Bachelor / Master Thesis: Stability Investigations on an Adjoint CFD Solver for Turbomachinery [...]

#UPLIFTY OURFUTURE

Over 13,000 people. 19 locations worldwide. One in three planes flies with our technology. Even more economical, lower in emissions and quieter. We shape the future of aviation. The only thing missing: You.

Our team is searching for you:

Bachelor / Master Thesis: Stability Investigations on an Adjoint CFD Solver for Turbomachinery Design

in Munich

Code – D MUC AEAB

Reduction of emissions, weight, and vibrations while increasing efficiency are goals for the design of the next generation of civil aircraft engines. To reach these technological goals, advanced aerodynamic designs of engine components need to be developed, which increase performance and yet fulfil multiple design constraints. While the single stage numerical optimization of a compressor or turbine is a mature technology, the optimization of multi-stages up to a full module or multiple modules is not. Due to the higher dimensional problem, it requires a gradient-based approach using the adjoint method to evaluate gradients of aerodynamic constraints and targets (efficiency, massflow, surge margin etc.) with respect to blade parameters (stagger angle, chord length, curvature etc.).

However, our algorithmically differentied CFD solver faces sometimes instabilities when solving the adjoint system of equations. For this reason, stability methods are a prime subject for this study.

In this project, a combination of various solution and stability methods must be studied on representative compressors and turbine test cases. The goal is to propose more efficient methods of stabilizing and accelerating the adjoint calculations while being as robust as the method currently in use. The researched methods would then be implemented in the adjoint CFD solver and their performance confirmed. Finally, a shape optimization employing the improved adjoint CFD solver will be carried out.

YOUR TASKS

• Get familiar with the adjoint method for aerodynamic optimization and stabilization methods for large PDE systems (e.g. Anderson Acceleration, Recursive Projection Method, Krylov Subspace methods, preconditioning)
• Benchmark existing methods using MTU’s computational and optimization tools
• Evaluate and compare methods such as RPM, GMRes, and Anderson Acceleration on compressor and turbine test cases
• Analyze performance and convergence behavior for each approach
• Develop and test improvements by tuning solver parameters or extending existing methods
• Implement the most promising stabilization method in our C++ adjoint CFD solver
• Validate the improved solver through benchmark and optimization studies

YOUR PROFILE

• Excellent mathematical basis (especially linear algebra)
• Good programming skills in C++ and Python
• Comfortable in English for work environment
• Comfortable in Linux Environment is mandatory
• Basic knowledge of aerodynamics and CFD simulation process

OUR OFFER

• Insights into the aviation industry and jobs with responsibility
• Team spirit, e.g. through student network & events
• Mentoring and individual support
• Flexible working time models with mobile working
• Canteen, café bar and sales shops
• Sports cooperation with EGYM Wellpass & company sports groups
• Excellent connections thanks to commuter bus, parking garage, e-charging station

READY?

Give your career a boost and send us your complete application (CV, Transcript of Records, High School Diploma).

We are looking forward to getting to know you!

Interested?

Recruitingteam

#J-18808-Ljbffr