Boundary layer, grid turbulence, and periodic wake effects on turbulent juncture flows

The spatial-temporal flow-field and associated surface heat transfer for a turbulent juncture flow were examined using a technique which combines Particle Image Velocimetry (PIV) and a Liquid Crystal (LC) based heat transfer measurement technique. Turbulent boundary layer “burst” events, which have been previously described as hairpin vortex “packets”, interact and periodically strengthen the Horseshoe Vortex (HV). The strengthening of the HV precipitates its movement upstream and toward the surface that induces the ejection of a Secondary Vortex (SV), following which the HV eventually weakens and repeats the cycle quasi-periodically. The dominant period of motion for the HV was qualitatively and quantitatively linked to the impinging turbulent boundary layer bursting frequency. Endwall surface heating is shown to increase the bursting frequency and thus the HV frequency.;The time-mean saddle point that develops on the approach endwall surface is determined to be a point of separation; however, the instantaneous saddle point can be, either a point of attachment or separation depending on the position and strength of the HV. Increasing endwall surface heating causes the saddle point to move downstream, closer to the bluff body.;Elevated free stream turbulence levels reduce the time-mean strength of the HV and can disrupt the quasi-periodic behavior of the HV, although the undisturbed HV frequency behavior can still be detected for approach turbulence intensities below 11%. Similarly, impinging periodic wakes are capable of eliminating the HV periodicity. However, when either the Strouhal number or turbulence intensity are sufficiently low, the undisturbed periodic HV behavior is still present. Impinging wakes also increase the juncture region surface heat transfer very near the cylinder by inducing short, high-energy down-washes along the face of the cylinder.;The temporal heat transfer behavior for a linear turbine cascade juncture was observed to be essentially the same as that for a simple bluff body, with impinging boundary layer bursting frequency again being the controlling parameter. The linear cascade also displays a similar temporal heat transfer behavior adjacent to the trailing edge of the pressure side of the blades, although a clear relationship to the leading edge behavior was not established.