Is there a max. number of wrapping ligaments during FDK?

Main Forum
1

In my model I can activate the wrapping for about 150 ligament strings, but then the FDK doesn’t work anymore.

If I deactivate the Wrapping only (not the ligaments) the FDK works fine
… and the other way around …
If I deactivate the FDK the Wrapping works fine.

Thanks
MMMPPPMMM

PS: I know that the wrapping algorithm is unstable, but I would like to try it.

2

Hi Fritz,

We were just discussing your issue. There is no max number of wrapping ligaments during fdk simulations. I guess in your case there are several factors of difficulty coming together.

How does the fdk model fail? Can it not find a solution?

Could you see something similar in a simple model that you could share with us?

3

BTW: the newest AMS version is 6.0.3

4

The study doesn't fail. The study works fine, but the FDK is not activ. There is no error or warning message.

At the moment I'm only working on this model. And the creation of a simple FDK model with 150 wrapping ligaments is presumably very time consuming.

I'm going to post the study logs for both cases tomorrow.

5

ok, thanks. This is really strange. We will discuss this again. >Wait for the logs.

6

With FDK without wrapping - FDK works fine - Simulation time: ca. 20 min.:


0.0) ...Ligament Length Adjustment finished.
0.1) ...Design variables of the calibration model have been updated.
1.0) InverseDynamics (Operation: Main.CarpusAnalyse.CarpusStudy.InverseDynamics): 
1.0.0) PreOperation (Operation: Main.CarpusAnalyse.CarpusStudy.InverseDynamics.PreOperation): 
1.0.0.0) InitialConditions (Operation: Main.CarpusAnalyse.CarpusStudy.InitialConditions): 
1.0.0.0) ...Design variables have been updated.
1.0.0.1) ...Load-time positions have been re-established.
1.0.0.2) ...Kinematic analysis completed. The kinematic constraints have been resolved.
1.0.0.3) ...Initial conditions are fully updated.
1.0) Inverse dynamic analysis...
WARNING(OBJ.MCH.MUS1) :   D:/...  :   CarpusStudy  :  The muscles in the model are not loaded due to kinetically over-constrained mechanical system.
NOTICE(OBJ1) :   D:/...  :   CarpusStudy.InverseDynamics  :  No muscles in the model.
1.99) ...Inverse dynamic analysis completed
2.0) save (Operation: Main.CarpusAnalyse.save):

With FDK with wrapping - No movement in the FDK DOF - Simulation time: ca. 2 min.:


0.0) ...Ligament Length Adjustment finished.
0.1) ...Design variables of the calibration model have been updated.
1.0) InverseDynamics (Operation: Main.CarpusAnalyse.CarpusStudy.InverseDynamics): 
1.0.0) PreOperation (Operation: Main.CarpusAnalyse.CarpusStudy.InverseDynamics.PreOperation): 
1.0.0.0) InitialConditions (Operation: Main.CarpusAnalyse.CarpusStudy.InitialConditions): 
1.0.0.0) ...Design variables have been updated.
1.0.0.1) ...Load-time positions have been re-established.
1.0.0.2) ...Kinematic analysis completed. The kinematic constraints have been resolved.
1.0.0.3) ...Initial conditions are fully updated.
1.0) Inverse dynamic analysis...
WARNING(OBJ.MCH.MUS1) :   D:/...  :   CarpusStudy  :  The muscles in the model are not loaded due to kinetically over-constrained mechanical system.
NOTICE(OBJ1) :   D:/...  :   CarpusStudy.InverseDynamics  :  No muscles in the model.
1.99) ...Inverse dynamic analysis completed
2.0) save (Operation: Main.CarpusAnalyse.save):

I think the FDK solver crashes at the beginning but there is not error message
Same Bug as:
http://forum.anyscript.org/showthread.php?t=3503

7

Hi,
I’ll forward the issue to the FDK guys in our group. We will discuss it and get back to you!

8

Hello,
Unfortunately, we cannot reproduce your problem. So it is difficult for us to tell you what exactly goes wrong without seeing the model and doing some debugging on it.
I can explain how we would start debugging such a case.

  1. You use a force tolerance of 10 seems very high, can you reduce that a little bit? I usually have 1N

  2. There are two options either try to reduce the wrapping number or the fdk DOFs to see when it works

That way I would see if there is a specific link that causes trouble.