research TEXSTAN

For more than 25 years TEXSTAN and its predecessor codes STAN5 and STANCOOL have been used worldwide in academia and industry. Research versions of this code have extended turbulence closure modeling that include full Reynolds stress models, turbulent heat flux models, and multiple-time-scale k-ε models. The performance of various two-equation models with transition is described in Harasgama et al.¹ for a highly-loaded transonic turbine guide vane data of Arts et al.² They discovered the models are generally adequate for the laminar and fully turbulent portions of the vane surfaces over the various Mach and Reynolds numbers and free stream turbulence levels of the experiments, but predictions indicate a somewhat inconsistent model performance for a given model in the start/end of transition region, especially at lower Reynolds numbers. To model stochastic surface roughness the Taylor-Coleman-Hodge³ discrete roughness model and its modification for heat transfer by Tarada⁴ have been incorporated into the research version of TEXSTAN. Tolpadi and Crawford⁵ used this model to predict the suction and pressure side heat transfer data of airfoils for heat transfer data by Abuaf, Bunker, and Lee⁶ at four Reynolds numbers and high free stream turbulence. The roughness elements were simulated using full-height right-circular cones, with parameters of the elements being matched to the airfoil's roughness statistical profile measurements. Both the heat transfer level and transition locations were in good agreement for all but the lowest of the Reynolds number cases. The film cooling models of research TEXSTAN have evolved since the early version described by Miller and Crawford.⁷ The revised film cooling models are derived based on the early work of Tafti and Yavuzkurt⁸ and the extensions to that work by Neelakantan and Crawford.^9,10 Those references show extensive validation of the models for flat plate conditions. For film-cooled airfoils, very little validation of the Crawford models have been carried out. Weigand, Bonhoff, and Ferguson¹¹ evaluated research TEXSTAN and a 3D Navier-Stokes code for predicting heat transfer on a fully film-cooled vane data for both film cooling effectiveness and heat transfer. They found the boundary layer program provided good agreement with the data, providing the appropriate boundary conditions are used for the velocity loading. Weigand, Crawford, and Lutum¹² evaluated research TEXSTAN for the combined effects of its roughness and film cooling models using the flat plate data included in Hoffs, Drost, and Boelcs.¹³ The combined effects were successfully modeled for both the effectiveness and heat transfer.

References