FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1956
1956 - 0294.PDF
292 FLIGHT HOT TURBINES An N.G.T.E. Engineer's Exposition of a Crucial Subject IN all types of gas turbine, and particularly in turboprops andsupersonic turbojets, a principal factor restricting perform-ance is the upper limit on turbine-inlet gas temperature. Last week an engineer who for several years has been associated withinvestigations into methods of increasing this temperature pre- sented his findings—insofar as security conditions permitted—as a main lecture of the Royal Aeronautical Society. The author was Mr. D. G. Ainley, B.Sc, A.M.IJvIech.E., A.F.R.Ae.S., of theNational Gas Turbine Establishment, Pyestock, Hants. It is undoubtedly one of the most useful all-round appraisals of thissubject from an aeronautical point of view yet published. The lecturer began by pointing out that since Sir FrankWhittle's first flight engine of 1941, maximum gas temperatures had risen scarcely more than 100 deg C, although improved con-structional techniques and axial compressor developments had increased thrust/weight ratios approximately threefold and thrustper unit frontal area roughly tenfold. The law of diminishing returns was tending to restrict further improvement and if furtherprogress was to be substantial, the problems of high temperature operation would have to be faced squarely. This called for newtechnological advances and Mr. Ainley's paper was broadly con- cerned with the years of preliminary study which, it was hoped,would clear much of what would seem to be the roughest ground ahead of engine development. It could be shown (Fig. 1) that, as flight Mach number wasincreased, the gas temperature for minimum specific consumption tended to rise. If temperatures were raised slightly above thelevel for minimum consumption, the latter factor rose only slightly, whereas the specific thrust increased appreciably, result-ing in a reduction in engine weight, frontal area and nacelle drag. Thus, the optimum gas temperature for a minimum total engine/fuel weight for a given mission would be slightly higher than the values shown in Fig. 1, the difference becoming more marked asthe range was increased. For example, a long-range aircraft cruis- ing at a Mach number of 2 to 2.5 would require optimum gastemperatures from 100-300 deg C higher than any contemporary "uncooled" engines. In the case of intercepters with afterburning engines increasinggas temperature could have a very beneficial effect in the non- reheat condition, whereas with the afterburner in action it couldbe argued that the advantage was less. In the case of turboprop and by-pass engines there was everything to gain by raising gastemperature. For example, in a turboprop flying at 400 kt in the stratosphere, raising the gas temperature from 1,100-1,400 deg Kcould increase the specific power by roughly 50 per cent while reducing the specific consumption by about 10 per cent. Of all the problems impeding the development of high-tempera-ture engines the worst was presented by die highly stressed turbine blade. Two possibilities were available: either newmaterials could be employed or the blade temperature could be maintained roughly at the present value by employing a systemof cooling. Problems chiefly associated with cooling also centred around the combustion chamber, jet-pipe and propelling nozzle,and in supersonic applications the problems of high gas tempera- ture would be aggravated by the very high stagnation intaketemperatures. Nevertheless, Mr. Ainley confined his paper primarily to a discussion of blading. Taking even an optimistic view, the lecturer thought it improb-able that further developments in nickel- or cobalt-base blade materials would lead to increases in blade temperature greaterthan from 50-100 deg C. Molybdenum alloys might allow further improvement, although their use postulated the develop-ment of a satisfactory protective surface coating. High-tempera- ture operation could also be achieved by various ceramic orceramic-metal materials, although these had so far not fulfilled all their inherent promise. Blade-cooling was therefore inevitableand, said Mr. Ainley, it was unlikely that future development of blade materials would displace blade-cooling; rather would itenhance its potentialities, since with cooled blades a 50-deg-C rise in permissible temperature would allow a substantially greaterincrease in gas temperature. In a cooled turbine heat was extracted from the gas flow at ahigh rate. This heat had to be removed by the coolant, and it resulted in a loss of power owing to the reduced gas temperaturein the later turbine stages and propelling nozzle. At the same time, the cooling system tended to increase weight, complexityand drag. Transfer of heat from a cooled blade was largely by forcedconvection, although relatively insignificant transfer might also be due to conduction or radiation.Fig. 2 depicted the variation of local heat transfer coefficient around a typical blade section. High rates occurred at the leadingedge where the boundary layer was thin and laminar. The value fell off appreciably over the upper surface back to the transitionpoint to turbulent flow, but in the latter region the coefficient increased sharply. On blades of different profile the mean heattransfer rates were roughly similar if the transition points were, similarly disposed, although in practice the transition point wasdependent on the profile. Thus, on a high-reaction blade (e.g., a stator, or nozzle guide vane) the pressure gradients were favour-able over most of the surface and a relatively large laminar region resulted in correspondingly low heat transfer. On low-reactionblades the less-favourable pressure gradients increased the turbulent region and hence also the mean rate of heat transfer.This was shown in Fig. 3, in which experimental data emphasized the correlation between mean transfer coefficient and bladereaction. The correlation was emphasized by the fact that the plotted points were drawn from a wide variety of sources inwhich none of the variables were necessarily related. The mean heat transfer coefficient (expressed as a Nusselt number) generallyvaried with Reynolds number, where x, the usual Reynolds number exponent, varied between 0.5 (wholly laminar) and 0.8 (whollyturbulent). Most of the existing heat transfer data had been obtained fromstationary cascade experiments. Gas flows in actual engines were more turbulent and unsteady, and it could be assumed that earliertransition and higher rates of heat transfer would therefore result. Approximate estimates had been made from N.G.T.E. work withan experimental air-cooled turbine from which the dotted lines of Fig. 3 had been "speculatively introduced." Nusselt numbers, being non-dimensional, gave little hint ofthe actual quantities involved. On typical engines the heat transfer rates ranged from about 0.02 to 0.2 CH.U./sq ft/sec/degC. In a 150 lb/sec (at sea level) engine the surface blade area Fig. 1 (Left) Variation of optimum gas temperature with flight Mach number. Fig. 2 (Below) Variation of local heat transfer coefficient .•~.~ around a typical turbine blade profile. Fig. 3 (Right) Correlation of experimental data for mean heat transfer rates, with results from actual testing shown as a dotted line. 6OO I O 2 O FLIGHT MACH NUMBER JO ..68**" If * 5OO in | 4OOV—-• I 3OO z 2OO ACTUAL TURBINE. ST;•ACE \ ^^," •^ • CASCADE TESTS .1 O2 O4 O6 O8BLADE INLET ANGLEGAS OUTLET ANGLE NO22LE BLADES i O 8 O-6 •
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
