PI(s) Ganapathysubramanian (Iowa State U.), DiCarlo (UCLA), Zola (Rutgers)
Controlling the shape and location of a fluid stream provides a fundamental tool for creating structured materials, preparing biological samples, and engineering heat and mass transport. Methods to manipulate the cross-sectional shape of fluids have focused on creating chaos and mixing by fluid twisting instead of ordering or structuring streams with precise sequences of fluid perturbations. The ability to engineer the cross-sectional shape of a fluid using the integrated inertial flow deformations induced by sequences of simple microstructures (i.e. pillars) was recently demonstrated. Discretization of single pillar operations followed by their programmed superposition allows for the hierarchical assembly of complex flow programs. Although this approach has allowed for the sculpting of complex fluid shapes, creating user-defined flow shapes important for practical applications currently requires laborious and time-consuming trial and error design iterations, and often complex fluid shapes of interest are not achievable in a reasonable time frame. The ability to create a user-defined flow shape and automatically determine a sequence of pillars that yields this shape is a significant and impactful advance, which is ideally addressed by computational approaches. This challenge motivates the objectives of the CDS&E project: (i) Computationally explore and create a library of pillar-induced transformations annotated at different levels of granularity to aid in computational selection using parallel CFD simulations. (ii) Develop efficient computational methods to solve the inverse problem and select pillar sequences for a set of desired flow transformation. This part of the project will explore mathematical and computational issues related to uniqueness of solution sequences, scalable approaches to deal with the large libraries of pillar transformations, and choice of cost-functionals to enable efficient solution to the design problem. (iii) Test the computational framework and associated solutions by fabricating microfluidic designs with the defined pillar sequences that address three transformative applications in medicine and materials, including fabricating tailored cross-sectional polymer fibers, and capturing biomolecules on microchannel surfaces.
The introduction of a general strategy to program fluid streams in which the complexity of the nonlinear equations of fluid motion are abstracted from the user can impact biological, chemical and materials automation in the same way that abstraction of semiconductor physics from computer programmers enabled a revolution in computation. As part of dissemination efforts, gaming and educational modules involving immersive simulations and directed rubix-cube like puzzles will be developed that will allow the public and interested parties to experiment with different pillar programs and learn about fluid mechanics and applications in a gaming environment. These outreach and workforce development activities will emphasize to the community the crucial role of computing in science and technology. This outreach will synergistically enable crowd-sourced design of complex flow transformations for applications that have a major impact on cell diagnostics, nano-materials fabrication, and thermal cooling.
