Novel multi-block strategy for CAD tools for microfluidics type applications

Abstract

A novel strategy to implement multi-block schemes in Computer aided design (CAD) tools for a variety of applications specifically dealing with multi-physics transport phenomena is presented here. Multi-block schemes are domain decomposition schemes in which the domain is split into several smaller sub-domains, and independent, synchronous calculations are performed in the sub-domains individually, retaining the global identity of the problem. These schemes are very appropriate for problems with complex geometries and physics, and problems that are computationally intensive. They also allow parallelization on distributed and shared memory parallel architectures. Therefore, for complex multi-physics applications like microfluidics, which is spotlighted here, multi-block strategies can offer many advantages over traditional finite element or finite volume schemes. However, from a CAD tool implementation point of view multi-block schemes are still in infancy. One important reason for this is the different kind of geometry and mesh builders required for multi-block schemes owing to the inherently different philosophy entailed. To circumvent this, in this work we present some novel strategies to utilize the commonly used geometry builders and meshing tool of finite element technique for multi-block solvers, and develop a versatile CAD tool based on domain decomposition technique. In particular a lucid methodology is presented using which rudimentary and widely available geometry builder and meshing tool for finite element calculations can be utilized to implement a powerful multi-block scheme using structured grids and finite-volume discretization, both of which are amenable to ease of implementation and numerical efficiency for the class of problems considered. The proposed multi-block methodology can also be used with finite difference or finite element calculations in individual blocks. It will be shown that this formulation will have the potential to be the bulwark of next generation of micro- and nano-scale fluidics type problems that have applications in areas like proteomics, lab-on-a-chip design, synthetic biomaterials, etc.