is considered as the O–C spectrum in macroflows, which is often visible under relatively high Re conditions, such as the Taylor microscale Reynolds number Re > 500 in turbulent flows. 

(100) However, the Re value in the studied system is less than 1 at the specific location and applied voltage. A secondary flow is also suggested to occur close to microchannel walls, being attributed to the increase of convective instability within the system.Despite the experimental phenomenon proposed by Zhao et al., 

(99) the range of effects induced by vital parameters of an EOF mixing system on the enhanced mixing results and mechanisms of disturbance generated by the turbulent-like flow instability is not further characterized. Such a gap in knowledge may hinder the adaptability and commercialization of the discovery of micromixers. One of the parameters for further evaluation is the conductivity gradient of the fluid flow. A relatively strong conductivity gradient (5000:1) was adopted in the system due to the conductive properties of the two fluids. The high conductivity gradients may contribute to the relatively large Rayleigh number and differences in EDL layer thickness, resulting in an unusual disturbance in laminar flow conditions and enhanced mixing results. However, high conductivity gradients are not always achievable by the working fluids due to diverse fluid properties. The reliance on turbulent-like phenomena and rapid mixing results in a large conductivity gradient should be established to prevent the limited application of fluids for the mixing system. In addition, the proposed system utilizes distinct zeta potential distributions at the top and bottom walls due to their difference in material choices, which may be attributed to the flow instability phenomena. Further studies should be made on varying zeta potential magnitude and distribution to evaluate their effect on the slip boundary conditions of the flow and the large shear rate condition close to the channel wall of EOF. Such a study can potentially offer an optimized condition in zeta potential magnitude through material choices and geometrical layout of the zeta potential for better mixing results and manipulation of mixing fluid dynamics. The two vital parameters mentioned above can be varied with the aid of numerical simulation to understand the effect of parameters on the interaction between electro-osmotic forces and electroviscous forces. At the same time, the relationship of developed streamlines of the simulated velocity and concentration field, following their relationship with the mixing results, under the impact of these key parameters can foster more insight into the range of impact that the two parameters have on the proposed phenomena and the microfluidic dynamic principles of disturbances.

In addition, many of the current investigations of electrokinetic mixers commonly emphasize the fluid dynamics of mixing for Newtonian fluids, while the utilization of biofluids, primarily viscoelastic fluids such as blood, and their distinctive response under shear forces in these novel mixing processes of LOC systems are significantly less studied. To develop more compatible microdevice designs and efficient mixing outcomes for the biomedical industry, it is necessary to fill the knowledge gaps in the literature on electro-osmotic mixing for biofluids, where properties of elasticity, dynamic viscosity, and intricate relationship with shear flow from the fluid are further considered.

5.2.2. Electro-osmosis Separation in LOC Systems

Particle separation in LOC devices, particularly in biological research and diagnostics, is another area where disturbances may play a significant role in optimization. 

(101) Plasma analysis in LOC systems under precise control of blood flow phenomena and blood/plasma separation procedures can detect vital information about infectious diseases from particular antibodies and foreign nucleic acids for medical treatments, diagnostics, and research, 

(102) offering more efficient results and simple operating procedures compared to that of the traditional centrifugation method for blood and plasma separation. However, the adaptability of LOC devices for blood and plasma separation is often hindered by microchannel clogging, where flow velocity and plasma yield from LOC devices is reduced due to occasional RBC migration and aggregation at the filtration entrance of microdevices. 

(103)It is important to note that the EOF induces flow instability close to microchannel walls, which may provide further solutions to clogging for the separation process of the LOC systems. Mohammadi et al. 

(104) offered an anti-clogging effect of RBCs at the blood and plasma separating device filtration entry, adjacent to the surface wall, through RBC disaggregation under high shear rate conditions generated by a forward and reverse EOF direction.

Further theoretical and numerical research can be conducted to characterize the effect of high shear rate conditions near microchannel walls toward the detachment of binding blood cells on surfaces and the reversibility of aggregation. Through numerical modeling with varying electrokinetic parameters to induce different degrees of disturbances or shear conditions at channel walls, it may be possible to optimize and better understand the process of disrupting the forces that bind cells to surface walls and aggregated cells at filtration pores. RBCs that migrate close to microchannel walls are often attracted by the adhesion force between the RBC and the solid surface originating from the van der Waals forces. Following RBC migration and attachment by adhesive forces adjacent to the microchannel walls as shown in Figure 7, the increase in viscosity at the region causes a lower shear condition and encourages RBC aggregation (cell–cell interaction), which clogs filtering pores or microchannels and reduces flow velocity at filtration region. Both the impact that shear forces and disturbances may induce on cell binding forces with surface walls and other cells leading to aggregation may suggest further characterization. Kinetic parameters such as activation energy and the rate-determining step for cell binding composition attachment and detachment should be considered for modeling the dynamics of RBCs and blood flows under external forces in LOC separation devices.

Figure 7. Schematic representations of clogging at a microchannel pore following the sequence of RBC migration, cell attachment to channel walls, and aggregation. (105) Reproduced with permission from ref (105). Copyright 2018 The Authors under the terms of the Creative Commons (CC BY 4.0) License https://creativecommons.org/licenses/by/4.0/.

5.2.3. Relationship between External Forces and Microfluidic Systems

In blood flow, a thicker CFL suggests a lower blood viscosity, suggesting a complex relationship between shear stress and shear rate, affecting the blood viscosity and blood flow. Despite some experimental and numerical studies on electro-osmotic non-Newtonian fluid flow, limited literature has performed an in-depth investigation of the role that applied electric forces and other external forces could play in the process of CFL formation. Additional studies on how shear rates from external forces affect CFL formation and microfluidic flow dynamics can shed light on the mechanism of the contribution induced by external driving forces to the development of a separate phase of layer, similar to CFL, close to the microchannel walls and distinct from the surrounding fluid within the system, then influencing microfluidic flow dynamics.One of the mechanisms of phenomena to be explored is the formation of the Exclusion Zone (EZ) region following a “Self-Induced Flow” (SIF) phenomenon discovered by Li and Pollack, 

(106) as shown in Figure 8(a) and (b), respectively. A spontaneous sustained axial flow is observed when hydrophilic materials are immersed in water, resulting in the buildup of a negative layer of charges, defined as the EZ, after water molecules absorb infrared radiation (IR) energy and break down into H and OH

+.

Figure 8. Schematic representations of (a) the Exclusion Zone region and (b) the Self Induced Flow through visualization of microsphere movement within a microchannel. (106) Reproduced with permission from ref (106). Copyright 2020 The Authors under the terms of the Creative Commons (CC BY 4.0) License https://creativecommons.org/licenses/by/4.0/.

Despite the finding of such a phenomenon, the specific mechanism and role of IR energy have yet to be defined for the process of EZ development. To further develop an understanding of the role of IR energy in such phenomena, a feasible study may be seen through the lens of the relationships between external forces and microfluidic flow. In the phenomena, the increase of SIF velocity under a rise of IR radiation resonant characteristics is shown in the participation of the external electric field near the microchannel walls under electro-osmotic viscoelastic fluid flow systems. The buildup of negative charges at the hydrophilic surfaces in EZ is analogous to the mechanism of electrical double layer formation. Indeed, research has initiated the exploration of the core mechanisms for EZ formation through the lens of the electrokinetic phenomena. 

(107) Such a similarity of the role of IR energy and the transport phenomena of SIF with electrokinetic phenomena paves the way for the definition of the unknown SIF phenomena and EZ formation. Furthermore, Li and Pollack 

(106) suggest whether CFL formation might contribute to a SIF of blood using solely IR radiation, a commonly available source of energy in nature, as an external driving force. The proposition may be proven feasible with the presence of the CFL region next to the negatively charged hydrophilic endothelial glycocalyx layer, coating the luminal side of blood vessels. 

(108) Further research can dive into the resonating characteristics between the formation of the CFL region next to the hydrophilic endothelial glycocalyx layer and that of the EZ formation close to hydrophilic microchannel walls. Indeed, an increase in IR energy is known to rapidly accelerate EZ formation and SIF velocity, depicting similarity to the increase in the magnitude of electric field forces and greater shear rates at microchannel walls affecting CFL formation and EOF velocity. Such correlation depicts a future direction in whether SIF blood flow can be observed and characterized theoretically further through the lens of the relationship between blood flow and shear forces exhibited by external energy.

The intricate link between the CFL and external forces, more specifically the externally applied electric field, can receive further attention to provide a more complete framework for the mechanisms between IR radiation and EZ formation. Such characterization may also contribute to a greater comprehension of the role IR can play in CFL formation next to the endothelial glycocalyx layer as well as its role as a driving force to propel blood flow, similar to the SIF, but without the commonly assumed pressure force from heart contraction as a source of driving force.

5.3. Challenges

Although there have been significant improvements in blood flow modeling under LOC systems over the past decade, there are still notable constraints that may require special attention for numerical simulation applications to benefit the adaptability of the designs and functionalities of LOC devices. Several points that require special attention are mentioned below:

1.The majority of CFD models operate under the relationship between the viscoelasticity of blood and the shear rate conditions of flow. The relative effect exhibited by the presence of highly populated RBCs in whole blood and their forces amongst the cells themselves under complex flows often remains unclearly defined. Furthermore, the full range of cell populations in whole blood requires a much more computational load for numerical modeling. Therefore, a vital goal for future research is to evaluate a reduced modeling method where the impact of cell–cell interaction on the viscoelastic property of blood is considered.
2.Current computational methods on hemodynamics rely on continuum models based upon non-Newtonian rheology at the macroscale rather than at molecular and cellular levels. Careful considerations should be made for the development of a constructive framework for the physical and temporal scales of micro/nanoscale systems to evaluate the intricate relationship between fluid driving forces, dynamic viscosity, and elasticity.
3.Viscoelastic fluids under the impact of externally applied electric forces often deviate from the assumptions of no-slip boundary conditions due to the unique flow conditions induced by externally applied forces. Furthermore, the mechanism of vortex formation and viscoelastic flow instability at laminar flow conditions should be better defined through the lens of the microfluidic flow phenomenon to optimize the prediction of viscoelastic flow across different geometrical layouts. Mathematical models and numerical methods are needed to better predict such disturbance caused by external forces and the viscoelasticity of fluids at such a small scale.
4.Under practical situations, zeta potential distribution at channel walls frequently deviates from the common assumption of a constant distribution because of manufacturing faults or inherent surface charges prior to the introduction of electrokinetic influence. These discrepancies frequently lead to inconsistent surface potential distribution, such as excess positive ions at relatively more negatively charged walls. Accordingly, unpredicted vortex formation and flow instability may occur. Therefore, careful consideration should be given to these discrepancies and how they could trigger the transport process and unexpected results of a microdevice.

Author Information

ARTICLE SECTIONS

Jump To


  • Corresponding Authors
    • Zhe Chen – Department of Chemical Engineering, School of Chemistry and Chemical Engineering, State Key Laboratory of Metal Matrix Composites, Shanghai Jiao Tong University, Shanghai 200240, P. R. China;  Email: zaccooky@sjtu.edu.cn
    • Bo Ouyang – Department of Chemical Engineering, School of Chemistry and Chemical Engineering, State Key Laboratory of Metal Matrix Composites, Shanghai Jiao Tong University, Shanghai 200240, P. R. China;  Email: bouy93@sjtu.edu.cn
    • Zheng-Hong Luo – Department of Chemical Engineering, School of Chemistry and Chemical Engineering, State Key Laboratory of Metal Matrix Composites, Shanghai Jiao Tong University, Shanghai 200240, P. R. China;  Orcidhttps://orcid.org/0000-0001-9011-6020; Email: luozh@sjtu.edu.cn
  • Authors
    • Bin-Jie Lai – Department of Chemical Engineering, School of Chemistry and Chemical Engineering, State Key Laboratory of Metal Matrix Composites, Shanghai Jiao Tong University, Shanghai 200240, P. R. China;  Orcidhttps://orcid.org/0009-0002-8133-5381
    • Li-Tao Zhu – Department of Chemical Engineering, School of Chemistry and Chemical Engineering, State Key Laboratory of Metal Matrix Composites, Shanghai Jiao Tong University, Shanghai 200240, P. R. China;  Orcidhttps://orcid.org/0000-0001-6514-8864
  • NotesThe authors declare no competing financial interest.

Acknowledgments

ARTICLE SECTIONS

Jump To


This work was supported by the National Natural Science Foundation of China (No. 22238005) and the Postdoctoral Research Foundation of China (No. GZC20231576).

Vocabulary

ARTICLE SECTIONS

Jump To


Microfluidicsthe field of technological and scientific study that investigates fluid flow in channels with dimensions between 1 and 1000 μm
Lab-on-a-Chip Technologythe field of research and technological development aimed at integrating the micro/nanofluidic characteristics to conduct laboratory processes on handheld devices
Computational Fluid Dynamics (CFD)the method utilizing computational abilities to predict physical fluid flow behaviors mathematically through solving the governing equations of corresponding fluid flows
Shear Ratethe rate of change in velocity where one layer of fluid moves past the adjacent layer
Viscoelasticitythe property holding both elasticity and viscosity characteristics relying on the magnitude of applied shear stress and time-dependent strain
Electro-osmosisthe flow of fluid under an applied electric field when charged solid surface is in contact with the bulk fluid
Vortexthe rotating motion of a fluid revolving an axis line

References

ARTICLE SECTIONS

Jump To


This article references 108 other publications.

  1. 1Neethirajan, S.; Kobayashi, I.; Nakajima, M.; Wu, D.; Nandagopal, S.; Lin, F. Microfluidics for food, agriculture and biosystems industries. Lab Chip 201111 (9), 1574– 1586,  DOI: 10.1039/c0lc00230eViewGoogle Scholar
  2. 2Whitesides, G. M. The origins and the future of microfluidics. Nature 2006442 (7101), 368– 373,  DOI: 10.1038/nature05058ViewGoogle Scholar
  3. 3Burklund, A.; Tadimety, A.; Nie, Y.; Hao, N.; Zhang, J. X. J. Chapter One – Advances in diagnostic microfluidics; Elsevier, 2020; DOI:  DOI: 10.1016/bs.acc.2019.08.001 .ViewGoogle Scholar
  4. 4Abdulbari, H. A. Chapter 12 – Lab-on-a-chip for analysis of blood. In Nanotechnology for Hematology, Blood Transfusion, and Artificial Blood; Denizli, A., Nguyen, T. A., Rajan, M., Alam, M. F., Rahman, K., Eds.; Elsevier, 2022; pp 265– 283.ViewGoogle Scholar
  5. 5Vladisavljević, G. T.; Khalid, N.; Neves, M. A.; Kuroiwa, T.; Nakajima, M.; Uemura, K.; Ichikawa, S.; Kobayashi, I. Industrial lab-on-a-chip: Design, applications and scale-up for drug discovery and delivery. Advanced Drug Delivery Reviews 201365 (11), 1626– 1663,  DOI: 10.1016/j.addr.2013.07.017ViewGoogle Scholar
  6. 6Kersaudy-Kerhoas, M.; Dhariwal, R.; Desmulliez, M. P. Y.; Jouvet, L. Hydrodynamic blood plasma separation in microfluidic channels. Microfluid. Nanofluid. 20108 (1), 105– 114,  DOI: 10.1007/s10404-009-0450-5ViewGoogle Scholar
  7. 7Popel, A. S.; Johnson, P. C. Microcirculation and Hemorheology. Annu. Rev. Fluid Mech. 200537 (1), 43– 69,  DOI: 10.1146/annurev.fluid.37.042604.133933ViewGoogle Scholar
  8. 8Fedosov, D. A.; Peltomäki, M.; Gompper, G. Deformation and dynamics of red blood cells in flow through cylindrical microchannels. Soft Matter 201410 (24), 4258– 4267,  DOI: 10.1039/C4SM00248BViewGoogle Scholar
  9. 9Chakraborty, S. Dynamics of capillary flow of blood into a microfluidic channel. Lab Chip 20055 (4), 421– 430,  DOI: 10.1039/b414566fViewGoogle Scholar
  10. 10Tomaiuolo, G.; Guido, S. Start-up shape dynamics of red blood cells in microcapillary flow. Microvascular Research 201182 (1), 35– 41,  DOI: 10.1016/j.mvr.2011.03.004ViewGoogle Scholar
  11. 11Sherwood, J. M.; Dusting, J.; Kaliviotis, E.; Balabani, S. The effect of red blood cell aggregation on velocity and cell-depleted layer characteristics of blood in a bifurcating microchannel. Biomicrofluidics 20126 (2), 24119,  DOI: 10.1063/1.4717755ViewGoogle Scholar
  12. 12Nader, E.; Skinner, S.; Romana, M.; Fort, R.; Lemonne, N.; Guillot, N.; Gauthier, A.; Antoine-Jonville, S.; Renoux, C.; Hardy-Dessources, M.-D. Blood Rheology: Key Parameters, Impact on Blood Flow, Role in Sickle Cell Disease and Effects of Exercise. Frontiers in Physiology 201910, 01329,  DOI: 10.3389/fphys.2019.01329ViewGoogle Scholar
  13. 13Trejo-Soto, C.; Lázaro, G. R.; Pagonabarraga, I.; Hernández-Machado, A. Microfluidics Approach to the Mechanical Properties of Red Blood Cell Membrane and Their Effect on Blood Rheology. Membranes 202212 (2), 217,  DOI: 10.3390/membranes12020217ViewGoogle Scholar
  14. 14Wagner, C.; Steffen, P.; Svetina, S. Aggregation of red blood cells: From rouleaux to clot formation. Comptes Rendus Physique 201314 (6), 459– 469,  DOI: 10.1016/j.crhy.2013.04.004ViewGoogle Scholar
  15. 15Kim, H.; Zhbanov, A.; Yang, S. Microfluidic Systems for Blood and Blood Cell Characterization. Biosensors 202313 (1), 13,  DOI: 10.3390/bios13010013ViewGoogle Scholar
  16. 16Fåhræus, R.; Lindqvist, T. THE VISCOSITY OF THE BLOOD IN NARROW CAPILLARY TUBES. American Journal of Physiology-Legacy Content 193196 (3), 562– 568,  DOI: 10.1152/ajplegacy.1931.96.3.562ViewGoogle Scholar
  17. 17Ascolese, M.; Farina, A.; Fasano, A. The Fåhræus-Lindqvist effect in small blood vessels: how does it help the heart?. J. Biol. Phys. 201945 (4), 379– 394,  DOI: 10.1007/s10867-019-09534-4ViewGoogle Scholar
  18. 18Bento, D.; Fernandes, C. S.; Miranda, J. M.; Lima, R. In vitro blood flow visualizations and cell-free layer (CFL) measurements in a microchannel network. Experimental Thermal and Fluid Science 2019109, 109847,  DOI: 10.1016/j.expthermflusci.2019.109847ViewGoogle Scholar
  19. 19Namgung, B.; Ong, P. K.; Wong, Y. H.; Lim, D.; Chun, K. J.; Kim, S. A comparative study of histogram-based thresholding methods for the determination of cell-free layer width in small blood vessels. Physiological Measurement 201031 (9), N61,  DOI: 10.1088/0967-3334/31/9/N01ViewGoogle Scholar
  20. 20Hymel, S. J.; Lan, H.; Fujioka, H.; Khismatullin, D. B. Cell trapping in Y-junction microchannels: A numerical study of the bifurcation angle effect in inertial microfluidics. Phys. Fluids (1994) 201931 (8), 082003,  DOI: 10.1063/1.5113516ViewGoogle Scholar
  21. 21Li, X.; Popel, A. S.; Karniadakis, G. E. Blood-plasma separation in Y-shaped bifurcating microfluidic channels: a dissipative particle dynamics simulation study. Phys. Biol. 20129 (2), 026010,  DOI: 10.1088/1478-3975/9/2/026010ViewGoogle Scholar
  22. 22Yin, X.; Thomas, T.; Zhang, J. Multiple red blood cell flows through microvascular bifurcations: Cell free layer, cell trajectory, and hematocrit separation. Microvascular Research 201389, 47– 56,  DOI: 10.1016/j.mvr.2013.05.002ViewGoogle Scholar
  23. 23Shibeshi, S. S.; Collins, W. E. The Rheology of Blood Flow in a Branched Arterial System. Appl. Rheol 200515 (6), 398– 405,  DOI: 10.1515/arh-2005-0020ViewGoogle Scholar
  24. 24Sequeira, A.; Janela, J. An Overview of Some Mathematical Models of Blood Rheology. In A Portrait of State-of-the-Art Research at the Technical University of Lisbon; Pereira, M. S., Ed.; Springer Netherlands: Dordrecht, 2007; pp 65– 87.ViewGoogle Scholar
  25. 25Walburn, F. J.; Schneck, D. J. A constitutive equation for whole human blood. Biorheology 197613, 201– 210,  DOI: 10.3233/BIR-1976-13307ViewGoogle Scholar
  26. 26Quemada, D. A rheological model for studying the hematocrit dependence of red cell-red cell and red cell-protein interactions in blood. Biorheology 198118, 501– 516,  DOI: 10.3233/BIR-1981-183-615ViewGoogle Scholar
  27. 27Varchanis, S.; Dimakopoulos, Y.; Wagner, C.; Tsamopoulos, J. How viscoelastic is human blood plasma?. Soft Matter 201814 (21), 4238– 4251,  DOI: 10.1039/C8SM00061AViewGoogle Scholar
  28. 28Apostolidis, A. J.; Moyer, A. P.; Beris, A. N. Non-Newtonian effects in simulations of coronary arterial blood flow. J. Non-Newtonian Fluid Mech. 2016233, 155– 165,  DOI: 10.1016/j.jnnfm.2016.03.008ViewGoogle Scholar
  29. 29Luo, X. Y.; Kuang, Z. B. A study on the constitutive equation of blood. J. Biomech. 199225 (8), 929– 934,  DOI: 10.1016/0021-9290(92)90233-QViewGoogle Scholar
  30. 30Oldroyd, J. G.; Wilson, A. H. On the formulation of rheological equations of state. Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences 1950200 (1063), 523– 541,  DOI: 10.1098/rspa.1950.0035ViewGoogle Scholar
  31. 31Prado, G.; Farutin, A.; Misbah, C.; Bureau, L. Viscoelastic transient of confined red blood cells. Biophys J. 2015108 (9), 2126– 2136,  DOI: 10.1016/j.bpj.2015.03.046ViewGoogle Scholar
  32. 32Huang, C. R.; Pan, W. D.; Chen, H. Q.; Copley, A. L. Thixotropic properties of whole blood from healthy human subjects. Biorheology 198724 (6), 795– 801,  DOI: 10.3233/BIR-1987-24630ViewGoogle Scholar
  33. 33Anand, M.; Kwack, J.; Masud, A. A new generalized Oldroyd-B model for blood flow in complex geometries. International Journal of Engineering Science 201372, 78– 88,  DOI: 10.1016/j.ijengsci.2013.06.009ViewGoogle Scholar
  34. 34Horner, J. S.; Armstrong, M. J.; Wagner, N. J.; Beris, A. N. Investigation of blood rheology under steady and unidirectional large amplitude oscillatory shear. J. Rheol. 201862 (2), 577– 591,  DOI: 10.1122/1.5017623ViewGoogle Scholar
  35. 35Horner, J. S.; Armstrong, M. J.; Wagner, N. J.; Beris, A. N. Measurements of human blood viscoelasticity and thixotropy under steady and transient shear and constitutive modeling thereof. J. Rheol. 201963 (5), 799– 813,  DOI: 10.1122/1.5108737ViewGoogle Scholar
  36. 36Armstrong, M.; Tussing, J. A methodology for adding thixotropy to Oldroyd-8 family of viscoelastic models for characterization of human blood. Phys. Fluids 202032 (9), 094111,  DOI: 10.1063/5.0022501ViewGoogle Scholar
  37. 37Crank, J.; Nicolson, P. A practical method for numerical evaluation of solutions of partial differential equations of the heat-conduction type. Mathematical Proceedings of the Cambridge Philosophical Society 194743 (1), 50– 67,  DOI: 10.1017/S0305004100023197ViewGoogle Scholar
  38. 38Clough, R. W. Original formulation of the finite element method. Finite Elements in Analysis and Design 19907 (2), 89– 101,  DOI: 10.1016/0168-874X(90)90001-UViewGoogle Scholar
  39. 39Liu, W. K.; Liu, Y.; Farrell, D.; Zhang, L.; Wang, X. S.; Fukui, Y.; Patankar, N.; Zhang, Y.; Bajaj, C.; Lee, J.Immersed finite element method and its applications to biological systems. Computer Methods in Applied Mechanics and Engineering 2006195 (13), 1722– 1749,  DOI: 10.1016/j.cma.2005.05.049ViewGoogle Scholar
  40. 40Lopes, D.; Agujetas, R.; Puga, H.; Teixeira, J.; Lima, R.; Alejo, J. P.; Ferrera, C. Analysis of finite element and finite volume methods for fluid-structure interaction simulation of blood flow in a real stenosed artery. International Journal of Mechanical Sciences 2021207, 106650,  DOI: 10.1016/j.ijmecsci.2021.106650ViewGoogle Scholar
  41. 41Favero, J. L.; Secchi, A. R.; Cardozo, N. S. M.; Jasak, H. Viscoelastic flow analysis using the software OpenFOAM and differential constitutive equations. J. Non-Newtonian Fluid Mech. 2010165 (23), 1625– 1636,  DOI: 10.1016/j.jnnfm.2010.08.010ViewGoogle Scholar
  42. 42Pimenta, F.; Alves, M. A. Stabilization of an open-source finite-volume solver for viscoelastic fluid flows. J. Non-Newtonian Fluid Mech. 2017239, 85– 104,  DOI: 10.1016/j.jnnfm.2016.12.002ViewGoogle Scholar
  43. 43Chee, C. Y.; Lee, H. P.; Lu, C. Using 3D fluid-structure interaction model to analyse the biomechanical properties of erythrocyte. Phys. Lett. A 2008372 (9), 1357– 1362,  DOI: 10.1016/j.physleta.2007.09.067ViewGoogle Scholar
  44. 44Xu, D.; Kaliviotis, E.; Munjiza, A.; Avital, E.; Ji, C.; Williams, J. Large scale simulation of red blood cell aggregation in shear flows. J. Biomech. 201346 (11), 1810– 1817,  DOI: 10.1016/j.jbiomech.2013.05.010ViewGoogle Scholar
  45. 45Johnson, K. L.; Kendall, K.; Roberts, A. Surface energy and the contact of elastic solids. Proceedings of the royal society of London. A. mathematical and physical sciences 1971324 (1558), 301– 313,  DOI: 10.1098/rspa.1971.0141ViewGoogle Scholar
  46. 46Shi, L.; Pan, T.-W.; Glowinski, R. Deformation of a single red blood cell in bounded Poiseuille flows. Phys. Rev. E 201285 (1), 016307,  DOI: 10.1103/PhysRevE.85.016307ViewGoogle Scholar
  47. 47Yoon, D.; You, D. Continuum modeling of deformation and aggregation of red blood cells. J. Biomech. 201649 (11), 2267– 2279,  DOI: 10.1016/j.jbiomech.2015.11.027ViewGoogle Scholar
  48. 48Mainardi, F.; Spada, G. Creep, relaxation and viscosity properties for basic fractional models in rheology. European Physical Journal Special Topics 2011193 (1), 133– 160,  DOI: 10.1140/epjst/e2011-01387-1ViewGoogle Scholar
  49. 49Gracka, M.; Lima, R.; Miranda, J. M.; Student, S.; Melka, B.; Ostrowski, Z. Red blood cells tracking and cell-free layer formation in a microchannel with hyperbolic contraction: A CFD model validation. Computer Methods and Programs in Biomedicine 2022226, 107117,  DOI: 10.1016/j.cmpb.2022.107117ViewGoogle Scholar
  50. 50Aryan, H.; Beigzadeh, B.; Siavashi, M. Euler-Lagrange numerical simulation of improved magnetic drug delivery in a three-dimensional CT-based carotid artery bifurcation. Computer Methods and Programs in Biomedicine 2022219, 106778,  DOI: 10.1016/j.cmpb.2022.106778ViewGoogle Scholar
  51. 51Czaja, B.; Závodszky, G.; Azizi Tarksalooyeh, V.; Hoekstra, A. G. Cell-resolved blood flow simulations of saccular aneurysms: effects of pulsatility and aspect ratio. J. R Soc. Interface 201815 (146), 20180485,  DOI: 10.1098/rsif.2018.0485ViewGoogle Scholar
  52. 52Rydquist, G.; Esmaily, M. A cell-resolved, Lagrangian solver for modeling red blood cell dynamics in macroscale flows. J. Comput. Phys. 2022461, 111204,  DOI: 10.1016/j.jcp.2022.111204ViewGoogle Scholar
  53. 53Dadvand, A.; Baghalnezhad, M.; Mirzaee, I.; Khoo, B. C.; Ghoreishi, S. An immersed boundary-lattice Boltzmann approach to study the dynamics of elastic membranes in viscous shear flows. Journal of Computational Science 20145 (5), 709– 718,  DOI: 10.1016/j.jocs.2014.06.006ViewGoogle Scholar
  54. 54Krüger, T.; Holmes, D.; Coveney, P. V. Deformability-based red blood cell separation in deterministic lateral displacement devices─A simulation study. Biomicrofluidics 20148 (5), 054114,  DOI: 10.1063/1.4897913ViewGoogle Scholar
  55. 55Takeishi, N.; Ito, H.; Kaneko, M.; Wada, S. Deformation of a Red Blood Cell in a Narrow Rectangular Microchannel. Micromachines 201910 (3), 199,  DOI: 10.3390/mi10030199ViewGoogle Scholar
  56. 56Krüger, T.; Varnik, F.; Raabe, D. Efficient and accurate simulations of deformable particles immersed in a fluid using a combined immersed boundary lattice Boltzmann finite element method. Computers & Mathematics with Applications 201161 (12), 3485– 3505,  DOI: 10.1016/j.camwa.2010.03.057ViewGoogle Scholar
  57. 57Balachandran Nair, A. N.; Pirker, S.; Umundum, T.; Saeedipour, M. A reduced-order model for deformable particles with application in bio-microfluidics. Computational Particle Mechanics 20207 (3), 593– 601,  DOI: 10.1007/s40571-019-00283-8ViewGoogle Scholar
  58. 58Balachandran Nair, A. N.; Pirker, S.; Saeedipour, M. Resolved CFD-DEM simulation of blood flow with a reduced-order RBC model. Computational Particle Mechanics 20229 (4), 759– 774,  DOI: 10.1007/s40571-021-00441-xViewGoogle Scholar
  59. 59Mittal, R.; Iaccarino, G. IMMERSED BOUNDARY METHODS. Annu. Rev. Fluid Mech. 200537 (1), 239– 261,  DOI: 10.1146/annurev.fluid.37.061903.175743ViewGoogle Scholar
  60. 60Piquet, A.; Roussel, O.; Hadjadj, A. A comparative study of Brinkman penalization and direct-forcing immersed boundary methods for compressible viscous flows. Computers & Fluids 2016136, 272– 284,  DOI: 10.1016/j.compfluid.2016.06.001ViewGoogle Scholar
  61. 61Akerkouch, L.; Le, T. B. A Hybrid Continuum-Particle Approach for Fluid-Structure Interaction Simulation of Red Blood Cells in Fluid Flows. Fluids 20216 (4), 139,  DOI: 10.3390/fluids6040139ViewGoogle Scholar
  62. 62Barker, A. T.; Cai, X.-C. Scalable parallel methods for monolithic coupling in fluid-structure interaction with application to blood flow modeling. J. Comput. Phys. 2010229 (3), 642– 659,  DOI: 10.1016/j.jcp.2009.10.001ViewGoogle Scholar
  63. 63Cetin, A.; Sahin, M. A monolithic fluid-structure interaction framework applied to red blood cells. International Journal for Numerical Methods in Biomedical Engineering 201935 (2), e3171  DOI: 10.1002/cnm.3171ViewGoogle Scholar
  64. 64Freund, J. B. Numerical Simulation of Flowing Blood Cells. Annu. Rev. Fluid Mech. 201446 (1), 67– 95,  DOI: 10.1146/annurev-fluid-010313-141349ViewGoogle Scholar
  65. 65Ye, T.; Phan-Thien, N.; Lim, C. T. Particle-based simulations of red blood cells─A review. J. Biomech. 201649 (11), 2255– 2266,  DOI: 10.1016/j.jbiomech.2015.11.050ViewGoogle Scholar
  66. 66Arabghahestani, M.; Poozesh, S.; Akafuah, N. K. Advances in Computational Fluid Mechanics in Cellular Flow Manipulation: A Review. Applied Sciences 20199 (19), 4041,  DOI: 10.3390/app9194041ViewGoogle Scholar
  67. 67Rathnayaka, C. M.; From, C. S.; Geekiyanage, N. M.; Gu, Y. T.; Nguyen, N. T.; Sauret, E. Particle-Based Numerical Modelling of Liquid Marbles: Recent Advances and Future Perspectives. Archives of Computational Methods in Engineering 202229 (5), 3021– 3039,  DOI: 10.1007/s11831-021-09683-7ViewGoogle Scholar
  68. 68Li, X.; Vlahovska, P. M.; Karniadakis, G. E. Continuum- and particle-based modeling of shapes and dynamics of red blood cells in health and disease. Soft Matter 20139 (1), 28– 37,  DOI: 10.1039/C2SM26891DViewGoogle Scholar
  69. 69Beris, A. N.; Horner, J. S.; Jariwala, S.; Armstrong, M. J.; Wagner, N. J. Recent advances in blood rheology: a review. Soft Matter 202117 (47), 10591– 10613,  DOI: 10.1039/D1SM01212FViewGoogle Scholar
  70. 70Arciero, J.; Causin, P.; Malgaroli, F. Mathematical methods for modeling the microcirculation. AIMS Biophysics 20174 (3), 362– 399,  DOI: 10.3934/biophy.2017.3.362ViewGoogle Scholar
  71. 71Maria, M. S.; Chandra, T. S.; Sen, A. K. Capillary flow-driven blood plasma separation and on-chip analyte detection in microfluidic devices. Microfluid. Nanofluid. 201721 (4), 72,  DOI: 10.1007/s10404-017-1907-6ViewGoogle Scholar
  72. 72Huhtamäki, T.; Tian, X.; Korhonen, J. T.; Ras, R. H. A. Surface-wetting characterization using contact-angle measurements. Nat. Protoc. 201813 (7), 1521– 1538,  DOI: 10.1038/s41596-018-0003-zViewGoogle Scholar
  73. 73Young, T., III. An essay on the cohesion of fluids. Philosophical Transactions of the Royal Society of London 180595, 65– 87,  DOI: 10.1098/rstl.1805.0005ViewGoogle Scholar
  74. 74Kim, Y. C.; Kim, S.-H.; Kim, D.; Park, S.-J.; Park, J.-K. Plasma extraction in a capillary-driven microfluidic device using surfactant-added poly(dimethylsiloxane). Sens. Actuators, B 2010145 (2), 861– 868,  DOI: 10.1016/j.snb.2010.01.017ViewGoogle Scholar
  75. 75Washburn, E. W. The Dynamics of Capillary Flow. Physical Review 192117 (3), 273– 283,  DOI: 10.1103/PhysRev.17.273ViewGoogle Scholar
  76. 76Cito, S.; Ahn, Y. C.; Pallares, J.; Duarte, R. M.; Chen, Z.; Madou, M.; Katakis, I. Visualization and measurement of capillary-driven blood flow using spectral domain optical coherence tomography. Microfluid Nanofluidics 201213 (2), 227– 237,  DOI: 10.1007/s10404-012-0950-6ViewGoogle Scholar
  77. 77Berthier, E.; Dostie, A. M.; Lee, U. N.; Berthier, J.; Theberge, A. B. Open Microfluidic Capillary Systems. Anal Chem. 201991 (14), 8739– 8750,  DOI: 10.1021/acs.analchem.9b01429ViewGoogle Scholar
  78. 78Berthier, J.; Brakke, K. A.; Furlani, E. P.; Karampelas, I. H.; Poher, V.; Gosselin, D.; Cubizolles, M.; Pouteau, P. Whole blood spontaneous capillary flow in narrow V-groove microchannels. Sens. Actuators, B 2015206, 258– 267,  DOI: 10.1016/j.snb.2014.09.040ViewGoogle Scholar
  79. 79Hirt, C. W.; Nichols, B. D. Volume of fluid (VOF) method for the dynamics of free boundaries. J. Comput. Phys. 198139 (1), 201– 225,  DOI: 10.1016/0021-9991(81)90145-5ViewGoogle Scholar
  80. 80Chen, J.-L.; Shih, W.-H.; Hsieh, W.-H. AC electro-osmotic micromixer using a face-to-face, asymmetric pair of planar electrodes. Sens. Actuators, B 2013188, 11– 21,  DOI: 10.1016/j.snb.2013.07.012ViewGoogle Scholar
  81. 81Zhao, C.; Yang, C. Electrokinetics of non-Newtonian fluids: A review. Advances in Colloid and Interface Science 2013201-202, 94– 108,  DOI: 10.1016/j.cis.2013.09.001ViewGoogle Scholar
  82. 82Oh, K. W. 6 – Lab-on-chip (LOC) devices and microfluidics for biomedical applications. In MEMS for Biomedical Applications; Bhansali, S., Vasudev, A., Eds.; Woodhead Publishing, 2012; pp 150– 171.ViewGoogle Scholar
  83. 83Bello, M. S.; De Besi, P.; Rezzonico, R.; Righetti, P. G.; Casiraghi, E. Electroosmosis of polymer solutions in fused silica capillaries. ELECTROPHORESIS 199415 (1), 623– 626,  DOI: 10.1002/elps.1150150186ViewGoogle Scholar
  84. 84Park, H. M.; Lee, W. M. Effect of viscoelasticity on the flow pattern and the volumetric flow rate in electroosmotic flows through a microchannel. Lab Chip 20088 (7), 1163– 1170,  DOI: 10.1039/b800185eViewGoogle Scholar
  85. 85Afonso, A. M.; Alves, M. A.; Pinho, F. T. Analytical solution of mixed electro-osmotic/pressure driven flows of viscoelastic fluids in microchannels. J. Non-Newtonian Fluid Mech. 2009159 (1), 50– 63,  DOI: 10.1016/j.jnnfm.2009.01.006ViewGoogle Scholar
  86. 86Sousa, J. J.; Afonso, A. M.; Pinho, F. T.; Alves, M. A. Effect of the skimming layer on electro-osmotic─Poiseuille flows of viscoelastic fluids. Microfluid. Nanofluid. 201110 (1), 107– 122,  DOI: 10.1007/s10404-010-0651-yViewGoogle Scholar
  87. 87Zhao, C.; Yang, C. Electro-osmotic mobility of non-Newtonian fluids. Biomicrofluidics 20115 (1), 014110,  DOI: 10.1063/1.3571278ViewGoogle Scholar
  88. 88Pimenta, F.; Alves, M. A. Electro-elastic instabilities in cross-shaped microchannels. J. Non-Newtonian Fluid Mech. 2018259, 61– 77,  DOI: 10.1016/j.jnnfm.2018.04.004ViewGoogle Scholar
  89. 89Bezerra, W. S.; Castelo, A.; Afonso, A. M. Numerical Study of Electro-Osmotic Fluid Flow and Vortex Formation. Micromachines (Basel) 201910 (12), 796,  DOI: 10.3390/mi10120796ViewGoogle Scholar
  90. 90Ji, J.; Qian, S.; Liu, Z. Electroosmotic Flow of Viscoelastic Fluid through a Constriction Microchannel. Micromachines (Basel) 202112 (4), 417,  DOI: 10.3390/mi12040417ViewGoogle Scholar
  91. 91Zhao, C.; Yang, C. Exact solutions for electro-osmotic flow of viscoelastic fluids in rectangular micro-channels. Applied Mathematics and Computation 2009211 (2), 502– 509,  DOI: 10.1016/j.amc.2009.01.068ViewGoogle Scholar
  92. 92Gerum, R.; Mirzahossein, E.; Eroles, M.; Elsterer, J.; Mainka, A.; Bauer, A.; Sonntag, S.; Winterl, A.; Bartl, J.; Fischer, L. Viscoelastic properties of suspended cells measured with shear flow deformation cytometry. Elife 202211, e78823,  DOI: 10.7554/eLife.78823ViewGoogle Scholar
  93. 93Sadek, S. H.; Pinho, F. T.; Alves, M. A. Electro-elastic flow instabilities of viscoelastic fluids in contraction/expansion micro-geometries. J. Non-Newtonian Fluid Mech. 2020283, 104293,  DOI: 10.1016/j.jnnfm.2020.104293ViewGoogle Scholar
  94. 94Spanjaards, M.; Peters, G.; Hulsen, M.; Anderson, P. Numerical Study of the Effect of Thixotropy on Extrudate Swell. Polymers 202113 (24), 4383,  DOI: 10.3390/polym13244383ViewGoogle Scholar
  95. 95Rashidi, S.; Bafekr, H.; Valipour, M. S.; Esfahani, J. A. A review on the application, simulation, and experiment of the electrokinetic mixers. Chemical Engineering and Processing – Process Intensification 2018126, 108– 122,  DOI: 10.1016/j.cep.2018.02.021ViewGoogle Scholar
  96. 96Matsubara, K.; Narumi, T. Microfluidic mixing using unsteady electroosmotic vortices produced by a staggered array of electrodes. Chemical Engineering Journal 2016288, 638– 647,  DOI: 10.1016/j.cej.2015.12.013ViewGoogle Scholar
  97. 97Qaderi, A.; Jamaati, J.; Bahiraei, M. CFD simulation of combined electroosmotic-pressure driven micro-mixing in a microchannel equipped with triangular hurdle and zeta-potential heterogeneity. Chemical Engineering Science 2019199, 463– 477,  DOI: 10.1016/j.ces.2019.01.034ViewGoogle Scholar
  98. 98Cho, C.-C.; Chen, C.-L.; Chen, C. o.-K. Mixing enhancement in crisscross micromixer using aperiodic electrokinetic perturbing flows. International Journal of Heat and Mass Transfer 201255 (11), 2926– 2933,  DOI: 10.1016/j.ijheatmasstransfer.2012.02.006ViewGoogle Scholar
  99. 99Zhao, W.; Yang, F.; Wang, K.; Bai, J.; Wang, G. Rapid mixing by turbulent-like electrokinetic microflow. Chemical Engineering Science 2017165, 113– 121,  DOI: 10.1016/j.ces.2017.02.027ViewGoogle Scholar
  100. 100Tran, T.; Chakraborty, P.; Guttenberg, N.; Prescott, A.; Kellay, H.; Goldburg, W.; Goldenfeld, N.; Gioia, G. Macroscopic effects of the spectral structure in turbulent flows. Nat. Phys. 20106 (6), 438– 441,  DOI: 10.1038/nphys1674ViewGoogle Scholar
  101. 101Toner, M.; Irimia, D. Blood-on-a-chip. Annu. Rev. Biomed Eng. 20057, 77– 103,  DOI: 10.1146/annurev.bioeng.7.011205.135108ViewGoogle Scholar
  102. 102Maria, M. S.; Rakesh, P. E.; Chandra, T. S.; Sen, A. K. Capillary flow of blood in a microchannel with differential wetting for blood plasma separation and on-chip glucose detection. Biomicrofluidics 201610 (5), 054108,  DOI: 10.1063/1.4962874ViewGoogle Scholar
  103. 103Tripathi, S.; Varun Kumar, Y. V. B.; Prabhakar, A.; Joshi, S. S.; Agrawal, A. Passive blood plasma separation at the microscale: a review of design principles and microdevices. Journal of Micromechanics and Microengineering 201525 (8), 083001,  DOI: 10.1088/0960-1317/25/8/083001ViewGoogle Scholar
  104. 104Mohammadi, M.; Madadi, H.; Casals-Terré, J. Microfluidic point-of-care blood panel based on a novel technique: Reversible electroosmotic flow. Biomicrofluidics 20159 (5), 054106,  DOI: 10.1063/1.4930865ViewGoogle Scholar
  105. 105Kang, D. H.; Kim, K.; Kim, Y. J. An anti-clogging method for improving the performance and lifespan of blood plasma separation devices in real-time and continuous microfluidic systems. Sci. Rep 20188 (1), 17015,  DOI: 10.1038/s41598-018-35235-4ViewGoogle Scholar
  106. 106Li, Z.; Pollack, G. H. Surface-induced flow: A natural microscopic engine using infrared energy as fuel. Science Advances 20206 (19), eaba0941  DOI: 10.1126/sciadv.aba0941ViewGoogle Scholar
  107. 107Mercado-Uribe, H.; Guevara-Pantoja, F. J.; García-Muñoz, W.; García-Maldonado, J. S.; Méndez-Alcaraz, J. M.; Ruiz-Suárez, J. C. On the evolution of the exclusion zone produced by hydrophilic surfaces: A contracted description. J. Chem. Phys. 2021154 (19), 194902,  DOI: 10.1063/5.0043084ViewGoogle Scholar
  108. 108Yalcin, O.; Jani, V. P.; Johnson, P. C.; Cabrales, P. Implications Enzymatic Degradation of the Endothelial Glycocalyx on the Microvascular Hemodynamics and the Arteriolar Red Cell Free Layer of the Rat Cremaster Muscle. Front Physiol 20189, 168,  DOI: 10.3389/fphys.2018.00168ViewGoogle Scholar

가상 경계 방법 (Immersed Boundary Method)

가상 경계 방법 (Immersed Boundary Method)

Flow Science의 CFD 엔지니어 인 개발자 및 Adwaith Gupta 인 Zongxian Liang이 블로그에 참여했습니다.

 

힘과 에너지 손실을 정확하게 예측하는 것은 오리피스 판에서 배출, 장애물을 지나는 흐름이나 갑작스런 수축이 있는 파이프에서의 흐름과 같이 많은 엔지니어링 문제를 모델링하는 데 중요합니다. 곧 출시 될 FLOW–3D v12.0 릴리스에는 새로운 수치 옵션 인 가상 경계 방법 (Immersed Boundary Method)이 있어 이러한 문제의 흐름을 정확하게 예측합니다.

가상 경계 방법(정확한 고스트 셀 기반)은 고체 유체 인터페이스에서 수치 자속 계산의 정확성을 향상시킵니다. Flow Science의 개발자 인 Zongxian Liang은 갑작스런 수축 파이프와 선박 선체에 대한 검증 예제를 제공합니다. 가상 경계 방법 및 고스트 셀 접근법에 대한 간단한 수학적 세부 사항은 블로그 끝에서 설명합니다.

 

갑작스럽게 수축되는 관

가상 경계 솔버의 정확도를 나타내는 유체 문제 중 하나는 그림 1과 같이 수축 파이프에서의 물의 손실을 추정하는 것입니다. 파이프는 직경 3m의 큰 섹션에서 1m의 작은 섹션으로 갑자기 수축됩니다. 대형 파이프 입구의 유량은 4m3/sec입니다.  속도 헤드는 수축 위치와 관련하여 업스트림 및 다운 스트림에서 3.5m로 측정됩니다. 문헌 [1,2]에서 사용 된 다른 가정에 기초하여, 머리 손실의 이론적 가치는 0.494m에서 0.711m의 범위에 있어야합니다. 이 시뮬레이션에서 하나의 직교 메쉬 블록이 전체 형상에 사용되고 압력 경계는 출구에서 정체 압력이 0으로 지정됩니다. 2 방정식 k-ω 모델은 최대 난류 혼합 길이가 0.07m로 설정되어 있습니다. 음의 x 방향으로 중력이 활성화됩니다. 완료 시간은 흐름이 일정하고 완전히 발달 된 것으로 간주되는 15 초로 설정됩니다.

그림 1. 수축 파이프의 모양, 속도 헤드 측정을위한 플럭스 표면의 위치 및 흐름 방향을 설명하는 회로도

갑작스런 수축 위치에서 세포 크기 및 메쉬 정렬의 효과를 조사하기 위해 2 변수 파라 메트릭 연구가 수행되었습니다. 메쉬의 셀 크기는 0.1m 및 0.05m입니다. 표 1에 나열된 4 가지 메쉬 정렬을 테스트했습니다.  “정렬 됨”은 메시 평면이 갑작스러운 수축 위치와 정렬되는 기준 사례를 나타내고, “Z- 시프트 : X %”는 메시 평면이 z- 방향으로 수축 위치로부터 셀 크기의 X %만큼 시프트되었음을 ​​나타냅니다.

가상 경계 솔버가 제공하는 예측은 모두 이론적으로 0.494에서 0.711 사이입니다. 특히 작은 셀 크기가 0.05m 인 경우 헤드 손실은 이론적 값의 중앙값의 4.6 % 인 0.603 이내입니다.

  그리드 정렬
  AlignedZ-shifted: 25%Z-shifted: 50%Z-shifted: 75%
Cell size (m)0.10.7060.7100.6470.666
0.050.6070.6200.5750.589

표1 : 가상 경계 솔버에 의해 예측되는 헤드 손실 (m)

 

선체 모델에 대한 저항력

외부 유동 역학 문제에서 힘의 정확한 예측은 일반적으로 설계 단계에서 중요합니다. 자유 표면의 외부 유동 문제의 예는 NAVY 선박 모델 선체에 대한 전체 저항력의 계산입니다.
이 경우 선체 길이는 5.72m이고 구배는 0.248m입니다. 선체 길이와 평균 유속 2.10 m/s 를 기준으로 시뮬레이션에서 레이놀즈 수는 약 12 ​​× 106입니다. 이 경우는 대칭이므로 선체의 절반 만 모델링됩니다. 계산 영역은 길이 30m x 폭 8m, 깊이 5.5m이며 가장 작은 셀 크기가 0.02m 인 3 개의 중첩 된 메쉬 블록을 갖습니다. 입구 및 출구 경계에는 각각 속도 및 유출 조건이 사용됩니다. 역동적인 난류 길이 스케일 계산을 사용한 RNG 난류 모델은 난류 흐름과 운동량 이류에 대한 2 차 단 조성 보존 체계를 모델링하는 데 사용됩니다.

자유 표면 근처의 압력 윤곽 및 저항력 이력을 보여주는 애니메이션.

실험에서 총 항력 계수 0.0423을 기준으로 선체 모델의 절반에 대해 총 저항력은 22.62N입니다. 시뮬레이션의 힘은 x 방향의 전단력과 압력 력의 합으로 평균 35에서 50 초입니다. 가상 경계 솔버는 22.43N으로 실험 결과보다 0.8 % 낮습니다.

 

결론

우리는 이 두 가지 검증에서 가상 경계 방법이 문헌에 제공된 이론적 범위 내에서 머리 손실과 힘을 정확하게 추정한다는 것을 알 수 있습니다. 가상 경계 방법은 매우 기본적인 수준에서 플럭스 추정의 정확도를 향상 시키며 대부분은 아니지만 대부분의 어플리케이션에 대한 시뮬레이션 결과를 향상시킬 것으로 예상됩니다. 고스트 셀 기반 가상 경계 방법 개발에 대한 자세한 내용은 계속 읽으십시오. 그렇지 않으면 다음 블로그 게시물을 계속 지켜봐 주시기 바랍니다!

 

고스트 셀 기반 가상 경계 방법

FLOW-3D(볼드 기울기)에서, 자유 슬립 경계 조건은 속도의 대류에 적용되어 분수 셀 영역과 고체 경계 근처의 체적에 의해 야기되는 수치 경계층을 제거합니다. 제어 체적에 대한 합리적인 플럭스를 추정하기 위해, 가상 경계 솔버는 그림 2와 같이 경계 조건을 암시적으로 만족시키는 고체 속의 유체 속도를 계산합니다. 고체의 유체 셀을 고스트 셀이라고하며 이 방법을 일반적으로 고스트 셀 방식이라고 합니다.

그림 2. 제어 체적 왼쪽면의 플럭스 (파란색 점선으로 묶음)는 솔리드 안의 고스트 셀 u_ (i-1)의 속도를 사용하여 계산됩니다.

경계 조건을 시행하기 위해 고스트 셀의 이미지 포인트 (IP로 표시되는 열린 다이아몬드)가 고스트 셀 (GC로 표시되는 빨간색 다이아몬드)에서 법선의 벽까지 선분을 연장하여 유체 영역에 생성됩니다.  GC와 IP 사이의 중간 점 (경계-절편 점으로 BI로 표시되는 열린 원)에서 벽과 교차합니다. BI의 비 침투 경계 조건과 GC 및 IP의 접선 속도가 벽 표면 속도와 같다고 가정하면 고스트 셀의 속도는 다음 방정식으로 계산됩니다.

여기서,는 고스트 셀에서의 유체 속도, 이미지-포인트 및 경계-절편 포인트이고, 경계에서의 단위 법선 벡터입니다. 우리는 3 차 보간법을 사용하여 주변 셀에서 유체 속도를 사용하여 속도 값을 근사합니다.

여기서 u1v1 그리고 w1 는 이미지 포인트를 둘러싼 보간 노드의 속도이고 α1, β1 및 γ1은 보간 계수입니다. 보간 계수 계산에 대한 자세한 내용은 Ref. 4. 찾을 수 있습니다.

보간은 다른 고스트 셀의 속도 값을 불러 와서 고스트 셀의 결합 시스템이 생성됩니다. 우리는 결과 시스템의 빠른 솔루션을 얻기 위해 수렴 가속 기술과 함께 Jacobi 기반 반복 방법을 사용합니다.

 

참고 문헌

1. White, F. M., Fluid Mechanics (McGraw-Hill Book Company, 2003).
2. Saleh, J., Fluid Flow Handbook (McGraw-Hill Professional, 2002).
3. Larsson, L., Stern, F. & Bertram, V., Benchmarking of computational fluid dynamics for ship flows: the Gothenburg 2000 workshop. Journal of Ship Research 47 (1), 63–81 (2003).
4. Mittal, R. et al., A versatile sharp interface immersed boundary method for incompressible flows with complex boundaries. Journal of computational physics 227 (10), 4825-4852 (2008).

FLOW-3D What’s New Ver.12.0

FLOW-3D v12는 그래픽 사용자 인터페이스 (GUI)의 설계 및 기능에서 매우 큰 변화를 이룬 제품으로 모델 설정을 단순화하고 사용자 워크 플로를 향상시킵니다. 최첨단 Immersed Boundary Method(침수경계 방법)은 FLOW-3D v12 솔루션의 정확성을 높여줍니다. 다른 주요 기능으로는 슬러지 침강 모델, 2-Fluid 2-Temperature 모델 및 Steady State Accelerator가 있으며,이를 통해 사용자는 자유 표면 흐름을 더욱 빠르게 모델링 할 수 있습니다.

Physical and Numerical Model

Immersed boundary method

힘과 에너지 손실에 대한 정확한 예측은 고체 주위의 흐름과 관련된 많은 엔지니어링 문제를 모델링하는 데 중요합니다. 새 릴리스 FLOW-3D v12에는 이러한 문제점 해결을 위해 설계된 새로운 고스트 셀 기반 Immersed Boundary Method (IBM)가 있습니다. IBM은 내 외부 흐름 해석을 위해, 벽 근처에서 보다 정확한 해를 제공하여 드래그 앤 리프트 힘의 계산을 향상시킵니다.힘과 에너지 손실의 정확한 예측은 고체 주위의 흐름을 포함하는 많은 공학적 문제를 모델링 하는데 중요합니다.

Two-field temperature for the two-fluid model

2 유체 열전달 모델은 각 유체에 대한 에너지 전달 방정식을 분리하기 위해 확장되었습니다. 각 유체는 이제 자체 온도 변수를 가지므로 인터페이스 근처의 열 및 물질 전달 솔루션의 정확도가 향상됩니다. 인터페이스에서의 열전달은 이제 시간의 표 함수가 될 수 있는 사용자 정의 열전달 계수에 의해 제어됩니다.

블로그 보기

Sludge settling model

새로운 슬러지 정착 모델은 수처리 애플리케이션에 부가되어 사용자들이 수 처리 탱크와 클래리퍼의 고형 폐기물 역학을 모델링 할 수 있게 해 줍니다. 침전 속도가 분산상의 액적 크기의 함수 인 드리프트-플럭스 모델과 달리, 침전 속도는 슬러지 농도의 함수이며 기능 및 표 형식으로 입력 할 수 있습니다.

개발노트 읽기

Steady-state accelerator for free surface flows

이름에서 알 수 있듯이 정상 상태 가속기는 정상 상태 솔루션에 대한 접근을 빠르게합니다.
이것은 작은 진폭 중력과 모세관 표면파를 감쇠시킴으로써 달성되며 자유 표면 흐름에만 적용 할 수 있습니다.

개발노트 읽기

Void particles

Void particles 가 기포 및 상 변화 모델에 추가되었습니다. Void particles는 붕괴 된 Void 영역을 나타내며, 항력 및 압력을 통해 유체와 상호 작용하는 작은 기포로 작용합니다. 주변 유체 압력에 따라 크기가 변하고 시뮬레이션이 끝날 때의 최종 위치는 공기 유입 가능성을 나타냅니다.

Sediment scour model

퇴적물 수송 및 침식 모델은 정확성과 안정성을 향상시키기 위해 정비되었습니다. 특히 퇴적물 종의 질량 보존이 크게 개선되었습니다.

개발 노트 읽기>

Outflow pressure boundary condition

고정 압력 경계 조건에는 압력 및 유체 분율을 제외한 모든 유량이 해당 경계의 상류의 유량 조건을 반영하는 ‘유출’옵션이 포함됩니다. 유출 압력 경계 조건은 고정 압력 및 연속 경계 조건의 하이브리드입니다.

Moving particle sources

시뮬레이션 중에 입자 소스를 이동할 수 있습니다. 시간에 따른 병진 및 회전 속도는 표 형식으로 정의됩니다. 입자 소스의 운동은 소스에서 방출 된 입자의 초기 속도에 추가됩니다.

Variable center of gravity

기변 무게중심은 중력 및 비관 성 기준 프레임 모델에서, 시간의 함수로서 무게 중심의 위치는 외부 파일에서 테이블로서 정의 될 수있다. 이 기능은 연료를 소비하고 분리 단계를 수행하는 로켓과 같은 모형을 모델링 할 때 유용합니다.

공기 유입 모델

가장 간단한 부피 기반 공기 유입 모델 옵션이 기존 질량 기반 모델로 대체되었습니다. 질량 기반 모델은 부피와 달리 주변 유체 압력에 따라 부피가 변화하는 동안 흡입된 공기량이 보존되기 때문에 물리학적 모델입니다.

Tracer diffusion

유동 표면에서 생성된 추적 물질은 분자 및 난류 확산 과정에 의해 확산될 수 있으며, 예를 들어 실제 오염 물질의 동작을 모방한다.

Model Setup

Simulation units

온도를 포함하여 단위 시스템은 완전히 정의해야하는데 표준 단위 시스템이 제공됩니다. 또한 사용자는 다양한 옵션 중에서 질량, 시간 및 길이 단위를 정의 할 수 있으므로 사용자 정의가 가능한 편리한 단위를 사용할 수 있습니다. 사용자는 압력이 게이지 또는 절대 단위로 정의되는지 여부도 지정해야합니다. 기본 시뮬레이션 단위는 기본 설정에서 설정할 수 있습니다. 단위를 완전히 정의하면 FLOW-3D 가 물리량의 기본값을 정의하고 범용 상수를 설정하여 사용자가 요구하는 작업량을 최소화 할 수 있습니다.

Shallow water model

Manning’s roughness in shallow water model

Manning의 거칠기 계수는 지형 표면의 전단 응력 평가를 위해 천수(shallow water) 모델에서 구현되었습니다. 표면 결함의 크기를 기반으로 기존 거칠기 모델을 보완하며 이 모델과 함께 사용할 수 있습니다. 표준 거칠기와 마찬가지로 매닝 계수는 구성 요소 또는 하위 구성 요소의 속성이거나 지형 래스터 데이터 세트에서 가져올 수 있습니다.

Mesh generation

하단 및 상단 경계 좌표의 정의만으로 수직 방향의 메시 설정이 단순화되었습니다.

Component transformations

사용자는 이제 여러 하위 구성 요소로 구성된 구성 요소에 회전, 변환 및 스케일링 변환을 적용하여 복잡한 형상 어셈블리 설정 프로세스를 단순화 할 수 있습니다. GMO (General Moving Object) 구성 요소의 경우, 이러한 변환을 구성 요소의 대칭 축과 정렬되도록 신체에 맞는 좌표계에 적용 할 수 있습니다.

Changing the number of threads at runtime

시뮬레이션 중에 솔버가 사용하는 스레드 수를 변경하는 기능이 런타임 옵션 대화 상자에 추가되어 사용 가능한 스레드를 추가하거나 다른 태스크에 자원이 필요한 경우 스레드 수를 줄일 수 있습니다.

Probe-controlled heat sources

활성 시뮬레이션 제어가 형상 구성 요소와 관련된 heat sources로 확장되었습니다. 히스토리 프로브로 열 방출을 제어 할 수 있습니다.

Time-dependent temperature at sources     

질량 및 질량 / 운동량 소스의 유체 온도는 이제 테이블 입력을 사용하여 시간의 함수로 정의 할 수 있습니다.

Emissivity coefficients

공극으로의 복사 열 전달을위한 방사율 계수는 이제 사용자가 방사율과 스테판-볼츠만 상수를 지정하도록 요구하지 않고 직접 정의됩니다. 후자는 이제 단위 시스템을 기반으로 솔버에 의해 자동으로 설정됩니다.

Output

  • 등속 필드 솔버 옵션을 사용할 때 유량 속도를 선택한 데이터 로 출력 할 수 있습니다 .
  • 벽 접착력으로 인한 지오메트리 구성 요소의 토크 는 기존 벽 접착력의 출력 외에도 일반 이력 데이터에 별도의 수량으로 출력됩니다.
  • 난류 모델 출력이 요청 될 때 난류 에너지 및 소산과 함께 전단 속도 및 y +가 선택된 데이터로 자동 출력됩니다 .
  • 공기 유입 모델 출력에 몇 가지 수량이 추가되었습니다. 자유 표면을 포함하는 모든 셀에서 혼입 된 공기 및 빠져 나가는 공기의 체적 플럭스가 재시작 및 선택된 데이터로 출력되어 사용자에게 공기가 혼입 및 탈선되는 위치 및 시간에 대한 자세한 정보를 제공합니다. 전체 계산 영역 및 각 샘플링 볼륨 에 대해이 두 수량의 시간 및 공간 통합 등가물 이 일반 히스토리 로 출력됩니다.
  • 솔버의 출력 파일 flsgrf 의 최종 크기 는 시뮬레이션이 끝날 때보 고됩니다.
  • 2 유체 시뮬레이션의 경우, 기존의 출력 수량 유체 체류 시간 및 유체 가 이동 한 거리는 이제 유체 # 1 및 # 2와 유체의 혼합물에 대해 별도로 계산됩니다.
  • 질량 입자의 경우 각 종의 총 부피와 질량이 계산되어 전체 계산 영역, 샘플링 볼륨 및 플럭스 표면에 대한 일반 히스토리 로 출력되어 입자 종 수에 대한 현재 출력을 보완합니다.
  • 예를 들어 사용자가 가스 미순환을 식별하고 연료 탱크의 환기 시스템을 설계하는 데 도움이 되도록 마지막 국부적 가스 압력이 옵션 출력량으로 추가되었습니다. 이 양은 유체가 채워지기 전에 셀의 마지막 간극 압력을 기록하며, 단열 버블 모델과 함께 사용됩니다.

New Customizable Source Routines

사용자 정의 가능한 새로운 소스 루틴이 추가되었으며 사용자의 개발 환경에서 액세스 할 수 있습니다.

소스 루틴 이름설명
cav_prod_cal캐비 테이션 생산 및 확산 속도
sldg_uset슬러지 정착 속도
phchg_mass_flux증발 및 응축에 의한 질량 흐름
flhtccl유체#1과#2사이의 열 전달 계수
dsize_cal2상 유동에서의 동적 낙하 크기 모델의 충돌 및 이탈율
elstc_custom.점탄성 유체에 대한 응력 방정식의 소스 용어

Brand New User Interface

FLOW-3D의 사용자 인터페이스가 완전히 재설계되어 사용자의 작업 흐름을 획기적으로 간소화하는 최신의 타일 구조를 제공합니다.

Dock widgets 설정

Physics, Fluids, Mesh 및 FAVOR ™를 포함한 모든 설정 작업이 형상 창 주위의 dock widgets으로 변환되어 모델 설정을 단일 탭으로 압축 할 수 있습니다. 이 전환을 통해 이전 버전의 복잡한 트리가 훨씬 깔끔하고 효율적인 메뉴 표시로 바뀌어 모델 설정 탭을 떠나지 않고도 모든 매개 변수에 쉽게 액세스 할 수 있습니다.

New Model Setup icons
With our new Model Setup design comes new icons, representing each step of the setup process.
New Physics icons
Our Physics icons are designed to be easily differentiated from one another at a glance, while providing clear visual representation of each model’s purpose and use.

RSS feed

새 RSS 피드부터 FLOW-3D v12.0 의 시뮬레이션 관리자 탭이 개선되었습니다 . FLOW-3D 를 시작하면 사용자에게 Flow Science의 최신 뉴스, 이벤트 및 블로그 게시물이 표시됩니다.

Configurable simulation monitor

시뮬레이션을 실행할 때 중요한 작업은 모니터링입니다. FLOW-3Dv12.0에서는 사용자가 시뮬레이션을 더 잘 모니터링할 수 있도록 Simulation Manager의 플로팅 기능이 향상되었습니다. 사용자는 시뮬레이션 런타임 그래프를 통해 모니터링할 사용 가능한 모든 일반 기록 데이터 변수를 선택하고 각 그래프에 여러 변수를 추가할 수 있습니다. 이제 런타임에서 사용할 수 있는 일반 기록 데이터는 다음과 같습니다.

  • 최소/최대 유체 온도
  • 프로브 위치의 온도
  • 유동 표면 위치에서의 유량
  • 시뮬레이션 진단(예:시간 단계, 안정성 한계)
Runtime plots of the flow rate at the gates of the large dam / Large dam with flux surfaces at the gates

Conforming mesh visualization

사용자는 이제 새로운 FAVOR ™ 독 위젯을 통해 적합한 메쉬 블록을 시각화 할 수 있습니다 .

Large raster and STL data

데이터를 처리하는 데 걸리는 시간으로 인해 큰 형상 데이터를 처리하는 것은 어려울 수 있습니다. 대형 지오메트리 데이터를 처리하는 데 여전히 상당한 시간이 소요될 수 있지만 FLOW-3D는 이제 이러한 대형 데이터 세트를 백그라운드 작업으로로드하여 사용자가 데이터를 처리하는 동안 완벽하게 응답하고 중단없는 인터페이스에서 계속 작업 할 수 있습니다.

Immersed Boundary Method

Immersed Boundary Method

이 기사에서 개발자인 Zongxian Liane박사는 곧 출시될 FLOW-3D v11.3에서 사용할 수 있는 새로운 Immersed Boundary Method에 대해 설명합니다.

힘과 에너지 손실에 대한 정확한 예측은 오리피스 판에서의 배출, 장애물을 지나가는 흐름 및 갑작스런 수축 관에서의 흐름과 같은 많은 엔지니어링 문제를 분석하는데 중요합니다. 셀 면적 및 부피 Method인 FAVORTM은 30년 전에 도입된 이래로 FLOW-3D의 표준 솔버로 적용되었으며 벽 근처의 운동량 fluxes를 근사화하는 간단한 방법을 사용했습니다 (Hirt and Sicilian 1985). 벽이나 자유 표면 근처에서 운동 이류항을 계산할 때 솔리드 또는 보이드 영역 내에 위치한 속도 값은  경계층의 모양을 제거하기 위해 0으로 설정됩니다. 물리적 관점에서 이 방법은 벽의 돌출부에 자유 미끄러짐(비침투)경계 조건을 적용하여 인공 경계층(Hirt1993)을 억제한다.

운동량 방정식에서 플럭스의 손실은 압력에 의해 보상됩니다. 특정 상황에서는 플럭스손실을 보상하는 압력의 비율이 시간에 따라 증가하며, 단일 유전물질로 표현되는 “세속적 불안정성”이라고 하는 수치적 불안정성을 야기할 수 있습니다. 속도의 증가 이러한 불안정성의 전개를 방지하기 위해, 경험적 기법을 사용하여 불안정성이 발생할 수 있는 위치에서 플럭스를 “보정” 했습니다. 그러나 이 방법은 선원으로부터의 플럭스 손실을 해결하지 못하며, 때때로 압력 변동과 같은 용액의 비정치적인 동작을 초래할 수 있습니다.

ghost – 내접 경계법 (Mittal et al., 2008)에 기초한 이류 항을 근사화하는 기법은 FLOW-3D v11.3을 위해 개발되었다. 이 내접 경계 방법 기술은 근본적으로이 문제를 해결하고보다 정확한 압력과 힘 예측을 제공합니다. ghost – 내접 경계법은 복잡한 형상을 포함하는 문제에서 전통적인 데카르트 그리드 근사법에서 강화 된 경계 처리로서 최근에 출현했다. 이 방법은 경계를 처리하는 수단 일 뿐이므로 기존의 해석기 구조가 비교적 적게 변경되어 기존의 FLOW-3D 해석기에 모델로 쉽게 추가 될 수 있으며 FLOW-3D의 다른 물리적 모델과 호환됩니다. 다양한 보간 방법과 함께 가중치 평균 프로브 기술을 사용하여 다른 지오메트리 구성을 처리합니다. 새 모델은 3D 메쉬 블록 또는 하이브리드 3D / 얕은 워터 메쉬 블록이있는 플로우에는 작동하지만 얕은 워터 메쉬에는 적합하지 않습니다.

Immersed Boundary Method Results

새로 도입된 경계 방법 모델의 간단한 예는 직경 1m의 원형 오리피스에서 물이 방출되는 것입니다. 물 용기의 길이는 10m, 폭은 10m, 오리피스 중앙부까지의 수위는 6m이다. 애니메이션에 표시된 것처럼 오리피스 Q에서 표고, h및 볼륨 유량의 강하는 각각 2차 곡선과 선형 곡선을 따릅니다. 

시뮬레이션에서 배출 Cd의 평균 계수는 0.660으로, 비대칭 값 0.611보다 약 8% 큽니다(SwameeandSwamee, 2010). immersed boundary solver 을 사용한 시뮬레이션은 이중 인터페이스(Xeon E5-2623 v3)에서 약 19시간이 소요된다. 반면에 the standard solver의 방전 계수와 벽-블록은 각각 0.800과 39시간이 소요된다.

또 다른 예는 NAVY 선박 모델 선체에 대한 총 저항력의 계산입니다. 이 경우, 선체 길이는 5.72m이고, 드래프트는 0.248m이다. 평균유속은 2.10m/s이고, 레이놀즈 수는 약 12 × 106입니다. 이 해석은 대칭이므로 선체의 절반만 모델링됩니다. 계산 영역은 길이 30m, 너비 8m, 깊이 5.5m입니다. 선체 절반에 대해 실험적으로 얻어진 총 저항력의 평균은 22.62N이다 (Larsson et al., 2003). the standard solver의 총 저항력의 평균은 24.41N이었으며 실험 결과보다 7.9 % 차이가 있으며 immersed boundary solver 경우 총 저항력의 평균은 22.43N이었고 0.8 % 더 낮았습니다 (오류가 8 개 줄었습니다. 또한 immersed 경계 솔버는 약 40 시간 만에 완성되었으며 표준 솔버보다 8 시간 빠릅니다).

References

Hirt, C., & Sicilian, J. (1985). A porosity technique for the definition of obstacles in rectangular cell meshes. International Conference on Numerical Ship Hydrodynamics, 4th. Washington, D.C.

Hirt, C. (1993). Volume-fraction techniques: powerful tools for wind engineering. Journal of Wind Engineering and Industrial Aerodynamics, 46 & 47, 327-338.

Mittal, R., Dong, H., Bozkurttas, M., Najjar, F., Vargas, A., & von Loebbecke, A. (2008). A versatile sharp interface immersed boundary method for incompressible flows with complex boundaries. Journal of computational physics, 227(10), 4825-4852.

Swamee, P., & Swamee , N., (2010). Discharge equation of a circular sharp-crested orifice. Journal of Hydraulic Research, 48(1), 106-107.

FLOW-3D 제품소개

About FLOW-3D


FLOW-3D 2022R2
FLOW-3D 2022R2

FLOW-3D 개발 회사

Flow Science Inc Logo Green.svg
IndustryComputational Fluid Dynamics Software
Founded1980
FounderDr. C.W. “Tony” Hirt
Headquarters
Santa Fe, New Mexico, USA
United States
Key people
Dr. Amir Isfahani, President & CEO
ProductsFLOW-3D, FLOW-3D CAST, FLOW-3D AM, FLOW-3D CLOUD, FlowSight
ServicesCFD consultation and services

FLOW-3D 개요

FLOW-3D는 미국 뉴멕시코주(New Mexico) 로스알라모스(Los Alamos)에 있는 Flow Scicence, Inc에서 개발한 범용 전산유체역학(Computational Fluid Dynamics) 프로그램입니다. 로스알라모스 국립연구소의 수치유체역학 연구실에서 F.Harlow, B. Nichols 및 T.Hirt 등에 의해 개발된 MAC(Marker and Cell) 방법과 SOLA-VOF 방식을 기초로 하여, Hirt 박사가 1980년에 Flow Science, Inc사를 설립하여 계속 프로그램을 발전시켰으며 1985년부터 FLOW-3D를 전세계에 배포하였습니다.

유체의 3차원 거동 해석을 수행하는데 사용되는 CFD모형은 몇몇 있으나, 유동해석에 적용할 물리모델 선정은 해석의 정밀도와 밀접한 관계가 있으므로, 해석하고자 하는 대상의 유동 특성을 분석하여 신중하게 결정하여야 합니다.

FLOW-3D는 자유표면(Free Surface) 해석에 있어서 매우 정확한 해석 결과를 제공합니다. 해석방법은 자유표면을 포함한 비정상 유동 상태를 기본으로 하며, 연속방정식, 3차원 운동량 보전방정식(Navier-Stokes eq.) 및 에너지 보존방정식 등을 적용할 수 있습니다.

FLOW-3D는 유한차분법을 사용하고 있으며, 유한요소법(FEM, Finite Element Method), 경계요소법(Boundary Element Method)등을 포함하여 자유표면을 포함하는 유동장 해석(Fluid Flow Analysis)에서 공기와 액체의 경계면을 정밀하게 표현 가능합니다.

유체의 난류 해석에 대해서는 혼합길이 모형, 난류 에너지 모형, RNG(Renormalized Group Theory)  k-ε 모형, k-ω 모형, LES 모형 등 6개 모형을 적용할 수 있으며, 자유표면 해석을 위하여 VOF(Volume of Fluid) 방정식을 사용하고, 격자 생성시 사용자가 가장 쉽게 만들 수 있는 직각형상격자는 형상을 더욱 정확하게 표현하기 위해 FAVOR(Fractional Area Volume Obstacle Representation) 기법을 각 방정식에 적용하고 있습니다.

FLOW-3D는 비압축성(Incompressible Fluid Flow), 압축성 유체(Compressible Fluid Flow)의 유동현상 뿐만 아니라 고체와의 열전달 현상을 해석할 수 있으며, 비정상 상태의 해석을 기본으로 합니다.

FLOW-3D v12.0은 모델 설정을 간소화하고 사용자 워크 플로우를 개선하는 GUI(그래픽 사용자 인터페이스)의 설계 및 기능에 있어 중요한 변화를 가져왔습니다. 최첨단 Immersed Boundary Method는 FLOW-3Dv12.0솔루션의 정확도를 높여 줍니다. 다른 특징적인 주요 개발에는 슬러지 안착 모델, 2-유체 2-온도 모델, 사용자가 자유 표면 흐름을 훨씬 더 빠르게 모델링 할 수 있는 Steady State Accelerator등이 있습니다.

물리 및 수치 모델

Immersed Boundary Method

힘과 에너지 손실에 대한 정확한 예측은 솔리드 바디 주변의 흐름과 관련된 많은 엔지니어링 문제를 모델링하는 데 중요합니다. FLOW-3D v12.0의 릴리스에는 이러한 문제 해결을 위해 설계된 새로운 고스트 셀 기반 Immersed Boundary Method (IBM)가 포함되어 있습니다. IBM은 내부 및 외부 흐름을 위해 벽 근처 해석을 위해 보다 정확한 솔루션을 제공하여 드래그 앤 리프트 힘의 계산을 개선합니다.

Two-field temperature for the two-fluid model

2유체 열 전달 모델은 각 유체에 대한 에너지 전달 공식을 분리하도록 확장되었습니다. 이제 각 유체에는 고유한 온도 변수가 있어 인터페이스 근처의 열 및 물질 전달 솔루션의 정확도를 향상시킵니다. 인터페이스에서의 열 전달은 시간의 표 함수가 될 수 있는 사용자 정의 열 전달 계수에 의해 제어됩니다.

슬러지 침전 모델 / Sludge settling model

중요 추가 기능인 새로운 슬러지 침전 모델은 도시 수처리 시설물 응용 분야에 사용하면 수처리 탱크 및 정화기의 고형 폐기물 역학을 모델링 할 수 있습니다. 침전 속도가 확산된 위상의 방울 크기에 대한 함수인 드리프트-플럭스 모델과 달리, 침전 속도는 슬러지 농도의 함수이며 기능적인 형태와 표 형태로 모두 입력 할 수 있습니다.

Steady-state accelerator for free surface flows

이름이 암시하듯이, 정상 상태 가속기는 안정된 상태의 솔루션에 대한 접근을 가속화합니다. 이는 작은 진폭의 중력과 모세관 현상을 감쇠하여 이루어지며 자유 표면 흐름에만 적용됩니다.

꾸준한 상태 가속기

Void particles

보이드 입자가 버블 및 위상 변경 모델에 추가되었습니다. 보이드 입자는 항력과 압력 힘을 통해 유체와 상호 작용하는 작은 기포의 역할을 하는 붕괴된 보이드 영역을 나타냅니다. 주변 유체 압력에 따라 크기가 변경되고 시뮬레이션이 끝난 후 최종 위치는 공기 침투 가능성을 나타냅니다.

Sediment scour model

침전물의 정확성과 안정성을 향상시키기 위해 침전물의 운반과 침식 모델을 정밀 조사하였다. 특히, 침전물 종에 대한 질량 보존이 크게 개선되었습니다.

Outflow pressure boundary condition

고정 압력 경계 조건에는 이제 압력 및 유체 비율을 제외한 모든 유량이 해당 경계의 상류에 있는 흐름 조건을 반영하는 ‘유출’ 옵션이 포함됩니다. 유출 압력 경계 조건은 고정 압력 및 연속성 경계 조건의 혼합입니다.

Moving particle sources

시뮬레이션 중에 입자 소스는 이동할 수 있습니다. 시간에 따른 변환 및 회전 속도는 표 형식으로 정의됩니다. 입자 소스의 운동은 소스에서 방출 된 입자의 초기 속도에 추가됩니다.

Variable center of gravity

중력 및 비 관성 기준 프레임 모델에서 시간 함수로서의 무게 중심의 위치는 외부 파일의 표로 정의할 수 있습니다. 이 기능은 연료를 소모하는 로켓을 모델링하고 단계를 분리할 때 유용합니다.

공기 유입 모델

가장 간단한 부피 기반 공기 유입 모델 옵션이 기존 질량 기반 모델로 대체되었습니다.  질량 기반 모델은 부피와 달리 주변 유체 압력에 따라 부피가 변화하는 동안 흡입된 공기량이 보존되기 때문에 물리학적 모델입니다.

Air entrainment model in FLOW-3D v12.0

Tracer diffusion / 트레이서 확산

유동 표면에서 생성된 추적 물질은 분자 및 난류 확산 과정에 의해 확산될 수 있으며, 예를 들어 실제 오염 물질의 거동을 모방합니다.

모델 설정

시뮬레이션 단위

이제 온도를 포함하여 단위계 시스템을 완전히 정의해야 합니다. 표준 단위 시스템이 제공됩니다. 또한 사용자는 선택한 옵션에서 질량, 시간 및 길이 단위를 정의하여 편리하며, 사용자 정의된 단위를 사용할 수 있습니다. 사용자는 또한 압력이 게이지 단위로 정의되는지 절대 단위로 정의되는지 여부를 지정해야 합니다. 기본 시뮬레이션 단위는 Preferences(기본 설정)에서 설정할 수 있습니다. 단위를 완벽하게 정의하면 FLOW-3D는 물리적 수량에 대한 기본 값을 정의하고 범용 상수를 설정할 수 있으므로 사용자가 필요로 하는 작업량을 최소화할 수 있습니다.

Shallow water model

천수(shallow water) 모델에서 매닝의 거칠기

Manning의 거칠기 계수는 지형 표면의 전단 응력 평가를 위해 천수(shallow water) 모델에서 구현되었습니다. 표면 결함의 크기를 기반으로 기존 거칠기 모델을 보완하며이 모델과 함께 사용할 수 있습니다. 표준 거칠기와 마찬가지로 매닝 계수는 구성 요소 또는 하위 구성 요소의 속성이거나 지형 래스터 데이터 세트에서 가져올 수 있습니다.

메시 생성

하단 및 상단 경계 좌표의 정의만으로 수직 방향의 메시 설정이 단순화되었습니다.

구성 요소 변환

사용자는 이제 여러 하위 구성 요소로 구성된 구성 요소에 회전, 변환 및 스케일링 변환을 적용하여 복잡한 형상 어셈블리 설정 프로세스를 단순화 할 수 있습니다. GMO (General Moving Object) 구성 요소의 경우, 이러한 변환을 구성 요소의 대칭 축과 정렬되도록 신체에 맞는 좌표계에 적용 할 수 있습니다.

런타임시 스레드 수 변경

시뮬레이션 중에 솔버가 사용하는 스레드 수를 변경하는 기능이 런타임 옵션 대화 상자에 추가되어 사용 가능한 스레드를 추가하거나 다른 태스크에 자원이 필요한 경우 스레드 수를 줄일 수 있습니다.

프로브 제어 열원

활성 시뮬레이션 제어가 형상 구성 요소와 관련된 heat sources로 확장되었습니다.  history probes로 열 방출을 제어 할 수 있습니다.

소스에서 시간에 따른 온도

질량 및 질량/모멘트 소스의 유체 온도는 이제 테이블 입력을 사용하여 시간의 함수로 정의 할 수 있습니다.

방사율 계수

공극으로의 복사 열 전달을위한 방사율 계수는 이제 사용자가 방사율과 스테판-볼츠만 상수를 지정하도록 요구하지 않고 직접 정의됩니다. 후자는 이제 단위 시스템을 기반으로 솔버에 의해 자동으로 설정됩니다.

Output

  • 등속 필드 솔버 옵션을 사용할 때 유량 속도를 선택한 데이터로 출력 할 수 있습니다.
  • 벽 접착력으로 인한 지오메트리 구성 요소의 토크는 기존 벽 접착력 출력과 함께 별도의 수량으로 일반 이력 데이터에 출력됩니다.
  • 난류 모델 출력이 요청 될 때 난류 에너지 및 소산과 함께 전단 속도 및 y +가 선택된 데이터로 자동 출력됩니다.
  • 공기 유입 모델 출력에 몇 가지 수량이 추가되었습니다. 자유 표면을 포함하는 모든 셀에서 혼입 된 공기 및 빠져 나가는 공기의 체적 플럭스가 재시작 및 선택된 데이터로 출력되어 사용자에게 공기가 혼입 및 탈선되는 위치 및 시간에 대한 자세한 정보를 제공합니다. 전체 계산 영역 및 각 샘플링 볼륨 에 대해이 두 수량의 시간 및 공간 통합 등가물이 일반 히스토리 로 출력됩니다.
  • 솔버의 출력 파일 flsgrf 의 최종 크기는 시뮬레이션이 끝날 때 보고됩니다.
  • 2 유체 시뮬레이션의 경우, 기존의 출력 수량 유체 체류 시간 및 유체 가 이동 한 거리는 이제 유체 # 1 및 # 2와 유체의 혼합물에 대해 별도로 계산됩니다.
  • 질량 입자의 경우, 각 종의 총 부피 및 질량이 계산되어 전체 계산 영역, 샘플링 볼륨 및 플럭스 표면에 대한 일반 히스토리 로 출력되어 입자 종 수에 대한 현재 출력을 보완합니다.
  • 최종 로컬 가스 압력 은 사용자가 가스 포획을 식별하고 연료 탱크의 배기 시스템 설계를 지원하는 데 도움이되는 선택적 출력량으로 추가되었습니다. 이 양은 유체로 채워지기 전에 셀의 마지막 공극 압력을 기록하며 단열 버블 모델과 함께 사용됩니다.

새로운 맞춤형 소스 루틴

새로운 사용자 정의 가능 소스 루틴이 추가되었으며 사용자의 개발 환경에서 액세스 할 수 있습니다.

소스 루틴 이름기술
cav_prod_calCavitation 생성과 소산 비율
sldg_uset슬러지 침전 속도
phchg_mass_flux증발 및 응축으로 인한 질량 플럭스
flhtccl유체 # 1과 # 2 사이의 열전달 계수
dsize_cal2 상 흐름에서 동적 액적 크기 모델의 응집 및 분해 속도
elstc_custom점탄성 유체에 대한 응력 방정식의 Source Terms

새로운 사용자 인터페이스

FLOW-3D 사용자 인터페이스는 완전히 새롭게 디자인되어 현대적이고 평평한 구조로 사용자의 작업 흐름을 획기적으로 간소화합니다.

Setup dock widgets

Physics, Fluids, Mesh 및 FAVOR ™를 포함한 모든 설정 작업이 지오 메트리 윈도우 주변에서 독 위젯으로 변환되어 모델 설정을 단일 탭으로 요약할 수 있습니다. 이러한 전환으로 인해 이전 버전의 복잡한 접이식 트리가 훨씬 깨끗하고 효율적인 메뉴 프레젠테이션으로 대체되어 사용자는 ModelSetup탭을 떠나지 않고도 모든 매개 변수에 쉽게 액세스 할 수 있습니다.

New Model Setup icons

새로운 모델 설정 디자인에는 설정 프로세스의 각 단계를 나타내는 새로운 아이콘이 있습니다.

Model setup icons - FLOW-3D v12.0

New Physics icons

RSS feed

새 RSS 피드부터 FLOW-3D v12.0의 시뮬레이션 관리자 탭이 개선되었습니다. FLOW-3D 를 시작하면 사용자에게 Flow Science의 최신 뉴스, 이벤트 및 블로그 게시물이 표시됩니다.

RSS feed - FLOW-3D

Configurable simulation monitor

시뮬레이션을 실행할 때 중요한 작업은 모니터링입니다. FLOW-3Dv1.0에서는 사용자가 시뮬레이션을 더 잘 모니터링할 수 있도록 SimulationManager의 플로팅 기능이 향상되었습니다. 사용자는 시뮬레이션 런타임 그래프를 통해 모니터링할 사용 가능한 모든 일반 기록 데이터 변수를 선택하고 각 그래프에 여러 변수를 추가할 수 있습니다. 이제 런타임에서 사용할 수 있는 일반 기록 데이터는 다음과 같습니다.

  • 최소/최대 유체 온도
  • 프로브 위치의 온도
  • 유동 표면 위치에서의 유량
  • 시뮬레이션 진단(예:시간 단계, 안정성 한계)
출입문에 유동 표면이 있는 대형 댐
Runtime plots of the flow rate at the gates of the large dam

Conforming 메쉬 시각화

용자는 이제 새로운 FAVOR ™ 독 위젯을 통해 적합한 메쉬 블록을 시각화 할 수 있습니다.Visualize conforming mesh blocks

Large raster and STL data

데이터를 처리하는 데 걸리는 시간 때문에 큰 지오 메트리 데이터를 처리하는 것은 수고스러울 수 있습니다. 대형 지오 메트리 데이터를 처리하는 데는 여전히 상당한 시간이 걸릴 수 있지만, FLOW-3D는 이제 이러한 대규모 데이터 세트를 백그라운드 작업으로 로드하여 사용자가 데이터를 처리하는 동안 완전히 응답하고 중단 없는 인터페이스에서 작업을 계속할 수 있습니다

FLOW-3D Features

The features in blue are newly-released in FLOW-3D v12.0.

Meshing & Geometry

  • Structured finite difference/control volume meshes for fluid and thermal solutions
  • Finite element meshes in Cartesian and cylindrical coordinates for structural analysis
  • Multi-Block gridding with nested, linked, partially overlapping and conforming mesh blocks
  • Conforming meshes extended to arbitrary shapes
  • Fractional areas/volumes (FAVOR™) for efficient & accurate geometry definition
  • Closing gaps in geometry
  • Mesh quality checking
  • Basic Solids Modeler
  • Import CAD data
  • Import/export finite element meshes via Exodus-II file format
  • Grid & geometry independence
  • Cartesian or cylindrical coordinates

Flow Type Options

  • Internal, external & free-surface flows
  • 3D, 2D & 1D problems
  • Transient flows
  • Inviscid, viscous laminar & turbulent flows
  • Hybrid shallow water/3D flows
  • Non-inertial reference frame motion
  • Multiple scalar species
  • Two-phase flows
  • Heat transfer with phase change
  • Saturated & unsaturated porous media

Physical Modeling Options

  • Fluid structure interaction
  • Thermally-induced stresses
  • Plastic deformation of solids
  • Granular flow
  • Moisture drying
  • Solid solute dissolution
  • Sediment transport and scour
  • Sludge settling
  • Cavitation (potential, passive tracking, active tracking)
  • Phase change (liquid-vapor, liquid-solid)
  • Surface tension
  • Thermocapillary effects
  • Wall adhesion
  • Wall roughness
  • Vapor & gas bubbles
  • Solidification & melting
  • Mass/momentum/energy sources
  • Shear, density & temperature-dependent viscosity
  • Thixotropic viscosity
  • Visco-elastic-plastic fluids
  • Elastic membranes & walls
  • Evaporation residue
  • Electro-mechanical effects
  • Dielectric phenomena
  • Electro-osmosis
  • Electrostatic particles
  • Joule heating
  • Air entrainment
  • Molecular & turbulent diffusion
  • Temperature-dependent material properties
  • Spray cooling

Flow Definition Options

  • General boundary conditions
    • Symmetry
    • Rigid and flexible walls
    • Continuative
    • Periodic
    • Specified pressure
    • Specified velocity
    • Outflow
    • Outflow pressure
    • Outflow boundaries with wave absorbing layers
    • Grid overlay
    • Hydrostatic pressure
    • Volume flow rate
    • Non-linear periodic and solitary surface waves
    • Rating curve and natural hydraulics
    • Wave absorbing layer
  • Restart from previous simulation
  • Continuation of a simulation
  • Overlay boundary conditions
  • Change mesh and modeling options
  • Change model parameters

Thermal Modeling Options

  • Natural convection
  • Forced convection
  • Conduction in fluid & solid
  • Fluid-solid heat transfer
  • Distributed energy sources/sinks in fluids and solids
  • Radiation
  • Viscous heating
  • Orthotropic thermal conductivity
  • Thermally-induced stresses

Numerical Modeling Options

  • TruVOF Volume-of-Fluid (VOF) method for fluid interfaces
  • Steady state accelerator for free-surface flows
  • First and second order advection
  • Sharp and diffuse interface tracking
  • Implicit & explicit numerical methods
  • Immersed boundary method
  • GMRES, point and line relaxation pressure solvers
  • User-defined variables, subroutines & output
  • Utilities for runtime interaction during execution

Fluid Modeling Options

  • One incompressible fluid – confined or with free surfaces
  • Two incompressible fluids – miscible or with sharp interfaces
  • Compressible fluid – subsonic, transonic, supersonic
  • Stratified fluid
  • Acoustic phenomena
  • Mass particles with variable density or diameter

Shallow Flow Models

  • General topography
  • Raster data interface
  • Subcomponent-specific surface roughness
  • Wind shear
  • Ground roughness effects
  • Manning’s roughness
  • Laminar & turbulent flow
  • Sediment transport and scour
  • Surface tension
  • Heat transfer
  • Wetting & drying

Turbulence Models

  • RNG model
  • Two-equation k-epsilon model
  • Two-equation k-omega model
  • Large eddy simulation

Advanced Physical Models

  • General Moving Object model with 6 DOF–prescribed and fully-coupled motion
  • Rotating/spinning objects
  • Collision model
  • Tethered moving objects (springs, ropes, breaking mooring lines)
  • Flexing membranes and walls
  • Porosity
  • Finite element based elastic-plastic deformation
  • Finite element based thermal stress evolution due to thermal changes in a solidifying fluid
  • Combusting solid components

Chemistry Models

  • Stiff equation solver for chemical rate equations
  • Stationary or advected species

Porous Media Models

  • Saturated and unsaturated flow
  • Variable porosity
  • Directional porosity
  • General flow losses (linear & quadratic)
  • Capillary pressure
  • Heat transfer in porous media
  • Van Genunchten model for unsaturated flow

Discrete Particle Models

  • Massless marker particles
  • Multi-species material particles of variable size and mass
  • Solid, fluid, gas particles
  • Void particles tracking collapsed void regions
  • Non-linear fluid-dynamic drag
  • Added mass effects
  • Monte-Carlo diffusion
  • Particle-fluid momentum coupling
  • Coefficient of restitution or sticky particles
  • Point or volumetric particle sources
  • Initial particle blocks
  • Heat transfer with fluid
  • Evaporation and condensation
  • Solidification and melting
  • Coulomb and dielectric forces
  • Probe particles

Two-Phase & Two-Component Models

  • Liquid/liquid & gas/liquid interfaces
  • Variable density mixtures
  • Compressible fluid with a dispersed incompressible component
  • Drift flux with dynamic droplet size
  • Two-component, vapor/non-condensable gases
  • Phase transformations for gas-liquid & liquid-solid
  • Adiabatic bubbles
  • Bubbles with phase change
  • Continuum fluid with discrete particles
  • Scalar transport
  • Homogeneous bubbles
  • Super-cooling
  • Two-field temperature

Coupling with Other Programs

  • Geometry input from Stereolithography (STL) files – binary or ASCII
  • Direct interfaces with EnSight®, FieldView® & Tecplot® visualization software
  • Finite element solution import/export via Exodus-II file format
  • PLOT3D output
  • Neutral file output
  • Extensive customization possibilities
  • Solid Properties Materials Database

Data Processing Options

  • State-of-the-art post-processing tool, FlowSight™
  • Batch post-processing
  • Report generation
  • Automatic or custom results analysis
  • High-quality OpenGL-based graphics
  • Color or B/W vector, contour, 3D surface & particle plots
  • Moving and stationary probes
  • Visualization of non-inertial reference frame motion
  • Measurement baffles
  • Arbitrary sampling volumes
  • Force & moment output
  • Animation output
  • PostScript, JPEG & Bitmap output
  • Streamlines
  • Flow tracers

User Conveniences

  • Active simulation control (based on measurement of probes)
  • Mesh generators
  • Mesh quality checking
  • Tabular time-dependent input using external files
  • Automatic time-step control for accuracy & stability
  • Automatic convergence control
  • Mentor help to optimize efficiency
  • Units on all variables
  • Custom units
  • Component transformations
  • Moving particle sources
  • Change simulation parameters while solver runs
  • Launch and manage multiple simulations
  • Automatic simulation termination based on user-defined criteria
  • Run simulation on remote servers using remote solving
  • Copy boundary conditions to other mesh blocks

Multi-Processor Computing

  • Shared memory computers
  • Distributed memory clusters

FlowSight

  • Particle visualization
  • Velocity vector fields
  • Streamlines & pathlines
  • Iso-surfaces
  • 2D, 3D and arbitrary clips
  • Volume render
  • Probe data
  • History data
  • Vortex cores
  • Link multiple results
  • Multiple data views
  • Non-inertial reference frame
  • Spline clip