This work is the result of an investigation based on numerous test data supplied by major compressor manufacturers in USA and in Europe. The main objective of the work is to propose improved formulae for the correction of the efficiency, the head, and the flow as influenced by the Reynolds number variation between workshop tests and specified conditions, carried out with the same machine. Tests on hand have shown that a sufficiently good correlation between measured and predicted values can be obtained with the proposed formulae. In addition a proposal is made for the allowable range, taking into account the inherent limitations for accurate testing at low Reynolds numbers. As a conclusion to this study it is recommended that the proposed formulae and allowable range be reviewed by the ASME, the ISO, or any other appropriate associations for adoption in revised test codes for centrifugal compressors.

No abstract

The results of the investigation of the flow in a transparent (quartz tube) downscaled (≈1:3) model of a lean-premix type gas turbine combustion chamber are presented and discused. The model was tested at atmospheric pressure in reacting conditions; flow measurements were taken by a two-channel fiber-optic laser doppler velocimeter, using Al2O3 seeding of the air flowrate. The measurements cover a wide flow field inside the combustion chamber, including flame development and recirculating regions. Long-time samples (10–20 s) were used in order to achieve a good accuracy in the measurement of average flow conditions over the whole flow field; this involved a limited capability of representation of high-frequency components of turbulence, which could be locally obtained with optimization of the data rate and seeding conditions. Fast measurements were also locally performed where the seeding conditions were favourable. Integral variables and power spectra for reacting conditions show some distinctive aspects for the turbulence structure of reacting turbulent flows in confined spaces. Further measurements cover the outlet throat section of the premix combustor, demonstrating the persistence of a radial flow component on account of wall curvature effects and a certain degree of asimmetry in the inlet velocity distribution.