FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1956
1956 - 0303.PDF
FLIGHT, 16 March 1956 299 Rocket-motor Combustion Chambers A British Interplanetary Society Lecture by Professor Baxter IN his lecture to the British Interplanetary Society on March3rd, Prof. A. D. Baxter, M.Eng., M.I.Mech.E., F.R.Ae.S.,F.Inst.Pet. (Professor of Aircraft Propulsion, College of Aero- nautics, Cranfield), reviewed the principles and practice of thedesign of combustion chambers for liquid propellants He pre- faced his lecture by saying that although much had been learnedabout combustion of rocket propellants since his last lecture to the Society five years ago, his general conclusions had not changed.These conclusions were that combustion chambers were developed by experience, empiricism and hard trial and error. The lecturer showed a cross-sectional drawing of a typicalregeneratively cooled combustion chamber (Fig. 1). It consisted of an inner and outer shell, between which one of the propellantswas passed to act as a coolant. The propellants entered the cham- COOLANT COOLING INJECTOR JACKET, HEAD fig. 1. Simplified dia-gram of typical re- generatively cooled combustion chamber, showing double walls. ber through a burner which served to mix and atomize thempreparatory to combustion, and the resultant gases were ejected through a convergent-divergent nozzle. Combustion chambers had been made in various shapes andproportions, varying from long thin cylinders to spheres; but, whatever the shape used, there was an optimum chamber size fora given thrust. If a chamber was too small, die reaction would not be complete before die gases left the chamber, and hence therewould be a loss due to combustion inefficiency; there was also a possibility of combustion instability occurring. With too largea chamber there would be higher frictional losses and die chamber surface area would be larger, thus increasing the total amount ofheat transferred to the coolant. As cooling was one of the major problems an increase in the cooled area could lead to serious diffi-culties. A common method of indicating the required chamber volume for a given propellant combination was the parameterknown as "characteristic kngth" (L*). This was defined as V L*=-A where V=combustion chamber volumeA=nozzle throat area. It could be shown that this parameter was a function of dietime of stay of die gases in die combustion chamber and hence was a measure of the time available for reaction. However, thesystem of injection also had a big effect on the required time of stay and hence an improvement in injector design reduced therequired value of L*. For example, die V.2 combustion chamber, burning 75 per cent ethyl alcohol and liquid oxygen, had an L*of 125in. The chamber of the Armstrong Siddeley Snarler, using radier similar propellants, had an L* of lOOin, and a modernchamber would need only about 60in. Thus die characteristic length was a useful parameter so long as it was used only to givean indication of chamber volume and not adhered to rigidly. It seemed probable that, as our knowledge increased, some othermore reliable design paramerer would be evolved. Scaling of combustion chambers also introduced problems. Forexample, if both L* and the chamber length/diameter ratio were kept constant, then, as the thrust increased, the throat diameterwould increase at a greater rate dian the chamber diameter until at a certain thrust level they became equal. Above thatfigure thechamber diameter would be smaller than die throat. Thus for a large increase in size this method of scaling would be unacceptable.Another method of scaling was to keep die chamber diameter/ diroat diameter ratio and the chamber length constant. This waseffectively the same as using a "bundle" of similar combustion chambers to produce a higher thrust. ... . Although the nominal time of stay based only on the gas phasewas usually 2.5 to 3.5 milliseconds, the real times—including the time of stay in the liquid phase—were much longer and werelargely influenced by the evaporation time of the propellants and the degree of mixing. Assuming that the droplet size and hence evaporation times were die same in different combustion chambers,then mixing remained the controlling factor. This could be improved by increasing the turbulence. Thus by increasing dieaverage gas velocity through the chamber a lower L* could be used. There was, however, a limitation to this line of thought,as indicated in Fig. 2. Curve A represented the variation of overall chamber efficiency (incorporating combustion efficiency and fric-tional losses) widi L* for a very squat chamber in which the gas velocity was low. The maximum efficiency was reached at a highvalue of L*. For the limiting "tubular" or "throadess" chamber the maximum efficiency occurred at a low value of L* (curve C),but was inherendy a lower absolute value than for diroated cham- bers. A series of such curves could be drawn for chambers ofvarying aspect ratio and die maximum of die envelope of diese curves would be the optimum design point (curve B). The droplet-size distribution depended on the injection system.Various methods of injection had been devised, largely on an ad hoc basis. Of diese, impinging jets, impingement on a targetplate, multiple swirl, shower-head and annular slit had been fairly widely used. In the shower-head system, the propellants wereinjected through a number of interspersed holes without impinge- ment, and in die annular slit diey were injected through slits a fewthousands of an inch in widdi Little work had been carried out on the effect of atomization on the combustion of rocket propel-lants, but it had been shown that in a gas turbine an increase in droplet size from 140 to 170 microns increased die burning timeby 50 per cent. A similar situation undoubtedly arose in die rocket motor and, as a result of the small droplets produced, a swirlinjector was, generally speaking, a better proposition than any of the other systems. The cross-sectional area of the chamber wasalso an important factor. As the droplets left the injector they slowed down, and if there was not room for them to spread out(a chamber area of at least 100 times the injector orifice area was required) coagulation occurred and the burning time was increased. The higher the combustion pressure the better the performance.The major effect was die increase in expansion ratio, which increased the thermodynamic efficiency of the cycle. Also, thehigher pressure reduced the degree of dissociation of the gases, hence increasing die combustion temperature, which gave asecondary increase in performance. For a constant thrust the volume of gas required varied inversely as die pressure; if die gasvelocity were unchanged then die cross-sectional area was also inversely proportional to the pressure. This meant a decrease inchamber weight for an increase in pressure, although in practice there was a limiting pressure above which the chamber weight waseffectively constant. Unfortunately, the nozzle increased in weight with pressure and the analysis of the opposing effects on weightand die varying performance with pressure were not easy to analyse. Few rocket motors operated at constant ambient pressure anda further problem was the choice of design altitude. A nozzle designed for correct expansion at ground level (100 per cent) wouldproduce 113.5 per cent dirust at 100,000ft, owing to the fact that the pressure of the exhaust gases was greater than the ambientpressure. However, a nozzle correctly designed for 100,000ft would produce 137 per cent thrust, although at ground level itwould be less dian 100 per cent. It would also be twice as heavy as the sea-level nozzle. The nozzle exit angle also affected performance, and here againtwo opposing factors were at play. The gas flow from the nozzle was, of course, three-dimensional and hence only the gas at the Fig. 2. Comparative efficiencies of various chamber shapes plotted against characteristic length L".
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
