Conceptual Hip Exoskeleton Design

Main Forum
1

Hello,

I would like to characterize the effect of applying an assistive torque at the hip joint during walking with a crutch device, using the musculoskeletal parameters. I have two musculoskeletal models developed based on mocap data: 1. Walking without a crutch (normal walking), and 2. Walking with a crutch at a matched gait speed. Now, my goal is to apply an assistive torque at the hip joint for the 2nd model (Walking with a crutch), and compare between the conditions to characterize the effect of an assistive torque. Finally, I would like to determine the optimum amount of torque needed to walk easier. I have a few questions when I am applying this torque. I would appreciate your feedback on this.

  1. I have gone through the AnyBody webinar on "Simulation-Driven Conceptual Design of Exoskeletons" from March 28, 2022. It shows how to apply the AnyReac Force and define the ideal amount of torque needed to obtain the optimum effect. Finally, 'reaction moment' (knee) is compared against 'knee angle' to determine the equation. Finally, this equation is used to apply the force directly to the model and simulate again (This force equation is similar to a torsional spring, and so we finally applied a spring force here. Please let me know if I understood the process correctly). I have been able to locate the knee reaction moment from model, but I don't see the hip reaction moment here (please see the screenshot attached below). Since, my goal is to do this for a conceptual hip exoskeleton, how can I add/find the hip reaction moment from the model? I would highly appreciate your feedback on how I can optimize the assistive torque for a simulation-driven conceptual hip exoskeleton.

Screenshot 2024-12-03 074152
Screenshot 2024-12-03 074152446×574 63.3 KB

  1. Since I am applying the torque to the model based mocap data for walking with the crutch, how do you see the effect of this torque when we compare between conditions? Since the mocap data will be the same, is this an ideal approach to characterize the effect of hip torque based on musculoskeletal parameters, e.g. joint angles from the inverse dynamics analysis? Whatever change we notice in the musculoskeletal parameters for this model, can we consider this as an effect of assistive torque only?

  2. I would like to confirm the units of assistance that we provide in the "AnyReac" or "Spring" force approach in the "Simulation-Driven Conceptual Design of Exoskeletons". Apparently, we are applying an assistive force here, so the units would be in Newtons. Please correct me if I am wrong.

I would highly appreciate your feedback. Please let me know if you have any questions. Thank you very much in advance.

Best Regards,
Arif

2

Hi Arif,

  1. It is a little bit hard for us to answer questions related to webcasts. Fortunately, this is one that I remember somewhat. I think your understanding of reaction moment is slightly incorrect and that's understandable because sometimes the nomenclature is confusing. What Prof. Rasmussen referred to in his webcast is the knee reaction moment that he applied using AnyReacForce, which basically determines the moment needed at the knee. This AnyReacForce must be added to your model to see how much knee moment is needed. The screenshot that you have attached is already existing in AMMR and collects some important measures for moments.
  2. I think, you are right. There is however a limitation that the presence of an assistive torque will probably affect how the motion will be performed in reality. This is not really possible to simulate in an inverse dynamics simulation. Given this limitation, you are right in assuming that the effect you will see will be purely due to the effect of the assistive torque. However, if you record the motion again with the assistive device, you will probably see a difference in motion that can also have an effect on the outputs.
  3. The units of inputs/outputs in AnyBody classes are SI units. So, Newtons for force and Newton-meter for moment or torque. However, you should pay attention if the user has sometimes created new variables with converted units. For example, in the AMMR human model, we use the mannequin.any file with angles in degrees for convenience. These are converted to radians before they are passed as inputs for kinematic drivers.

If you have further questions regarding a webcast, it's perhaps best to reach out to the presenter of the webcast.

Best regards,
Dave