Session: 35-18 Turbine Methods 2

Paper Number: 103064

103064 - Reynolds Number Impact on High Pressure Turbine Flows 

Recent advances in the availability of large-scale computer systems that employ GPU acceleration, as well as algorithmic choices that can take advantage of the power of these systems, have opened the door to a wide range of simulations that have until recently been considered too costly for engineering analysis. In the field of turbomachinery computational fluid dynamics, this applies to the class of Large Eddy Simulation problems that allow for high-fidelity simulation of flows around entire turbine components, such as blades.

During operation, turbine blades in aircraft engines can experience flows over a wide range of Reynolds numbers. This range is impacted by engines of different sizes, as well as different operating altitudes, all of which add to the overall range of Reynolds numbers that are relevant for aircraft engine design. Unfortunately, increasing flow Reynolds number has a significant impact on the mesh resolution required for accurately simulating wall-resolved LES models due to the mesh resolution required to capture the near-wall flow physics. During the design process, providing engineers a detailed understanding of the flow physics phenomena, and how they change with Reynolds number, could lead to more efficient and durable turbine designs.

In this paper we present the study of flows around an uncooled high-pressure turbine blade. The simulations are completed with the high-order GENESIS code, which has been successfully applied to thread-parallel compute architectures, such as large computational systems with GPU capability. The simulations cover a range of Reynolds numbers in order to study the impact of Reynolds number increase over a fixed geometry. This allows for examining the impact on flow physics of interest, such as boundary layer evolution, laminar-to-turbulent transition, and wake shedding. Understanding the behavior of the flow as the Reynolds number increases potentially has broad implications for turbine design, and how the designs are evaluated. Trends examined will include the impact of incoming turbulence intensity on laminar-to-turbulent transition and pressure loss, along with the impact of increasing Reynolds number on the same quantities. Recent experimental studies have shown Reynolds number-related trends in airfoil loss that can have significant performance implications, and the current computational study provides a detailed flowfield representation over a similar Reynolds number sweep, providing the ability to interrogate detailed flow features.

Presenting Author: Michal Osusky GE Research

Presenting Author Biography: Dr. M. Osusky’s research focus is in the area of heat transfer for turbomachinery, with a focus on simulation of the hot gas path of the turbine. Dr. Osusky’s current role of Senior Engineer in the Thermosciences Organization of GE Research involves leveraging parallel codes on various HPC systems, including GE-owned Cray and Linux machines, as well as external resources (such as the OLCF Summit system), to conduct large domain CFD simulations. Dr. Osusky served as PI for the ENG111 project under a 2020 INCITE grant, which demonstrated the capability of a high-order GPU-accelerated lEs solution algorithm for solving full 3D turbomachinery problems.

Authors:

Michal Osusky GE Research
Gustavo Ledezma GE Research
Greg Sluyter GE Aerospace
Sriram Shankaran GE Aerospace