Experimental/computational study of viscous flow in a contracting rectangular elbow

Abstract

Experimental results have been obtained for flow through a contracting, 90-deg, rectangular Stanitz elbow with thickened endwali boundary layers to generate passage vortices and strong secondary flow. This includes Pitot-static and LDV measurements for inflow boundary conditions. The data were used as the basis for three- dimensional viscous code validation based on a two-equation q-u turbulence model using both integration to the wall and a wall-function treatment. Wall functions were found to be reliable for the complex three-dimen- sional flow considered. Eddy viscosity well away from walls appears to be over predicted. HE Stanitz elbow was originally introduced to test certain aspects of secondary flow development in a turbine vane, namely, that associated with passage vortex formation and intensification in a contracting turn.1 The geometry was de- signed to avoid adverse pressure gradients and hence the detri- mental influence of separation. Experiments were conducted with the configuration shown in Fig. 1, including the option of spoilers upstream to simulate thickened endwali boundary layers which are characteristic of turbomachiner y.2 Subse- quent studies on this rectangular duct at our organization will be the focus of this paper. The experiment has been repeated to broaden the database, including wall pressure measurements on the endwali and operation at high subsonic Mach number. Definition of inflow boundary conditions is also emphasized due to significance of these in code validation. Computational studies assess model- ing capability for a complex three-dimensional turbulent flow- field. This is based on Coakley's two-equation g-u model,3 using both integration to the wall and a wall-function treat- ment. The latter performs well even in the presence of strong secondary flow, and could lead to significant reduction in gridding requirements for three-dimensional viscous cascade flow simulation. Representative results from this combined experimental/theoretical study are summarized below. Code Validation Experiment The experiments were conducted in the Aerodynamic Test Facility4 at Avco Research Laboratory (ARL). This facility can be operated in two modes: continuous operation at mass flow of up to 2.3 kg/s at 9.5 atm, or blowdown operation with mass flow up to 18 kg/s at 32 atm for 20 s test duration. The facility is fully computer controlled, including automated traversing equipment for Pitot-static and .Pilot probes, and is capable of aggregate data-collection rates of 1 MHz into a MASSCOMP computer. This allows extensive data collection in a matter of seconds and display of results to assess data quality in minutes. Reliable and repeatable blowdown perfor- mance was established by conducting the Stanitz elbow experiment, an underlying reason for choosing a configura- tion that had been studied previously.