For hypersonic compressible turbulence, the correlations with respect to the density fluctuation must not be neglected. A Reynolds averaged K-ε model is proposed in the present paper to include these correlations, together with the Reynolds averaged Navier-Stokes equations to describe the mean flowfield. The K-equation is obtained from Reynolds averaged single-point second moment equations which are deduced from the instantaneous compressible Navier-Stokes equations. Under certain hypotheses and scales estimation of the compressible terms, the K-equation is simplified. The correlation terms of the fluctuation field appearing in the resulting K-equation, together with a conventional form of the ε-equation, are thus correlated with the variables in the average field. The new modeling coefficients of closure terms are optimized by computing the hypersonic turbulent flat-plate measured by Coleman and Stollery [J. Fliud Mech., Vol. 56 (1972), p. 741]. The proposed model is then applied to simulate hypersonic turbulent flows over a wedge compression corner angle of 34 degree. The predicting results compare favorably with the experimental results. Also, comparisons are made with other turbulence models. Additionally, an entropy modification function of Harten-Yee’s TVD scheme is introduced to reduce artificial diffusion near boundary layers and provide the required artificial diffusion to capture the shockwaves simultaneously.

An improved K-ω model, which allows for compressible corrections, is proposed in this paper. Numerical scheme was established utilizing the improved Total Variation Diminishing (TVD) scheme and applying implicit scheme to the negative source terms of the turbulence model. Hypersonic flat-plate boundary-layer flows and hypersonic compression ramp flows marked with separation, reattachment and shock/boundary layer interactions are then computed. Comparisons between the computational results, the experimental results and the semi-empirical formulations show that the compressible correction term of the K-ω turbulence model is a pressure-dilatation correlation. In addition, for flow with separation and without separation, calculation results of wall pressures, friction coefficients and wall heat transfer rate distributions using the improved model and established scheme agree better with the experimental results than that using the original K-ω model.

An improved Shear Stress Transport(SST) model, which allows for the compressible corrections, is proposed in this study. Numerical scheme was established by taking advantage of the improved Total Variation Diminishing (TVD) scheme and by applying implicit scheme to the negative source terms of the turbulence model. Hypersonic flat-plate boundary-layer flows and hypersonic compression ramp flows marked with separation, reattachment and shock/boundary-layer interactions are then computed. The comparisons between the computational results, the experimental results and the semi-empirical formulations show that the compressible correction term of the SST turbulence model is the scalar product of the weighted density average of the turbulent fluctuating velocity and the pressure gradients of the average flow field correlation quantities. In addition, for flow with separation and without separation, calculation results of wall pressures, friction coefficients and wall heat transfer rate distributions using the improved model and established scheme agree better with the experimental results than that using the original SST model.

A numerical insight was accomplished to optimize the scramjet combustor configuration based on orthogonal experimental design. Parametric modeling of combustor configurations was performed by the orthogonal array with 13 factors at 3 levels. Numerical simulations were proceeded by k-ε standard turbulence model and eddy-dissipation model in the combustion process. The performance indexes of combustion efficiency, total pressure recovery coefficient and thrust gain coefficient were evaluated. Detailed comparison with the effect of the factors on the performance was also carried out to demonstrate the main factors and determine the optimal configuration. The analysis of the extreme differences of the factors indicates that the main factors affecting combustion efficiency were the length of the wedge, the length depth ratio of the cavity, the depth of the cavity, and the length of the expanding section; The main factors affecting total pressure recovery coefficient are the angle of the primary combustor, the length of the expanding section, and the thickness of the strut; The main factors affecting thrust gain coefficient are the thickness of the strut, the length of the expanding section, and the angle of the secondary combustor. Validation of the optimal configuration is then confirmed that its performance is higher than the rest of the configurations, with the combustion efficiency of 0.915 and the total pressure recovery coefficient of 0.486, which are 31.5% and 65.9% higher than the experimental results, respectively.