question about the use of contact geometry

[Linjie Wang]

Hi!

Is there anyone who can help explain the rule of using contact geometry in the musculoskeletal model?

I applied the meshed femoral, tibial implants and patella (namely obj files) to my musculoskeletal model in order to calculate the contact force between them during gait simulation. I am confused by the use of this element whether these mesh geometries would be used in the analysis of Inverse Kinematics, Residual Reduction Algorithm, Computed Muscle Control. Or only in Forward Dynamics will these mesh geometries be invoked for calculation?

I noticed the force result produced by the Static optimization includes the contact forces and moments which could be affected by the values of contact stiffness but have nothing to do with the external force (e.g. ground reaction force).

Thanks in advance,
Linjie

[Dimitar Stanev]

Hi,

First of all, you would add a contact model to this system if you have more realistic joint model (e.g. the femur-tibia and patella-femoral joint are 6DOFs each). If you have a pin joint for the knee and a prescribed patella-femoral joint you don't gain anything.

These forces are estimated from the kinematics of the bodies and are applied at the dynamic stage. If the algorithm is purely kinematic (e.g. IK) then these forces are not accounted. Thus SO, RRA, CMC and FD compute the contact forces and apply them to the bodies.

Best

[Linjie Wang]

Hi, Dimitar.

Thanks for your reply.

Can I ask one more question about the algorithm of elastic foundation force? Will the contact stiffness that is input in contact parameters change with the increment of time intervals, or remain still for whole calculation? Will the penetration amounts between two contact components (e.g. femur and tibia) be adjusted, or directly used as detected by system (simply contact force = penetration amount times contact stiffness)? Because there is motion definition between femur and tibia (three translation degrees and three rotation degrees), the penetration between two parts is inevitable. I dont know if my understanding is right.

Thanks,
Linjie

[Dimitar Stanev]

Hi,

I am not sure if it is possible to update these parameters during simulation since they may be topologically related (defined during initialization), but you can try.

If you have high values in the stiffness parameter then large forces will be applied due to the high penetration. That is a model's problem if the motion makes the two bodies penetrate, assuming that you have measured this behavior correctly. You would have to adjust the model so that the penetration is normal or small.

Typically contact forces are used in the context of predicting a movement and a force. For example if you don't have GRF measurements, but you have the kinematics. What you will do is to add a contact surface and contacts points on the feet. Unfortunately you will have to adjust the height of the contact surface so that it can be in agreement with the kinematics. It is inevitable that the two will be affected (F <-> ma) e.g. the reaction force is affected by the kinematics (and the position of the surface) and the simulated movement is affected if large forces are applied (e.g. you may not be able to reproduce the measured kinematics). This holds true for the forward dynamics case. If you perform inverse dynamics, then the motion is respected thus the generalized forces must be high enough so that they satisfy the desired motion and the reaction forces due to the contact model. Since the contact model affects the joint troques this is a source of uncertainty in your model and requires further validation.

https://simtk.org/docman/view.php/47/15 ... 2-p241.pdf

Hope this helps

[Linjie Wang]

Hi, Dimitar.

Thank you very much for your kind help.

Linjie