GRF Prediction

Provide easy-to-use, extensible software for modeling, simulating, controlling, and analyzing the neuromusculoskeletal system.
POST REPLY
User avatar
Melanie Hook
Posts: 9
Joined: Fri Mar 05, 2021 10:02 am

GRF Prediction

Post by Melanie Hook » Thu Jul 08, 2021 6:57 am

Hello everyone,
I have been trying to predict ground reaction forces and moments during walking gait using the zero moment point method described in this paper: https://www.sciencedirect.com/science/a ... 9015005278.

In the paper, they run ID without ground reaction forces applied then transform the resultant forces and moments at the pelvis to the ground. I have been using the raw data values from ID (pelvis tx, ty, tz, tilt, list, rotation) in the calculations described in the paper but am not sure if this is correct. Using the gait2392 model and comparing my calculated results to the provided ground reaction data, my GRF results seem reasonable but GRM do not. I know that residual actuators need to be added to the pelvis for RRA so maybe a similar process is required in this case? I am not very familiar with the concept of residual actuators or how to implement them. If anyone has any suggestions on how to obtain the correct resultant forces and moments at the pelvis, please let me know!

Also, if you know of an easier way to predict GRF, please help! I have mainly been focusing on this paper but also have looked into these methods:

Fluit et al. : https://www.sciencedirect.com/science/a ... via%3Dihub#! - not sure if contact spheres with Hunt-Crossley forces are sufficient or what to do once the contact model is implemented

Contact force static optimization plugin: https://simtk.org/projects/statopt-contact - model won't show up in OpenSim visualizer once contacts are added but no error messages are provided

Thank you so much for your help!
Melanie Hook
melaniehook@vt.edu

User avatar
Thomas Mokadim
Posts: 4
Joined: Thu Jul 22, 2021 4:13 am

Re: GRF Prediction

Post by Thomas Mokadim » Thu Dec 08, 2022 7:12 am

Hi Melanie,

I am also working on such topic. Did you manage to get trough your issue ? Finally what solution is the best to estimate GRF regarding the original dataset ?

Thank you so much.

Regards,
Thomas

User avatar
Johannes Amend
Posts: 6
Joined: Fri Oct 20, 2023 12:45 am

Re: GRF Prediction

Post by Johannes Amend » Mon Nov 13, 2023 6:54 am

Hey guys,

did you find a solution yet? I have a similar problem.

Regards, Johannes

User avatar
Adrián Schmedling
Posts: 13
Joined: Tue Dec 06, 2022 11:36 am

Re: GRF Prediction

Post by Adrián Schmedling » Fri Nov 17, 2023 2:58 am

Hello,

I am also researching this topic, I have found the following related posts on the forum, maybe you can find some information here:

viewtopicPhpbb.php?f=91&t=7095&p=34452&start=0&view=
viewtopicPhpbb.php?f=91&t=11946&p=33745&start=0&view=
viewtopicPhpbb.php?f=91&t=8675&p=23903&start=0&view=
viewtopicPhpbb.php?f=91&t=8457&p=23214&start=0&view=
viewtopicPhpbb.php?f=91&t=7785&p=20709&start=0&view=
viewtopicPhpbb.php?f=91&t=6840&p=17734&start=0&view=

Apparently there are 3 ways to solve the problem of obtaining the GRF from a known motion.
1) Applying the calculus described in the article above. Using the ID tool, obtain the forces applied on the CoM and translate them to the ZMP (CoP).
2) Create a ground contact model of the feet with the ground and use the FD tools.
3) Add a constraint of the foot with the ground to obtain the generated forces. This case is only valid for studies where the foot is not moving (doing squats, evaluating balance, weight lifting, etc).

User avatar
Johannes Amend
Posts: 6
Joined: Fri Oct 20, 2023 12:45 am

Re: GRF Prediction

Post by Johannes Amend » Mon Nov 20, 2023 1:42 am

Hello Adrian,
Thank you for your answer, i will give it a look.

User avatar
Adrián Schmedling
Posts: 13
Joined: Tue Dec 06, 2022 11:36 am

Re: GRF Prediction

Post by Adrián Schmedling » Fri Dec 15, 2023 2:23 am

adriansch wrote:
Fri Nov 17, 2023 2:58 am
3) Add a constraint of the foot with the ground to obtain the generated forces. This case is only valid for studies where the foot is not moving (doing squats, evaluating balance, weight lifting, etc).
It seems this is technically correct, but it's very tricky to deal with in practice. So the first two options are better approaches

POST REPLY