Low-speed centrifugal compressor

Application Challenge 6-02 © copyright ERCOFTAC 2004

Introduction

The LSCC is an experimental facility designed to duplicate the essential flow physics of high-speed subsonic centrifugal compressor flow fields in a large size, low-speed machine.

A complete description of the experimental facility for the LSCC case is provided by Hathaway et al. (1993, 1995). A considerable amount of data including static pressure profiles, total temperature and total pressure profiles, flow angles are provided at different span locations as well as static pressure contours on blade surfaces, velocity vectors near blade surfaces, throughflow velocity contours at different streamwise stations.

Relevance to Industrial Sector

The LSCC application challenge is a state-of-the-art flow problem that addresses the impact of physical models and grid point distribution on the accuracy of the predicted flow solution which is dominated by complex secondary flows and wakes.

Design or Assessment Parameters

The design or assessment parameters are represented by two indicators, the overall pressure ratio and the adiabatic efficiency. The DOAPs are detailed below:

• The overall pressure ratio: This DOAP is calculated from the plenum total pressure (equal to the static pressure at standard conditions p_std) and the energy-average of the pitch averaged spanwise total pressure distribution at survey station 2 (Figure 2).

Pressure ratio=<P_t>/p_std

• The adiabatic efficiency: This DOAP is calculated based on the average total pressure at station 2 and the mass average of the pitch averaged spanwise distribution of the total temperature at station 2. Pressure and temperature at standard conditions are used as reference values:

\eta _{\text{ad}}=\left({(<P_{t}>/P_{\text{std}})}^{\gamma /\gamma -1}-1\right)/{(<T_{t}>/t_{\text{std}}-1)}

p_std = 101325 Pa

t_std =288.15 K

Flow Domain Geometry

The test impeller (see figure1) is a backswept impeller with a design tip speed of 153 m/sec. The impeller has 20 full blades with a backsweep of 55°. The inlet diameter is 0.870 m and the inlet blade height is 0.218 m. The exit diameter is 1.524 m and the exit b1ade height is 0.141 m. The clearance between the impeller blade tip and the shroud is a constant 2.54 mm from the impeller inlet to the impeller exit. This clearance is 1.8 percent of the blade height at the exit of the impeller. The blade surfaces are composed of straight-line elements from hub to tip. This feature allowed the laser anemometer optical axis to be directed parallel to the blade surface, thereby facilitating laser anemometer measurement of velocities close to the blade surfaces. The impeller is followed by a vaneless diffuser that generates an axisymmetric outflow boundary condition, which is desirable for CFD analysis of an isolated blade row. The original vaneless diffuser was modified to eliminate a region of reverse flow that occurred on the back wall of the diffuser (Hathaway, Wood, and Wasserbauer (1992) as seen in figure2. This modification ensured that there would be no backflow at stations downstream of the impeller.

The full geometry can be obtained at Blade Geometry definition and Hub and Casing Geometry definition courtesy of Numeca Int.

Figure 1. - The test impeller

Figure 2.- Meridional view of the LSCC flow path Station 0: z=-76.581 cm; Station 1: z=-20.373 cm; Station 2: r=81.28 cm; Station 3: r=167,64 cm

Flow Physics and Fluid Dynamics Data

A significant progress in the understanding of centrifugal compressor flows has been obtained in last decades, from the jet/wake concept introduced by Dean and Senoo (1960) and Dean (1971), and confirmed later by Moore (1973), Eckardt (1976), and Johnson and Moore (1983a). This model recognizes that viscous and secondary flow phenomena have turned the largely inviscid dominated inlet flow field into an exit flow where high velocities are found along the pressure surface and lower velocities appear in the shroud/SS region. It is found, however, that the position of the wake (low energy region) is not always located in the shroud/SS corner but changes with the impeller geometry and Rossby number. As reported by Eckardt (1979), the wake in an impeller with a radial exit angle is in the shroud/SS corner, while for an impeller with backswept blades it is away from the corner. Johnson and Moore (1983b) found, from their experimental studies in a shrouded impeller with radial discharge, that the wake is shifting towards the pressure side with increasing flow rate. A similar observation has been reported by Chriss et al. (1996), based on their experimental and numerical data. Hirsch et al. (1996 a and b) concluded, in analyzing the numerical and experimental data of a high subsonic compressor impeller and a low speed pump impeller, that the location of the wake results from a balance between the various secondary vortices, and the tip leakage flow. A sketch of the different components of the secondary vortices is shown in Figure 3.

It consists essentially of the endwall passage vortex, PV, generated by the end wall boundary layers and the blade-to-blade streamline curvature, the blade surface vortex, BV, due to the meridional curvature of the flow channel and the blade surface boundary layers, and CV due to the Coriolis force and the endwall boundary layers in the radial parts. However, the wake position variation with geometry and Rossby number remain unclear, especially in an unshrouded impeller.

Figure 3. Combination of passage and blade surface vortices, after Hirsch et al. (1996b) PVH&PVS: passage vortices at shroud and hub

BVS&BVP: blade surface vortices at SS and PS

CV : Coriolis passage vortex

The influence of flow rate on the three-dimensional viscous flow in a centrifugal impeller with tip clearance, by means of an analysis of the experimental and numerical results of the NASA Large Scale Centrifugal Compressor (LSCC) impeller with vaneless diffuser has been described by Kang and Hirsch (1996, 1999a, 1999b, 2001). Emphasis is put on the influence of the impeller geometry and the Rossby number on the position of the throughflow wake in the discharge flow.

The governing non-dimensional parameter (GNDP)and fluid properties:

The working fluid is air and the Sutherland law was used to evaluate the viscosity. The specific heat at constant pressure and the specific heat ratio are:

C_p = 1004.5 J/kg/K

γ=1.4

The governing non-dimensional parameters are:

• Prandtl number : Pr=0.72

• Reynolds number: Re = 1.6 107(based on tip speed, 153 m/s, and tip radius:1.524m)

• Mach number : M= 0.45 (based on tip speed and standard conditions)

Contributors: Nouredine Hakimi - NUMECA International

Site Design and Implementation: Atkins and UniS