Researchers from the University of Melbourne, in collaboration with GE Aerospace, are working to rapidly enhance jet engine performance using advanced computer simulations on the Frontier supercomputer at Oak Ridge National Laboratory. The focus is on high-pressure turbine (HPT) engines, which are critical for the efficiency and durability of both commercial and military aircraft.
Oak Ridge National Laboratory reported that the Melbourne-based team has conducted simulations of unprecedented scale, examining how surface degradation affects turbine blades using models with 10–20 billion grid points and 10¹⁷ degrees of freedom. These detailed simulations allow researchers to accurately predict the mechanical impacts and changes in aerothermal performance resulting from long-term wear and tear.
Dr. Richard Sandberg, chair of Computational Mechanics at the University of Melbourne, told ORNL that simulating degradation is challenging due to the differences in time and length scales. He explained that while turbine blades are large, the surface changes occur at a microscopic level, requiring extremely large simulations to understand their impact on overall performance. The results of these simulations are expected to play a crucial role in developing more efficient and durable turbine engines for future aerospace applications.
Greg Sluyter, a senior engineer at GE Aerospace, emphasized the significance of the research, noting that Dr. Sandberg and his team are advancing understanding of how surface roughness and its distribution influence turbine performance. The findings are expected to inform the design of turbines that better withstand surface degradation, leading to engines that are more fuel-efficient and require less maintenance.
Findings published in the ASME Journal of Turbomachinery highlight the complexity of these simulations and how they challenge previous assumptions about surface roughness based on simpler systems when applied to the intricate geometries of turbine engines. Thomas Jelly, a lead researcher at the University of Melbourne, noted that conventional modeling approaches developed over the years are not suitable for these more complex, industrially relevant flow conditions.
By partnering with ORNL and leveraging the Frontier supercomputer, Dr. Sandberg’s team can perform highly detailed simulations that would have taken over a thousand years on conventional computers. Paul Vitt, chief consulting engineer at GE Aerospace, highlighted the benefits of this collaboration and high-performance computing, noting that DOE’s INCITE program has allowed them to accelerate the use of advanced simulation techniques and drive innovation more quickly.
Looking ahead, the research team, backed by a three-year allocation of computing time from DOE’s INCITE program, plans to further refine their simulations to better understand how cooling films interact with degraded turbine blade surfaces. These insights are expected to improve the performance of current engines and support the development of more efficient and advanced next-generation aerospace technologies.
