Hi,
I try to test an assembly in AnyBody that should be combined with the AnyBody human model.
Before to do it, I tried to test the assembly itself alone.
In the attached file, there is an assembly that contains only two parts.
The mainframe part is fixed to the ground. And the right handle part is connected to the main frame by revolute joint. So this is very simple structure.
But the reference node for the revolute joint on the main frame part is skewed with respect the local reference frame of the main frame.
For kinematic analysis of this assembly, I selected the kinematically over-determinate solver type.
But when I tried to run the kinematic analysis, I got the following errors:
Failed to resolve kinematic constraints. Newton relaxation too small. (final kin. error = 3.331485E-005)
Constraint no. 17 above error tolerance 0.000001, error = 0.000030.
Constraint no. 18 above error tolerance 0.000001, error = 0.000033.
Constraint no. #3 in ‘Main.Model.EnvironmentModel.Joints.Joint8.Constraints’ above error tolerance 0.000001, error = 0.000030.
Constraint no. #4 in ‘Main.Model.EnvironmentModel.Joints.Joint8.Constraints’ above error tolerance 0.000001, error = 0.000033.
18) …Kinematic analysis terminated
If there are some closed-loops or redundant constraints, I may understand that this situation can be happened.
But this assembly is simple structure and it doesn’t have any closed-loop.
When I saw the error message, the two Y-axes of the two reference nodes make the opposite direction.
Of course I could perform the kinematic analysis by loosing the kinematic analysis tolerance.
But I think that this method is a temporary solution for this kind of problems.
So I’d like to know whether the skewed reference nodes for a revolute joint may cause some problems for kinematic analysis.
I’d very appreciate if you can give me your advice.
Best regards,
Moonki