Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Accelerated Computational Fluid Dynamics Simulations of Microfluidic Devices by Exploiting Higher Levels of Abstraction

Michel Takken, Robert Wille

TL;DR

The paper tackles the heavy compute costs of CFD simulations for microfluidic devices and proposes a hybrid, two-stage acceleration that leverages higher-abstraction models where flow is organized. By partitioning the device into Omega_high (high-abstraction) and Omega_low (CFD) regions and coupling them at interface Gamma via the fixed-point iteration q^n = f(q^{n-1}), the method preserves fidelity while dramatically reducing runtime. Implemented with LBM for Omega_low and MNA for Omega_high, and validated on four continuous-channel networks, the approach achieves speedups up to orders of magnitude with close agreement in pressure and velocity fields compared to full CFD. The work points to extensions to 3D domains and additional phenomena (diffusion, heat, droplets), offering a scalable path for rapid, fidelity-preserving microfluidic simulations.

Abstract

The design of microfluidic devices is a cumbersome and tedious process that can be significantly improved by simulation. Methods based on Computational Fluid Dynamics (CFD) are considered state-of-the-art, but require extensive compute time - oftentimes limiting the size of microfluidic devices that can be simulated. Simulation methods that abstract the underlying physics on a higher level generally provide results instantly, but the fidelity of these methods is usually worse. In this work, a simulation method that accelerates CFD simulations by exploiting simulation methods on higher levels of abstraction is proposed. Case studies confirm that the proposed method accelerates CFD simulations by multiple factors (often several orders of magnitude) while maintaining the fidelity of CFD simulations.

Accelerated Computational Fluid Dynamics Simulations of Microfluidic Devices by Exploiting Higher Levels of Abstraction

TL;DR

The paper tackles the heavy compute costs of CFD simulations for microfluidic devices and proposes a hybrid, two-stage acceleration that leverages higher-abstraction models where flow is organized. By partitioning the device into Omega_high (high-abstraction) and Omega_low (CFD) regions and coupling them at interface Gamma via the fixed-point iteration q^n = f(q^{n-1}), the method preserves fidelity while dramatically reducing runtime. Implemented with LBM for Omega_low and MNA for Omega_high, and validated on four continuous-channel networks, the approach achieves speedups up to orders of magnitude with close agreement in pressure and velocity fields compared to full CFD. The work points to extensions to 3D domains and additional phenomena (diffusion, heat, droplets), offering a scalable path for rapid, fidelity-preserving microfluidic simulations.

Abstract

The design of microfluidic devices is a cumbersome and tedious process that can be significantly improved by simulation. Methods based on Computational Fluid Dynamics (CFD) are considered state-of-the-art, but require extensive compute time - oftentimes limiting the size of microfluidic devices that can be simulated. Simulation methods that abstract the underlying physics on a higher level generally provide results instantly, but the fidelity of these methods is usually worse. In this work, a simulation method that accelerates CFD simulations by exploiting simulation methods on higher levels of abstraction is proposed. Case studies confirm that the proposed method accelerates CFD simulations by multiple factors (often several orders of magnitude) while maintaining the fidelity of CFD simulations.
Paper Structure (15 sections, 6 equations, 8 figures, 3 tables)

This paper contains 15 sections, 6 equations, 8 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Simulation Methods for Microfluidic Devices
  3. Review of CFD Methods
  4. Review of Methods on High Abstraction Levels
  5. Accelerating CFD Simulations
  6. Continuous Channel-Based Microfluidics
  7. The Proposed Accelerated Method
  8. Implementation Details
  9. Step 1: Identifying the Required Fidelity
  10. Step 2: Defining the Communication
  11. Case Studies and Evaluation Results
  12. Setup: Considered Cases and Parameters
  13. Obtained Results
  14. Discussion
  15. Conclusion

Figures (8)

  • Figure 1: Example network for continuous channel-based microfluidics. The pressure $p$ and velocity fields $\bm{u}$ are shown in detail for a crossing (red) and a straight channel section (green).
  • Figure 2: The example network in the proposed method. (a) The separation of the example network into $\Omega_\text{low}$ and $\Omega_\text{high}$. (b) The network during the initial iteration; $\Omega_\text{low}$ is replaced by fully connected graphs, and the resulting network is used to find $q^0$.
  • Figure 3: Communication schemes for the pressure and flow fields. (a) Communicate the flow rate from $\Omega_\text{high}$ to $\Omega_\text{low}$ and the pressure vice versa. Here, the flow field information must be extrapolated from the communicated flow rate. (b) Communicate the flow field from $\Omega_\text{low}$ to $\Omega_\text{high}$ and the pressure vice versa.
  • Figure 4: The networks of the considered case studies. The number of separate regions in $\Omega_\text{low}$ increments with the case studies, starting at one region (a) for Network 1 and ending at four regions (a--d) for Network 4.
  • Figure 5: The pressure and velocity fields obtained from the CFD simulations (left) and the proposed method (right) for Region 1a in Network 1 with $l\text{ is }1,2,3\text{ and }4$.
  • ...and 3 more figures