There were four project areas: rocket motor internal flows (Cucci, Iaccarino, Najjar, Moser); compressible and transitional flow (Lien, Kalitzin); buoyant heat transfer (Tieszen, Ooi); and non-local pressure effects (Manceau, Wang). The rocket internal flow is treated as duct flow with mass injection through the wall. The regime of interest is high injection rate, so the near-wall region is far from equilibrium. Turbulence in the injection flow is thought to have a significant effect. RANS computations of channel flow with wall injection do a good job of predicting the significant departures from equilibrium. RANS simulations of a nozzleless rocket are complicated by the transitional nature of the flow. The rocket group also preformed an initial DNS of channel flow with injection through both walls. A force was added to the momentum equation to obtain streamwise homogeneity. High injection velocities were found to produce large near-wall structures, that are not present with impermeable walls. The question of whether elliptic relaxation is applicable to transonic flow with shock waves was answered in the affirmative by Lien and Kalitzin. In the V2F model, the elliptic f-equation affects the mean flow only indirectly through the eddy viscosity; so it is no more implausible to use elliptic relaxation than to use other turbulence closures in transonic flow. Transition from laminar to turbulent flow occurs in many experiments to which RANS is applied. Lien and Kalitzin looked into modifications of the turbulence model that might simulate transition. The idea is that the model bifurcates from a laminar to a turbulent solution branch at some point in the flow. The model coefficients control that locations. Transition is notoriously present in buoyancy driven flows. Tieszen and Ooi have examined the effect of adding buoyant production terms to the turbulent kinetic energy equation. In some cases the main effect of such terms was to control the location of transition. Good predictions of the boundary layer on a vertical plate and of recirculation in a closed cavity were obtained --- provided that transition occurred at the correct position. Non-local pressure effects are associated with velocity-pressure gradient correlations in the Reynolds stress transport equations. Closing these correlations is a primary focus of Second Moment Closure. Elliptic relaxation is a non-local treatment of such terms. There is a degree of arbitrariness in the derivation of the Yukawa type of equation for elliptic relaxation. Manceau and Wang have exploited a DNS database in order to evaluate the assumptions inherent in this derivation. Although the basic method was found consistent with DNS data, it does not account for several sources of anisotropy. Modifications of elliptic relaxation that might treat these effects are discussed.