Relationship contact geometry and computeStateVariableDerivatives

Provide easy-to-use, extensible software for modeling, simulating, controlling, and analyzing the neuromusculoskeletal system.
POST REPLY
User avatar
Lars Boers
Posts: 14
Joined: Thu Sep 06, 2018 2:23 am

Relationship contact geometry and computeStateVariableDerivatives

Post by Lars Boers » Thu Sep 05, 2019 2:31 am

Hi Guys,

Would adding contact geometry such as HuntCrossley Spheres influence the resulting computations of the computeStateVariableDerivatives function? Would think it does because it is invoked from realizeAcceleration() right?

Regards,

Lars

Tags:

User avatar
Nicos Haralabidis
Posts: 196
Joined: Tue Aug 16, 2016 1:46 am

Re: Relationship contact geometry and computeStateVariableDerivatives

Post by Nicos Haralabidis » Thu Sep 05, 2019 6:45 am

Hi Lars,

From my own experiences with computeStateVariableDerivatives and the H-C contact model, the resulting accelerations take into account the contact forces. A simple test to convince yourself of this behaviour would be to prescribe some states and controls (when the model is near the ground); include the contact model spheres and call computeStateVariableDerivatives; you could then do the same again but exclude the contact model spheres.

Kind regards,

Nicos Haralabidis

User avatar
Ayman Habib
Posts: 2248
Joined: Fri Apr 01, 2005 12:24 pm

Re: Relationship contact geometry and computeStateVariableDerivatives

Post by Ayman Habib » Fri Sep 06, 2019 11:39 am

Hello,

The addition of contact geometry by itself does nothing unless you also add a contact force that uses this contact geometry. If you do then state realization, force computation will take this contact force into account.

Hope this helps,
Ayman

User avatar
Lars Boers
Posts: 14
Joined: Thu Sep 06, 2018 2:23 am

Re: Relationship contact geometry and computeStateVariableDerivatives

Post by Lars Boers » Mon Sep 09, 2019 1:28 am

Hello all,

Thanks for the replies! Would it be possible to add contact geometries purely for recording the 'would be' forces, without actually adding said contact forces to the system?

Regards,

Lars Boers

User avatar
Ayman Habib
Posts: 2248
Joined: Fri Apr 01, 2005 12:24 pm

Re: Relationship contact geometry and computeStateVariableDerivatives

Post by Ayman Habib » Mon Sep 09, 2019 2:29 pm

Hello,

I don't think you can compute "Would-be" forces without an actual force element mainly because the state need to be integrated over time. You can add forces, perform analyses and then disable them when you don't want the contact forces applied.

Hope this helps,
-Ayman

POST REPLY