| Applies To | |||
| Product(s): | STAAD.Pro | ||
| Version(s): | All | ||
| Environment: | N/A | ||
| Area: | XSTAAD.Pro PDELTA Analysis | ||
| Subarea: | How does STAAD.Pro check for convergence in a Pdelta Analysis | ||
| Original Author: | Bentley Technical Support Group | ||
How does STAAD.Pro check for convergence in a Pdelta Analysis ?
The convergence is checked by comparing the displacements along each dof between successive analysis iterations. If the change in the displacements are lesser than a certain tolerance value, the solution is considered as converged. In general the default tolerance used should be good enough. If one ever needs to change the tolerance value, there is a SET PDELTA TOLERANCE xx command that can be inserted before the JOINT COORDINATES where xx denotes the tolerance in current units of length. The default value for this tolerance is 0.01 inch.
There is also an alternative type of tolerance check which computes the change in Euclidean norm of displacements between two consecutive iterations and compares that to a preset tolerance value. This tolerance can be set using the SET DISPLACEMENT xx command which again has to be placed before the JOINT COORDINATES command. The default value for this tolerance is maximum structure dimension divided by 120. Generally speaking, it is very difficult to set an appropriate value for that sort of a tolerance and so it is more convenient to go with the first option.
More details on these set commands can be found at section 5.5 of the technical reference titled Set Command Specification. In addition, the section 5.37.2 of the technical reference manual titled P-Delta Analysis Options contains more details on the convergence methods discussed above.