Supercomputing Challenges for Research in Aeronautics

Chair:

Prof. Dr. Isabelle Terrasse, Dean of College of Experts, EADS Innovation Works, France

The aeronautics industry enters in a new era with new products like Airbus A380, Boeing 787 and upcoming Airbus A350. Air transport demand is predicted to triple in 20 years time. This offers a significant opportunity for the aerospace industry, but also major challenges need to be solved with regard to environmental effects, safety, security and affordability. For example, the target set by ACARE for the greening of air transport means developing technologies to reduce the environmental impact of aviation with the aim of halving the amount of carbon dioxide (CO2) emitted by air transport, cutting specific emissions of nitrogen oxides (NOx) by 80% and halving perceived noise. To adapt to such a fast changing environment, aircraft manufacturers have to shorten the time to market cycles, thereby keeping development costs under control.

Supercomputing simulation is one of the most promising approaches to optimize the performance of aeronautical products and therefore meet those objectives. Three benefits in the design phase can be expected through systematic use of simulation in physics and systems engineering:

1. The lead time to market can be reduced thanks to simulation in the design phase. Main issues in this area come from the capability to model complex scenarios in which couplings can be at stake (e.g. the aeronautical industry is confronted with new couplings configurations and electromagnetic, acoustic or thermal constraints, due to the recent introduction of composite structures)
2. Differentiating technologies can be demonstrated using simulation as a large number of configurations can be investigated during the design phase. Upon obtaining a promising configuration, simulation is then used to verify the stability of the solution via intensive parametric analysis in order to achieve a robust design.
3. Effectiveness of tests can be significantly enhanced if prepared by a well managed simulation process (e.g. if sensor’s position is optimized to obtain more useful data, more accurate extrapolation is achievable because of a better match between tests and computation results on well known subsets of parameters; this leads to a much reduced number of costly full scale tests, which are performed only for the very compulsory ones)

However, to meet those appealing benefits, the following constraints should be addressed:

• Models are available at the appropriate levels of accuracy, maturity and validation,
• Simulation is made affordable in time, cost and complexity.

Most of these topics will be covered in this Aerospace session.

The first part of the session will focus on the current technologies and capabilities used in aeronautical research and industrial applications both for numerical simulations and data analysis. Norbert Kroll from DLR will present current computing platforms and state of the art multi-physics numerical methods. Then, Eric Seinturier from Turbomeca SAFRAN will describe the challenges for greener aircraft and explain how simulation helps in developing next generation turboshaft engines. Rainald Löhner, George Mason University, will address in his presentation supercomputing needs and trends in aero- and hydrodynamics thereby also considering large-scale multidisciplinary simulations. Gabriel Staffelbach from CERFACS will outline in his speech high-performance computing related propulsion systems challenges. Alexander Kuleshov from the Russian Academy of Science will conclude this part of the session by addressing various simulation problems which can be effectively solved using data-based models requiring large data processing capabilities.

The second part of the session will mostly focus on the challenges that need to be solved in the next 10-20 years in the aerospace industry. Martin Küssner from Abaqus will present recent progresses in HPC methods for the simulation of large aircraft structures using composite materials. Detlef Müller-Wiesner, EADS, will focus in his presentation on the advantages high-performance computing do imply for innovative developments in the EADS divisions. Susan Ying from Boeing will then review the different challenges the aeronautics industry is facing with the impressive pace of advance in computational hardware. She will emphasize the role of education in order to take advantage of new approaches and tools in the aerospace industry as soon as possible. Axel Flaig from Airbus will conclude the seminar by demonstrating how innovative high performance simulation platforms will help Airbus to virtually “fly test” the next aircraft generations.

The program will be highly relevant for members of the academic, corporate, and government research communities involved in, or interested in pursuing research and development in the field of aerospace design – particularly their implication and application in the field of multi-physics numerical simulation. Those interested in an introductory look at the next challenges for the aeronautics industry may also wish to attend especially to understand the new requirements in term of IT and especially HPC.