Why acceleration values obtained using Body Kinematics are not correct ?

Provide easy-to-use, extensible software for modeling, simulating, controlling, and analyzing the neuromusculoskeletal system.
POST REPLY
User avatar
Muhammad Muaaz
Posts: 6
Joined: Tue Dec 04, 2018 1:47 pm

Why acceleration values obtained using Body Kinematics are not correct ?

Post by Muhammad Muaaz » Thu Jan 17, 2019 3:46 am

Hi.

I am interested in getting position, velocity, and acceleration of bodies in XYZ axis. From documentation, I found I can use body or point kinematics analysis for this purpose. However, when I should use Analysis tool in the Opensim workflow. For example, I load the model, Scale it, and then run Inverse Kinematics on it. Now, is it correct to use Body kinematics or point kinematics on Inverse Kinematics results, or for the sake of argument can I run body kinematics after running RRA.

I have tried running body Kinematics on Inverse Kinematics results, the analysis tool did not complain, but I am not sure if this is correct usage because my results seem bit strange. I get position, velocity, and acceleration data after running body kinematics. When I differentiate position values, I get same velocity as that tool has given me, but when I differentiate velocity values, I do not get similar acceleration values which are given by the analysis tool. I have attached pictures of these plots.
Capture1.PNG
Capture1.PNG (33.83 KiB) Viewed 436 times
Capture2.PNG
Capture2.PNG (33.04 KiB) Viewed 436 times
Therefore, I wanted to ask, what is the correct way of running body/point kinematics ?

Tags:

User avatar
Ryan Byrne
Posts: 28
Joined: Thu Jan 22, 2015 8:33 am

Re: Why acceleration values obtained using Body Kinematics are not correct ?

Post by Ryan Byrne » Tue Jan 22, 2019 1:50 pm

Hi Muhammad,

I have seen this issue brought up a few times and have gone through it myself. Here's a link that should give you some insight. viewtopicPhpbb.php?f=91&t=3221&p=11248&start=0&view=
Basically, it seems that accelerations derived form BodyKinematics are not simply the double derivatives of the body's positions, but instead are calculated using F=ma. So that may be where you are running into issues.

I have observed this to be true by changing bushing forces in my model - while prescribing the same motion - and observing a change in acceleration values.

User avatar
Muhammad Muaaz
Posts: 6
Joined: Tue Dec 04, 2018 1:47 pm

Re: Why acceleration values obtained using Body Kinematics are not correct ?

Post by Muhammad Muaaz » Mon Feb 04, 2019 3:27 am

Hi Ryan,

Thanks for the answer, I have checked and found this is also true when running Point Kinematics as well. I feel OpenSim documentation for Body and Point Kinematics need to be more elaborative like other topics e.g., IK or RRA. Is there any way to request OpenSim core for this?

POST REPLY