This content is not included in your SAE MOBILUS subscription, or you are not logged in.

Validation of a 1D Compressor Model for Performance Prediction

Journal Article
ISSN: 1946-3936, e-ISSN: 1946-3944
Published September 08, 2013 by SAE International in United States
Validation of a 1D Compressor Model for Performance Prediction
Citation: De Bellis, V., Bozza, F., Bevilacqua, M., Bonamassa, G. et al., "Validation of a 1D Compressor Model for Performance Prediction," SAE Int. J. Engines 6(3):1786-1800, 2013,
Language: English


In the present paper, a recently developed centrifugal compressor model is briefly summarized. It provides a refined geometrical schematization of the device, especially of the impeller, starting from a reduced set of linear and angular dimensions. A geometrical module reproduces the 3D geometry of the impeller and furnishes the data employed to solve the 1D flow equations inside the rotating and stationary ducts constituting the complete device. The 1D compressor model allows to predict the performance maps (pressure ratio and efficiency) with good accuracy, once the tuning of a number of parameters is realized to characterize various flow losses and heat exchange.
To overcome the limitations related to the model tuning, unknown parameters are selected with reference to 5 different devices employing an optimization procedure (modeFRONTIER™). A unique set of tuning constants is identified by the optimizer, which is able to provide, in most of cases, a good agreement with the experimental maps of each compressor.
To verify the reliability of the previously selected parameters, the tuned model is employed to compute the performance map of an additional compressor of different geometry, not included in the tuning procedure. The predicted map shows a good agreement with the experimental one.
The proposed model can hence be very helpful in providing with good accuracy a first estimate of the performance maps of a centrifugal compressor, only basing on its geometry, and in absence of manufacturer data.