Barrier constraints are not implemented as recommended by GenOpt - need to use multiplicative inverse. Furthermore, too many constraints are included in the barrier. I think that we should consider allowing heat exchange feasibility to be violated during optimization search, which (without switching to iterative solver) could be simply reflected in an infinite total UA value.
Barrier constraints are not implemented as recommended by GenOpt - need to use multiplicative inverse. Furthermore, too many constraints are included in the barrier. I think that we should consider allowing heat exchange feasibility to be violated during optimization search, which (without switching to iterative solver) could be simply reflected in an infinite total UA value.