Help needed: Problems evaluating contact forces in sitting MoCapModel

Main Forum
1

Dear colleagues,

I courteously ask for your help!!

I’m dealing with a MoCap model consisting in the full mannequin (AMMR v.1.6.3) sitting on a chair. The mannequin position is defined by using own .c3d data acquired from a sitting subject (according to VICON protocol). Mannequin height and weight in ‘TrialSpecificData.any’ are that of the acquired subject.

I used the model in the following AMMR repository, containing a chair model: :\AnyBody\AMMR\Application\Validation\WilkeSpinalDiscPressure\SpinePressureSeatingRelaxed. (Anybody version is 6.0.4).

Then, I simplified the chair model (just seat and backrest). Feet are placed on the ground. The chair placement matches that of the real chair where the acquired subject sited.

I need to evaluate the contact forces between the mannequin and the chair surfaces. Contact points were defined for the vertebrae, for pelvis and for thighs (‘SeatSupport.any’ and ‘BackrestSupport.any’).

Model runs correctly (no errors), but something that seems not correct happens in the results of the contact forces.

I upload the files. Parameter optimization step was already done (the resulting files are in the \Input folder) and you can thus run the InverseDynamic step directly (or otherwise you can repeat the Optimization step if you prefer).

As you will see running the model, the contact forces result concentrated in single contact points instead being distributed along all the contact points that are close to the seat surfaces.

I tried to change the limits for the strength measure function (in ‘SeatSupport.any’ and ‘BackrestSupport.any’) but nothing changed.

Important notices:

  • differently from the original sitting model (that from the AMMR repository), in my model no joints are defined to connect the mannequin and the chair since the mannequin position is calculated according to the .c3d data.
  • the model form the AMMR repository accounted for the InverseDynamic step only whereas my model (MoCapModel) accounts for Optimization and then InverseDynamic.

Do you have any helping or solving idea??? Am I making mistakes??
Thanks a lot in advance!
Tito

2

Hi Tito,

The explanation might be that you have too few contact points or some contact elements do not initiate contact for one or another reason. For example, you do not have feet and thigh support - the whole body weight is carried by those contact elements (possibly due to the horizontal proximity to the CoM or largest moment arm to counter balance back support force). If you add more points on the thighs and the feet ground reaction force prediction objects - the picture changes to something like that:

The pressure is now distributed over the surface of the thighs. I would try adding more points at the expected area of contact. Possibly more points on the back

I am attaching modification of your model from the screenshot back.

Kind regards,
Pavel

3

Thanks a lot Pavel for your precious help!

I’ll try to work in that direction.

Kind regards,
Tito