simtk.org

Hello teachers, I am having some problems running RRA. I used the gait2354 model. Refer to the Visual3D to OpenSim_4_0 file. Runs RRA directly after running scale. But I found some issues after running the RRA (as shown in the picture), the scaling model was fine in the left support phase, but dropped in the left swing phase. I tried to run the sample file and it didn't happen, which confused me. Do any of the teachers know what's going on?

Hi, RRA works well when all external forces exist and are applied to the model. If this is a walking task, three force plates must be assigned (one to the left leg and two to the right leg in your case). Is this condition satisfied?

kernalnet wrote: ↑

Sun Oct 29, 2023 9:00 am

Hi, RRA works well when all externalfo rces exist and are applied to the model. If this is a walking task, three force plates must be assigned (one to the left leg and two to the right leg in your case). Is this condition satisfied?

It is indeed a walking task. Unfortunately, we only have two force plates. During the collection process, it is impossible to ensure that the left and right feet are fully stepped on the two force plates. You mean that this situation is due to the lack of complete force data on the right side, right? Is it OK if I only select the left support time range when running RRA? Or do you have any better suggestions?

we only have two force plates. During the collection process, it is impossible to ensure that the left and right feet are fully stepped on the two force plates.

These explain why the model drops. When using RRA, always check the residual forces/moments as well as the kinematics errors. Please see: https://simtk-confluence.stanford.edu:8 ... ourResults:

You can crop the time range to include the frames in which all external forces exist (single limb support).

But in your case, you may want to use Static Optimization as an alternative to RRA+CMC.

kernalnet wrote: ↑

Wed Nov 08, 2023 11:33 am

we only have two force plates. During the collection process, it is impossible to ensure that the left and right feet are fully stepped on the two force plates.

These explain why the model drops. When using RRA, always check the residual forces/moments as well as the kinematics errors. Please see: https://simtk-confluence.stanford.edu:8 ... ourResults:

You can crop the time range to include the frames in which all external forces exist (single limb support).

But in your case, you may want to use Static Optimization as an alternative to RRA+CMC.

I followed the tutorial on the Visual 3D website and ran the RRA directly after scale. The tutorial may be a little easier than the OpenSim website, eliminating some intermediate steps. But according to your suggestion, I may have to do this according to the official website scale-IK-ID-Static Optimization, which needs to adjust the markers, which may be more complicated than the previous method. So I want to know what effect crop the time range and use Static Optimization as an alternative to RRA+CMC have on the muscle force obtained in the end?

Hello Jing,

Have you taken a look at the tutorial 'Computing Joint Moments With Experimental Data'? https://simtk-confluence.stanford.edu:8 ... ental+Data The Inverse Dynamics/RRA steps and associated questions may be of relevance to you here.

Hope that helps.

Nicos

If using the visual3d to opensim workflow. you already have all the required files for running Static Optimization:

• Scaled model (.osim),

• IK file (.mot),

• GRF file (.xml),

• CMC actuator file (xml).

Please see: https://simtk-confluence.stanford.edu:8 ... ion-Inputs
So, you can continue the workflow with Static Optimization after scaling the model.