No abstract

No abstract

surface diffusion ratio (value above which {6/0)2 becomes excessive), blades should be designed for low friction coefficient (high Reynolds number and low relative surface roughness) and minimum laminar flow (high Reynolds number and high free-stream turbulence). The experimental correlation between wake momentum thickness and suction-surface diffusion ratio was made applicable for conventional blade-design use by empirically establishing an equivalent diffusion ratio expressible in terms of only the design velocity triangle and the blade solidity. The possibility of blade stall at minimum loss and greater angles of attack, as evidenced by a sharp rise in wake momentum thickness, occurred when the equivalent diffusion ratio attained a value of about 2. With the use of the equivalent diffusion ratio, calculations were made of the total-pressure loss and unstalled range of operation of conventional cascade sections as functions of solidity, air inlet angle, and air turning angle.

DiscussionW. A. Benser. 2 The paper presents a very interesting discussion of off-design operating problems of multistage axial-flow compressors. Stage-matching techniques, such as used by the author, are extremely helpful in obtaining and understanding stall and surge problems and in evaluating various design philosophies and modifications to improve part-speed performance. As he points out, publications on this subject have been limited. Several unclassified papers based on stage-stacking techniques have been published and are considered pertinent.Stage-stacking studies done at the NACA in general verify the conclusions drawn in the subject paper. Messrs. H. B. Finger and J. F. Dugan examined the effect of the shape of the stage-characteristic curves on off-design performance and obtained results similar to those predicted by Mr. Stone. Messrs. Finger and Dugan also examined the effects of stage rematching by means of area change as well as stator-blade resetting. Area change gave a marked improvement in part-speed efficiency; but when scaled to the same design weight flow, indicated no gain in regard to the surge limits. Improvements in low-speed efficiency, however, are beneficial in avoiding low-speed stall in gas-turbine-type engines. These results are presented in detail elsewhere. 3 Studies also were made on the effects of stage interactions on part-speed compressor operation. In this analysis, a depreciation of performance of several stages was assumed to result from stall of the inlet stages. This study also indicated the double-valued performance characteristics and potential surge-line knee discussed by the author. They also indicated the necessity of careful component test procedures to insure obtaining all possible potential operating conditions for any given compressor. These results have been reported by the writer. 4 The low-speed operating region where stall exists in the front stages of a multistage matching is of extreme importance from the standpoint of blade vibration. In this region of compressor operation, rotating stall may exist which, if in resonance with the blades, can cause severe blade vibrations even though the resulting flow fluctuations are not sufficiently large to cause severe penalties in compressor efficiency. This problem was discussed by M. C. Huppert and W. A. Benser. 6 In this paper it was assumed that rotating stall occurred at the point of inlet-stage stall. However, E. L. Costilow and M. C. Huppert 6 point out that the occurrence of rotating stall is affected by stability and that adjacent blade rows influence the initial point for rotating stall. Stalling of the inlet stages is a necessary condition for the occurrence of rotating stall at low compressor speeds and, therefore, analysis of firsf^stage stall limits are valuable in regard to determination of the region of potential blade vibrations resulting from this unsteady-flow phenomenon.The author points out the potential use of stage stacking in compressor development. He states that changes in stage charac...