The facility matched both engine im

The facility matched both engine impingement and mainstream Reynolds numbers, and provided Biot number similarity to within approximately 15%. The core wind tunnel flow was heated, whilst the coolant was fed at ambient temperature. Coolant mass flow rate was controlled via a flow regulator and critical flow Venturi nozzle, with the flow rate presented using the definition of non-dimensional coolant mass flow (m*) which is repeated in equation 33. Note that here, the value of h is that obtained using the average of the local Nusselt number obtained from the correlation in equation 34 [26]. The work in the current paper considers four double-wall geometries. Three of these shared a basic repeat unit (termed the unit-cell) comprising diamond pedestals. The fourth comprised a repeating unit with circular pedestals. The precise form of the repeating building blocks, along with specific dimensions, are provided in Figure 4. Fully conjugate CFD simulations were developed that modelled the experimental setup verbatim. Further elaboration on these can be found in [2]. Briefly, the simulations utilised boundary conditions obtained during the experimental runs. The meshes varied between 40 – 65 million elements. The simulations utilised the realisable k-ε turbulence model with enhanced wall treatment and, like the experiments, were steady- state in nature. As shown by Figure 5, which demonstrates area- averaged film wall overall effectiveness, the conjugate simulations very successfully captured the solid temperature profiles in the double-wall geometries thus validating the CFD procedure.