Computational study of jet-in-crossflow and film cooling using a new unsteady-based turbulence model

This paper documents a computational investigation of the unsteady behavior of jet-in-crossflow applications. Improved prediction of fundamental physics is achieved by implementing a new unsteady, RANS-based turbulence model developed by the authors. Two test cases are examined that match experimental efforts previously documented in the open literature. One is the well-documented normal jet-in-crossflow, and the other is film cooling on the pressure side of a turbine blade. All simulations are threedimensional, fully converged, and grid-independent. High-quality and high-density grids are constructed using multiple topologies and an unstructured, super-block approach to ensure that numerical viscosity is minimized. Computational domains include the passage, film hole, and coolant supply plenum. Results for the normal jet-in-cross flow are for a density ratio of I and velocity ratio of 0.5 and include streamwise velocity profiles and injected flow or "coolant" distribution. The Reynolds number based on the average jet exit velocity and jet diameter is 20,500. This represents a good test case since normal injection is known to exaggerate the key flow mechanisms seen in film-cooling applications. Results for the pressure side film-cooling case include coolant distribution and adiabatic effectiveness for a density and blowing ratio of 2. In addition to the in-house model that incorporates new unsteady physics, CFD simulations utilize standard, RANS-based turbulence models, such as the "realizable" k-epsilon model. The present study demonstrates the importance of unsteady physics in the prediction of jet-in-cross flow interactions and for film cooling flows that exhibit jet liftoff.