Anybody to Abaqus

I need someone’s help.
I applied the force generated from GaitFullBody of AnyBody to custom femur of which diaphysis was replaced with prosthesis. I found the coupling in proximal and distal femur was failed.
I don’t know why. It wouldn’t occurred in project of normal femur,which I have done before.
Coul someone do me a favor?

Dear Yao Tommy,

It is not really clear what went wrong here. We would need a little more information.
I have only one idea - did you try to use a diaphysis mesh (without proximal and distal parts) for converting the loads by chance? Would it somehow be possible to share the input data with us so I could try?

Secondly, it looks like you have a mismatch between the muscle attachments and the surface. Is that intended?


I just had another thought. Are the nodes in the mesh deck somehow distributed over a few sections? Was the mesh file generated by the Abaqus CAE itself?

The converter handles a single section of nodes (it will find the first one) and create constraints to the first section only. It seems that this is the case here.


Thank you.
No muscle attachment can be observed in distal and proximal part of the model. But actually, I have exported all the forces and adjusted their local reference coordinate to my model (with Hypermesh).
“Are the nodes in the mesh deck somehow distributed over a few sections?”----I dno’t know.
The mesh generated from a 3D reconstruction software——Simpleware 5.0.
I shared my initial data to you, and would preciate your preciate help.
Tip: The two inp file should be copied to “D:\model” of your computer, and you have to import the file “model_56.inp” to load in.
Thank you again.

Dear Tommy,

There is something wrong with the automated conversion - for one another reason the exterior surface is not properly selected by the converter. Could be due to a little complex definiton of the mesh. But what you can do - you could delete the following section from the generated file:

*elset, elset=anybodysurfaceset, generate
1,35162, 1
anybodysurfaceset, S1
anybodysurfaceset, S2
anybodysurfaceset, S3
anybodysurfaceset, S4
anybodysurfaceset, S5
anybodysurfaceset, S6

or better simply modify this section and manually select exterior surface of the bone. The code should pick it up and apply the loads properly.

But it will only work if the muscle forces were output in the CT-scan ref. frame. For that you could do the following - you should create an AnyRefNode, which positioned so that putting the geometry (an STL) in CT scan inside of this node, will put it on top of the ANyBody geometry. Please have a look at the ANSYS tutorial, which may demonstrate how this is done for ANSYS.

And after that the output code will look like:

AnyMechOutputFileForceExport FEAOutputObject = 
  FileName = "../Output/FEAOutput.xml";
  UseRefFrameOnOff = On;
  //AllSegmentsInStudyOnOff = Off;
  XMLformatOnOff = On;
  //ForceObjectExclude = ;
  //ForceObjectList = ;
  AnyRefFrame &rfout = Main.HumanModel.BodyModel.Left.ShoulderArm.Seg.Scapula.FEOutputNode;
  AnySeg &seg = Main.HumanModel.BodyModel.Left.ShoulderArm.Seg.Scapula;

Where the red code represents the newly added reference frame.

Kind regards,

Dear Pavel,
Thank you very much. I am so sorry that I haven’t figured it out.
I did finish what the tutorials say and have the successful case of normal femur (Fig.1). There must be something I have missed, I think.
Please look at this case—A femur, recontructed from dicom pictures, was separated into two parts (cortical bone and cancellous bone). The local coordinate was well accommodated. However, it seems that loads were only coupled with the surface of cancellous bone instead of cortical bone (Fig.2). I don’t know why? Can you help me?

Best regards!!

Dear Tommy,

I did not understand what the problem is in the first image - it looks like a correctly applied configuration.

But there is a 2 step solution to the problem in the second picture:

  1. you need to recompute influence radii for the bone - the easiest way is to create a surface mesh (only exterior cortical bone) and run the converter on this mesh.

  2. secondly, for the actual model you need to specify the volumetric mesh file instead of the surface mesh file, but there will be one change. You need to redefine ‘anybodysurfaceset’ to be the exterior cortical surface of the femur. It should be rather easy to do from the GUI.

Once you do these two steps i would expect everything to work.

Does this make sense?

Kind regards,