The ROBOFLEX project “Robust Turbomachines for Flexible Use” included a sub‑project titled “Gap Influence in the Compressor” carried out at the Institute for Radiation Drives and Turbomachines of RWTH Aachen University (IST). The work spanned from 1 January 2020 to 30 June 2023 and was funded under the grant code O3EE5013A. The industrial partner MT U Aero Engines AG supplied the industrial context, data, and validation facilities, while IST performed the research, numerical simulations, and experimental rig development. The sub‑project comprised four work packages: (1) fundamentals and a generic test case, (2) an academic rig, (3) a multi‑stage compressor, and (4) additional improvement measures.
The scientific effort focused on the impact of the radial gap between rotor blades and cantilever stators on loss production and stability in axial compressors. The gap creates a high‑velocity vortex that interacts with the main flow, producing significant pressure losses. Conventional Reynolds‑averaged Navier–Stokes (RANS) models, which rely on semi‑empirical turbulence closures, are sensitive to the chosen model because the vortex is highly anisotropic. To quantify this sensitivity, a systematic RANS parameter study was performed, varying grid resolution, geometry mapping, and turbulence model settings. The study revealed that loss predictions can differ by up to 30 % depending on the turbulence model, underscoring the need for higher‑fidelity data.
To provide a reliable reference, a high‑order simulation was executed using Large Eddy Simulation (LES) and the Improved Delayed Detached Eddy Simulation (IDDES) approach. These simulations were carried out on the Virginia Tech cascade, which offers the most extensive experimental database among the considered test cases. The cascade operates at Reynolds numbers ranging from 97 000 to 1.1 million and features gap sizes of 0.5 % to 3.3 % of the channel height. LES and IDDES results were compared against RANS predictions, showing that the former capture the vortex structure and loss mechanisms more accurately, with pressure loss coefficients within 5 % of the experimental values.
The academic rig developed in work package 2 enabled systematic variation of the gap geometry and relative motion between blade and stator. By conducting a series of controlled experiments, the rig identified the most influential parameters: gap size, blade clearance, and relative motion. These findings guided the design of the multi‑stage compressor in work package 3, where the optimized gap configuration reduced total pressure loss by approximately 8 % compared to the baseline design.
Work package 4 explored additional improvement measures, such as modified blade profiles and labyrinth seals, to further mitigate gap‑induced losses. Numerical tests indicated that a 1.6 % gap reduction combined with a streamlined blade tip could lower loss coefficients by an additional 3 %. The combined effect of the optimized geometry and advanced turbulence modelling promises a more robust compressor that maintains high efficiency under off‑design operating conditions.
Overall, the sub‑project delivered a comprehensive understanding of gap‑induced loss mechanisms, validated high‑fidelity simulation tools, and identified practical design modifications that enhance compressor performance. The collaboration between IST and MT U Aero Engines AG ensured that the research addressed real‑world engineering challenges while advancing the scientific knowledge base for robust turbomachinery.