The paper describes experimental and computational fluid dynamics analyses of the non-uniform static pressure distortion caused by the discharge volute in a high pressure, centrifugal compressor. The experiments described in this paper were done using a heavily instrumented gas re-injection compressor operating at over 6000 psia discharge. Instrumentation was installed to measure static, total, and dynamic pressure as well as impeller strain and mechanical vibrations. A brief description of the compressor and instrumentation are provided. Concurrent with the experimental work, CFD runs were completed to study the reasons for the pressure non-uniformity. The CFD pressure profile trends agreed well with the experimental results and provided analytical corroboration for the conclusions drawn from the test data. Conclusions are drawn regarding: a) the response of the non-uniformity to changing flow rates; b) the extent to which the non-uniformity can be detected upstream of the impeller; and c) the mechanical influences of the non-uniformity on the impellers.

This paper reviews test results from a rotatable low solidity vaned diffuser (RLSD). The device was installed in a single stage test vehicle consisting of a pseudo-return channel inlet, a subsonic centrifugal impeller (approx. 2:1 pressure ratio), the rotatable diffuser, a return channel and a dump collector. Static taps, total pressure probes, and thermocouples were located in critical areas throughout the stage. Manual traverse probes measured the pressure and angle profiles at RLSD leading and trailing edges. Results for various stagger angles, leading edge radius ratios, etc. are presented in terms of pressure recovery (Cp) and loss coefficient (LC). Comments are made regarding the applicability of the RLSD in production units.

This paper addresses the use of 5-hole probes in the testing of industrial centrifugal compressors. The 5-hole probes utilized for this work are of the conical-tip type and were used in a non-nulling configuration (i.e., the probes do not need to be rotated or moved in any way during the tests). These 5-hole probes proved to be fairly robust, making them practical for a nonlaboratory setting such as an industrial multistage compressor test stand. A discussion of 5-hole probes and how they function is provided, including an overview of the mathematical formulations and calibrations required to translate the pressure data gathered from the 5 holes into static and total pressures, velocities and flow angles. A method to transform these variables from a probe-based coordinate system to a machine-based coordinate system is also presented and schematics of this process are provided to aid the reader’s understanding. The testing performed on a prototype multistage centrifugal compressor using 5-hole probes is also discussed, showing that the probes provided valuable insight into the flowfield exiting the impellers and at the return bend. The hub-to-shroud velocity profile exiting an impeller was found to be more skewed than expected and was contributing to poor performance in the downstream stationary components. The measured flowfield from one of the tests is also compared against 3-D CFD results and comments are offered regarding the agreement between the analytical and measured results. Advantages and disadvantages of 5-hole probes as compared to more conventional instrumentation are presented. Finally, conclusions are drawn regarding the value of 5-hole probe data in the development and/or troubleshooting of high performance turbomachinery and in the validation/calibration of design and analysis tools.

The paper is a sequel to an earlier work by Sorokes et al. 1998, “Investigation of the Circumferential Static Pressure Non-Uniformity Caused by a Centrifugal Compressor Discharge Volute.” The earlier work described experimental and computational fluid dynamics analyses of the non-uniform static pressure distortion caused by the discharge volute in a high pressure, centrifugal compressor with vaneless diffusers. This paper describes additional testing and analytical work done using low solidity vaned diffusers (LSD’s) in place of select vaneless diffusers to determine the alternate diffuser’s effectiveness in eliminating or reducing the magnitude of the non-uniform pressure field. As in the earlier studies, the experiments described in this paper were done using a heavily instrumented gas re-injection compressor operating at over 6000 psia discharge pressure. Instrumentation was installed to measure static, total, and dynamic pressure as well as impeller strain and mechanical vibrations. A brief description of the compressor and instrumentation are provided. Concurrent with the experimental work, CFD runs were completed to study the effect of the alternate vaned diffusers. The CFD pressure profile trends agreed well with the experimental results and provided analytical corroboration for the conclusions drawn from the test data. Conclusions are drawn regarding: a) the effectiveness of the LSD’s on the pressure non-uniformity; b) the associated effects on the measured dynamic strains in the impellers; and c) the usefulness of computational fluid dynamics (CFD) in assessing the aerodynamic forces associated with the non-uniformity.

The paper addresses the use of a rib style (partial height) vaned diffuser to improve the flowfield downstream of a high flow coefficient centrifugal impeller. Empirical and analytical (3-D CFD) results are presented for both the original vaneless diffuser and the replacement rib configuration. Comparisons are made between the CFD results and the data obtained through single stage rig (SSTR) testing. Comments are offered regarding the qualitative and quantitative agreement between the empirical and analytical results.