Choice of test command

[brag006]

I am facing convergence problems and I developed some code to try out different soultion algorithms and in most cases that works fine. However for a small number of simulations I am still not converging even after decreasing the timestep. In some cases changing the convergence test resulted in convergence. So I am wondering what is the guidance in terms of choosing the most appropriate test command? Or does it work like the solution algorithm, the idea being to try what works.

[fmk]

try what works .. if you have to lower the tolerance, do so for only a few steps and tighten it up again .. you can of course go marching on in the wrong direction so it is also advisable to open up some text file and spit out to it some message about having problems .. when done go back and have a look at what is happening to the solution around those points.

of course if you are a student and have time play around .. what happens if the tolerance is not tight to the solution?, what happens if i just have a linear solution algorithm and accept the unbalnce at each step and hope it ggoes away at the next, what happens to the solution if i have a tight tolerance but play with the input parameters, maybe i don't need a tight tolerance after all and need to spend more time worrying about the model than the analysis.

[mhscott]

You can also try the Newton with initial stiffness on the first iteration. Sometimes, the tangent stiffness on the first iteration can take you to an unrecoverable state, particularly when unloading during cyclic response.