European Turbomachinery Conference 2013

08 April 2013

The 10th European Conference on Turbomachinery Fluid Dynamics and Thermodynamics ETC 10 ( will be held on April 15 – 19, 2013 in Lappeenranta, Finland. PCA Engineers will be represented by Professor Mick Casey, who has co-authored three papers.  


M. V. Casey, D. Krähenbuhl, C. Zwyssig

This paper describes recent experience in the development and application of several ultra-high-speed miniature centrifugal compressors with an impeller diameter less than 30 mm and using high speed electric motors to provide rotational speeds between 200,000 and 600,000 rpm. In a growing number of applications at low flow rates such micro-compressors can be successfully used to replace much larger positive displacement devices or to replace larger centrifugal compressors operating at lower rotational speeds. A range of applications for heat pumps, fans and general air supply are considered in which the typical features of the stages developed are the low size, very low mass flows and high speeds. Dimensional analysis is used to show how the scaling laws for the compressor match those of the motor. These also show that a range of designs of different design styles are required; from low flow coefficient stages with two-dimensional geometry to high flow coefficient mixed flow stages. In addition to a description of the applications, and the stages developed for these, some information is provided on the design strategy, design tools and performance prediction methods used in the design process. Test data from a range of devices demonstrates that an overall efficiency around 65% can be achieved. The measured performance is shown to agree well with the predicted performance of the stages to validate the design techniques used.


Fabian Dietmann and Michael Casey

An empirical correlation (due to Casey and Robinson (2011)) for the effect of Reynolds number and roughness on the performance of compressors is examined and refined. The original correlation quantifies the inefficiency due to friction losses with an empirical coefficient Bref which differs for different machine types. The variation of this coefficient was originally given as a function of the specific speed and determined from extensive test data for many compressors, in which the Reynolds number, roughness and size were varied. In this paper the value of the empirical coefficient is determined from a theoretical analysis and the experimental data is reanalysed and extended to provide an improved correlation.


Two theoretical approaches are given. The first determines the inefficiency using a method due to Traupel to give the frictional dissipation losses as a function of flow coefficient for a range of radial compressor stages. The second uses a well-known correlation for global efficiency, which is corrected for the non-frictional losses to derive the inefficiency due to the friction losses alone. The original test data and some additional data has also been analysed with a slight modification of the equation for the friction factor as a function of Reynolds number and roughness, and this removes some of the scatter in the experimental data.


On this basis three independent methods for determining the relevant coefficient are now available. A suggestion is made for an improved correlation based on these three approaches.

A.Grönman, F. Dietmann, M. Casey,  J. Backman

Low solidity diffusers are sometimes used in radial compressors where they promise a wide operating range close to that of vaneless diffusers but with efficiency levels that approach those of vaned diffusers. A literature survey has identified that despite the large number of papers on this subject there are no simple guidelines available for their design.However, general guidelines can be drawn for some of the design parameters from the available research data. Additionally, this paper proposes some simple design guidelines based on classical diffuser design parameters, such as length-to-width ratio and area ratio, whereby these parameters need to be specially defined in the case of very low solidity diffusers where the blades do not overlap. Experimental diffuser data is supported with 1D diffuser calculations to suggest suitable design guidelines.