Search

Two extended models for the calculation of rough wall transitional boundary layers with heat transfer are presented. Both models comprise a new transition onset correlation, which accounts for the effects of roughness height and density, turbulence intensity, and wall curvature. In the transition region, an intermittency equation suitable for rough wall boundary layers is used to blend between the laminar and fully turbulent states. Finally, two different submodels for the fully turbulent boundary layer complete the two models. In the first model, termed KS-TLK-T in this paper, a sand roughness approach from (Durbin, et al., 2001, 'Rough Wall Modification of Two-Layer k-[epsilon],' ASME J. Fluids Eng., 123, pp. 16-21), which builds on a two-layer k-[epsilon]-turbulence model, is used for this purpose. The second model, the so-called DEM-TLV-T model, makes use of the discrete-element roughness approach, which was recently combined with a two-layer k-[epsilon]-turbulence model by the present authors. The discrete-element model will be formulated in a new way suitable for randomly rough topographies. Part I of this paper will provide detailed model formulations as well as a description of the database used for developing the new transition onset correlation. Part II contains a comprehensive validation of the two models, using a variety of test cases with transitional and fully turbulent boundary layers. The validation focuses on heat transfer calculations on both the suction and the pressure side of modern turbine airfoils. Test cases include extensive experimental investigations on a high pressure turbine vane with varying surface roughness and turbulence intensity, recently published by the current authors, as well as new experimental data from a low pressure turbine vane. In the majority of cases, the predictions from both models are in good agreement with the experimental data. [DOI: 10.1115/1.2992512]

Two extended models for the calculation of rough wall transitional boundary layers with heat transfer are presented. Both models comprise a new transition onset correlation, which accounts for the effects of roughness height and density, turbulence intensity, and wall curvature. In the transition region, an intermittency equation suitable for rough wall boundary layers is used to blend between the laminar and fully turbulent states. Finally, two different submodels for the fully turbulent boundary layer complete the two models. In the first model, termed KS-TLK-T in this paper, a sand roughness approach from Durbin et al. (2001, 'Rough Wall Modification of Two Layer k-[epsilon],' ASME J. Fluids Eng., 123, pp. 16-21), which builds upon a two-layer k-[epsilon]-turbulence model, is used for this purpose. The second model, the so-called DEM-TLV-T model, makes use of the discrete-element roughness approach, which was recently combined with a two-layer k-[epsilon]-turbulence model by the present authors. The discrete-element model will be formulated in a new way, suitable for randomly rough topographies. Part I of the paper will provide detailed model formulations as well as a description of the database used for developing the new transition onset correlation. Part H contains a comprehensive validation of the two models, using a variety of test cases with transitional and fully turbulent boundary layers. The validation focuses on heat transfer calculations on both the suction and the pressure side of modern turbine airfoils. Test cases include extensive experimental investigations on a high-pressure turbine vane with varying surface roughness and turbulence intensity, recently published by the current authors as well as new experimental data from a low-pressure turbine vane. In the majority of cases, the predictions from both models are in good agreement with the experimental data. [DOI: 10.1115/1.2992511]

The present study deals with the unsteady flow simulation of trailing edge film cooling on the pressure side cut back of gas turbine airfoils. Before being ejected tangentially on the inclined cut-back surface, the coolant air passes a partly converging passage that is equipped with turbulators such as pin fins and ribs. The film mixing process on the cut back is complicated. In the near slot region, due to the turbulators and the blunt pressure side lip, turbulence is expected to be anisotropic. Furthermore, unsteady flow phenomena like vortex shedding from the pressure side lip might influence the mixing process (i.e., the film cooling effectiveness on the cut-back surface). In the current study, three different internal cooling designs are numerically investigated starting from the steady RaNS solution, and ending with unsteady detached eddy simulations (DES). Blowing ratios M=0.5; 0.8; 1.1 are considered. To obtain both, film cooling effectiveness as well as heat transfer coefficients on the cut-back surface, the simulations are performed using adiabatic and diabatic wall boundary conditions. The DES simulations give a detailed insight into the unsteady film mixing process on the trailing edge cut back, which is indeed influenced by vortex shedding from the pressure side lip. Furthermore, the time averaged DES results show very good agreement with the experimental data in terms of film cooling effectiveness and heat transfer coefficients.

The present study deals with trailing edge film cooling on the pressure side cutback of gas turbine airfoils. Before being ejected tangentially onto the inclined cut-back surface the coolant air passes a partly converging passage that is equipped with turbulators such as pin fins and ribs. The experiments are conducted in a generic setup and cover a broad variety of internal cooling designs. A subsonic atmospheric open-loop wind tunnel is utilized for the tests. The test conditions are characterized by a constant Reynolds number of [Re.sub.hg] = 250 000, a turbulence intensity of [Tu.sub.hg] = 7%, and a hot gas temperature of [T.sub.hg] = 500 K. Due to the ambient temperature of the coolant, engine realistic density ratios between coolant and hot gas can be realized. Blowing ratios cover a range of 0.20 < M < 1.25. The experimental data to be presented include discharge coefficients, adiabatic film cooling effectiveness, and heat transfer coefficients in the near slot region (x/H < 15). The results clearly demonstrate the strong influence of the internal cooling design and the relatively thick pressure side lip (t/H = 1) on film cooling performance downstream of the ejection slot.

A new model for predicting heat transfer in the transitional boundary layer of rough turbine airfoils is presented. The new model makes use of extensive experimental work recently published by the current authors. For the computation of the turbulent boundary layer, a discrete element roughness model is combined with a two-layer model of turbulence. The transition region is modeled using an intermittency equation that blends between the laminar and turbulent boundary layer. Several intermittency functions are evaluated in respect of their applicability to rough-wall transition. To predict the onset of transition, a new correlation is presented, accounting for the influence of freestream turbulence and surface roughness. Finally, the new model is tested against transitional rough-wall boundary layer flows on high-pressure and low-pressure turbine airfoils. [DOI: 10.111511.2750675]

This paper summarizes the experimental results of a comprehensive study on the heat transfer and aerodynamic losses of a highly loaded turbine blade with surface roughness. A few hundred test cases conducted at several Reynolds numbers, freestream turbulence levels, and different deterministic roughness geometry have been examined. Some of these results have been published in two previous papers, showing a strong effect of roughness on laminar-turbulent bypass transition on the airfoil suction side. Beside roughness height, roughness anisotropy has turned out to be one of the major influencing factors. The airfoil heat transfer distribution of these measurements is used for detecting the transition onset. Additionally, further transition onset data from the literature is reevaluated. Thus, important roughness (geometry) parameters are identified and a new correlation for the transition onset is deduced, including roughness parameters along with freestream turbulence. Moreover, a method to extract the relevant roughness param-eters from realistic surface roughness is presented. Additional heat transfer and aerody-namic measurements are conducted for two different real surface roughness types. Calculations with a ID-boundary layer code on these surfaces are presented in order to validate the new model. [ABSTRACT FROM AUTHOR]

The present experimental study is part of a comprehensive analysis accounting for heat transfer and aerodynamic losses on a highly loaded low pressure turbine blade with varying surface roughness. Whereas Part 1 focuses on heat transfer measurements at airfoil midspan with different deterministic surface roughnesses, Part II investigates surface roughness effects on aerodynamic losses of the same airfoil. A set of different arrays of deterministic roughness (the same as used in Part I) is investigated in these experiments. The height and eccentricity of the roughness elements are varied, showing the combined influence of roughness height and anisotropy on the losses produced in the boundary layers. It is shown that the boundary layer loss is dominated by the suction side. Therefore, the investigations focus on measurements of the suction side boundary layer thickness at midspan directly upstream of the trailing edge. The experiments are conducted at several freestream turbulence levels (Tu1 =1.4-10.1%) and different Reynolds numbers. The measurements reveal that suction side boundary layer thickness is increased by up to 190% if surface roughness shifts the transition onset upstream. However, in some cases, at low Reynolds numbers and freestream turbulence, surface roughness suppresses boundary layer separation and decreases the trailing edge boundary layer thickness by up to 30%. [ABSTRACT FROM AUTHOR]

The present experimental study is part of a comprehensive heat transfer analysis on a highly loaded low pressure turbine blade and endwall with varying surface roughness. Whereas a former paper (Lorenz et al., 2009, "An Experimental Study of Airfoil and Endwall Heat Transfer in a Linear Turbine Blade Cascade--Secondary Flow and Surface Roughness Effects, " International Symposium on Heat Transfer in Gas Turbine Systems, Aug. 9-14, Antalya, Turkey) focused on full span heat transfer of a smooth airfoil and surface roughness effects on the endwall, in this work further measurements at the airfoil midspan with different deterministic surface roughness are considered. Part I investigates the external heat transfer enhancement due to rough surfaces, whereas part II focuses on surface roughness effects on aerodynamic losses. A set of different arrays of deterministic roughness is investigated in these experiments, varying the height and eccentricity of the roughness elements, showing the combined influence of roughness height and anisotropy of the rough surfaces on laminar to turbulent transition and the turbulent boundary layer as well as boundary layer separation on the pressure and suction side. It is shown that, besides the known effect of roughness height, eccentricity of roughness plays a major role in the onset of transition and the turbulent heat transfer. The experiments are conducted at several freestream turbulence levels (Tu1 = 1.4-10.1%) and different Reynolds numbers. [ABSTRACT FROM AUTHOR]

A new model for predicting heat transfer in the transitional boundary layer of rough turbine airfoils is presented. The new model makes use of extensive experimental work recently published by the current authors. For the computation of the turbulent boundary layer a discrete element roughness model is combined with a two-layer model of turbulence. The transition region is modeled using an intermittency equation that blends between the laminar and turbulent boundary layer. Several intermittency functions are evaluated in respect of their applicability to rough-wall transition. To predict the onset of transition, a new correlation is presented, accounting for the influence of freestream turbulence and surface roughness. Finally, the new model is tested against transitional rough-wall boundary layer flows on high-pressure and low-pressure turbine airfoils. [ABSTRACT FROM AUTHOR]