The focus of this paper is to report on measurements from and simulation of Cranfield University’s 3-stage high-speed axial compressor test rig. This newly built rig is supported by European Commission funding and has tested a set of conventionally stacked 2D rotor and stator blades (Reference 1). The results were used to evaluate and to assess the performance of several commercially available CFD codes leading to the collaborative design of an advanced three-dimensional blade set. The philosophy behind the advanced design is described. The datum test results show that the state of the art, highly loaded, datum compressor is well matched with limited potential for loss reduction. A comparison is made between the measured results and a series of numerical analyses using the various CFD codes. Although the codes showed reasonable qualitative agreement with each other and the measured data, there were significant differences in the predicted performance of the datum build. Further the codes were unable to grade candidate redesigns consistently on a quantitative basis and therefore increased the difficulty of selecting suitable ‘3d’ features. Generic studies involving sweep, lean and recambering are used to evolve a design philosophy for the advanced three-dimensional design. Over cambering of the end-wall sections, coupled with a suitable stack of the blades, enables the blade count to be reduced. In the presence of a clearance combinations of sweep and lean are used to modify the loading in the clearance gap, thereby influencing the associated losses. The application of three-dimensional features redistributes the flow. The opportunity is therefore taken to rematch the sections based on the predicted results of the CFD codes. The above philosophy is used in the redesign of the datum compressor. Overall characteristics and exit traverse results from the test of the advanced build are compared to those from the datum build.

This paper, in two parts, reports measurements from, and simulation of, Cranfield University’s 3-stage high-speed axial compressor. Using this newly built rig, supported by European Commission, a consortium of gas-turbine companies have tested a set of conventionally stacked 2D rotor and stator blades. The results from this experiment were used to evaluate and assess the performance of several commercially available CFD codes leading to the collaborative design of an advanced three-dimensional blade set seeking, if possible, a 2% efficiency gain. The limited axial spacing between the measurement planes and the blade rows required the design of a unique seven probe assembly and traverse mechanism able to yaw and pitch the probes and to control the insertion depths. This mechanism was designed to accommodate different probes, such as cobra, fast response (pneumatic) and temperature measuring probes, and deliver area traverses between rotor and stators throughout the compressor. For probe calibration a high speed wind tunnel section was designed to accommodate this mechanism enabling calibrations for Mach numbers up to 0.78, as well as for a wide range of pitch and yaw angles values. This mechanism combined with a post processing programme incorporating a mapping technique for the relative offset of the measurement points on the probe secured very detailed results throughout the compressor. Measurements show the complex three dimensional flow structure and secondary flows associated with tip-leakage, endwall boundary layers, wake transportation and blade row interactions. The importance of a rigorous mapping procedure was particularly useful where the wake thickness was small and pressure gradients high in comparison to the probe size.

There is a growing interest in seismic surveys in arctic areas. Normally 2D surveys can be carried out with limited risk, as long as the area is reasonably free of ice. However, 3D seismic surveys are an essential tool for exploration in order to de-risk prospective areas ahead of expensive and challenging drilling operations. Acquisition of 3D surveys, with multiple streamers, is far more difficult than single streamer 2D surveys, as the amount of in-sea equipment is an order of magnitude higher and the data density for a given area covered is far greater: the physical footprint of a 3D equipment spread being towed behind a vessel can be about a kilometer wide by several kilometers long. This significantly increases the risk of equipment damage due to ice. This paper summarizes experiences from several 3D surveys in the Arctic, and addresses how the use of new equipment and techniques can reduce such risks to acceptable levels.