Objectives will be measured, verified and monitored using specific Success Criteria metrics during the project:
- End of Phase 1: developments validated
- End of Phase 2: developments implemented in MDO processes
Success Criteria metrics will be organised around three main pillars:
Specific objectives and Success Criteria
Increase the CAPABILITY of MDO with HiFi simulations to reinforce its impact on current and future design objectives:
- Adjoint of multi-physics equations and solvers:
SC1: Multi-physics adjoint sensitivities verified (maximum relative error of 5% in the norm of the gradient)
- Manufacturability oriented design to faster integrate new materials or new manufacturing processes (UQ techniques and direct modelling of the manufacturing process)
SC2: Impact of manufacturing criterion in the MDO process on performance.
- Efficient exploration of large design space
SC3: Removal of constraints on design space exploration (by using topology optimisation while assuring manufacturing constraints and by using mesh deformation schemes allowing large deformations).
Improve the EFFICIENCY of the MDO process based on adjoint solvers to obtain results in reduced time
- Fast and robust adjoint-based MDO capability for large industrial test cases and time dependent CFD on next generation HPC infrastructure
SC4: MDO adjoint solvers as robust and fast as direct disciplinary solvers even in the presence of flow separation on heterogeneous HPC systems, mesh morphing less than 10% of primal solver in terms of computational time.
- Industry-compatible development time
SC5: Reduction of MDO development time for industrial deployment by a factor of 10.
Enhance USABILITY of key technologies that will reduce the time and complexity for the user or designer to setup a complete optimisation problem:
- Physics-based parameterisations and graphical environment to facilitate the definition of geometric constraints (link with CAD (Computed-Aided Design) or FEM (Finite Element Model) for instance)
SC6: MDO parameterisations defined with and approved by industrial aircraft and engine designers.
- Appropriate end-user formulations for MDO problem settings exploiting shape gradients and adapted to aircraft and engine design
SC7: Comparison between results obtained through coupled adjoint-based MDO and sequential/uncoupled single objective optimisations on various industrial test cases and definition of best practices and guidelines.