asdeasoft.net

Hello to everyone, I have some problems with my model, it is a gerber joint of a prestressed reinforced concrete bridge. I'm doing the analysis on a computer that has 24 logical processors, I'm therefore using an analysis on multiprocessors with openseesmp (24 partitions), however, the model has a huge number of elements, and I cannot reach an accurate level of discretization. At the moment I choose a global mesh of 10 cm (and I get some results), but I want to use a finer one. I tried to use a global finer mesh but the analysis failed, so I tried to use a finer mesh only at the junction area, but I can't get any results. Do you have any suggestions? Do you think there's something wrong with my model?

Hi mari,

Sorry we haven't answered you yet. We haven't had time to had a look at your model. What I can tell you is that the optimum combination between a number of elements and the number of processors can be reached when you assign around 7000 elements to each processor.

Did you make any further attempts?

Francesca