Automated Design Optimisation
A turbomachinery blade is automatically parameterised with 20-30 variables. Over the course of design iterations, the numerical optimiser refines the parameter set. The simultaneous testing of numerous test parameter sets is driven by a Tabu search algorithm which offers several advantages:
- Global convergence
- Climbs out of local optima
- Parallel execution
- Adapts to available resources
- Progressive fidelity
- Robust to error
| TurboOptimiser |
|
|
| Geometry Model |
From the rest parameter sets, the test geometries are derived and a structured mesh created
|
|
 |
| Flow Solver |
Each test geometry is run simulataneously through your choice of commercial or in-house flow solver
|
|
 |
|
 |
| Numerical Optimiser |
An optimisation engine examines the flow metrics from the interim solutions and derives new sets of test parameters
|
|
 |
| Flow Metrics |
Flow solutions are examined and the flow metrics relevant to the objective function are derived. Any figure of merit may be chosen
|
|
|
| Candidate Design |
The candidate design can be supplied to TurboOptimiser as an STL file, a structured or unstructured mesh file, or even a point cloud from an X-Ray scan
|
|
|
| Optimised Design |
The optimised design emerges, ready for rig testing or examination in the client's validated CFD process
|
|
