Issue Type
Description
We're trying to run a structural optimization to minimize mass subject to stress constraints for the nacaWing in order to size the thicknesses of each shell. Several components of the dstress/dthickness gradient are very large while others are small. I am currently using PSQP algorithm from PyOptSparse. The optimization algorithm currently steps way too far in the design space, pushing many of the thicknesses to the minimum thickness. It might be because the optimization begins in the infeasible region due to high stresses. Maybe I should allow a large maximum stress originally.
Example
NacaWing case
Issue Type
Description
We're trying to run a structural optimization to minimize mass subject to stress constraints for the nacaWing in order to size the thicknesses of each shell. Several components of the dstress/dthickness gradient are very large while others are small. I am currently using PSQP algorithm from PyOptSparse. The optimization algorithm currently steps way too far in the design space, pushing many of the thicknesses to the minimum thickness. It might be because the optimization begins in the infeasible region due to high stresses. Maybe I should allow a large maximum stress originally.
Example
NacaWing case