|
Highly compressible pulsating flows are often encountered in devices where knowledge of the flow rate is required but elimination of pulsations is not an option. The current work is a continuation of a previous investigation that characterized the orifice discharge coefficient C d as a function of dimensionless groups based on pulsation characteristics. The experimental apparatus has been rebuilt in the current work to mitigate temperature and vibration problems, allowing pressure and ΔP measurements to be made very close to the test section with 159-mm of nylon tubing. Data was acquired for 77 operating conditions spanning a range of pulsation frequencies, mass flow rates and system pressures. They confirm previously reported low C d 's in 0.20 range (calculated from time-averaged pressures) at some high-pressure low-flow operating conditions. Computational Fluid Dynamics (CFD) simulations of 12 of these data points suggest that the low C d 's result from reverse flow. Flow direction changed several times during each pulsation cycle closely tracking the orifice ΔP. A ‘core-and-sheath’ phenomena was observed for reverse-flow operating conditions: a positive core flow surrounded by a sheath of negative flow transitioned to a negative core and positive sheath several times during each pulsation cycle. The simulations also suggested that velocity profiles at the orifice stay stable and similar to steady-state profiles except for periods of rapid transitions. Based on these results a data-based quasi-steady method of estimating pulsating flow has been proposed. A pair of forward and reverse flow C d 's chosen by the data are used to predict instantaneous forward and reverse flows using the steady-state orifice discharge equation for compressible flow. The instantaneous values are then summed up over the pulsation cycle to estimate average mass flow rate. Average prediction errors were within 6%. A previously proposed method where regression was used to model C d as a function of dimensionless groupings was shown to produce similar results. Both methods are designed to extract information from experimental data in order to overcome theoretical limitations and experimental error. The data is available upon request for further understanding of the flow physics.
|
