Truss modeling - Link8

[jlankes]

I am currently working on integrating a suspension onto a chassis for torsional analysis. My chassis is modeled using Beam188 elements, and I am using Link8 elements to model the A-arms, without having to model the spherical bearings at each end. Each time I run the model, I get a small equation solver pivot term encountered error, and the truss elements appear to be severely buckled. Even when all the links but 1 are switched to Beam188 elements, the remaining truss element still kicks an error, and severely buckles.

I have seen that generally this error is due to an unconstrained model, however, I have run model with the frame constrained in all DOF at 4 points near the suspension to prevent movement, and all the links modeled as beams, and I still have the problem. I am not sure if there is possible a compatibility issue between Beam188 and Link8, or what the problem is.

 

[Drej]

Your problem is probably due to a degree of freedom incompatibility between the link (3 DOF - translational) and the beam (6 DOF). Either consider tying the DOF together between the link/beam (using the CE command) or consider using a beam instead of the link with the end moments released.




------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[lankesjo]

Just for reference the problem was that the lines that the LINK8 elements were applied were meshing as a series of small lines joined end to end, with 3 DOF between each small link. The solution was to clear the line segment, then use the lesi command to make the line segment all one. After that the model ran properly while still allowing the 3 DOF of the link to simulate a spherical joint.