Inverse dynamics unexpected result - am I missing something?

[Erica Montefiori]

Dear OpenSimers,

I am currently struggling with understanding the output of my ID simulations.
I have an MRI-based lower-limb model and I am trying to feed my ID with some extra external forces other than the GRF collected from the force plate.

I define my external loads in the "External loads settings" file (screenshot attached) and I have the actual loads in an mot file, structured accordingly.
This latter file includes 9 columns for the GRF, 9 columns for the loads applied to the talus (internally at the talus/tibia interface), 9 columns for the loads applied to the tibia (this are equal to the loads applied to the talus but with opposite direction as they are intended to represent internal loads).

I would need your feedback on the resulting joint moments reported for the hip, knee and ankle in the sagittal plane (my model has a hinge at both ankle and knee) in the following cases: 1. only GRF applied, 2. GRF + internal torques at the talus/tibia interface, 3. GRF + internal forces at the talus/tibia interface, 4. GRT + both internal forces and torques at the talus/tibia interface (see "ID results" attached). I have simulated the above by modifying the "External loads settings".

When looking at the ID results in the different scenarios, I would expect to see a change in the ankle plantar/dorsi flexion moment but no changes at the knee and hip. This because applying forces and torques that are equal and opposite at talus/tibia should simulate an internal load and therefore these loads should not be accounted for when computing the moments at the knee.

The results don't match my expectations, so I am obviously missing something here.

Would anyone have an idea or comment on this?

Thank you very much

Erica

[Erica Montefiori]

Just to give an update on this, when the internal forces applied to talus and tibia are expressed in ground, everything works as expected, and knee and hip moments are insensitive to the internal forces at the ankle.

This would suggest an error in changing reference system from ground to tibia and talus local systems. However, when visualising the two internal forces (expressed in the talus and tibia local refsys) in the GUI, they are exactly where they should be (see attached), equal and opposite at the talus/tibia interface.

Has anyone experienced any issues with changing reference systems? Any idea of what is happening here?

Thank you very much

[Erica Montefiori]

For those interested in this post, I have apparently solved the problem as follow:
instead of calculating the loading on the tibia independently, I have simply changed sign to the loadings on the talus and applied them to the tibia but now defining them in the talus reference system (see xml file, I have highlighted the change).

The resulting forces are exactly the same (they overlap perfectly when visualised in the GUI) but, having defined everything in the same reference system, changed the results somehow. Now we have the correct moment on the tibia and no effects on knee and hip, as one would expect from internal loads.

Similarly, everything works fine if I have all the loads defined in tibia and applied to tibia and talus bodies accordingly.

Does anyone have a clue why this happened?

Thank you