Detailed velocity and heat transfer measurements of an advanced insert for impingement cooling

This work reports the results of paired experiments for the aft section of a complex internal cooling flow within a gas turbine vane using Magnetic Resonance Velocimetry (MRV) and steady-state Infrared (IR) thermography. The aft cooling insert for the vane with two-pass impingement was designed at a scale five times larger than the original and built using stereolithography (SLA) fabrication methods. The MRV technique was used to measure the three-dimensional, three-component velocity field for a test case with a Reynolds number range of 2,600 to 7,600 based on diameter of the impingement holes. Flow distribution, impingement hole performance, and cross flow effects are discussed for the experiment in which a dilute aqueous copper sulfate solution was used as the working fluid. A paired experiment with a geometrically similar design employed electrical heating of a thin stainless steel shim to model a constant heat flux boundary condition of an interior wall of the turbine vane. The modeled vane insert was then operated with air as the working fluid at two test conditions with Reynolds numbers in the range of approximately 1,200 to 7,600 based on the diameter of the impingement holes. An IR camera was used to measure the surface temperature of the shim. Using energy balances and the known heat flux, the temperature data were used to determine heat transfer characteristics of the impinging jets for the pressure and suction side surfaces, including the Nusselt number. The MRV and IR data sets provide detailed insight into the surface effects of the flow distribution and the result on the local and area-averaged heat transfer performance. A strong coupling between the velocity field and temperature data provide insight into design feature performance, and serve as a validation data set for matched computational simulations. Finally, a comparison with internal heat transfer correlations is presented using the data from Florschuetz et al. [1]. The results showed a lack of agreement with Florschuetz, leading to the development of a novel methodology for estimating the heat transfer performance of an impinging hole with crossflow. Measurement uncertainty was estimated to be ±5% for velocity and ±5% for the spatially averaged Nusselt number distributions.