IODA API for gradient-enabled CAD-based optimisation
IODA has progressed substantially with CAD-based optimisation. This was achieved by developing approches for the efficient computation of shape sensitivities, ie the derivatives of surface positions wrt to changes in design variables, as well as constructing gradient-enabled approaches to impose geometric constraints.
A recurring issue is the lack of standards to express the relevant data and to interface gradient-enabled CAD tool, adjoint flow solver and optimiser. The prevalent STEP and IGES file standards do not cover gradient data.
IODA has therefore proposed a unifying API which defines relevant interface calls to raise and exchange the requried data. Adhering to the standard will allow seamless exchange of CAD tools in an optimisation loop. The interface is defined in python and documented here.