I’m new to using AnyBody, and I would really appreciate some guidance. I am currently working on a hip joint simulation and planning to replicate the motion in Abaqus. Specifically, I am simulating a case with 70° hip flexion and 20° internal rotation in Abaqus, and I would like to use outputs from AnyBody to drive the model.
At the moment, I am using the standing model in AnyBody, and my goal is to extract: Joint Reaction Forces (JRF), Joint Reaction Moments (JRM), Joint angles, so that I can apply them in Abaqus (I apply the forces at the femoral head center). I have a few questions:
For joint moments, should I use HipNetMoment or HipNetMomentMuscle to avoid over-constraining the FEA model?
If I use HipNetMoment, how can I control or change the output reference frame? It seems that by default, HipNetMoment is expressed in the pelvis frame. Since I apply loads at the femoral head in Abaqus, I am not sure if this is consistent. What is the exact difference between expressing HipNetMoment in the femur frame vs. pelvis frame? Is it only a sign convention difference, or can the magnitudes also change?
Any advice or recommended workflow for transferring AnyBody outputs to Abaqus would be very helpful.
The standing model is probably not the best for a case of 70 degrees hip flexion / 20 internal rotation. But i guess you are just testing the load transfer.
HipNetMoment will contain all loads, not just muscles. The moment should also account for knee reaction forces and so on.
HipNetMoment variable is used for reporting the values, but if you want a more specific output - you can copy this block of code and use a different ref. frame of interest. Just copy the code of HipNetMoment (based on AnyForceMomentMeasure2) and update to your needs. The moments will be computed automatically in the requested ref. frame.
My main goal is to simulate hip motion for impingement cases (FAI) and compare the shear results with a normal hip. Could you please advise which AnyBody model you would recommend for larger joint angles (e.g., 70° flexion, 20° internal rotation)?
I somehow got an impression that you use the model without modification. But I guess you drove the hip flexion to 70 degrees and to 20 degrees of internal rotation? And it does load the hip joint and provides required kinematics. Other options are to start from a seated human or modify the standing model to mimic a stair ascend by positioning leg in needed posture, moving CoM and moving ground reaction force prediction object to the “stair”.
I forgot to answer about Abaqus: it is probably easiest to apply loads as a 6 dof tabular data, since you are only looking for resultant forces/moment and not looking for all loads (muscles, etc.) being applied to the femur.
In fact, I modified the standing model and drove the hip to 70° flexion and 20° internal rotation to extract the outputs.
For the Abaqus model, I am currently prescribing joint angles together with JRF in a tabular form, rather than using full 6-DOF loading, since that approach can be more error-prone and may require a more complex dynamic setup. However, I am not fully sure whether using joint angles + JRF can reliably reproduce results comparable to using JRF + JRM.
At this stage, I am mainly investigating clinical impingement test positions (e.g., FADIR) rather than dynamic activities, and my goal is to compare shear behavior between impingement cases and a normal hip. I am therefore also unsure whether including ground reaction forces is necessary in this type of simulation.
Your FEA approach sounds reasonable. Maybe you could use some some components of the JRM with partial constraints, corresponding to the remaining components. Maybe this will help the stability.
Since your comparison is relative, maybe it indeed does not matter. But I guess a loaded state would result in higher muscle activation and, in turn, higher shear stresses.
How does your unhealthy differ from normal? Different geometries?