This study is an experimental and numerical investigation of the flow field in the fully developed portion of a square duct with dimple and pimple turbulators applied to one wall. Two geometries are investigated at channel Reynolds numbers ranging from 160,000 to 275,000, based on a hydraulic diameter of 0.0508 m. The test section is 25 hydraulic diameters long. The smooth, unfeatured bottom wall and side walls are fabricated out of acrylic. The featured (top) walls are 3D printed using poly lactic acid (PLA). Both the positive and negative features have a spherical shape. Side walls are instrumented with pressure taps, and pressure drop measurements are obtained along the channel for both the geometries, in addition to the unfeatured top wall (completely smooth channel). It has been observed that Geometry 1 has ~8-13% higher pressure drop as compared to Geometry 2 for the current range of Reynolds numbers. Friction factor follows the same trend as the pressure drop. Particle image velocimetry (PIV) is utilized to obtain detailed measurements of the flow field in the region near the dimple and pimple features. Stereoscopic PIV measurements are taken on the spanwise plane at a location of 18.6 hydraulic diameters downstream of the inlet, where the flow becomes hydrodynamically fully developed. Average of various fluid flow quantities such as velocities, turbulent kinetic energy (TKE) is analyzed. PIV measured flow field in both the cases look quite similar qualitatively, and magnitude of velocities increases with increasing Reynolds numbers. However, peak TKE values are ~10-40% higher for Geometry 1 as compared to Geometry 2, and turbulence generated by the features on the top wall penetrates deeper into the core flow for Geometry 1 as compared to Geometry 2. Steady and unsteady RANS CFD simulations are performed for the same geometries with same flow conditions as in the experiment, and it is found that flow fields are comparable to PIV measurements. Pressure drop and friction factor are under-predicted by CFD simulations as compared to measurements in the experiments. Based on CFD calculations, Geometry 2 has higher heat transfer augmentation and lower friction factor augmentation as compared to Geometry 1 which results into 0-16.5% better thermal performance. PIV data is used to validate observed mean flow characteristics from CFD simulations giving confidence that the flow field very close to the wall is also well captured by the simulation, which is crucial for accurate prediction of wall heat transfer.
展开▼
