RRA assembler error
Posted: Fri Mar 07, 2025 5:17 am
Hi everyone,
I am currently working on an ACL injury prevention project with a model from Xu et al. (2014) including knee ligaments, with change of direction motions. This version of the model includes trunk and lower body actuators and marketset. I am working on the RRA step but I have been receiving the same error every time I try to run it:
Method Assembler::track() failed because:
Unable to achieve required assembly error tolerance.
Assembly error tolerance achieved: 2.1955073314927631e-09 required: 1e-10.
Model unable to assemble: AssemblySolver::assemble() Failed: SimTK Exception thrown at Assembler.cpp:897:
Method Assembler::assemble() failed because:
Unable to achieve required assembly error tolerance.
Assembly error tolerance achieved: 2.1709696262917078e-09 required: 1e-10..Model relaxing constraints and trying again.
Requested COM adjustment time range 0.23 to 0.42, clamped to nearest available data times: 0.23 to 0.42.
Computing average residuals between 0.23 and 0.42...
Storage: read data file = D:\These\OpenSim\Test_modele_2\4_Dynamique_inverse\test_motion_2_grf.mot (nr=1010 nc=19)
ExternalForce::externalforce Data source being set to GRF
Storage: read data file = D:\These\OpenSim\Test_modele_2\4_Dynamique_inverse\test_motion_2_grf.mot (nr=1010 nc=19)
ExternalForce::externalforce Data source being set to GRF
I started with an assembly error tolerance of 2.1e-08, and already improved it to 2.2e-09 by fixing anatomical markers, and by locking subtalar and MTP ankle DOF for both legs.
I have seen that modifying kp and kv gains could help reduce the error, could it help in this case? I would also appreciate explanations about these kp and kv values, what do they relate to and how do they help reduce the error on a model?
Finally, could the error also be due to the absence of upper body modelling?
Thanks in advance for your help,
Leo-Paul
I am currently working on an ACL injury prevention project with a model from Xu et al. (2014) including knee ligaments, with change of direction motions. This version of the model includes trunk and lower body actuators and marketset. I am working on the RRA step but I have been receiving the same error every time I try to run it:
Method Assembler::track() failed because:
Unable to achieve required assembly error tolerance.
Assembly error tolerance achieved: 2.1955073314927631e-09 required: 1e-10.
Model unable to assemble: AssemblySolver::assemble() Failed: SimTK Exception thrown at Assembler.cpp:897:
Method Assembler::assemble() failed because:
Unable to achieve required assembly error tolerance.
Assembly error tolerance achieved: 2.1709696262917078e-09 required: 1e-10..Model relaxing constraints and trying again.
Requested COM adjustment time range 0.23 to 0.42, clamped to nearest available data times: 0.23 to 0.42.
Computing average residuals between 0.23 and 0.42...
Storage: read data file = D:\These\OpenSim\Test_modele_2\4_Dynamique_inverse\test_motion_2_grf.mot (nr=1010 nc=19)
ExternalForce::externalforce Data source being set to GRF
Storage: read data file = D:\These\OpenSim\Test_modele_2\4_Dynamique_inverse\test_motion_2_grf.mot (nr=1010 nc=19)
ExternalForce::externalforce Data source being set to GRF
I started with an assembly error tolerance of 2.1e-08, and already improved it to 2.2e-09 by fixing anatomical markers, and by locking subtalar and MTP ankle DOF for both legs.
I have seen that modifying kp and kv gains could help reduce the error, could it help in this case? I would also appreciate explanations about these kp and kv values, what do they relate to and how do they help reduce the error on a model?
Finally, could the error also be due to the absence of upper body modelling?
Thanks in advance for your help,
Leo-Paul