Parallel Numerical Airplane Wing Design

Development and application of multi-disciplinary optimization capabilities based on high-fidelity methods

“…DLR has a long tradition in shape optimization based on high-fidelity methods and mainly two approaches have been followed so far: the development of an in-house frameworks in C or in Fortran with a system call to the analysis module written in UNIX shell script language 39,40 or the use of commercial frameworks involving shell or Python scripts 41,42 . These systems proved their capability in solving aerodynamic design problems, but their structures were either not flexible enough to be able to easily integrate new modules, like the adjoint approach, or with limited networking capabilities sometimes mandatory in MDO context involving complex processes on various operating systems.…”

Aerodynamic Inverse Design Framework Using Discrete Adjoint Method

“…DLR-AS has a long tradition in shape optimization based on CFD and mainly two approaches have been followed so far: the development of frameworks in C or in Fortran with a system call to the analysis module written in shell [1,13] or the use of commercial frameworks involving shell or Python scripts [4]. These systems proved their capability in solving aerodynamic design problems, but the structures were either not flexible enough to be able to easily integrate new modules, like the adjoint approach, or difficult to be deployed on various operating system or limited by the number of simultaneous use due to the license policy and price.…”