Multidisciplinary Design Optimization (MDO) of Transonic Fan Blade

摘要: RESUME La conception de pales d’un rotor est une tâche complexe et difficile en raison l’ecoulement transsonique, du large espace design l’implication plusieurs disciplines l'ingenierie dans le but d’augmenter les performances metriques multidisciplinaire tels que l'efficacite, rapport pression, stress. Pour faire face a tous ces defis, comparaison d’approches pour les optimisations aerodynamiques multidisciplinaires automatises (MDO) des soufflante transsonique presentee. Le processus propose integre methode parametrisation geometrique pales, modelisation CAO outils d’analyse hautefidelite pour l'aerodynamique, la structure dynamique. Une multi-niveau ete utilisee pour modifier efficacement geometrie pale avec un faible nombre variables conception. modele construit CATIA afin d'utiliser commun analyses dynamiques. equations Navier-Stokes (RANS) tridimensionnelles moyennees au logiciel commercial CFD ANSYS CFX, utilise l'analyse aerodynamique tandis qu’un elements finis (EF) implemente sur realiser Des algorithmes d'optimisation heuristiques hybrides sont utilises resoudre probleme forme pales. La verification codes methodes effectuee comparant resultats calcules donnees experimentales disponibles litterature NASA Rotor 67, cas test representatif d'ecoulement complexes trois dimensions. Afin verifier faisabilite automatise l'optimisation, optimisation visant maximiser l'efficacite point tout maintenant debit massique pression constant, elabore execute redessiner 67. En outre, ce aide selectionner l'algorithme adapte resolution explorer l'espace Cependant, la inevitablement pluridisciplinaire qui necessite participation nombreuses telles structure, dynamique, etc., cours differentes etapes procedures actuelles impliquent dynamique apres l’optimisation aerodynamique.----------ABSTRACT The of current transonic fan blades is complex and challenging task due to multifaceted flow field, space involvement many engineering specialists increase performance on multidisciplinary metrics such as efficiency, pressure ratio, To tackle all these challenges, comparison approaches for the automated aerodynamic optimizations developed. The developed process integrates geometrical parameterization method, CAD modeling highfidelity analysis tools aerodynamics, dynamics disciplines. A multi-level method blade was utilized efficiently modify geometry with low number variables. model built in CATIA, use common dynamic analyses. three-dimensional Reynolds-Averaged based commercial software CFX used rotor; whereas Finite Element (FE) Mechanical conduct Heuristic hybrid optimization algorithms are employed solve problem. The capability methodologies validated by comparing computed results experimental data available open literature case representative structures problems. In order verify feasibility integrated working flow, an aiming maximize efficiency while maintaining mass rate formulated executed redesign case. It further helped select suitable algorithm explore space. However, inevitably which requires dynamics, during different stages process. In addition, involved after discipline i.e. sequential optimization. main drawback this procedure that good might not satisfy structural requirements make iterative