Innovative optimization and design techniques for modern aircraft (manned or UAV/UCAV) and engine systems aiming at maximum performance in a multidisciplinary context (aerodynamic efficiency, safety, drag, losses, weight, strength, heat fluxes, emission, noise, …), are now rapidly moving from research labs to industrial real and virtual platforms. To reach concurrently this level of excellence, emergent optimization methodologies require more and more robust and efficient associated software for a daily use in industrial collaborative design environments.

This course intends to provide the basic concepts and tools behind this technology, both in single discipline (single point or multi point design) and multidisciplinary (fluid-structure interaction, fluid- acoustics, conjugate heat transfer, …) context. Subjects which will be treated in detail include: gradient based and steepest descent methods, adjoint methods, one shot or goal oriented methods, evolutionary/differential evolution algorithms on parallel environments, game strategies like Pareto Fronts and Nash Equilibrium, parameterization, surrogate and reduced-order modeling (Radial Basis functions, Artificial Neural Networks, Kriging …), multifidelity modeling approaches, robust design, …

The content of this Course is oriented towards junior and experienced engineers and researchers involved in the field of multi disciplinary design and looking for innovative numerical solutions - or set of solutions- for complex multi criteria optimization problems.

The Lecture Series directors are Prof. Jacques Périaux, from CIMNE/Univ. Politecnica de Catalunya and Prof. Tom Verstraete from the von Karman Institute.

Monday 20 April 2020

8:30 Welcome

9:00 Introduction to optimization and multidisciplinary design I

10:30 Coffee break

11:00 Introduction to optimization and multidisciplinary design II

12:30 Lunch

14:00 Aerodynamic shape optimization of turbomachinery components - an industrial perspective I

15:15 Coffee break

15:45 Aerodynamic shape optimization of turbomachinery components - an industrial perspective I

17:00 Reception

Tuesday 21 April 2020

09:00 Theoretical background for aerodynamic shape optimization

10:30  Coffee break

11:00  Industrial applications of aerodynamic shape optimization

12:30 Lunch

14:00 Efficient sensitivity computation using automatic differentiation

15:15 Coffee break

15:45 CAD-based parametrisations for adjoint optimisation

Wednesday 22 April 2020

09:00  Hybridized  evolutionary optimization with game strategies for design. Applications to multi objective design in aeronautics

10:30   Coffee break

11:00    The continuous adjoint method in aerodynamic optimization, with applications

12:30  Lunch

14:00   Adjoint approaches in aerodynamic shape optimization and  MDO context, Part I

15:15    Coffee break

15:45    Adjoint approaches in aerodynamic shape optimization and  MDO context, Part II

Thursday 23 April 2020

09:00 CAD-based multidisciplinary optimization of turbomachinery components by gradient-based methods

10:30 Coffee break

11:00  Low-cost evolutionary optimization, with engineering applications

12:30  Lunch

13:30  Visit of the von Karman Institute laboratories

14:00 Differentiable shape optimization

15:15 Coffee break

15:45    CAD-based multidisciplinary optimization of turbomachinery components by gradient-free methods

Friday 24 April 2020

09:00 Adaptive surrogate modelling for global optimization in aerodynamic design I

10:30 Coffee break

11:00 Adaptive surrogate modelling for global optimization in aerodynamic design II

12:30 Lunch

14:00  Aircraft multidisciplinary design via numerical optimization, part 1

15:15 Coffee break

15:45 Aircraft multidisciplinary design via numerical optimization, part 2

17:00   End of the lecture series