Differentiation of the TFS Package

Area: Computational Fluid Dynamics
At Aachen University of Technology, a team of engineers, mathematicians, and computer scientists is investigating the fluidstructure interaction at airplane wings to further advance scientific knowledge of the aerodynamics of cruise and high lift configurations. One of the projects aims at optimizing an airfoil with respect to certain design parameters. Traditionally, finding a suitable set of parameters is carried out by running the simulation code over and over again with perturbed inputs. However, this approach may consume enormous computing time and may also require an experienced user to select suitable sets of parameters just to achieve improvement, even without optimality. Here, numerical optimization techniques can help reduce the number of simulation runs, but in particular provide more goaloriented computational support for a design engineer. When embedding a pure simulation code in an optimization framework, a crucial ingredient to any gradientbased optimization algorithm is the derivative of the output of the simulation with respect to the set of design parameters. Here, the Adifor tool is applied to a computational fluid dynamics code called TFS developed at the Aerodynamics Institute, Aachen University of Technology, to get accurate derivatives of the flow field with respect to certain design parameters.
Related ADTool:
ADIFOR
Reference:
People involved:
Reference:
C. H. Bischof, H. M. Bücker, B. Lang, A. Rasch, E. Slusanschi
Efficient and accurate derivatives for a software process chain in airfoil shape optimization
Article in Future Generation Computer Systems, 2005
H. Martin Bücker, Bruno Lang, A. Rasch, Christian H. Bischof
Computation of Sensitivity Information for Aircraft Design by Automatic Differentiation
Conference proceeding, Computational Science  ICCS 2002, Proceedings of the International Conference on Computational Science, Amsterdam, The Netherlands, April 2124, 2002. Part II, Springer, 2002
H. M. Bücker, B. Lang, A. Rasch, C. H. Bischof
Sensitivity Analysis of an Airfoil Simulation Using Automatic Differentiation
Conference proceeding, Proceedings of the IASTED International Conference on Modelling, Identification, and Control, Innsbruck, Austria, February 1821, 2002, ACTA Press, 2002
H. Martin Bücker, A. Rasch, E. Slusanschi, Christian H. Bischof
Delayed Propagation of Derivatives in a Twodimensional Aircraft Design Optimization Problem
Conference proceeding, Proceedings of the 17th Annual International Symposium on High Performance Computing Systems and Applications and OSCAR Symposium, Sherbrooke, Québec, Canada, May 1114, NRC Research Press, 2003
Efficient and accurate derivatives for a software process chain in airfoil shape optimization
Article in Future Generation Computer Systems, 2005
H. Martin Bücker, Bruno Lang, A. Rasch, Christian H. Bischof
Computation of Sensitivity Information for Aircraft Design by Automatic Differentiation
Conference proceeding, Computational Science  ICCS 2002, Proceedings of the International Conference on Computational Science, Amsterdam, The Netherlands, April 2124, 2002. Part II, Springer, 2002
H. M. Bücker, B. Lang, A. Rasch, C. H. Bischof
Sensitivity Analysis of an Airfoil Simulation Using Automatic Differentiation
Conference proceeding, Proceedings of the IASTED International Conference on Modelling, Identification, and Control, Innsbruck, Austria, February 1821, 2002, ACTA Press, 2002
H. Martin Bücker, A. Rasch, E. Slusanschi, Christian H. Bischof
Delayed Propagation of Derivatives in a Twodimensional Aircraft Design Optimization Problem
Conference proceeding, Proceedings of the 17th Annual International Symposium on High Performance Computing Systems and Applications and OSCAR Symposium, Sherbrooke, Québec, Canada, May 1114, NRC Research Press, 2003
People involved: