FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1957
1957 - 1832.PDF
*0 70 40 V VX s\ SINGLE STAGE >5; *" STAG!: co >^= MmOS) ««. | MULTI-STAGE COMWESSOR '* 1— N Aim UDC TEST F | iNT NOTE-REYNOLDS NUMBER 6ASC0 ON MEAN BLADE CHORD. AND MEAN BLAC SPEED. /N ON-SYMMETRIC '—— 922 FLIGHT oi Ol O4 OS OI O-« CO REYNOLDS NUMBER • K>~ 6 7 MO MO 1*42 H*4 l*t» N4« I«SO KS3 Fig. 3. Temperature rise per stage in axial compressors. GAS TURBINE DEVELOPMENT... return-circuit, variable-density wind tunnel in which tests can be made over a one-hundredfold range of Reynolds Number (Fig. 1). With this equipment it has been possible to study the effects of density both on the aerodynamic and flutter performance of axial compressor stages. The history of the aerodynamic effects has been slightly unusual in that the answers have been obtained by successive approxima- tions. Over the years we have obtained, or been presented with, various data alleged to represent Reynolds Number effects, with most of which we have been more or less dissatisfied. It was not until we had developed the 114 plant that we were able to get sufficiently controlled conditions and accurate results to make us believe that we had at last obtained a correct answer (Fig. 2). The flutter investigation showed that a reduction in the air density tended to move the blade away from the flutter region and there- fore, in general, to reduce the vibrational stress. Although great assistance has been given to our analytical work by testing other people's compressors, we have from time to time found a need for a more specific check on some detail point than these usually afford. Under these circumstances we usually con- struct and test a compressor of our own design. A good example is our number 109, which we first tested in 1946. Our general theory of compressor design had suggested that it should be possible, by maintaining a high Mach number throughout the compressor and using blade details rather unusual at that time, to reduce considerably the number of stages necessary to obtain a required pressure ratio. Accordingly, this compressor was laid out to give a pressure ratio of nearly 5 :1 in six stages, about half the number normally required (Ref. 4). The compressor achieved its high designed performance and for many years remained a milestone in compressor achievement (Fig. 3). Fig. 2. Reynolds Number effects in compressors. More recently, in 1956, the same compressor has been used to check another design point, on blade profiles. Our general practice had been to use the same profiles throughout the compressor, but we had been gradually collecting data from our cascade tests that suggested that an improvement in performance would be obtained by varying the profile. It appeared that blading with a parabolic profile with its larger throat area passed a greater airflow but with some loss in efficiency at high speed, while blading of circular arc profile gave a higher efficiency at a lower flow. We therefore rebladed the 109 compressor, which originally had parabolic blading throughout, with blades changing progressively from circular arc in the first stage to parabolic in the last. A consider- able improvement in the part-load performance, with an increase in efficiency and a straightening of the surge-line "kink" resulted (Fig. 4). Turbines. In gas-turbine research there has always been an unusual degree of co-operation and exchange of information. Individual contacts and formal organizations, such as the sub- committees of the Gas Turbine Collaboration Committee, have sometimes resulted in such close relations between different bodies working on the same problem as to give the impression that the work was being done by a single unit. The evolution of the aero- dynamic design of turbines is a good example. In the early designs of turbine the highest possible loading per stage was aimed at, the limit being set by the avoidance of recompression at the blade root. It was not until near the end of the war that it was realized that turbine efficiencies were lower than had been expected, and were mainly responsible for the poor specific fuel consumption of that time. Immediately after the war we carried out the first tests on the turbine components of gas-turbine engines (Ref. 5) in this country, and backed them up by a comprehensive range of cascade testing (Fig. 5). As a result of this work we came to the conclusion that the principal cause of poor efficiency in contemporary turbines was their high aerodynamic loading, and we pressed for the development of multi-stage turbines (Ref. 6). By 1950 the blade loading parameter for stages without exhaust swirl, fePAT/il72, which had previously risen to over 4.0, had dropped to between 2.5 and 3.0, and substantial gains in efficiency had resulted. The importance of a high degree of reaction had been appre- ciated from steam turbine experience. Others had demonstrated the gains resulting from a reduction in blade clearance and flow leakage. It remained for the Establishment to analyze the com- bined experience of many and produce in 1951 a systematic method for predicting overall turbine performance (Refs. 7 and 8). -r_- (To be continued) References (1) Aeroplane gas turbines: Proc. Inst. Mech. E., Vol. 157 (1947).—H. Constant. (2) The present basis of axial-flow compressor design; Part I—Cascadetheory and performance: R & M 2095—A. R. Howell. (3) Fluid dynamics of axial compressors. Design of axial compressors:Proc. Inst. Mech. E., Vol. 153 (1945)—A. R. Howell. (4) The design and testing of an axial compressor having a mean stagetemperature rise of 30 deg C: R & M 2985 (1953)—A. D. S. Carter, S. J. Andrews and E. A. Fielder. (5) Overall performance characteristics of a four-stage reaction turbine:R & M 2416 (1946)—D. G. Ainley, S. E. Petersen and R. A. Jeffs. (6) An approximate method for the estimation of the design pointefficiency of axial flow turbines: A.R.C. Current Paper No. 30 (1950) —D. G. Ainley. (7) An examination of the flow and pressure losses in blade rows of axialflow turbines: R & M 2891 (1951)—D. G. Ainley, G. C. R. Mathieson. (8) A method of performance estimation for axial flow turbines:R & M 2974 (1951)—D. G. Ainley, G. C. R. Mathieson. Fig. 4 (left). Per- formance of the No. 109 compressor. Fig. 5 (right). Charac- teristics of first turbines tested imme- diately after the war. MASS FLOW - L» / SCC
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
