Cabot’s Near-wall Function for Turbulence Closure Models

This 24 months research topic was an industrial driven project jointly conducted by RSAF-Defence Science & Technology Agency, NTU and Pratt & Whitney (PWA)United Technology Research Centre. The first phase was held in PWA mainly on field investigation of stall flutter in fan blades through the RSAF’s On-Job-Training Programme. The second phase was held in NTU for code development in simulation of flowfield event. The 2 phases were interdependent as expert knowledge and engine data from PWA has complimented the theoretical aspects and validation of test data at NTU. Computational fluid dynamics is becoming an increasing powerful tool in the aerodynamic design of aerospace vehicles as a result of improvements in numerical algorithms as well as in the processing speed and storage capacity of new generations of computers. For a long time, application of turbulence closure model is limited to region without the presence of a solid wall. Most practical flows however, are wall bounded. Over the years, many suggestions have been made for the extension of turbulence closure models to enable their use at low-Reynolds (Re) numbers and to describe the flow close to a solid wall. Basically, the near-wall effects are treated as low-Re number effect. Nevertheless, near-wall effects are different from low-Re number effects due to wall blocking that is absent in the case of unbounded turbulence [1]. This wall blocking effect gives rise to increased anisotropy of the turbulence, as the wall is approached [2]. Furthermore, application of such near-wall treatment will require large expenditure of computational resources since the near-wall viscous and buffer regions of the wallbounded flows have to be resolved adequately. These dense grids will severely reduce the time-step due to the Courant-Friedricks-Levy (CFL) condition in order to attain stability and/or accuracy. The aim of this work was to introduce the wall treatment method as proposed by Cabot for large eddy simulation (LES) into the RANS turbulence closure model [3].