A Novel Facility for Low-Cost Aerodynamic Testing of Single-Stage Centrifugal Compressors, 18-R8024Printer Friendly Version
Inclusive Dates: 01/01/09 10/30/09
Background - Experimental testing plays a vital role in developing high-performance centrifugal compressors. Despite the vast advances in computational fluid dynamics (CFD) and decades of design experience, the complex nature of the flow in centrifugal compressor components continues to be a challenge. The preferred design approach is a combination of CFD analysis, traditional design strategies and high-quality experiments. However, testing in traditional experimental facilities tends to be expensive, often limiting the number of experiments to less than the desired amount because of cost. Traditional tests are expensive, in part, because of the means used to measure compressor power, which involves temperature rise measurements. The need to measure the temperature rise with sufficient accuracy to calculate the performance with high fidelity generally means that test facilities must operate at elevated pressures and circulate the working fluid in a closed loop. These features, as well as the generally large amount of power needed to drive the compressor when pressures are elevated, raise the cost of experimental testing considerably. In addition, the high pressures and loads on the components require the compressor parts to be machined as if they were going into a production unit, even though they will only be used for a few hours.
Approach - The focus of the current project is developing a facility that will allow high-quality testing to be done at relatively low cost in a small, low-power test stand that uses atmospheric air as the working fluid. The approach to achieving this is two-fold. First, an innovative means of direct measurement of the compressor power input is incorporated in this design (an invention disclosure is in preparation). This method of measuring power does not involve the traditional temperature rise measurement and, hence, can dispense with the need for operation at elevated pressure in a closed loop. This innovation is a core technology of this facility. Second, the low pressures and resulting lower power levels involved will allow the non-rotating components of the experimental compressor stage to be made from plastic material in a rapid-prototyping facility. A side benefit of dispensing with a pressurized closed-loop is that it will allow easier access to the test section for making geometry alterations and working on instrumentation, leading to faster turnaround time. The combination of these benefits should lower testing costs considerably, bringing high-quality compressor testing into the realm of affordability to customers with limited R&D budgets.
Accomplishments - Considerable progress in the design of the test stand has been made during the reporting period. The facility will be used with a range of compressor diameters, rotating speeds and flow rates. The wide variation of these parameters is driven by the varying needs of industrial applications that employ centrifugal compressors. Sound test rig design requires rotordynamic stability. In the design of a facility that will accommodate the desired range of machine sizes, one needs to furnish a rotor shaft and bearings that are narrow enough to accommodate the aerodynamic flow path for the smaller impellers. At the same time, it needs to retain rotordynamic stability when fitted with impellers of various weights and inertial properties. A satisfactory design of rotor shaft and bearings has been completed. This is an important accomplishment because the rotor shaft and its associated supports dictate the dimensions of much of the rest of the design. Conceptual design and layout of the major components of the facility are shown in the illustration. In addition, good progress has been made in obtaining the necessary data acquisition system and in programming it for compressor testing. Finally, analytical tools required for the design of compressor components and for enhanced interpretation of experimental results have also been worked on during the reporting period.