I would like to start running FSI simulations for cardiovascular applications. More precisely, I'm interested in the behaviour of aortic valve under physiological flow conditions in which the interaction of fluid and structure must taken into account and the valve undergoes very large deformations. I read that there are two approaches to couple fluid and solid solvers: segregated or monolithic. Apparently seregated solver has drawbacks in terms of accuracy and stability. Is any of the 2 coupling ways more advised for certain type of setups?

While searching online I have find several opensource codes such as Fenics, Oomphlib, ElmerFEM and LifeV (I have used LifeV in the past). Besides this, I have access to commercial software, but would like to have an open source alternative as well. This is for two reasons: self-education (I can look at the source code and learn more about the numerical method) and flexilibility - if I have some special requirement, I could eventually implement it.

Unfortunately I don't have the time to develop an FSI solver from scratch and I need to start running simulations fairly quickly before I dig into the source code. Which solver and method would you recommend so that the initial learning curve is not too steep?


People typically agree that the monolithic approach is the conceptually better one, but of course more difficult to implement.

If you want to use a code that already does most of what you need, I would suggest to look at this paper and its accompanying code: http://journals.ub.uni-heidelberg.de/index.php/ans/issue/view/1244


Fenics has support for FSI via Unicorn, that is an extension for the flexible handling of PDE systems problems.

You should check out the freely available Fenics book that contains a detailed exposition of the topic [1]. You should be able to find example code for FSI problems as well, that come with Unicorn.

[1] Logg, Anders, Kent-Andre Mardal, and Garth Wells, eds. Automated solution of differential equations by the finite element method: The FEniCS book. Vol. 84. Springer Science & Business Media, 2012.; Chapter 28 Turbulent flow and fluid–structure interaction

URL: https://launchpadlibrarian.net/83776282/fenics-book-2011-10-27-final.pdf


The monolithic approach is not necessarily better, as stated in another answer.

Of course, if you only want to solve one specific problem in one way, then the monolithic approach could maybe be more robust. But why not get a more general solution you can?

The partitioned approach has some very important advantages:

  • You can use existing single-physics solvers for each domain. This is not only easier, as you don't re-invent the wheel, but you may also already have and know some very good, special, or simply closed-source solvers that you want to reuse. You can use completely different meshes, numerical solvers etc. in each subdomain. They don't need to restrict each other.

  • You can achieve better performance in a cluster, since you are free to assign e.g. less computing resources to the (usually faster) solid simulation and more to the fluid. This way, you can also minimize the communication time, since you only need to exchange information on the coupling times.

Read more in the section 1.2 (page 4 / pdf-page 9) of [1].

There are tools that can couple single-physics solvers to a multi-physics simulation. More precisely, since you are looking for open-source tools, you can try preCICE (on GitHub). There are quite a few codes already coupled (see also [2]).

More precisely, see the 1D FSI example: flow through a deformable artery.

It looks like some people are already using preCICE for the evaluation of heart valve biomechanics.

[1] Uekermann, B., Partitioned Fluid-Structure Interaction on Massively Parallel Systems, PhD Thesis, Institut für Informatik, Technische Universität München, 2015.

[2] Uekermann, B., Bungartz, H.-J., Cheung Yau, L., Chourdakis, G., Rusch, A., Official preCICE Adapters for Standard Open-Source Solvers, in Conference Proceedings at the 7th GACM Colloquium on Computational Mechanics for Young Scientists from Academia, 2017.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.