Hi everyone I am looking to produce a structural model for a fuel cell stack. I would like to look at the load distribution throughout the assembly and at the seal deformation and stress. Does anyone have any experience of modeling something similar? It could be heat exchangers or anything with multiple layers of seals. I have tried using COMSOL to do this but it fails miserably to model just a basic seal in a groove and their support have said it can not be done. I know people have used ANSYS for hyperelastic modeling of seals in grooves but was wondering if anyone had taken it further or what is the best software package for this? Thanks Mike
Although I have done plenty of structural analyses using Nastran, Catia and Ansys, I have never modelled hyperelastic elements. You could try and look on the Ansys website. There are plenty of webinars on various applications. Maybe one's about what you have in mind.
Hi Mike, I provide analysis services and can quote this for you. Let me know how you'd like to proceed. Do you have solid models yet of the design? Regards, David Haag, M.E.
yo can use Abaqus for modelling seals. There are hybrid elements available in Abaqus for modelling incompressible materials.
Hi David Thanks for your reply This is a tool that we will need a lot so it is worthwhile for me to develop it. Thanks for your offer. I will be in touch if I get stuck.
Hi Ghouse Thanks for the reply, I will look into now. I rember seeing on the COMSOL forum how someone had switched to Abaqus for a solution to a similar problem. Mike
I have not actually addressed the problem you are considering, but you may want to have a look at some of the packages included in the OpenSource CAELinux distro.
The CFD package seems incredibly interesting giving the codes it is sourced from. Do you know if it run smoothly in a virtual box ? Regarding the seal model, if I remind well you can work with the time base btw each iterations. I might be able to do it by myself one more time if you decide to externalize this project. But that Comsol can't achieve some kind of results here is surprising.
srdfmc- Yes- CAELinux can run in a virtual machine (if the virtual machine is 64 bit- all the latest versions of CAELinux are 64 bit. You have to go back a few versions to find a 32 bit compatible version). Or, you can run it from a Live DVD or usb stick. Or dual boot (except maybe unless your primary operating system is Windows 7- I understand MS has opened up the boot sector to some commercial houses to allow them to write anti-piracy guards where owners can't find it. Apparently, this tends to really screw up the boot sector used for dual boots. I don't know for sure- my latest version of Windows is XP, although I still prefer 98). I personally prefer to run it as the primary operating system- faster. Also, it can be a memory hog, as one would expect- that might put some limits on it in a virtual machine. It is not only the CFD packages that are phenomenal- I especially like the Salome platform for geometry/meshing/post processing, although it uses a unique file format- Code_Aster (for FEA) and Code_Saturne (for CFD) can be run from within the Salome platform. There are also some really intriguing special-purpose analyses programs (Impact, Gerris, MBDyn, Dynela, Dolfin/Fenics, etc.- Dynela, a non-linear explicit dynamics package may be appropriate for the deforming seal problem), other meshing options (GMESH, netgen, tetgen, Engrid), and a pretty good suite of mathematics (Scilab, Octave, R, Maxima). Of course, all the other stuff one expects from a typical Linux distro (OpenOffice, audio/video, etc).
@ CWarner : Thx for the detailed advice. I will download it and hve a try on my WIN7 Pro 64 OS Definitively, Win98 was the best OS MS ever released. But 7 Pro is quite stable and have proved (so far) to be bugless on my PC. Would be interesting to run it in parallel with Flows from SW an see how it behave.
