Sign In
 

 

 

Aerodynamics and multidisciplinar optimization<img alt="" src="http://webtest.cira.it/PublishingImages/mefl-APMD-FIG1a_1.png" width="434" style="BORDER:0px solid;" />https://www.cira.it/en/competences/fluid-mechanics/__subnav/analisi-e-progettazione-multidisciplinare/Aerodynamics and multidisciplinar optimizationAerodynamics and multidisciplinar optimization<h3>Mission:</h3><p>The group multidisciplinar analysis and design has been created to improve and develop the large know-how produced in the last two decades by participating to research project on optimization and aircraft components design.<br></p><h3>Goals:</h3><p></p><p style="text-align:justify;">The main objective of the group is to develop and use methods for aerodynamics optimization.</p><p style="text-align:justify;">Typical applications are the aerodynamic design of wing configuration in high lift (European projects EUROLIFT I and II, DESIREH), the natural laminar flow aerodynamics design in transonic and supersonic regime (EPISTLE, NACRE, SUPERTRAC, TELFONA, CESAR, JTI-GRA), innovative configurations design (flying wing design in the VELA project, backward swept wing and high aspect ratio wing in the NACRE project), helicopter components optimization (JTI-GRC project), robust optimization in uncertain conditions (UMRIDA project). Starting from the work performed in the AEROSHAPE project several development lines have been introduced in several projects some of which are still active in the group.</p><h3>Research Topics:</h3><p></p><p></p><p style="text-align:justify;">The group basic experience can be so summarized: </p><p></p><ol><li><p>Algorithm: development of numerical optimization algorithm able to efficiently manage multi-objective and multi-point problems (problems with several design conditions) and, eventually strongly constrained problems. This research line is mainly based on genetic algorithm and their hybridization with classical deterministic optimization procedures (such as gradient based) to have a more efficient optimal solution identification. The GAW/ADGLIB, developed within the group, is the main platform. <br></p></li><li><p>Meta-model: development of accurate and variable fidelity model based on reduced order models to accelerate the optimization process. Such technique are of particular interest when the optimization require the calculation of objective function that are particularly expensive from the computational point of view. The use of meta-models allows the use in an effective way of all the available information (even if obtained from low fidelity tools) before using additional computational efforts. In this contest, the use of specific criteria to sample, with an intelligent procedure, the design space has a crucial role and therefore the identification, implementation and development of these criteria is an important research theme in this field. The MANGO code, developed in-house from the group is the main result of the group in this area.<br></p></li><li><p>Robust and reliable design: development of techniques to introduce uncertainties quantification within optimization algorithms. This allow to obtain a more robust (less dependent from stochastic variation of operating conditions and/or design parameters) and more reliable (more far from a probabilistic point of view from malfunction) design. The optimization library GAW/ADGLIB is the main development platform).<br></p></li><li><p>Shapes parametrical representation<span lang="EN-US">: development of advanced techniques for complex geometries (such ha wing fuselage intersection) management and for the reduction of the number of parameters required for the complete geometrical description. To describe complex details It is possible to use multi-level and hierarchic parametrization techniques based on a local approach (such as free-form or mesh deformation) and, at an higher level, global techniques (such as class-shape transformation, NURBS, CAD model) to represent the full configuration. To reduce the number of parameters the “Principal Component Analysis” (PCA) allow to classify the design variables using an effectiveness criteria and, to eliminate the less important parameters. This allow a drastic reduction of the design area in the optimal solution research. </span><br></p></li></ol><p></p><p></p><p><br></p><p><img src="http://webtest.cira.it/PublishingImages/mefl-APMD-FIG2a_1.png" alt="" style="margin:5px;width:814px;" /><br></p><div style="text-align:center;">Multiobjective optimization of a supersonic laminar wing</div><br>

 Activities