The Cosmos knowledge base had this to say:
"The FFE solver issues this message when it is not converging after 5000 iterations. This means that the problem is hard to converge. Many causes can lead to this situation, including but not limited to: Large difference between Young's modulus of the materials, portions of the model with very low stiffness compared to other portions, etc...
You may want to use a different solver like the FFEPlus solver, or the Sparse solver. The sparse solver being a direct solver, convergence problems are irrelevant. But the required memory and time are much higher."
So there you go.