Then, we used following code to output the loading magnitudes for FE analysis:
AnyOperationShellExec ConvertToAbq_L4L5Exclude=
{
OPERATION_DISPLAY_PRIORITY(PriorityHigh);
FileName = "../Result/FEA/AnyFE2Abq/AnyFE2Abq.exe";
Arguments = "-i ../Result/FEA/L4L5_Exclude.xml -o ../Result/FEA/spine_Exc_FE.inp -m ..\Result\FEA\Mspine.inp";
WorkDir="./";
};
We successfully output the files, however, the inp files seem to be probamatic in following aspects.
while we have 111509 elements, the anybodysurfaceset only contains 104042 elements as shown below:
*elset, elset=anybodysurfaceset, generate
1,104042, 1
We found that many loadings of the same muscles (Fx,FY,FY,MX,MY,MZ) have exact same magnitudes through the entire stride. For instance, for the muscle (Erector spinae, LTptT11Sacrum), the forces and moments are all the same as shown below. I guess there might be an issue in the Anybody converting tool.
I can't remember, but the converter has a simplistic reader for Abaqus files - if you have a somewhat complex file, it will not be able to parse (it might miss some elements). OR there might also be a numbering difference, e.g. your elements are numbered different and you think you have more than the converter writes. It is intended for bringing simple meshes from segmentation tools in. Please check or share how your mesh input file looks like before we do complex checks
The similar forces sound suspicious... Could we get the files to check them? I don't think we've seen this problem before.
Could you also mention which version of the Abaqus converter you are using?
Thanks a lot for your response. I have attached the files for your reference. I couldn't find the version of the Abaqus converter but the "date modified" of the "AnyFE2Abq.exe" is 2013-11-13 and its size is 22 KB. I also saw a newer version called "AnyFE2Abq.exe" with a "date modified" of 2017-12-22 and a size of 184,805 KB. However, the newer version did not work (it showed a black screen for 5-10 seconds and then closed without any messages). FEA.7z (5.5 MB)
As i said, the mesh looks quite processed and a lot of numbering is irregular - the parser is simplistic and cannot handle it. One way to go around would be to redefine the "anybodysurfaceset" in your model that would correspond to the boney surface. This way you go away from the parsing problem.
The erector spinae repetition is probably caused by your exporting forces in L5 ref. frame, where the export target is L4 - in the XML file it shows as a sliding force and creates many instance of the same. I am not 100% certain why the values are the same, but i think it might not be important since the whole model is not properly defined.
Last, but not least, you are not accounting for the units. I could see that the converter was not given a unit conversion factor. Your mesh is in mm, your AnyBody output is in meters. You get a crazy behavior.
Thanks a lot for your response. Yes, I noticed that the mesh number is not continuous and regular. Thanks for your suggestions to resolve this parsing issue. Sorry for the confusion of mesh unit, yes, it should be in meters.
Regarding the export of muscle forces and moments to the inp file, we actually followed your previous suggestions in this post (Modeling L4-L5 and intervertebral disc).
I don't think this would be the cause of same magnitudes issue. As I checked the xml file, I found that for muscle ErectorSpinae_LTptT11Sacrum, the force in XYZ was different, and moment in XYZ was 0 as shown below.
For your reference, such repetition of forces and moments not only happen in erector spinae, it actually, happen in many muscles (e.g., PsoasMajor [both right and left], QuadratusLumborum (left but not right)). My guess is that there might be some issues in the AnyFE2Abq.exe.
Please let me know if you have any different thoughts. Thank you.
