These results suggest that PDE 5 inhibitors may play a more important role in early postoperative skin flap viability rather than at later time points and may be beneficial for skin flap viability as shown in the rat model. PDE 5 inhibitors may reduce the extent of necrosis after reconstructive surgeries.

The combination of sildenafil and VEGF decreases the extent of avascular and stasis zones in skin flaps. The skin flap improvement seen with the combination treatment was similar to the sildenafil treatment alone suggesting that enhanced flap survival was due solely to the effect of sildenafil.

This paper describes the effect of the state of the inlet boundary layer (laminar or turbulent) on the structure of the endwall flow on two different profiles of low-pressure (LP) turbine blades (solid thin and hollow thick). At present the state of the endwall boundary layer at the inlet of a real LP turbine is not known. The intention of this paper is to show that, for different designs of LP turbine, the state of the inlet boundary layer affects the performance of the blade in very different ways. The testing was completed at low speed in a linear cascade using area traversing, flow visualization and static pressure measurements. The paper shows that, for a laminar inlet boundary layer the two profiles have a similar loss distribution and structure of endwall flow. However, for a turbulent inlet boundary layer the two profiles are shown to differ significantly in both the total loss and endwall flow structure. The pressure side separation bubble on the solid thin profile is shown to interact with the passage vortex, causing a higher endwall loss than that measured on the hollow thick profile.

Research by de la Rosa Blanco et al. (“Influence of the State of the Inlet Endwall Boundary Layer on the Interaction Between the Pressure Surface Separation and the Endwall Flows,” Proc. Inst. Mech. Eng., Part A, 217, pp. 433–441) in a linear cascade of low pressure turbine (LPT) blades has shown that the position and strength of the vortices forming the endwall flows depend on the state of the inlet endwall boundary layer, i.e., whether it is laminar or turbulent. This determines, amongst other effects, the location where the inlet boundary layer rolls up into a passage vortex, the amount of fluid that is entrained into the passage vortex, and the interaction of the vortex with the pressure side separation bubble. As a consequence, the mass-averaged stagnation pressure loss and therefore the design of a LPT depend on the state of the inlet endwall boundary layer. Unfortunately, the state of the boundary layer along the hub and casing under realistic engine conditions is not known. The results presented in this paper are taken from hot-film measurements performed on the casing of the fourth stage of the nozzle guide vanes of the cold flow affordable near term low emission (ANTLE) LPT rig. These results are compared with those from a low speed linear cascade of similar LPT blades. In the four-stage LPT rig, a transitional boundary layer has been found on the platforms upstream of the leading edge of the blades. The boundary layer is more turbulent near the leading edge of the blade and for higher Reynolds numbers. Within the passage, for both the cold flow four-stage rig and the low speed linear cascade, the new inlet boundary layer formed behind the pressure leg of the horseshoe vortex is a transitional boundary layer. The transition process progresses from the pressure to the suction surface of the passage in the direction of the secondary flow.

In this paper, we investigate the feasibility of using DNS data and machine learning algorithms to assist RANS turbulence model development. High-fidelity DNS data are generated with the incompressible Navier–Stokes solver implemented in the spectral/hp element software framework Nektar++. Two test cases are considered: a turbulent channel flow and a stationary serpentine passage, representative of internal turbo-machinery cooling flow. The Python framework TensorFlow is chosen to train neural networks in order to address the known limitations of the Boussinesq approximation and a clustering based on flow features is run upfront to enable training on selected areas. The resulting models are implemented in the Rolls-Royce solver HYDRA and a posteriori predictions of velocity field and wall shear stress are compared to baseline RANS. The paper presents the fundamental elements of procedure applied, including a brief description of the tools and methods and improvements achieved.