Showing results for 
Search instead for 
Did you mean: 

Hard convergence in sol200


I am performing an optimization solution (sol 200) to get an optimal solution for
a model with composite material properties (plies thicknesses and orientation angles
as the variables).
The solution terminates with hard convergence to the best comprise infeasible design.
i understood that the constraint was violated for certain areas that 's why the
run didn't converge to an optimal solution, i did simplify the model and re increase
the number of iteration but the run is still terminated due to hard convergence
to a best compromise infeasible design at a cycle number lower than the maximum
number allowed.
how to control the run to continue til the last number of cycles allowed?
and why it 's always terminating due to hard convergence for an infeasible solution?

Re: Hard convergence in sol200

If the optimizer can't find a feasible design it's going to converge no matter how many iterations you give it. Usually the best approach is to relax the constraints until it can find a feasible design, and then slowly tighten them back down. Alternatively you can construct an objective function out of the constraints, and have it try to minimize that objective function.

Optimization isn't magic, and if it can't find a design that'll meet your constraints it gets stuck.
Paul Blelloch, Ph.D.
Director, Aerospace Analysis
ATA Engineering, Inc.
11995 El Camino Real
San Diego, CA 92130
(858) 480-2065