Abaqus from AnyBody

Main Forum
1

Hello!

I'm trying to make an FE-simulation i Abaqus. I have used the converter from AnyBody to Abaqus, it worked fine.

When importing the output file from AnyBody into Abaqus all the forces are a bit displaced. They are not aligned with the bone.

How can we fix this?

Thank you!

/Sofia

2

Hi Sofia,

You were supposed to export muscle forces in the FE mesh ref. frame. Normally you construct a CT scan ref. frame inside the relevant segment and use this one for export. If you place your CT-based STL in this ref. frame, it should match with the AnyBody bone.

Kind regards,
Pavel

3

Hello again,

We are using the same bone in both Abaqus and AnyBody so they have the same reference system. We exported the bone from AnyBody as a stl. file and then used it in Abaqus but the forces are as I said before are not aligned....

/Sofia

4

Then something is not the same. Try adding exported STL back to the model as a drawing object.

Main.HumanModel.BodyModel.Right.Leg.Seg.Thigh  = {
  AnyDrawSurf draw_exported_surf = {
    FileName = "my_export_femur.stl";
  };
};

The 2 bones should be aligned. If they are not - there is mismatch. Secondly, could your mesher somehow displace the mesh or export it in a different ref. frame?
Pavel

5

We tried this and they are aligned.

We did our mesh in Abaqus and visually the coordinate system looks like it's in the same place as in AnyBody...

Can there be forces coming from that some muscles attach to the fibula?

Sofia

6

Are you exporting tibia? Then yes. And even in other bones there are some muscle forces are a little away from the surface - they represent via points. Could you post a picture of the problem?

Kind regards,
Pavel

7

Yes, we are exporting the Tibia..

Here is a picture.

rcI3sfFQRBOwZe0ytqrrng
rcI3sfFQRBOwZe0ytqrrng3024×4032 2.49 MB

8

it could be okay, a little difficult to say from this projection. But you could also see what objects these are. Do joint nodes appear to be in the right place? I believe you could also check what forces away from the surface and then find corresponding entities in AnyBody.

9

We checked interactions in Abaqus and compared the with corresponding entities in AnyBody, it looks like they are all connected to Fibula.

Here is another picture of how they are attached.

oAfJjLFvROSY8nQTDwH15g
oAfJjLFvROSY8nQTDwH15g3024×4032 2.08 MB

Sofia

10

Okay, then it is probably okay. The fibula is included into the tibial segment. And the rest of the nodes seem to be matching well. I think the results are as expected.
Regards,
Pavel

11

Thank you for your help.

I have another question also. When we receive the output file it's only during one second.... It should be 2.25 seconds...

Do you have any ideas on what this can depend on?

Sofia

12

Are you looking at the load step time in Abaqus? Or actual output in the input files? I would expect everything to be there, but it might differ from the load step time, which is not set by the converter. Could you check this? You can just open generated files and see that the amplitudes are created for the whole analysis.

Kind regards,
Pavel

13

We were looking at the actual output in the input files but we solved it.

I know that this Abaqus is not your area but I'm asking anyway :slight_smile:. How do we now what period time we should have? And also increment size?

Sofia

14

Okay, sounds good.

The time - it probably depends what you try to analyze. Are you interested in peak stresses? Or in average loading over a cycle of motion? I think Abaqus may have an automatic increment decision system depending on the loads and selected parameters.

Try reading Abaqus ref. manuals for the associated keywords?

Kind regards,
Pavel

15

Thank you very much. We solved all our problems :smiley:

Sofia

1 Like
16

This topic was automatically closed 125 days after the last reply. New replies are no longer allowed.