Optimization

CFD-driven optimization (let alone RANS-based optimization) used to be a distant dream. But with AFT’s streamlined and automated RANS method, performance optimization is finally becoming a practical goal. AFT’s approach is amenable to either conventional objective functions or multi-objective optimization. In the former, RANS is treated as a ‘black box’ to take a design as input and return a given performance criteria as output. Multi-objective applications are handled by building a database of designs to span a given range of variables, complete a RANS simulation on each, and then perform the optimization as a post-processing function.

The time required to perform a given optimization depends on the number of variables or size of the design database. While a one or two variable study can be completed quite quickly, some of our longer optimizations have taken many weeks of computer time. Click on a menu item at left to see examples of how AFT has used optimization to aid customers’ design projects.