This paper presents the simulation results for the NASA C608 low-boom supersonic aircraft test case from the 3rd AIAA Sonic Boom Prediction Workshop. The unstructured finite-volume Navier-Stokes solver UNS3D was used to predict the turhulent nearfield flow. Nearfield solutions extracted three body lengths form the aircraft along a 50° azimuthal arc were propagated to the ground using NASA's augmented-Burgers nearfield propagation tool, sBoom, to predicted the sonic boom carpet. Four combinations of the convective flux scheme, gradient reconstruction method, and solution limiter were used to assess their impact on the nearfield solution. It was found that the flux function choice was not as important as the choice of the limiter and the gradient reconstruction. Using a combination of weighted LSQR gradient reconstruction method and modified Venkatakrishnan solution limiter led to solution instability on the coarsest grid and divergence on the remaining grids. Using the Dervieux limiter with weighted LSQR gradients improved overall convergence and allowed for flow predictions on all grids. However, the Dervieux limiter introduced too much dissipation into the solutions, and therefore were not used for sonic boom loudness predictions. The Green-Gauss gradient reconstruction method with the modified Venkatakrishnan limiter predicted the flow on all four of the grids. Nearfield and ground signature predictions were compared to an ensemble data set created using a subset of the workshop participants submissions. The authors created this set of data so it only includes results generated on the same type and size grid. The nearfield signature was found to correlate well with the workshop ensemble data. However, the resultant sonic boom carpet was quieter than the ensemble sonic boom carpet. A sensitivity study identified the component of the nearfield signature responsible for the variance observed in the sonic boom carpet.
展开▼
