Low Speed Wind Tunnel Flow Diagnostics and Benchmark Cases for Thermal Fluids CFD Validation Efforts

Abstract

A low speed, closed loop wind tunnel at Texas A&M University is presented for the study of turbulent mixing produced by a variety of flows types. Anticipated experiments range from canonical “unit flows” to more complex combinations of flows and geometries. Originally located at the University of Pittsburgh, the facility has since been re-located to the Thermal Hydraulics Verification and Validation (THVV) laboratory at Texas A&M University. The tunnel has undergone considerable modification and updated diagnostics prompting renewed interest in flow quality assessment. This includes a thorough mapping of the tunnel inlet velocity profile provided by Particle Image Velocimetry (PIV) measurements. Additional temperature and gage pressure measurements complete the assessment of system capabilities. These preliminary diagnostics yield empirically determined boundary conditions and fluid property correlations necessary for Computational Fluid Dynamics (CFD) model validation. The article concludes with the presentation of two unit flow types, including flow past a cylinder, with three distinct cross sections, and a single round jet in cross flow at three velocity ratios. The unit flows serve as initial benchmarks for THVV simulation efforts. Key validation metrics are presented for each benchmark including ensemble averaged velocities, Reynolds stresses, and proper orthogonal decomposition (POD) eigenvectors.

Towards Reconstruction of Complex Flow Fields Using Unit Flows

“…Instantaneous and time-averaged velocity magnitude profiles of the xz plane at y/D = 1 have been observed previously in [15,44] and are reproduced here in Figures 3 and 4. The single jet profiles in Figure 3a-c appear amorphous, as a result of the aggregation of Both the crossflow and jets are seeded sequentially using a TSI Six Jet Atomizer (TSI Incorporated, Shoreview, MN, USA) with Dioctyl Sebacate tracer particles.…”

“…Instantaneous and time-averaged velocity magnitude profiles of the xz plane at y/D = 1 have been observed previously in [15,44] and are reproduced here in Figures 3 and 4. The single jet profiles in Figure 3a-c appear amorphous, as a result of the aggregation of several mechanisms induced by the two perpendicular flow streams and boundary layer interactions with the wall leading up to the jet.…”

“…The composites, however, must also consider the effect of segmenting and weighting the data. The authors have previously applied POD to the original single jet in crossflow data sets [15,44] and found the corresponding τ values to be nearly identical to that of the tandem jet, thereby confirming that the major culprit here is the segmentation and/or weighting. These cutoff values, however, do not take away from the success of the composite in replicating the eigenvalue magnitudes, cumulative energy distributions, and potential compression of the measured data set as displayed previously in Figure 5 and here in Table 2.…”

“…Full details regarding the experimental arrangement have been outlined previously [15,44] with pertinent details reviewed here. The experiment consists of a low speed wind tunnel test section with three round jets in a tandem formation issuing vertically downward from the ceiling of the test section as outlined in Figure 2.…”