Hi Hadi and deal.II community, It appears as if the underlying issue with your input deck has been found, which is great! Thanks Timo and Wolfgang for taking the time to troubleshoot. I suspect that you could put any material ID in there and it wouldn't matter unless you write code that specifically uses that ID retrieved from a Cell_Accessor by the call cell->material_id() .
The "mesh converter", as Timo suggested, is not very robust at all. It was developed with no reference to documentation on how the ABAQUS input file is laid out, so I made it such that, when outputting from CUBIT (as stated in the documentation and program output), it gives the result one wants. Unfortunately for you, since I wrote it to solve my specific problem at the time and not deal with the issue in general (sorry...), I suspect that there sort of problems would be encountered when trying to convert any input deck from any other source. It turns out that there are many subtilties to the ABAQUS format, and they're impossible to account using the low-level approach which was used to develop the converter. In fact, CUBIT themselves revised the formatting of the file they output, a clear indication of the flexibility of the format and, hence the requirement for a switch to account for which CUBIT version was used to generate the file to be converted. That being said, poor / short-cut coding is no excuse for stagnating a project and I think that it *might* be possible to fix the problem you encountered and make the program a little better. Perhaps you'd be willing to send me the unaltered file that you tried to generate the UCD mesh from so that I can see if it can be done. One day I'll get to rewriting it to be more flexible or perhaps incorporating the functionality directly into deal.II. Regards, J-P On 25 August 2011 00:51, Hadi Rahemi <[email protected]> wrote: > > I dont have abaqus either. I am using output files from "IA-FEMesh". This > software creates the hexahedral mesh and output it as abaqus input file. I > am not sure which version it supports. I will create a single mesh tomorrow > and try to see if I can put material id in somewhere. > > Thanks, > > > On Wed, Aug 24, 2011 at 3:42 PM, Wolfgang Bangerth <[email protected] > > wrote: > >> >> > > I've actually never looked at that code. Is the question what to do >> with >> > > the material id if the input file doesn't specify it? Then yes, >> setting >> > > it to zero sounds like the right choice. >> > >> > No, that is clear to me. I was asking if you want to do it. :-) >> >> If you have already looked at the code I'll let either you or Jean Paul >> make >> that change. >> >> >> > There are some examples from Jean Paul in contrib/mesh_conversion/mesh >> >> I see. If you want to put Hadi's file in there as well go ahead! >> >> Best >> W. >> >> ------------------------------------------------------------------------- >> Wolfgang Bangerth email: [email protected] >> www: http://www.math.tamu.edu/~bangerth/ >> > >
_______________________________________________ dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
