SINOPSIS

Compressible and under-expanded jets with a nozzle-pressure ratio (NPR) up to 3 are computed with a two-component-limit second-moment closure and modifications to account for effects of compressibility on turbulence. The results are compared to experimental data from LDA measurements. Good agreement is observed for the baseline turbulence model beyond the potential core. In contrast, for the range of NPRs studied, model modifications, specifically designed to replicate the effects of compressibility in shear layers, tend to worsen the results for the jet flows. Al-though the model predicts well the shock-cell structure in the potential core at high NPRs, the core tends to be too long, and a much higher level of discharge turbulence than that measured is required to procure reasonable agreement. Differences in the core in respect of turbulence intensity are argued to partially reflect experimental limitations.