Dear all,I'm now processing on the arm26 model. Now I'm using ID tool to generate the joint torques but the results turn out to be something strange. The angles of arms are shown in the figure below. I expect that the torque should be zero at the initial position, in other words, the time at zero, as the angle is zero. But surprisingly there is a offset torque value in the beginning. So this confused me because there are no any setting about the offset in setup file or others documents to cause this result. What's more, I also use the same approaching model with the same properties to simulate the same trajectory in ansys, it turn out that there isn't any offset. So can anyone can tell me what's going on or what might be the reason causing this?
thanks a lot!
ID result has an offset !
ID result has an offset !
- Attachments
-
- this is the trajectory.
- arm_angle.png (15.83 KiB) Viewed 289 times
-
- this is the output torque calculated by opensim.
- arm_torque.png (25.58 KiB) Viewed 289 times
-
- this is the output torque calculated by ansys.
- arm_torque_ansys.png (34.42 KiB) Viewed 289 times
Last edited by Gary Lu on Mon Oct 28, 2019 11:42 pm, edited 1 time in total.
Tags:
-
Thomas Uchida
- Posts: 1800
- Joined: Wed May 16, 2012 11:40 am
Re: ID result has a offset !
I suggest running a Forward Dynamic simulation with zero muscle excitations (you can do this easily using the Simulation Controls, https://simtk-confluence.stanford.edu/d ... im/Toolbar). The joint angles do not remain at zero. Thus, the joint moments required to maintain zeroed joint angles must be nonzero.
Re: ID result has an offset !
But my purpose is to let the arm follow my trajectory as its initial angle is zero,in order word,just let the arm freely point downward without any torque applied at the beginning, I didn't mean to let the arm start from the position which the muscle has zero muscle excitation. I guess you're saying that when the arm is at its zero-muscle-excitation position, the angle is not zero, so the arm joint torque shouldn't be zero, am I right? Or is there anything wrong about my statements.tkuchida wrote: ↑Mon Oct 28, 2019 9:02 amI suggest running a Forward Dynamic simulation with zero muscle excitations (you can do this easily using the Simulation Controls, https://simtk-confluence.stanford.edu/d ... im/Toolbar). The joint angles do not remain at zero. Thus, the joint moments required to maintain zeroed joint angles must be nonzero.
-
Thomas Uchida
- Posts: 1800
- Joined: Wed May 16, 2012 11:40 am
Re: ID result has an offset !
Correct. Your hypothesis is that the angles are zero when muscle activations are zero. You can test your hypothesis by running a forward dynamic simulation with zero muscle excitations.I guess you're saying that when the arm is at its zero-muscle-excitation position, the angle is not zero