CFD Challenges in Power & Propulsion
CFD has become an established, valuable analysis tool in the power and propulsion industries, driving innovation to bring efficient, economical solutions to the world’s energy and transportation systems. Continual advances in high-performance computing systems and numerical methods provide engineers with the ability to efficiently implement design improvements based on characterization of pressure ratios, efficiency, reliability and critical flow features like boundary layers and separation. Incorporation of CFD in the design-cycle of power and propulsion systems has reduced the need for expensive experimental studies, bringing cost-efficient solutions to fruition with much shorter design cycle times.
The dramatic growth of CFD in power and propulsion has also brought attention to the difficulties in modeling complex flow phenomena in turbomachinery and the compromises that remain between physical accuracy and numerical limitations. Solvers must be fast and robust in all flow regimes spanning from subsonic to hypersonic, must accurately capture shock phenomena and boundary layer behavior and reliably predict machine efficiency and pressure ratios. Efficient and novel codes are needed to reduce computational requirements for multi-simulation parameter studies and grid-independence evaluations of transient flow fields in rotating machinery. Development in the area of turbulence modeling must improve predictions of transient, turbulent flow fields with secondary flows and vortex separation. Aerodynamic codes should also encompass modeling aero-acoustics for noise reduction, an effort that can conflict with aerodynamic optimization and remains difficult achieve. These CFD design tools should enable all of the above within a user-friendly, intuitive interface.
While significant development continues, CFD analysis has proven to be a crucial research, development and design tool in advancing power and propulsion technology and reducing costs and turnaround time. Quality CFD results reduce dependence on expensive experimental studies and allow for design optimization early in the design phase of a project. The major challenges of the future will be producing accurate, reliable and fast solver codes that continue to increase the value of CFD.