Search

In this numerical study, supersonic combustion of kerosene in three model combustor configurations is investigated. To this end, 3-D, compressible, turbulent, nonreacting, and reacting flow calculations with a single step chemistry model have been carried out. For the nonreacting flow calculations, the droplet diameter distribution at different axial locations, variation of the Sauter mean diameter, and the mixing efficiency for three injection pressures are presented and discussed. In addition, the effect of turbulent dispersion on the mixing efficiency is studied using a stochastic model in conjunction with the two-equation shear stress transport ?-? turbulence model. For the reacting flow calculations, contours of heat release and axial velocity at several axial locations are used to identify regions of heat release inside the combustor. Combustion efficiency predicted by the present results is compared with earlier predictions for all the combustor models. Furthermore, the predicted variation of static pressure along the combustor top wall is compared with experimental data reported in the literature. Calculations show that the penetration and spreading of the fuel increases with an increase in the injection pressure. Predicted values of the combustion efficiency are more realistic when the spray model is used for modelling the injection of the fuel. The importance of the mixing process, especially for a liquid fuel such as kerosene, on the prediction of heat release is discussed in detail.

Turbulent impinging air flow is investigated numerically by using the ANSYS-CFXA (R) code. All computations are performed by considering three-dimensional, steady, and incompressible flow. Three different Reynolds averaged Navier-Stokes (RANS) turbulence models and two Reynolds stress models (RSM's) are employed. Furthermore three different laminar to turbulent transition (LTT) models are employed with the shear stress transport (SST) and the baseline (BSL) models. Results show that predictions of the SST and two RSM's are very close each other and these models' results are in better agreement with the experimental data when all Reynolds numbers used in this study are considered. Secondary maxima in Nusselt number can be seen only if the LTT formula is employed with SST and BSL models.

The present study deals with the assessment of different turbulence models for heated wall jet flow. The velocity field and thermal characteristics for isothermal and uniform heat flux surfaces in the presence of wall jet flow have been predicted using different turbulence models and the results are compared against the experimental data of Wygnanski et al. (1992), Schneider and Goldstein (1994), and AbdulNour et al. (2000). Thirteen different turbulence models are considered for validation, which include the Standard k-? (SKE), Realizable k-? (RKE), shear stress transport (SST), Sarkar &amp, So (SSA), v 2-f, Reynolds stress Model (RSM), and Spalart Allmaras (SA) models. Both standard wall function (swf) and enhanced wall treatment (ewt) options available in a commercial CFD solver have been used for near wall treatment for the high Reynolds number models. From the study, it is observed that only a few models could accurately predict the complex flow and thermal features of the heated wall jet. The near wall velocity profile captured using Realizable k-? (RKE) with enhanced wall treatment (ewt) shows the best agreement with the experimental data as compared to the other models. Considerable deviation has been observed using SKE with standard wall function (swf) whereas the models of v 2-f show good prediction of velocity and temperature profiles in the near field region. However, the v 2-f model is found to deviate from the data in the downstream region where the velocity profiles exhibit similarity. In the prediction of heat transfer coefficient, RSM followed by SA and RKE with ewt, is found to be in closer agreement with the experimental data compared to the rest of the models. The computational time required for RSM is substantially higher than that of the other RANS models. Therefore, in the case of gas turbine combustor, since flow field is much more complex, the RKE with ewt would be the preferred choice over the SA model.

In this study computational fluid dynamic simulations of a turbine blade with endwall film cooling were compared to measurements of both aerodynamic and thermal performance. The experimental data were collected at low Mach number (Ma2is = 0.3) in a linear cascade arrangement with 7 blades which geometry is typical of first stage high pressure turbine. A junction between the blade hub and the platform is provided by a 3D fillet. Coolant is injected through ten cylindrical holes distributed along the blade pressure side. Coolant to mainstream mass flow ratio was set to assure an inlet blowing ratio of M1 = 2.4 and M1 = 3.2. The simulations were carried out using the Shear Stress Transport (SST) k-ω turbulence model. Numerical predictions were compared against experimentally measured secondary flows and endwall film cooling effectiveness, at different injection conditions. Simulation results agreed with the experiments for what concerns the general shape and the location of secondary flows. However, some limitations in the modeling were highlighted when going into the details of loss computation and vortex structure. Predictions overestimated both secondary and midspan blade wake losses. Moreover, the effect of the fillet on the aerodynamic flow features was not fully captured. Predicted film cooling results showed the sweeping of coolant across the passage in agreement with experiments even though jets persistency was higher than that measured. Levels of adiabatic effectiveness were generally well simulated.

Conjugate heat transfer from a flat circular disk is investigated both computationally and experimentally with a constant heat flux imposed on its bottom surface and a shower head of air jets impinging on the top surface. The shower head consists of a central jet surrounded by four neighboring perimeter jets. Measured temperature data at twelve locations within the plate are compared with the conjugate heat transfer result obtained at the same locations computationally by Shear Stress Transport (SST) ?-? turbulence model. The spacing to orifice diameter ratio (H/d = 1 to 4), the jet Reynolds number (7115 to 10674) and the plate thickness to diameter ratio (0.25, 1 and 2) are varied. Measured temperature data within the solid plate agrees well with the computationally obtained conjugate heat transfer result. Primary peaks of the Nusselt number at the stagnation point and secondary peaks at the interaction points of the wall jets are observed. Local variation of heat transfer rate with varying H/d is significant but found to be less sensitive to the thickness ratio. When compared with a single jet of equal mass flow rate and Reynolds number, the shower head jets provide more uniform temperature distribution with higher heat transfer rate.

Numerical simulations of compressible, turbulent jets at M = 0.75 using the Shear Stress Transport (SST) k-? model have been carried out for baseline nozzle and chevron nozzles with 4, 6 and 8 lobes and three different penetration angles (0�, 5� and 10�). The predicted far field noise level at several observer locations are compared with experimental data. Overall sound pressure levels at far field observer locations have been calculated using Ffowcs Williams-Hawkings equation. Numerical prediction of aerodynamic quantities like centreline velocity, stagnation pressure, turbulent kinetic energy and axial vorticity are compared among the nozzles. The Unsteady Reynolds Averaged Navier Stokes (URANS) calculations are able to predict the trends in overall SPL, even though the absolute values are slightly under-predicted. Copyright � 2010 Inderscience Enterprises Ltd.

In the present work, the influence of the turbulence model on the numerical predictions of supersonic combustion of hydrogen in a model combustor is investigated. Three dimensional, compressible, turbulent, reacting flow calculations using the Shear Stress Transport (SST) ? - ? model have been carried out. The results are compared with earlier results obtained using the Spalart-Allmaras model. The effect of detailed chemistry on the predictions of heat release and combustion efficiency is also investigated for one injection scheme. The calculations show that the mixing and hence the heat release is over predicted by the SA model. Whereas, the peak values for pressure and temperature are predicted better by the SST ? - ? model. This investigation demonstrates the importance of the use of a two equation turbulence model over a one equation model for studying such complex flow phenomena. Copyright � 2009, Inderscience Publishers.

This paper presents a computational fluid dynamics approach to study the forces and moments generated on the parent body due to attached fins in marine vehicles and also the velocity profile behind the fins. The conservation equations of mass and momentum are solved with Finite Volume Method by use of the software FLUENT. The turbulence closure model used is k-? SST (shear stress transport). The study is carried out for single fin at various angles of attack and four fins attached to a cylinder.