Optimization of the DLR SpaceLiner inside the integration environment RCE

Abstract

In this paper, we would like to present the multidisciplinary analysis and optimization of the preliminary DLR SpaceLiner design concept using the distributed workflow-driven integration environment, called RCE (Remote Component Environment). As the considered disciplines together build a coupled system and a change of input in one simulation tool may affect the output of another one, MultiDisciplinary Optimization (MDO) techniques have to be applied. To be able to efficiently analyze and optimize the overall system of the SpaceLiner, the described software and simulation tools have been integrated as a process chain inside the integration environment RCE. RCE comes with a number of integrated ready-to-use algorithms, but more importantly, it is furthermore possible to integrate third-party and user-provided algorithms. The described optimization problem of minimizing the mass of the DLR SpaceLiner design concept is solved using several publicly available software codes which we compare in terms of quality of the solution and in terms of number of function evaluations. Most of the integrated engineering simulation tools do not provide derivatives of the objective and constraint functions such that only optimization methods which do approximate or do not need derivatives at all can be applied in our case.