Session: 36-02 Surrogate Model-based applications

Paper Number: 153946

Aerothermal Design and Metamodel-Assisted Optimization of High-Speed Drive Turbine 

This work explores the development of a high-speed drive system for scaled-down electrified shockwave turboreactors tailored to the experimental laboratory pilot unit requirements. With the shockwave turboreactors operating at extremely high rotational speeds and elevated environmental temperatures, a direct electric motor drive is not feasible due to thermal limitations. To tackle this, a drive-turbine is proposed as an alternative for this scale of drive system. The primary focus of the study is to design a drive turbine capable of maintaining the necessary power. The preliminary 0D, 1D analysis led to detailed 3D design and simulations using Computational Fluid Dynamics (CFD), further incorporating design adjustments imposed by manufacturing constraints. Additive manufacturing (AM) has been selected for manufacturing, thereby imposing additional constraints that adversely impact the performance of the drive turbine. Inspections of the manufactured components using microscopy and profilometry are discussed to assess the impact of manufacturing constraints. A key challenge in laboratory facilities is the limited availability of pre-heated mass flow. To address this, the drive turbine was optimized using the in-house tool Computer Aided Design Optimization (CADO) to operate efficiently at lower mass flow rates. CADO employed a meta-model-assisted differential evolution optimization algorithm to achieve the optimal design while maintaining power requirements and additive manufacturing constraints. Structural feasibility through Finite Element Analysis (FEA) ensured safe operation at high rotational speeds. The one-to-one drive system configuration led to a new volute design. The volute design criteria are discussed, and its impact on the turbine performance was investigated, highlighting the impact of flow angle distribution and incidence angles. A detailed analysis of the influence of the volute on the overall system performance shows the reduction in efficiency. This necessitated the trade-offs between design complexity, manufacturability, and performance. This work validated the feasibility of the drive turbine concerning strength, quality, and performance through a multidisciplinary design process. The optimized drive system operates at the lowest possible mass flow while maintaining the required power, demonstrating a practical solution for driving high-speed micro-turbomachines in high-temperature environments.

Keywords: Drive system, Computational Fluid Dynamics (CFD), meta-model assisted optimization, volute design, additive manufacturing.

Presenting Author: Rejish Lal Johnson Von Karman Institute for Fluid Dynamics

Presenting Author Biography: Rejish Lal Johnson is a PhD candidate at the von Karman Institute for Fluid Dynamics, Belgium, in collaboration with Ghent University. His research focuses on the design and optimization of turbomachines using CFD, as well as aerothermal chemical reactor design and realization using additive manufacturing. He previously worked as a CFD Simulation Analysis Engineer at Volvo Group Trucks Technology, Sweden and completed his master’s thesis at Ariane Group, Munich. His expertise lies in fluid dynamics simulations, and he has contributed to key projects in the chemical, automotive and aerospace sectors.

Authors:

Rejish Lal Johnson Von Karman Institute for Fluid Dynamics
Nikolaos Rafail Lotsios Von Karman Institute for Fluid Dynamics
Mike Bonheure Ghent University
Johan Prinsier Von Karman Institute for Fluid Dynamics
Mohamed Hassanine Aissa Von Karman Institute for Fluid Dynamics
Kevin M. Van Geem Ghent University
Tom Verstraete Von Karman Institute for Fluid Dynamics