Problems with running finer meshes and higher orders

[ciaradunlop97]

Hello,

When I am running topology optimisation at finer meshes and higher orders I am not able to run many iterations. It doesn't create an error in the terminal as such, it just says the current mass is 'nan' (which I think means 'not a number'). When viewing in paraview no results are available for these iterations and the model is grey. I have attached some images of the screen in these situations.

Is this normal? Or can the my laptop/usb not handle the higher computational expense as a result of using finer meshes and higher orders?

Many thanks,

Ciara

[Karol Lewandowski]

Can you attach your journal file, mesh file and command line?

[ciaradunlop97]

Hi,

I dont have journal files or command lines saved but I have attached my cubit files for my three mesh sizes. For the medium mesh size I cannot get past the 1st iteration with 2nd order, using the following input to the terminal:

./topology_op -my_file Cube_Cantilever_16.cub -ksp_type gmres -pc_type lu -pc_factor_mat_solver_package mumps -ksp_monitor -my_order 2 -ksp_atol 1e-10 -ksp_rtol 1e-10 -optimize 1 -nb_iter 3 -penalty 3 -young 1e3 -poisson 0.1 -move 0.2 -volume_frac 0.35 -filter_r 0.05

Ciara