Direction des Relations et Internationales (DRI)

Programme INRIA "Equipes Associées" 

I. DEFINITION

La proposition en bref

Présentation de l'Équipe Associée

(environ 2 pages)

1. Présentation du coordinateur étranger 

Employment

Since March 2009

Adjunct Instructor at Oregon Institute of Technology, Mathematics Department

Since September 2001

Applied Mathematician, Computational Sciences, The Boeing Company, Seattle WA. “Associate Technical Fellow” since Nov. 2008.

September 1999 – August 2001

Assistant Professor, Mathematics Department, Ohio University, Athens, OH.

October 1994 - August 1999

Staff Scientist, ICASE, NASA Langley Research Center, Hampton, VA

Education

1983-1986 B.Sc. in Physics and Mathematics, The Hebrew University of Jerusalem, Israel.

1987-1989 M.Sc. in Theoretical Solid State Physics, Tel-Aviv University, Israel. 

1990-1994 Ph.D. in Applied Mathematics at the Weizmann Institute of Science, Rehovoth, Israel. 

Technical Interest at the Boeing Company

·      Aerodynamic Optimization, computational aspects, adjoint and Hessian approximation.

·      Unsteady CFD and its application to flutter analysis, far-field boundary conditions and downstream Greens function.

·      Dynamic and static aeroelasticity.

Awards

·      2006 Integrated Defense Employee Involvement award for elimination of flutter wind tunnel test.

·      2008 Boeing Commercial Airplanes (BCA) Pride@Boeing exceptional performance award (details proprietary).

·      Awarded “Associate Technical Fellow” on Nov. 2008. 

Author of 60 journal papers and conference communications. 

2. Historique de la collaboration

• 2.1. entre les équipes :  Eyal Arian a été invité pendant un mois à l'INRIA Bordeaux-Sud Ouest en 2009. Il a séjourné à Bordeaux deux semaines en 2008, une semaine en 2007 et deux semaines en 2005. Angelo Iollo éa été invité à Boeing deux semaines en 2009, deux semaines en 2007, deux semaines en 2006. Il ya déjà plusieurs publications communes ( Journal of Computational Physics en 2009, un article invité pour CFD Review 2008 (Ed. Mohammed Hafez), un article paru en Lecture Note in Physics (Proceeding of the ICCFD 2008)).  •            

• 2.2. entre l'INRIA et l'organisme partenaire : A notre connaissance, il n'y a jamais eu de relations avec Boeing. Eyal Arian a donné un séminaire en 1997 à l'INRIA Sophia Antipolis.

3. Impact : 

• 3.1. Sur la collaboration existante. Il s'agit pour nous de travailler sur des problèmes réel pour faire en sorte que nos méthodes puissent passer à l'échelle sur des problèmes issus de l'aéronautique. Comme indiqué dans le texte ci-dessus, nos compétences sont complémentaires. Il s'agit de coupler les modèles d'intéraction fluide-corps rigide et fluide-structure développés à MC2 avec les configurations typiques d'un avionneur. Un point important sur lequel nous désirons insister est que Boeing est prêt à investir dans cette collaboration dès la première année en dégageant du temps de travail des ingénieurs sur cette collaboration. 
• 3.2. Sur la collaboration avec d'autres projets INRIA : 
• 3.3. Sur la collaboration avec d'autres équipes de l'organisme étranger partenaire. Philippe Spallart, un expert mondial de modélisation de la turbulence, Senior Technical Fellow à dans l'équipe de CFD de Boeing, est intéressé à participer à ce projet. Son expertise peut nous apporter un atout pour mettre un place un modèle de turbulence basé sur le level-set et un maillage non conforme au bord.
• 4. Divers : Voici une liste des personnels de Boeing qui participeront à la collaboration:

Eyal Arian, David Young (Technical Fellow, expert écoulements transoniques), Philippe Spalart (Senior Technical Fellow, expert en modélisation de la turbulence). Pour l'INRIA le personnel potentiellement concerné est : Angelo Iollo, Heloise Beaugendre, Michel Bergmann, Lisl Weynans pour MC2 

4. Pourquoi Boeing Research & Technology : 

Cette collaboration peut nous apporter l'interaction avec une équipe de 50 mathématiciens appliquées et de près de 100 ingénieurs qui développent les outils de simulation pour les fluides et le structures d'un grand avionneur. Au sein de cette équipe on trouve un patrimoine unique de connaissance en aérodynamique appliquée et en calcul des structures . Les grands défis qui sont posés par cette communauté sont la simulation d'un avion en manœuvre, l'optimisation topologique, le vol en formation, une maîtrise complète du flutter.  Notre but ce n'est pas de développer des codes industriels pour eux, mais de comprendre les points durs de ces défis et d'apporter des réponses sur la base des outils que nous développons. Dans le futur,  cette interaction nous permettra de valoriser nos résultats dans un contexte ouvert à l'innovation.

II. PREVISIONS 2010-2011

Ci-dessous le programme établi avec notre partenaire. 

Introduction

The main field of interest is classical: fluid dynamics. However we aim at solving original problems rising in aeronautical context by an innovative paradigm that is based on cartesian grids, penalisation and level sets. The use of cartesian grids contours the meshing issue in complex geometries and moreover allows extensions to higher order accuracy in a natural and simple way. Penalisation is an efficient alternative to explicitly imposing boundary conditions so that body fitted meshes can be avoided, making fluid structure interactions, multi fluid/multi physics flows and iterative shape design procedures easy to set up and simulate. Level sets describe the geometry in a non-parametric way so that geometrical and topological changes due to physics or optimisation are straight forward to follow. It is our objective to solve realistic problems, and therefore the simulation codes will be conceived in order to take advantage of massive parallelisation. This scope can be pursued more easily thanks to cartesian grids and penalisation.

Programme de travail

The state of the collaboration is the following. There is one open collaboration concerning shape optimization. Simulation on Cartesian meshes and level-sets in compressible flows: this activity is of interest to Boeing. Another possible interaction is turbulence modelling.

We will focus on the following areas:

1.   Efficient methods for large movement time-domain CFD simulations.

INRIA has developed innovative techniques to simulate efficiently flow involving large movements, such as flow over moving blades in a turbine, and three fish swimming where it is demonstrated that the fish behind is advancing faster (related with flight formation). We are interested in exploring these ideas in the context of high-Reynolds number compressible flow.  At first these ideas will be demonstrated using codes developed by INRIA.  Our goal will be to work towards the following demonstrations:

An oscillating airfoil (flutter applications)

An airfoil approaching a solid surface (landing)

The expected result of this activity is to adapt the codes available at MC2 to solve compressible flow around obstacles of aeronautical interest. A criterion of success of this activity would be the positive comparison of our code to codes existing in the literature.  The comparison criteria would be the accuracy of the results, the fastness of the computations allowed by the code, the ability to take into account complex geometries.

2.   Topological shape optimization methods involving fluid-dynamics equations INRIA has demonstrated a capability that we do not currently have in topological optimization where one starts for example with two circles which evolve automatically into a single shape that is aerodynamically optimal in some sense.  Such a capability could allow automatic shape design of new concepts using CFD, where the general shape of the planform or engines placement for example is not known apriori.  Also here, at first these ideas will be demonstrated using codes developed by INRIA.  Here we will work towards an automatic emergence of an airfoil starting from some arbitrary shape.

The main expected result for shape optimization is to demonstrate the potential benefits coming from using a numerical method that couples cartesian mesh, level set and penalty methods. Indeed, the level set and penalization methods allow keeping the same mesh during the whole optimization process, avoiding a very costly remeshing. Moreover, a massive parallelization could be simply implemented thanks to the Cartesian mesh and the PETSc tools. Usually, only laminar flows around bluff bodies are considered. All the numerical tools cited above would allow us to optimize more complex flow configurations, such 3D turbulent flows, which are in fine a major stake for industrials. Our goal is thus to highlight the potential economical and ecological benefits that can be achieved using both shape optimization and active flow control (since, for instance, drag reduction leads to a fuel consumption reduction).

Budget et Plan pour 2010

We will focus on following key issue: in order to treat boundary layers with a cartesian grid it will be necessary to resort either to massive parallelization or to multi-block grids. This requires additional development in our codes. In conjunction with our collaborators in Boeing we will device the solution that is best adapted to industrial flow configurations. Yannick Gorsse and Jessica Hovnanian, two INRIA PhD students will be involved in this project and will be visiting Boeing for a month each. Angelo Iollo, Michel Bergmann and Heloise Beaugendre will also visit Boeing at these occasions. Boeing personnel will visit INRIA to possibly develop additional collaborations in the area of turbulence modelling with Heloise Beaugendre and Michel Bergmann (both MC2 researchers). We plan also to organize a workshop around the themes of the collaboration.  

1. Co-financement  Boeing (sujet à être definitivement approuvé)  et estimations de dépense pour l'INRIA

Here is the plan of how to spend the total budget composed of both Boeing and INRIA contributions:

1.   Four weeks of budget to cover time charged by Boeing employees:  $30,000. Budget BOEING 

2.   Two visits of students from INRIA to Boeing, each of a month long, for a total of $10,000. (topological shape optimization). Budget INRIA

3.   Two months contract with a software developer for $10,000.  (Demonstrate time-accurate simulation of an oscillating airfoil), Budget BOEING.

4.   Two visits of a Boeing employee to INRIA:  $8,500 travel expanses (coach). Budget BOEING.

5.   Two Visits of INRIA scientists to Boeing: $8,500 travel expanses. Budget INRIA.

6.   A one day workshop at INRIA $10,000. Budget INRIA

Total: $72,000

INRIA:  2+5+6   20,000 € ~ $28,500

Boeing:   1+3+4                  $48,500

© INRIA - mise à jour le 28/08/2007