Thermal closure of the engine casing is widely used to minimize undesirable blade tip leakage flows thus improving jet engine performance. This may be achieved using an impingement cooling scheme on the external casing wall, provided by manifolds attached to the outside of the engine. The assembly tolerance of these components leads to variation in the standoff distance between the manifold and the casing and its effects on casing contraction must be understood to allow build tolerance to be specified.
For cooling arrangements with promising performance, the variation in closure with standoff distance of z/d = 1–6 were investigated. A cooling manifold, typical of that adopted by several engine companies, incorporating three different arrays of short cooling holes (chosen from previous study by Choi et al. (2016)) and thermal control dummy flanges were considered. A series of heat transfer tests using a transient liquid crystal technique were undertaken to measure spatially resolved heat transfer coefficient of a baseline sparse jet array. The experimental heat transfer results validated the extensive numerical predictions using RANS realizable k-epsilon turbulence model. The associated casing contraction was inferred from a finite element analysis using these distributions as the external casing thermal boundary condition. The flow in the system can be modulated to match the closure at different engine operating conditions, the relationship between thermal closure and coolant mass flow rates, inferred from the averaged jet Reynolds numbers assuming uniform distribution between cooling holes was predicted. Typical contractions of 0.5–2.2mm are achieved from the 0.02–0.35kg/s of the current casing cooling flows.
The variation in heat transfer coefficient observed with standoff distance is much lower for the sparse array investigated compared to a previous designs employing arrays typical of blade cooling configurations. The reason for this is explained through interrogation of the local flow field and resultant heat transfer coefficient. This implies acceptable control of the circumferential uniformity of case cooling can be achieved with relatively large assembly tolerance of the manifold relative to the casing.