Hello,
when the dreaded "Soil body seems to collapse" error comes up, at the moment the only choice one has is to go through the Output of the failed phase and inspect deformation, stress, strain plots, etc.
Sometimes (quite often in my experience) what is reported as "collapse" has got nothing to do with an actual failure mechanism, but it is rather caused by failure of the software to converge in a particular position in the mesh. It could be a single point "failing" thus stopping the software from progressing further. The user has no other choice than spend an awful lot of time trying to figure out where convergence is failing.
I know of other pieces of software that can highlight where in the mesh (nodes, or elements) convergence fails by tracing a rectangle around the affected area. The software should be able to identify where this failure occurs. Pointing the user to the location where this happens would be very, very useful and to concentrate efforts directly where they are needed.
I have been using PLAXIS for a while now and I have always been hoping this fundamental feature would be included in one of the updates but so far, no luck. I'm sure someone must have pointed this out before.
Are the developers considering the possibility of adding this feature? If not, why?
Dear Francesco,
I understand your problem and I totally agree that sometimes you need to spend a lot of time (especially with a 3D model) investigating what is going on.
In general, we recommend inspecting the following plots to determining what is the cause of this error:
A combination of these plots will usually give insight into what is happening.
Apart from that, we are working on different solutions on providing feedback on the progress of the calculation so that a user can find it easier to pinpoint why a convergence issue occurs.
Stay updated with PLAXIS CONNECT Edition as many new features and improvements are added with every version.
Thanks Stefanos for this reply!
Any updates on those new features?
Thanks
Hello Omar,
Indeed, you can check this article: Calculation Kernel Convergence log
Thanks Stefanos!