CMC Tool

[Dalia Di Fraia]

Hello,

I am learning to use OpenSim for my thesis project, and I have problems with the tool CMC applied to the following models of upper limb: ``Upper Extremity Dynamic model (MoBL_ARMS_Upper_Limb_Model_OpenSim)'' ``Delft Shoulder and Elbow Model (das3_0)'' ``Dynamic Arm Simulator (DAS3)''.
I would use these three models of shoulder, already implemented in OpenSim, to evaluate muscle forces, providing as input a movement simulated in laboratory.
I do not yet have the file with the experimental kinematics; so I created a file ``.mot'' simulating a simple abduction movement (no external load). The file .mot provided is a matrix all whose entries are zero, whith the exception of the ``time'' column, and of the column corresponding to the abduction degree of freedom; the entries of this last column gradually increase from 0 to 90, both extremes included.
I provided as input to the tool CMC the following three files:
- the previous file .mot simulating the simple abduction movement (to be replaced with the file containing the
experimental kinematics, once available),
- the files ``Tasks.xml'' and ``Actuators.xml'', generated by copying the respective files of lower limb example
``Gait2354_Simbody'', and entering the coordinates of the joints of the upper limb model.

I applied the Tool CMC to each of the three models of upper limb ``'MoBL_ARMS_Upper_Limb_Model_OpenSim' ``das3_0'' ``DAS3'', and got the following results:

- ``'MoBL_ARMS_Upper_Limb_Model_OpenSim': it is suggested using the ``slow optimization target'' instead of the
``fast optimization target''. After the change, the tool printed the output files with the
results;
"`CMC.computeControls: WARN- The optimizer could not find a solution at time = 0.000000.
If using the fast target, try using the slow target.
Starting at a slightly different initial time may also help."'

- ``das3_0'': only the following error messages
"`CMC.computeControls: ERROR- Optimizer could not find a solution.
Unable to find a feasible solution at time = 0.5.
Model cannot generate the forces necessary to achieve the target acceleration.
Possible issues: 1. not all model degrees-of-freedom are actuated,
2. there are tracking tasks for locked coordinates, and/or
3. there are unnecessary control constraints on reserve/residual actuators."'

- ``DAS3'': the tool does not start calculating and does not report any error in the messages window.

Can someone tell me which mistakes I am making?
Is there someone who has already successfully applied the tool CMC to one of these models of upper limb? Which files were provided as input?
Providing as input a file .mot with many zeros can be a problem?

Sorry for the long post. Hope someone can help me.

Thank You,

Dalia

[jimmy d]

Hi Dalia-

Thanks for your detailed post. Those models all have a history of not working well in CMC simulations. This is more to do with the way the models are constructed. Have you managed to solve the issue or are you still struggling to get it to work?

-james

[Dalia Di Fraia]

Hi James.

First, thanks for your answer. I have tried to change muscle models of all upper limb models with the Millard's one (" Millard2012EquilibriumMuscle"), but it was not enough: I still have problems with the tool CMC.

Now, I am working to create the file ".mot" with the experimental kinematics, that will be provided as input to one of the selected upper limb models.

I am also considering the idea to estimate muscle forces by using static optimization Tool with suitable reserve actuators, instead of CMC Tool.

Would you have any suggestions for us to solve the issue?
Whatever advice would be helpful.

Thank you very much,

Dalia

[Dimitar Stanev]

Hi Dalia,

About MoBL_ARMS_Upper_Limb_Model_OpenSim, the mass properties of the bodies are not specified. I had tuned them from Hamner's full body. I also can provide you with the inverse kinematics for a simple motion. I think that I managed to run CMC when I corrected the mass properties, but I am not sure.

https://dl.dropboxusercontent.com/u/50567323/MoBL.rar

Hope it helps

[Dalia Di Fraia]

Hi Jim,
Thanks for your answer and your help!
In my MoBL_ARMS_Upper_Limb_Model_OpenSim the mass properties of bodies are specified. I have however noticed that in your model the ground and the thorax bodies (as in other upper limb models) have been separately handled. I am going to carefully compare your model with mine.

I will inform you whether I find a solution.

Any other tips are and will be well accepted.

Dalia




Again, thank you very much

[Eduardo Sáenz Aldea]

Hello all,

I'm also struggling to run CMC in MoBL_ARMS_Upper_Limb_Model_OpenSim when I activate the fast optimization target, it keeps saying this:

Code: Select all

SimTK Exception thrown at interiorpointoptimizer.cpp:264:
  Optimizer failed: Ipopt: Not enough degrees of freedom (status -10)

I've checked the model, the tasks file and the control constraints file and they all seem to be ok.

Dimitar, could you send again the model you modified? I'll try to compute CMC with that model and see if it works.

Thank you for your help