Learning Forum

[SULUN SHI]

Hi,

I am doing a nonlinear analysis for the pretensioned concrete beam. The key focus is transfer length. After analyzing the beam, I checked several results. However, some parts were confirmed that were different from the predicted data.

My teacher suggested I achieve a tighter convergence in the analysis.

In addition to increasing the maximum number of iterations, I can use the CNVTOL command to set the convergence value for the nonlinear analysis. I have some questions:

1. Convergence based issue: As far as the structure is concerned, which method has more convergence and better accuracy of the overall strain curve? Force or displacement?
2. Is a lower tolerance value better?
3. Minimum reference value. I don't understand what it is, Does this represent the best value of convergence achievable? 
4. What is the difference between VALUE and minimum reference value?
5. Should the convergence value be as low as possible (e.g. 10^(-10)) or as high as possible (e.g. (1,000,000,000)) in order to achieve higher accuracy of the results?

[Erik Kostson]

Duplicate - use this post:

/forum/forums/topic/nonlinear-convergence-criteria-in-ansys-apdl/