Experimental flow data recorded on a low speed centrifugal compressor rig are compared with theoretical predictions of the diffuser field. Results for four diffuser geometries are presented covering variation of the blade number, stagger angle, and radial separation of the rotor and diffuser. The theoretical predictions are based on an extension of the Martensen surface singularity theory. The diffuser is found to play a major role in establishing the circumferential variation in the mean flow at impeller outlet. Large pitch wise variations are observed, which may be explained by consideration of the flow round the leading edges of the diffuser blades. Experiments and potential flow theory show good agreement, except for cases where the diffuser shows signs of separation. These are clearly identified from the theoretical predictions by Carter's lift coefficient criterion. Although the absolute flow at rotor outlet is markedly time-dependent owing to the passage of the rotor, it would appear that the averaged flow conditions may be predicted by the steady-state singularity program. Unstalled diffuser flows are accurately predicted and stalled flows, which are generally to be avoided, may be anticipated from the lift coefficient criterion. The designer thus has the option of reducing pitchwise velocity and pressure variations at rotor outlet to acceptable amplitudes.

The Annual EPSRC/IMechE Expert Meeting brought together some 44 experts to consider sources of uncertainty in computational fluid dynamics (CFD). Presentations and discussions covered modelling, numerical solution techniques, boundary conditions, evaluation protocols and QA (quality assurance) procedures. The principal conclusions to emerge were: (a) the need for additional collaborative validation studies; (b) the desirability of introducing appropriate QA procedures, possibly based on the CFD Community Club initiative; (c) the need for additional postgraduate training, possibly based on the IGDS principle; (d) the value of continuing work in modelling and error estimation techniques for numerical schemes.

Aeroderivative gas turbines intercooled by water spray injection systems have recently entered service. A calculation framework is presented which permits the effect of water spray injection for both intercooling and inlet chilling to be evaluated and compared with conventional cooling techniques for a range of cycles. The calculations are based on representative performance maps for compressors and turbines, focusing upon the actual performance benefits that can be realized from existing turbomachinery. The principal conclusions are that:(a) spray intercooling can give a greater power boost than conventional intercooling for a given compressor operating envelope, (b) spray inlet chilling can give performance benefits comparable with absorption chilling and (c) intercooling is more attractive at ambient temperatures below 15°C, whereas inlet chilling is preferable at higher temperatures.

No abstract

The Fifth Joint Engineering and Physical Sciences Research Council and Institution of Mechanical Engineers Expert Meeting was held in Bournemouth on 27-29 November 1994. The Fifth Joint Engineering and Physical Sciences Research Council and Institution of Mechanical Engineers Expert Meeting was held in Bournemouth on 27–29 November 1994.