User login

Navigation

You are here

Convergence problems in Abaqus-Fortran

Hi mechanicians:

I decided to post because I have an inconvenience that is not shown in other posts.

I am working in the Abaqus - Fortran environment in a thermal analysis. After implementing a new procedure (which actually works), the .sta file shows me that it has problems in its unique step, more or less in the 60th iteration (it is supposed to get until 500). Twenty iterations after that gets into interruptions because the code cannot converge.

With these lines it is easier to understand my point:

1       83      1        0        5        5     0.735         0.735         0.02705   
1       84      1        0        5        5     0.762         0.762         0.02705   
1       85      1U     0        6        6     0.762         0.762         0.02705   
1       85      2        0        9        9     0.769         0.769         0.006761  
1       86      1        0       10       10    0.776         0.776         0.006761   

The incrementation goes smaller every time and I would need years until the point I want to get.

I have changed the equations in Fortran to simplify them, change the variables in the .inp file such as: *HEAT TRANSFER, *CONTROLS, PARAMETERS=FIELD, FIELD=TEMPERATURE, *CONTROLS, PARAMETERS=LINE SEARCH, *SOLUTION TECHNIQUE,TYPE=QUASI-NEWTON,REFORM KERNEL=20 but it does not converge properly after this point. Checking the xy data in the odb file I realized the code is giving results but the processing is not working.

Any suggestion?

Subscribe to Comments for "Convergence problems in Abaqus-Fortran"

More comments

Syndicate

Subscribe to Syndicate