Investigating How Turbine Blade Surface Degradation Affects Jet Engines

Why It Matters

Accurate modeling of blade surface degradation enables significant fuel savings, longer component life, and stronger aerospace competitiveness.

Key Takeaways

• Frontier enabled 10‑20 billion‑cell turbine simulations
• Traditional roughness models inaccurate for real turbine geometries
• Microscale roughness increases aerodynamic loss and heat flux
• HiPSTAR runs weeks, not millennia, on Frontier
• Insights inform GE’s next‑gen HPT and HyTEC projects

Pulse Analysis

The advent of exascale platforms such as Frontier is reshaping computational fluid dynamics for aerospace propulsion. By harnessing 2,500 nodes and AMD GPU acceleration, the University of Melbourne team executed direct‑numerical simulations with 10–20 billion grid points—orders of magnitude larger than any prior HPT study. This resolution captures micron‑scale surface roughness while still representing the full blade geometry, a feat previously impossible due to the mismatch between blade length and roughness features. The new HiPSTAR solver eliminates turbulence‑model assumptions, delivering physics‑based insight into laminar‑to‑turbulent transition and heat‑transfer pathways.

The simulations revealed that conventional roughness correlations, derived from simple pipe or flat‑plate flows, dramatically underestimate losses in a high‑pressure turbine environment. Microscopic degradation amplifies viscous drag and wall heat flux, directly degrading fuel efficiency and accelerating maintenance cycles. GE Aerospace is already feeding these data into the design loop for its next‑generation HPTs and the NASA‑backed HyTEC program, aiming to create airfoils that tolerate roughness without sacrificing performance. By quantifying the exact penalty of surface wear, engineers can optimize film‑cooling strategies and material selections, potentially cutting fuel burn by several percent.

Beyond immediate engine gains, the work aligns with broader energy‑security and emissions targets. More efficient turbines translate into lower CO₂ per flight, supporting airline sustainability pledges and national competitiveness in high‑tech manufacturing. The success also demonstrates the value of DOE’s INCITE allocations, proving that leadership‑class supercomputers can accelerate industrial R&D cycles that would otherwise span decades. As exascale resources become more widely available, similar high‑fidelity approaches are expected to spread across aerospace, automotive, and power‑generation sectors, ushering in a new era of data‑rich, physics‑first design.