[Dog model] Inconsistent values for joint reaction analysis

Provide easy-to-use, extensible software for modeling, simulating, controlling, and analyzing the neuromusculoskeletal system.
POST REPLY
User avatar
Maxence Lavaill
Posts: 6
Joined: Thu Mar 07, 2019 10:00 pm

[Dog model] Inconsistent values for joint reaction analysis

Post by Maxence Lavaill » Thu Mar 07, 2024 6:00 pm

Hi everyone,
I've had an issue for quite some time now and struggle to explain why the joint reaction analysis tool gives me stupidly high values (>15 000N) with the dog model from Stark et al., 2021 (https://simtk.org/projects/dogmodel) through all joints of the left forelimb.

When muscles are included in the SO computation, resultant joint forces are crazily high, whereas when I disable all muscles and only use actuators, the JRFs seem coherent with the ground reaction force. See pictures below
Do you know what is happening here?
WithMusclesAndActuators.png
WithMusclesAndActuators.png (70.12 KiB) Viewed 778 times
OnlyActuators.png
OnlyActuators.png (63.73 KiB) Viewed 778 times
I've also attached the model and data I use to run SO and JRF at the same time. Note that I've tried running this using OpenSim 3.3 and 4.4.

I tried one more thing without success but it gives hints to where the analysis may fail. First, I locked all degrees of freedom and iteratively unlocked them from the most distal to the most proximal. The JRF values seemed consistent with GRF until I unlock the dofs between the humerus and the scapula, where everything starts acting weird.

Any thoughts?
Thank you very much
Max Lavaill
Attachments
Dog_Example.zip
(36.07 KiB) Downloaded 45 times

User avatar
Heiko Stark
Posts: 11
Joined: Tue Jan 05, 2016 6:59 am

Re: [Dog model] Inconsistent values for joint reaction analysis

Post by Heiko Stark » Mon Mar 11, 2024 8:09 am

Seems to be a bug with Output Precision, see:

viewtopicPhpbb.php?f=2032&t=17653

Tested with OpenSim 4.5_24_01_10

Best regards,
Heiko

User avatar
Maxence Lavaill
Posts: 6
Joined: Thu Mar 07, 2019 10:00 pm

Re: [Dog model] Inconsistent values for joint reaction analysis

Post by Maxence Lavaill » Mon Mar 11, 2024 5:38 pm

It's been fixed and it works, thank you very much Heiko!

Max

POST REPLY