• Registered: 2010-03-09
  • Posts: 26

Topic: Convergence of result

In terms of convergence of an analysis in ATENA, may I know what is the % accuracy of the convergence?

In ATENA, the convergence is it built in?

Can we know for example, a beam with 1000 elements, what is the value when the result shows convergence?

In other words, how do we know that the convergence is the best results, with the number of elements that we used?



TQ

Quote Post 1

  • dpryl
  • CC software developer
  • Offline
  • Registered: 2007-07-13
  • Posts: 1,107

Re: Convergence of result

Dear Irene,
I am not sure if I understand your question. I assume you have already read the related sections of the ATENA Troubleshooting Manual, sections
2.2.2 I get the message "The execution is killed due to violation of stop iteration criteria", what does it mean?
and
2.2.5 Floating point problems - Division by zero, Loss of >7 digits on the matrix diagonal, Floating point exception: Multiple floating points traps, Zero or negative jacobian, InvertA: zero determinant, CCMaterials Extended AExc: CC5ParamYield F: project_stress_on_f_division by zero, and similar
and
2.2.6 I get the warning "Convergence criteria (within line search) not satisfied due to eta limit, continuing super-ordinate iteration" during analysis, what does it mean?
- you can find detailed explanation about the convergence measures, including the definition formulas of the 4 criteria being used, in the ATENA Theory Manual, section 4.2 Full Newton-Raphson Method.

In general, if there are errors >10% in several steps in a row, the results are most likely just numerical noise. On the other hand, a single step with error about 2% is absolutely no problem in most cases.

dpryl's Website

Quote Post 2