Issue with ID Results

[Josée Mallah]

Hello,
I am using the dataset of C. Fukuchi et al. with the gait2392 model, which I modified by removing the torso and related muscles as I only have makers on the lower body. I scaled the model with the remaining mass, and ran inverse kinematics - I got very low RMS and max marker errors as virtual and experimental markers were quite close. Then, I ran inverse dynamics and visualised the GRF, which looks about right.
Looking into the data, the IK hip and knee angle seem fine, and while the ankle angle has an OK trend, the values seem to have a positive offset (compared with plots of C. Fukuchi). The main issue however is with the ID joint moments, which are all the same and inaccurate.
I am attaching a spreadsheet of the data+plots, as well as the ID setup and external forces file (which mention a subject01.osim, while I am using a scaled subject29, but I don't think this is the issue though).
Any thoughts on what is going wrong?
Thank you