For engines operating at high turbine entry temperatures it is increasingly important to cool the high pressure nozzle guide vane (HP NGV) endwalls. This is particularly so for low NOx combustors operating with flatter outlet temperature distributions. Double-row arrangements of film/ballistic cooling holes upstream of the NGV passage have been employed in production engines. Optimisation of such systems is non-trivial, however, due to the complex nature of the flow in the endwall region. Previous studies have reported that strong cross passage pressure gradients lead to migration of coolant flow and boundary layer flow within the passage. In addition the vane potential field effects lead to non-uniform blowing ratios for holes upstream of the vanes. It has also been reported that inlet total pressure and turbulence profiles have a significant effect on the development of the film cooling layer. In this study, endwall film cooling flows are studied experimentally in a large-scale low-speed cascade tunnel with engine-realistic combustor geometry and turbulence profiles. At very low blowing ratios mild cross-passage migration effects are observed. At higher blowing ratios more realistic of the engine situation no cross-passage migration is observed. This finding is somewhat contrary to the classical view of endwall secondary flow, which is presented as significant at the scale of the vane passage by several authors. The difference arises in part because of the thinning of the boundary layer due to strong acceleration in the vane inlet contraction. The findings are further supported by CFD simulations. Methods of improving conventional double-row systems to offer improved cooling of the endwall are also discussed.

Progress in the computing power available for CFD predictions now means that full geometry, 3 dimensional predictions are now routinely used in internal cooling system design. This paper reports recent work at Rolls-Royce which has compared the flow and htc predictions in a modern HP turbine cooling system to experiments. The triple pass cooling system includes film cooling vents and inclined ribs. The high resolution heat transfer experiments show that different cooling performance features are predicted with different levels of fidelity by the CFD. The research also revealed the sensitivity of the prediction to accurate modelling of the film cooling hole discharge coefficients and a detailed comparison of the authors’ computer predictions to data available in the literature is reported. Mixed bulk temperature is frequently used in the determination of heat transfer coefficient from experimental data. The current CFD data is used to compare the mixed bulk temperature to the duct centreline temperature. The latter is measured experimentally and the effect of the difference between mixed bulk and centreline temperature is considered in detail.

An instrument for the measurement of heat flux from a surface with a nearly uniform temperature is described. This instrument contains a thin-film electrical resistance heater embedded in a copper cone which is thermally isolated from the surrounding walls. A differential thermocouple between the copper cone and the wall is nulled such that the electrical power becomes a direct measure of the surface heat flux. The advantage of this design over earlier sensors is its modular characteristic and its ability to be flush mounted in an external surface or mounted in the wall of a duct. It has been used to measure the local time-average heat transfer coefficient inside a circular duct. The time constant in this application was 43 s. For these measurements an uncertainty analysis is presented which shows that this instrument has an uncertainty of ±3.6% for a convective heat flux of 342 W/sq m. The major source of uncertainty was the surface area.

In this paper we experimentally evaluate the impact of in-service deterioration on the aerodynamic performance of heavily film cooled high-pressure nozzle guide vanes from large civil jet engines. We study 15 mid-life to end-of-life parts removed from operational engines, and compare their performance to those of new parts. Deterioration features included: increased surface roughness; thermal barrier coating spallation; damaged film cooling holes; and trailing edge burn-back. We characterize and present statistics for the surface roughness. Aerodynamic measurements were performed in the high technology readiness level Engine Component AeroThermal (ECAT) facility at the University of Oxford, at engine-representative conditions of exit Mach number, exit Reynolds number, coolant-to-mainstream pressure ratio and turbulence intensity. We present detailed experimental measurements of the coolant capacity characteristics, downstream loss, and downstream flow structures. The results show that service time significantly increases the surface roughness of high-pressure nozzle guide vanes, leads to reduced coolant flow capacity, increased overall aerodynamic loss, and greater downstream flow angle variation. This is one of the first significant studies of its type in the open literature, and is an important step towards whole-life engine performance assessment.

No abstract