OpenFOAM as a Model Sharing Environment for Macro-Microscale Biomedical Simulation

Although the micro-to-molecular scale biomedical research community has recently evolved numerous successful model sharing paradigms, there has been less progress at the organ-level macro-scales. This is because “model sharing” carries a significantly different set of implications at the micro- and molecular scales than in multi-dimensional macro-scale bio-physics modeling where high level software packages for medical image and geometry processing, mesh generation, CFD, CSM, mesh adaption/motion and their coupling, define the analysis tool space. Accordingly, “model sharing” in this context can involve specification of software components that have widely varying interfaces, source code availability, licensing and operating system requirements, data structures, mesh topologies, parallel execution frameworks, GUIs, required user training and other software engineering elements. In short, they are hard to “share”.

Recently, the open-source, open-development, and freely extensible OpenFOAM (Open Field Operation and Manipulation) toolbox, has emerged, which has features that give it the potential for greatly expanding the ability of macro-scale biomedical modeling researchers to share their models directly. OpenFOAM is an object-oriented framework written in C++, for customization and extension of numerical solvers for continuum mechanics problems including CFD, CSM, FSI, electromagnetics, neutronics and other multi-physics disciplines. The model development approach using OpenFOAM is that each sub-model (e.g., turbulence model, particle deposition model) or high level component (e.g., FSI solver, DNS solver) can be developed and installed modularly alongside current and future OpenFOAM installations which are freely available. This architecture is therefore inherently amenable to model sharing, since once users have adopted the richly-capable baseline version of OpenFOAM (relevant here: CFD, CSM, mesh motion/adaption, arbitrary polyhedral unstructured, high order numerics, MPI parallel), they can easily integrate the sub-models or high level components of any other group in the community and straightforwardly share their models with the community simply by downloading/posting the software from/to the OpenFOAM Wiki or one of numerous open discussion forums.

This open-source, sharing-based software development environment has led to OpenFOAM’s explosive growth in recent years among macro-scale multi-physics simulation research communities with over 3,000 users worldwide. It has only just begun to build acceptance within the biomedical macro-scale physics modeling communities, where model sharing, dissemination and reuse could lead to more rapid research progress and more efficient expenditures of sponsor support.

In this presentation, I will provide an overview of OpenFOAM, and our recent experience with the toolkit, with emphasis on how this approach can be effectively adopted in the bio-medical macro-scale simulation community.

{Return to Working Group 3}