288 THE ENG INEE R OoT. 12, 1945

• The President of the Air Council was a. party to the agreement which r esulted in the forma• The Whittle Jet Propulsion * tion of , and the was a. shareholder from the start in that a proportion By F. WHITTLE, C.B.E ., R.A.F., M.A. , Hon. M.I . Mech. E.t of the shares alloted to me was held in trust for the Department. No. I During the negotiations leading to the forma· tion of Power J ets, I was working on the designs of an experimental engine. Messrs. 0. T. Falk INTRODUCTION AND GENERAL OUTLINE a Flight Cadet at the R.A.F. College, . and Partners placed an order with the British HE main argument against the gas turbine Each term we had to write a science thesis, and in my fourth term I chose for my subject Thomson-Houston Company, Ltd., for engi. Twas that the maximum temperatures per• neering and design work in accordance with my missible with materials available, or likely to the future development of . Amongst other things, I discussed the possibilities of jet requirements in advance of the formation of be available, was such that the ratio of positive the new company. Power Jets placed the order to negative work in the constant-pressure cycle propulsion and of gas turbines ; but it was not until ·eighteen months later, when on an for the manufacture of the engine (except the could not be great enough to allow of a reason• combustion chamber, instruments, and some able margin of useful work to be obtained, after instructors' course at the , Wittering, that I conceived the idea of using a accessories), with the British Thomson-Houston allowing for the losses in the turbine and com• Company in June, 1936. pressor. There seemed to be a curious tendency gas turbine for jet propulsion. I applied for my first patent in January, 1930. The principal The engine was to be a simple jet propulsion to take it for granted that the low efficiencies gas turbine having a single-stage centrifugal of turbines and compressors commonly cited drawing of the patent specification as filed is reproduced in Fig. 1. It may be seen that I compressor with bilateral intakes, driven by a. • were inevitable. I did not share the prevalent tried to include the propulsive duct, or single-stage turbine directly coupled. Com• pessimism because I was convinced that big bustion was to take place in a single combustion improvements in these efficiencies were possible, " athodyd," as it has since been called, but this had been anticipated at least twice, so the upper chamber through which the working fluid passed from the compressor to the turbine. ftG. l . drawing and relevant descriptive matter had to be deleted from the specification. We were going beyond all previous engineer• The idea was submitted to the Air Ministry, ing experience in each of the major organs. We but was turned down on the ground that as it were aiming at a pressure ratio of about 4/1 in a single-stage centrifugal blower when at the time, so far as we knew, a ratio of 2t/1 had not 7 been exceeded. We were aiming at a breathing c .. ~ . D capacity in proportion to size substantially ~ 1\ NOZZl E THROAT \ lli S'~ A BS., ~ 1,938 FT. PE R SEC ))-8 SQ. IN . flG. 2., g ~ \ V B c ...... - ~ I ~ ... J MAIN NOZZl E ..a ~ UTH ~ \ EF ...« ...... • 1 ...... _ ' I'.. A G

I - 0 100 lio )CO 400 ~ lo 600 7lo elo " VOLUME-CV. FT. S WAI N Se .. Sea. level cycle. f tc• 0 Assumptions :-Coml?ressor efficiency, 70 per cent. SW"'f N Se. \0 Turbme efficiency, 70 per cent. Axia.l velocit y a.t turbine exhaust, 1020ft. FIG. 3- .Assembly of the First 111odel of E xperiTMntat per second. Engine E fficiency of final expansion, 90 per cent. Weight of air, 26 lb. per second. Weight of fuel, 0 · 3635lb. per second . greater than had previously been attempted. y = 1· 4 for compression= 1· 379 for ex- • The combustion intensity we aimed to achieve t pa.ns10n. Kp = 0 · 24 for compression= 0 · 26 for was far beyond anything previously attempted. I combustion and el.."P&nsion. Finally, we had to get over 3000 S.H.P. out of a Static thrust, 1240 lb. a single-stage turbine wheel of about 16lin. FIG. 2- The Pressure- Volume Diagram, Design A ssump• outside diamet er, and to do it with high \0 tiotl8, &, c. , on which the Initial Design of the efficiency. E xperimental Engine was Based At first, our intention was to do the job stage by stage that is, to make a compressor and $ W AI,.. Se. was a gas turbine the practical difficulties in the t est it ; then to add a combustion chamber to The upper drawing-the thermo-propulsive duct--had way of the development were too great. the compressor; then to test a turbine alone; to be deleted from the specification During 1930 I tried to interest various firms and finally to build an engine but we very FIG. !- Reproduction of Drawings Illustrating Briti,sh in the scheme, but met with no success ; for soon realised that this was likely to be a. long Patent No. 341,206,flled January 16th, 1930 the most part they thought the same way as and costly pro ess and we decided to go for a u.nd, in the application of jet propulsion to the Air Ministry. It is probably also true that in complete engine right away. aircraft, I realised that there were certain their view the prevalent industrial depression This first engine was based on a design for favourable factors not present in other applica• made it anything but a favourable moment for flight, but was not intended for flight ; and tions, namely :- expensive ideas of this sort. though it was designed to b j very light by normal Nothing very much happened for a few engineering standards, we did not put forth ( 1) The fact that the low temperatures at years. I gave up hope of ever getting the idea special efforts to make it as light as possible. high altitudes made possible a greater ratio to the practical stage, but continued to do paper I was fairly confident in the compressor a.nd of positive to negative work for a given work at intervals, until, in May, 1935, when I turbine elements, but felt rather out of my depth maximum cycle temperature. was at Cambridge as an engineer officer taking with the combustion problem, and so (in 1936) (2) A certain proportion of the com• the Tripos Course, I was approached by two I visited the British Industries F air with a pression could be obtained at high efficiency ex-R .A.F. officers (Mr. R. D. Williams and view to enlisting the aid of one of the oil burner by the ram effect of forward speed, thereby Mr. J. C. B. Tinling}, who suggested that they firms, but the requirements I specified were raising the average efficiency of the whole should try to get something started. Though I considered to be far too stringent by most of compression process. had allowed the original patent to lapse them until I met Mr. Laidlaw, ofLaidlaw, Drew ( 3) The expansion taking place in the through failure to pay the renewal fee, and and Co. Though he recognised that we were turbine element of such an engine was only though I regarded thein as extremely opti• aiming at something far in advance of previous that which was necessary to drive the com• m istic, I agreed to co-operate. I thought that experience in this field he considered the target pressor ; and therefore only part of the there was just a bare chance that something possible of attainment, and so it was with his expansion process was subject to turbine might come of it. help that we attacked the combustion problem. losses. We eventually succeeded in coming to an While the engine was in course of design and Outline.- ! first started thinking about this arrangement with a firm of investment bankers manufacture, we carried out a number of com• general subject in 1928, in my fourth term as (Messrs. 0. T. Falk and Partners}, which led to bustion el..'Periments on the premises of the the formation of Power J ets. Ltd., in March, British Thomson-Houston Company, with • First James Clayton Lecture. Institution of 1936. Before Power Jets was formed, 0. T. apparatus supplied by Laidlaw, Drew and Co., Mechanical Engineers, October 6th. Abstract. t Special Duty List, R.A.F., attached to Power Jets Falk and Partners obtained the opinion of a until we considered that we had enough informs.· (Research and Development), Ltd., Whetstone, Leicester. consulting engineer (Mr. M. L. Bramson), who tion to design a combustion chamber. P ower The statements and opinions expressed herein are the gave a wholly favourable report. The initial Jets therefore placed the contract for the design personal views of th& lo c t~ e r, and ar_e .not to be t ake? as in any way re~r e sentmg the opm10ns. of the A1r sum subscribed was £2000, and with this we and manufacture of the combustion chamber Ministry or the Min1stry of Aircraft Product10n. cheerfully went ahead. with Laidlaw, Drew and Co. OcT. 12, 1945 THE ENGINEER 289

By this time the Tripos Examinations at impeller as manufacturing limitations would assembly which contained one of the two water Cambridge were over, and the Air Ministry permit, in order to keep the blade loading as jackets for the turbine cooling could not be he.d agreed that I should stay for a post• low as possible. In particular, it was hoped split in the diametral plane. These considera• graduate year. This was really a device to that by keeping the pitch-chord ratio of the tions governed the rotor design. enable me to continue work on the engine, and rotating guide vanes small we should avoid Testing .-As already mentioned, the testing of so a considerable proportion of my time was stalling at the intake, as I believed then-and the engine under its own power began on April spent at the British Thomson-Houston Com• still believe that this is one of the main sources 12th, 1937. The very :first attempt to start pany's works in Rugby. of loss in centrifugal compressors. was successful, in that the engine ran under its Testing of the engine commenced on April No diffuser blades were fitted to the blower own power, but it accelerated out of control 12th, 1937, and continued intermittently until casing at :first. Two stages of diffusion were up to about half its designed full speed. This August 23rd. These early tests made it clear aimed at. The intention was to obtain partial happened several times, and altogether it was a that the combustion problem was by no means diffusion in the bladeless vortex space between very alarming business, so much so that in the solved, and that the compressor performance the impeller tips and the scroll, and to convert early days the individuals in the vicinity did was far below expectations. Nevertheless, they more running than the engine. were sufficiently encouraging to show that we The starting procedure was as follows :- The were on the right track. engine was motored at about 1000 r.p.m. and the pilot jet (which injected an atomised spray) THE DESIGN AND TESTING OF THE IWODiltJ was switched on and ignited by means of a. EXPERIMENTAL ENGINE 11 ,----'' 1 I (--- _, PILOT JET sparking plug and hand magneto. The motor• The Design and Testing of the First Model: ing speed was then raised to about 3000 r.p.m., Design.-The first engine was designed with PILOT JET R and the main control opened slowly. The engine CONTROL a specific target in mind. It was a very optimis• u' • fUEL PUMP would then accelerate under its own power ; tic one, but, nevertheless, it formed the starting but, as I say, not always under control. The point and is worth recording. The aim was to explanation of the first few uncontrolled accele• propel a very " clean " little aeroplane of about rations was simple when we found it, and may 2000 lb. " all up '' weight at a speed of 600 be understood by reference to the diagram of t~AIN ...... CONTROl (MEI\CENCY the early fuel system, shown in Fig. 5. If the 1\ELIEF VAlVE 8Y·PASS spill line from the burner was not full of fuel the needle valve of the burner would be forced :,OWAI,.. S C o into the fully open position when the fuel FIG. 5 The Fuel System Used in the Early T uts of the pump ran. We were frequently breaking the First 111odel of the Experimental Engine fuel line and doing various motoring tests, so that often, unknown to us, there was a " lake " of fuel in the combustion chamber. Other most of the remaining kinetic energy into pres• uncontrolled accelerations were caused by the sure in the " honeycomb " di.ffuser through sudden opening of the burner needle valve after which the air passed from the compressor scroll initial sticking ; by loss of temper in the burner 001 HC CNGI,..t(.A • to the combustion chamber. spring through overheating, &c. Fortunately, The turbine nozzle arrangement was very none of these uncontrolled accelerations took FIG. 4- T e.

TABLE I.-Leading Particulars of First Edition of Experimental Engine Dimensions are in inches. Compreseor impeller• Tip diameteT ...... 19 Tip width ...... ,. 2 Outer diameter of eye ... 10·75 Inner diameter of eye 5·5 Number of blades ... 30 Materia l ...... Hiduroinium RR 56 Compressor ce.sing- Inner diameter of scroll . . . 31 :Material ...... Hiduminium RR 55, DTD. 133 B Turbine-a- Mean diameter of blo.des... 14 Blade length ...... 2 · 4 Material of blades ...... Firth-Vickers Stayblade Material of disc...... Firth-Vickers Stayblade Blade chord ...... 0 · 8 Number of blades ...... 66 Maximum speed- Revolutions per minute ... 17,750 Figs. 3, 4, and 5 illustrate various features of FIG. 6 First M odel of the First Experimental Eng~ne the design, which are further amplified in Figs. 6, 7, 8, and 9. The assumption of 80 per cent. adia,batic now, I thought it a good idea at the time. It is system up to maximum speeds of about 8500 efficiency for a running of considerable interest in retrospect, because r.p.m., but the combustion was so bad that this at a tip speed of l470ft. per second was very it became evident later than I had not succeeded speed could not Le exceeded. Any further optimistic indeed, and received a good deal of in conveying to others what I had in mind. opening of the control seemed to result only in criticism, but I felt confident that we could Air tests were made on a half-scale model of the burning of more fuel aft of the turbine. design to avoid many of the losses which were this nozzle, and though very rough they seemed Many attempts to improie the combustion occurring in all centrifugal compressors of to show that it would behave approximately as were made by a series of modifications to the which I had knowledge at the time. The general expected. combustion chamber, and some improvement view was that we should be fortunate if we got The design of the rotor assembly needed was achieved-we managed to get up to a speed 65 per cent. adiabatic efficiency. much careful thought. It was considered of 11,000 r.p.m. for 10 min. This series of runs We went for the double-sided compresso_r necessary to use unbored forgings for both the ended when the compressor impeller fouled the because we wanted to get the greatest possible turbine wheel and compressor impeller, also to casing after running for 4 min. at 12,000 r.p.m. breathing capacity in proportion to size. I also use an overhung turbine rotor, as it was thought The d.a.mage to the compressor and impeller counted on this featW'e to give a reduced pro• that the provision of a satisfactory bearing casing was only slight, but as it had now become portion of skin friction losses. support on the exhaust side of the turbine clear that the delivery pressure from the com• We aimed at having as many blades on the would be very difficult. The bearing housing pressor was much below expectations, we 290 THE ENGINEER OcT. 12, 1945 decided to fit diffuser blades in an effort t o the burner now b eing insulated against over• to the national standards bodjes in Britain. improve tho blower performance before further heating by using a fuel-cooling arrangement. America, and Canada, so that what is termed testing. " 'o th n managed to attain a speed Combustion appeared to be improved, and for an" A. B. C." standard will be speedily approved. of 13,000 r.p.m., but tho compressor perform• the first time no part of the casings reached Mutual under ~ tanding was reached on specifica• ance still left much to be desired. Moreover, the glow heat at speeds of up to 12,000 r.p.m. tions for sma11 screws and various screws and combustion had deteriorated, and as this was Testing was now suspended for the following threads u ed in the optical and scientific in. tru• believed to be due to th e nature of the flow reasons :- First, because the speed restriction ment trades, and on buttress thread forms. from the compressor scroll to the combustion of 12,000 r.p.m. made it necessary to :find a new Considerable progress was made in gathering chamber, many modifications to improve this site for running at higher speeds, and, secondly, data on high-duty studs in light alloys, but a great deal more exploratory work is, it is felt, required. Drawing practice and its tmi:fication wero discussed, and it was agreed that thi:-: subject be actively pursued. Pipe threads wore dealt with and an invitation to the British and Canadian representatives to continue discnR• sions of this subject at the November me e ting~:~ of the American Petroleum Institute were given. On the question of limits and fits, a further meeting with the British delegates will be held in New York before the return of the delegates. Suggestions were forthcoming on precision and gauging m ethods, with a view to the co-ordination of future practice. Propo. ed specifications for screw threads and connections for gas cylinders were presented. On the wholo. thic; third Conference exemplified the spirit of collaboration which prevails among the engi• neering profc. sions of the thrE"e EngliRh ·Spf'aking countries.

FIG. 7-Rotor Assembly of First Experimental Engine flow were made, but without noticeable improve• becau::;e it was decided to make major modifica• Institution Programmes ment. There wa ::~ evidence of a flow reversal tions to the general arrangement. in the elbow, bnl, we did not realise how severe The tests so far had been very disappointing, Pn.oGRAMM:ES for the session 1945-46 have now this was until (in one test late in the series) but there were m any encouraging features. We been issued by the fo1lowing institutions and :fltlmes were seen through a small hole which had demonstrated that there was no particular societies :- • had been drilled in the n E"ck of the blower difficulty in starting or in control. There was INSTITUTE OF MARINE ENGINEERS casing scroll. also plenty of evidence which suggested that November 13th, 1945 : " Development in Tur• A series of rapid modifications to the diffuser the whole scheme was well worth developing, bine Blading," by J. G. Monypenny. though it had become obvious that much harrl D ecember 11th: '· Deck Machinery, with work lay ahead. Particular Reference to Latest Developments," by

''' OIA Tho principal defects of that particn lnr T. Brown. anangement were shown to be:- January 8th, 1946 : " Elements of Propulsive Efficiency," by F. H. Todd, B.Sc., Ph.D., M.I.N.A. (1) Poor compressor efficiency. February 12th : '· The Development of the (2) Excessive prehcating of the air to the Opposed-Piston Marine Oil Engine, with Special rear intake, owing to the disposition and tem• Reference to Engine Dynamics," by \V. Ker \Yi1 on, pt-rature of the combustion chamber. D .Sc., Ph.D., 'Wh. Ex. (3) Very unsatisfactory combustion. March 12th: "Strain Gauges," by Dr. F. ( 4) Excessive frictional loss in the n n• Aughtie. orthodox turbine nozzle scroll. April 9th : " Engine-Room Lay-Out, with Special Reference to Pipe W ork," by Lient.-Colonel No reading of sufficient reliability had bePn G. Rochfort, M.C., D.S.O. obtained from which any estimat e of the effi• April 30th : " Lubricating Oils a.nd their ciency of the turbine could be macle, but it Charactel'istics," by C. Lawrie nnd Colonel S. J. M. eemed practically certain that it wa,:-: wc.>ll bP low Auld. FIG. 8-The Uncotwentional BLadeless Turbine Nozzle thA-t ft,RRUm <'