Degenerate mass distribution on follower side of a joint !

I am trying to model a reduced order flexible solid by reduced Mass and Stiffness matrices from nx.

I have created the block with right matrices and I am trying to connect its ports to joints in order fix translational DOFs in some of the boundary points (F6, F8, ...) - for now lets assume just F6 or any other port(the issue happens with all boundaries).

when I connect F6 to world frame with spherical joint (fix translational DOFs ) and try to run the simulation or try to do linear analysis, then I get the following error for the spherical joint:

when I replace the joint with a fixed weld joint or with a revolute joint , then I dont get any error; but of course I want to have my model with correct boundary conditions.

[sina sarneizehdoost]

I removed the rigidbody modes and frequencies from the reduced system. the error is still there.

Answer 1

[Martijn]

This is somewhat of a catch-all error in SimMechanics, but there are a couple common factors that can lead to this:

Diagnosis 1

• If you have a Joint block with no Solid block on one or both ends (Base/Follower), this error will be thrown.
• In other words, you cannot connect two Joint blocks in sequence without mass and inertia in between.
• To create multiple degrees of freedom between bodies, you should use more complex joints with more degrees of freedom, if possible.
• If this is not possible, place a small "dummy" mass in between the two Joints. This should really only be used as a last resort measure, though.

Diagnosis 2

• You have two co-ordinate frames that coincide in position, but not in orientation. For example, the z axes in the Base and Follower are facing in opposite directions.
• SimMechanics will try to rotate the entire assembly so that the z axes are aligned, which will sometimes cause this error to arise.
• If this error comes up, cut or comment out the Joint block and check the relative positions and orientations of the Base and Follower coordinate frames in the Mechanics Explorer.

Diagnosis 3

• The Inertia matrix is not positive semi-definite (ill-conditioned), e.g. entering Inf, NaN, or very small values for density, mass, or inertia
• It is possible for SimMechanics to derive all mass values from the geometry for all solid blocks to resolve this error.

If none of these situations appear to apply, please submit an actual support case such that we can assist you further.

[Gianluca Mammone]

It happens in my models if I select a minimum time step too big.

It seems Simscapes returns that error if the solution didn't converge and/or the time step is too big to reach it. I solved by selecting a variable step and setting a minimum time step as auto, although simulation performance deteriorate accordingly

[snegaa js]

actual support

[Martijn]

To submit an actual support case, please go to:

https://www.mathworks.com/support/servicerequests/index.html