3 ms·
Reviewing Ulstein’s marketing material got me to naive wondering - what’s the state of generative design in this field? Why not just throw all the parameters of
by erikig 4y ago
Reviewing Ulstein’s marketing material got me to naive wondering - what’s the state of generative design in this field? Why not just throw all the parameters of the problem (expected vessel usage, capacity, range, geography etc) into a model and have the system generate the most optimal designs?
- rkangel 4y agoIn order to do that, the 'test' for each iteration needs to be relatively quick/cheap. To see how good a bow design is you need to do some computationally expensive CFD, making iteration incredibly computationally expensive. I've read some papers about generative design for wind turbines and antennas and in most cases the special sauce is an approximation of the simulation that can be run more quickly.