FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1961
1961 - 0140.PDF
140 FLIGHT, 3 February Missiles and Spaceflight LIQUID-OXYGEN PROBLEMS TIMELY, in view of the recent spillage of some 600gal ofthe fluid at an RAF Thor emplacement (reported last week),is this condensed presentation of a paper entitled Problems with Liquid Oxygen in Large Rocket Engines, which was readbefore the second Cranfield Symposium on rocket propulsion last month. The author was Air I. E. Smith, chief performanceengineer (rockets), Rolls-Royce Ltd, who is careful to point out that his paper deals only with the difficulties; the fact that liquidoxygen is employed in the largest rocket engines is, he said, recommendation in itself. Mr Smith's paper highlights problems arising in the fluid'scontrol and management. To do this [he said] it is proposed to follow the course of an engine test from the initial filling of thetanks to the final analysis of the data. Such an engine consists of a regeneratively cooled chamber fed by two centrifugal pumpsmounted on a common shaft and driven by a turbine. Power for the turbine is derived from the combustion of a small proportionof the main propellant flows in a separate gas-generator, the power output from which is maintained constant by means of a regulatorin the liquid-oxygen feed line. The engine is started by separate starting tanks which supply propellants to the gas-generator only.Liquid oxygen is transported at its normal boiling temperature of 90 °K, and in this state fed into the main tank on the test standor vehicle. This can be done either by addition of heat to the supply tank or by means of a pump. Once the main tank has beenfilled further boiling is suppressed by slight pressurization, and the surface/volume ratio is usually such that the temperature risesby only a few tenths of a degree per minute. But the liquid in the line to the engine warms up much more rapidly, since the surface/volume ratio is much higher than in the tank itself. The liquid soon reaches its boiling point, and, in the absence of particlescontaining absorbed gas, undergoes superheating. When boiling finally takes place the bubbles expand very rapidly, extracting heatfrom the liquid until the system is once more in thermal equili- brium. At a superheat of only 1°C liquid oxygen will producetwice its own volume of vapour at pressures close to atmospheric, and the pressure surge as the gas escapes into the tank and theliquid falls back has been known to cause serious structural damage. The remedy adopted is to provide a continuous cir-culation of cold liquid, either from the tank or from a separate Schematic cross-section of the RZ.2, tints indicating the flows of liquid oxygen and kerosine. The design of this engine was based upon the early Rocketdyne unit designed tor the Navaho III booster KEROSINE FROM MAIN TANK LIQUID OXYGEN (LOX) FROM MAIN TANK LOX REGULATOR / / / START TANK SUPPLY MAIN f UEL VALVE NON-RETURN VALVE MAIN LOX VALVE INJECTOR THRUST CHAMBER A Rolls-Royce RZ.2 rocket engine. This unit, half the pro- pulsion system of Blue Streak, is that on which Rolls-Royce have cut their teeth in the big-rocket field source. When the preparations for firing are complete, the tank is pressurized to about 2 or 3 ata, to prevent boiling during the test. When the main valves are opened and liquid oxygen enters theengine, vigorous boiling ensues until the metal surfaces have cooled to liquid-oxygen temperatures. Any attempt to start the engine atonce would lead to violent and unpredictable transients, but the main passages can be precooled simply by allowing liquid oxygento flow under tank head for a short while before starting. It is in the smaller lines, such as the gas-generator feed system, thatthe greatest difficulties are encountered; and this particular system controls the behaviour of the entire engine. Precooling of the linesup to the last point of control, the gas-generator valve, may be achieved by providing a bleed to some point of low pressure inorder to allow a small flow to pass (care has to be taken in selecting the return point, in order not to bleed off too much during steadyrunning). In the engine illustrated, in which the oxygen is sup- plied to the gas-generator by a constant-pressure regulator, acertain tolerance may be allowable, but for a constant-resistance control a bleed of uncertain magnitude may introduce more prob-lems than it eliminates. In such cases the designer has no option but to close the bleed lines with a solenoid valve once they haveserved their purpose. Problems of this type apply to unprimed lines in general, although the difficulties are greatly increased bya propellant which can, so to speak, generate its own gas pocket. Not only the surfaces in direct contact with the liquid oxygenare cooled during this phase, but also such others as valve spindles, control cylinders and shaft bearings. Stringent precautions mustbe taken to avoid the presence of condensible gases and vapours which would cause seizure. Nitrogen has a lower boiling pointthan liquid oxygen, and may safely be used for purging. How- ever, water vapour must be avoided at all costs if the freezingof controls is to be prevented. In order to err on the safe side, heaters are generally fitted wherever freezing is liable to occur. With a non-cryogenic propellant the starting pressure at thepump inlet is essentially tank head minus the head required to accelerate the fluid along the line. For liquid oxygen the corres-ponding pressure is at least an atmosphere less, owing to the vapour pressure of the liquid, and in practice the deficit is usuallyrather greater. Moreover, designers frequently place the liquid oxygen tank ahead of the fuel tank in order to minimize e.g.shift in flight, which means that the oxygen line becomes at least as long as the fuel tank, and the acceleration of the propellant fflthe pipe normally outweighs the gain in static head. As a result an initial 601b/sq in absolute at the bottom of the tank maydwindle near to zero at the pump inlet, leading to severe cavi-ratio" at the pump and turbine overspeed. The answer may be to minimize acceleration loss by employ-ing as large an inlet line as practicable. Methods of ensuring that
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
