Pugazhenthi S K, Research Scholar of VIT University, working on Automotive Biomechanics.
we are working on FEA (nonlinear explicit analysis) exploring FSI simulation to attain real biomechanical application in crash conditions. The study tool is LS Dyna and Abaqus for the analysis.
It is as complex to me to handle FSI as I newbie to handle in LS Dyna. Further, I hope Precise is the exact platform to run the simulation. Here also I am a newbie. Is anyone helps me out to sort the possibility and methods suitable for my analysis?
I hope your valuable input/ guidance or software coding for the coupling process (Multiphysics) will definitely be a torch-bearer for the study.
curious to explore, have lack of sources and guidance, hope your support.
Hi @pugal456 welcome to the forum!
Lets try to understand your application and the coupling part of it a bit better. You are trying to set up a FSI simulation using LS-Dyna and Abaqus. Is LS-Dyna the structural part and Abaqus the fluid part? Are you experienced with both codes?
There are mainly two ways you can couple a solver using preCICE. One way is to set up the simulation case and then add some preCICE API commands to hand over some control to preCICE so that it can perform various coupling steps. The second way is to first check if the solver you are working with has an adapter in the preCICE project. If we already provide an adapter then we recommend you try to use it. The purpose of an adapter is to wrap preCICE API in a way that a user experienced in the solver will be able to integrate it rather quickly in their simulation setup.
In the current list of adapters you will not find LS-Dyna or Abaqus, but coincidentally at the moment we are working on developing an adapter for LS-Dyna. This adapter is still not in a working state as we are currently developing it. Are you familiar with LS-Dyna? We can then discuss further regarding this adapter.
I have no experience with Abaqus, but coupling a Abaqus case using preCICE should not be impossible if you know Abaqus well enough.
Now, I am working in LS DYNA solver for study. Iam interested to use any type adapter which is suitable and user friendly to handle, since I am a newbie to this I request your recommendation of solvers.
Our research team is using LD Dyna for FE study and we have licence. If you prefer any tool which cable to import ANSA MESH (output) with errors I am interested to benefit it.
Instead of CSD - LS DYNA, CFD-OPEN FOAM(or any other) tool you suggest I am ready to explore, as its is Biomechanics application your perspective and guidance will be definitely insightful for detailed complex model with 50K solid elements
Thanking you in anticipation
Pugazhenthi S K
For your case if you are already using LS-DYNA for your study then I would suggest to stick with it and try to build up an adapter for it. Usually it is you as a user deciding which solver you want to use in your research and then we see if there is an adapter for it. If an adapter for it does not exist then we just try to couple your code directly. Coupling directly means adding preCICE API commands at specific places. We have an example: Couple your code | preCICE - The Coupling Library which I had tagged before too.
I would suggest going through this example and trying to couple your code and then we can work on it.
Really I am grateful for your guidance! Thanks a lot.
I will follow the couple codes.
I need a few clarifications,
Is it possible to use LS-Dyna key file direct for example mesh file or *.key file. in Precise.?
because the model has been for the past two years of teamwork, we have validated it in LS Dyna, so
- should know whether Could I use mesh file or model data in any particular format to avoid redo from the beginning)
- Any specific mesh-like moving mesh/ Immersed boundary Method should do for FSI study even for Precise.(could you brief)
I do not fully understand what you mean by the
.key file from LS-DYNA. Does this file have information regarding the mesh? Like coordinates of points? Or is it something else?
A typical workflow of defining a mesh as a coupling mesh in preCICE would be as follows: first you define the mesh in your solver, which in your case is LS-DYNA. Then you tell preCICE of the mesh you want it to use for mapping data during coupling. You do this using the
setMeshVertices() command, which is just one of the many API commands. To define the mesh in preCICE you do not need any specific format, you do it manually.
preCICE works with stationery meshes. So during the coupling the meshes cannot move. This is done as having moving meshes for a partitioned black-box coupling involves dealing with several complexities. But currently we are working on this. What users typically do is that in their solvers they have a mapping from the stationery mesh to the moving mesh and the coupling is done over the stationery mesh and then an additional computation is done to map the quantities to the moving mesh. Is this approach suitable for your case?