Virtual Tour
The Buildings
Urban Exploration
Pyestock Diaries

Engine Test Facilities - Issue Four

home | documentation | literature | engine test facilities - issue four | engine test facilities

Engine Test Facilities

3.1 The Engine Test Facility

The Engine Test Facility has five altitude Test Cells which can arrange engine and component tests at ground level or under simulated altitude conditions. Also included in the Facility is a general purpose sea level test bed.

The Cells can test a wide range of engines at various flight speeds and altitudes. Engineering experience in the gas turbine field has shown that performance presentation of altitude, speed and range can be somewhat misleading because the limits do not depend on plant alone, but are also related to the engine demand; consequently, any performance diagram relates to a particular engine choen to present the case and a different test envelope may result if another engine is chosen to show plant output. Where limited information on performance is given in this brochure it should be relasied that the information provides the broad trend of Cell capacity rather than specific detail of total N.G.T.E. capability and further information should be sought, as indiciated in Section 1. An outline of the testing capacity of the five different Test Cells is given in Table 3.1.

Table 3.1
Capacity Of Plant At N.G.T.E. For Engine Testing

Generally, Cell 1 carries out free jet testing and Cell 2 connected jet testing; both Cells having altitude capabilities of 3.5 inches of mercury absolute. The two Cells are sited adjacent to each other and have many common services. Cell 3 carries out connected testing over a much wider range of inlet air and altitude conditions and has air processing equipment which can reproduce temperatures from -70°C to 600°C. Cell 3 West is also used extensively for simulating icing conditions on full-scale helicopter fuselages. Wet and dry icing clouds can be produced.

Cell 4 carries out free jet tests on engines with complete supersonic intake configurations.

Figure 2.3 shows a diagrammatic arrangement of the air distribution network and the relative position of the test cells, while Section 2.1 gives a description of the various air supply installations.

3.2 Cell 1 and 2 Test Plant

Test Cells 1 and 2 are used for full-scale testing of turbojet and ramjet engines as well as for component development work under sea level and simulated altitude conditions. Cell 2 has also been used for the testing of solid propellant rocket motors at altitude. Figure 3.1 shows the layout and it can be seen that the two cells are sited adjacent to each other. Cell 1 is for free jet testing and is provided with a means of varying the incidence of the supersonic blowing nozzle as shown in Figure 3.2; Cell 2 is normally used for connected rig or combustion chamber testing and Figure 3.3 shows this type of installation when used in conjunction with sea level exhaust conditions.

Fig. 3.1 Layout of E.T.F. Cells 1 and 2

Fig. 3.2 Cell 1 set up for high altitude conditions

Fig. 3.3 Cell 2 installation of turbo-jet engine

The compressed air supplies from either the Air House G.E.C. sets or the Plant House Metropolitan-Vickers sets are connected to the Cells as shown in Figure 2.3. The pipeline connecting the Air House to the Cells is 54 inch diameter and this restricts the air capacity in

Cells 1 and 2 to the output from three G.E.C. sets. The standard arrangement is to use air from either 1, 3, 5 or 7 machines, but if these particular sets are not available, it is possible to make special plans to supply Cells 1 and 2 from the other compressor/exhausters using the interconnecting cross-over main originally provided to allow the G.E.C. sets to operate in cascade at 1:81 pressure ratio. This improvisation restricts the simultaneous operation of Cells 1 and 2 with Cell 3. Figures 3.4 and 3.5 show photographically the general arrangement of the air mains and cells.

Fig. 3.4 View of Cells 1 and 2 showing air mains

Fig. 3.5 General view of Cells 1 and 2

In addition to supplying the needs of engines under test, pressure air is used to operate ejectors which create altitude conditions inside the Cells. Sometimes, when high altitudes are simulated, the pressure of the air entering the test engine is less than atmospheric and consequently the engine requirements are drawn from atmosphere. For this purpose the Cell plant incorporates a silenced air-intake through which the engine draws its incoming supply.

To drive the ejector plant, air of 9:1 pressure ratio from the G.E.C. compressors is used to supply the primary side of four separate air ejectors. The individual performance of one of these ejectors is given in Table 3.2 and is shown graphically in Figure 3.6. Each ejector consumes 100 lb/s of G.E.C. driving air. The overall altitude performance of the ejector plant, with all four units operating, is shown in Figure 3.7.

Table 3.2
Performance on Single Ejector No.1 and No.2 Cells

Cell pressure inch Hg abs. 12 10 8 6 4 3.5 3.2
Air weight flow lb/s 36 32 25 19 10 8 4
Four similar ejectors installed.

Fig. 3.6 Performance of Cells 1 and 2 ejector plant

Fig. 3.7 Altitude performance of Cells 1 and 2

The test envelope shown has been calculated using the demand of an Olympus 320 engine. When test conditions so demand, the engine inlet air can be preheated in a 3 MW heater. The unit is built in two 1.5 MW sections each of which can operate independently. The total heater performance is shown in Figure 3.8, the maximum air delivery temperatures being 350°C. Inlet air temperatures to the heater depend on the mode of operation used, but when G.E.C. air is involved they can be as high as 210°C.

Fig. 3.8 Calibration of Metropolitan-Vickers 3 MW air heater

When air is supplied direct to Cells 1 and 2 from the Air House, the minimum temperature that can be obtained from the G.E.C. after-coolers is 70°C. However, by indirectly routing the air through the Cell 3 cold air plant, a lower temperature of 30°C can be obtained; with wintertime low ambient temperatures this figure can be reduced to 10°C. The maximum air weight flow at this low temperature is 200 lb/s.

Both Cells have an air filter fitted in the inlet pipework to handle a throughput of up to 450 lb/s at a maximum air temperature of 250°C. The filter's elements are constructed wither perforated steel cages supporting glass fibre filtration media. The design is suitable for temperatures ranging between -70°C to 250°C and removed 99.5% of all 0.5 micron particles. The filter elements are normally changed when the pressure drop reached 6 lb/in2. The filter pressure vessel is capable of withstanding 12 atmospheres.

The plant fuel system, as distinct from the test engine fuel system, installed in Cells 1 and 2 delivers fuel at flows up to 12,000 gal/h at pressures up to 1,400 lb/in2. A secondary plant fuel system is also available which delivers 6,000 gal/h at pressures up to 2,000 lb/in2.

Fig. 3.9 A typical engine installation within the free jet cell

Fig. 3.10 The main control room for Cells 1 and 2

Engine exhaust cooling is achieved by direct water injection through spray nozzles. Two systems are available giving maximum injection rates of 180,000 gal/h at 80 ft head and 190,000 gal/h at 125 ft head respectively. The maximum amount of cooling water injected for a particular test is governed by evaporation losses and the capacity of the Cell water extract pumps, currently 216,000 gal/h. A total water storage capacity of 330,000 gallons is available and the Cells operate on a re-circulatory basis so that only evaporation losses need to be made good.

Engine operation during transient acceleration cycles can be measured on U.V. recorders in addition to the usual performance observations under steady state conditions. All engine test information from Cells 2 is normally fed through to an on-line ICL1904S* computer via a PDP 11/10 computer for data acquisition. Full details of the data gathering systems are available in Section 4.

3.3 Cell 3 Test Plant

Unlike Cells 1 and 2 which are driven by ejectors, Cell 3 uses the Air House G.E.C. and the Parsons No. 9 and No. 10 exhausters to produce altitude conditions. Most tests also demand that compressed air shall be blown at the engine under test and this service is also provided by the Air House plant in its compressing role (see para 2.2.) An extensive system of air pipework circuits closely integrates Cell 3 with the Air House installation and, as shown in Figure 2.3, there are different circuits for compressed air supply and sub-atmospheric exhaust gas streams. The diagram particularly shows how the high pressure pipes supply compressed air to the forward end of Cell 3 while the suction mains are connected to the opposite end. For each of the eight compressor/exhauster sets there is a pressure main of 54 inches diameter and a suction main of 84 inches diameter. In every test a certain number of compressors (according to engine demand) supply inlet air to the engine under test whilst a matched capacity of exhausters suck away the exhaust gases. No. 1 G.E.C. compressor is not connected directly to Cell 3 but this particular limitation is overcome, if machine availability makes it essential to use the No. 1 set, by routing the air through a cross-over main and using this circuit to supply the pressurised air to Cell 3.

To reduce the emitted noise the whole structure of Cell 3 is submerged below ground level in a concrete trench as shown in the photographs and diagrams of the area and thus there is little outward appearance of the complicated engineering installation which forms Cell 3.

Figure 3.11 shows a cross section through the Cell from air inlet to suction exhaust manifold and a general view of the area is illustrated in Figure 3.12. The total length of the cell is 318 ft of which the working section accounts for 56 ft, the diffusing section 64 ft, the cooler and flame trap sections 90 ft, the rest being accounted for by the suction main manifold. The diameter of the working section with its inlet plenum chamber is 20 ft and it has a removable section in the roof so that test engines and associated equipment can be lowered into the Cell. The diameter of the diffuser and cooler sections is 10 ft and 30 ft respectively.

Fig. 3.11 Cross-section of Cell 3 test plant

Fig. 3.12 General view of Cell 3 area

When originally designed, the air inlet ducting of Cell 3 limited the throughput to 400 lb/s; however, modifications to the inlet ducting have since increased the throughput to 600 lb/s. The maximum allowable temperature has been reduced for this higher flow becuase the total cooling capacity of the gas cooler remains unaltered.

As described in para 3.1, the operating performance of a test cell is intimately associated with the engine demand. However, Figure 3.13 shows the test boundaries which can be obtained for Cell 3 for a typical 400 lb/s gas turbine engine with reheat. This case will be very close to that which would be obtained with the Olympus 593 engine for Concorde. Pictorial illustrations of engines as mounted in the Cell are shown in Figures 3.14 and 3.15 whilst Figure 3.16 shows the engine control room layout.

Fig. 3.13 Typical test envelope for a 400 lb/s engine with reheat

Fig. 3.14 View of Cell 3 engine installation

Fig. 3.15 Concorde Olympus 593 mounted in simulated altitude nacelle in Cell 3

Fig. 3.16 General view of control room for Cell 3

Cell 3 is designed to cater for a wide range of turbine engine performance and hence facilities are provided to give an extensive air temperature range. Both hot and cold air processing plants are installed which can be used to set up the required air temperature within the working range. Figure 3.17 shows the air temperature circuits diagrammatically. The inlet air temperature field that can be obtained in Cell 3 is shown in Figure 3.18.

Fig. 3.17 Diagrammatic arrangement of Cell 3 air mains

Fig. 3.18 Cell 3 inlet air temperature field

The air supplied through the various routes to the engine intake is filtered and additional filtration using N.G.T.E. developed elements, was installed in 1979 for use with advanced engines which employ very small air cooling passages. The filter elements are constructed with perforated steel cages supporting glass fibre filtration media. The design is suitable for temperatures ranging between -70°C to 250°C and removes 99.5% of all 0.5 micron particles.

Engine thrust can be measured in Cell 3 and Figure 3.19 shows the oil borne bearing frame which supports the engine, permitting the movement on which thrust measurement depends.

Fig. 3.19 Cell 3 engine test frame

A major part of Cell 3 installation is devoted to engine exhaust gas cooling. The gas cooler consists of 20 similar elements arranged radially within the 30 ft diameter pressure shell, as shown in Figure 3.20. The pressure shell is 1.25 inches thick and designed to withstand full vacuum and an explosion pressure of 127 lb/in2. The engine exahust gases are first passed through a kerosene burning torch system which ensures than any unburnt fuel is ignited, thereby eliminating the explosion risk. Then the exhaust is cooled in an indirect cooler which can handle a throughput of 320 lb/s from 1,700°C to 150°C at pressures between atmospheric and 4.5 lb/in2 abs. At low flow the cooler will handle 35 lb/s at 0.5 lb/in2 abs., at this condition the design pressure drop is 1.4 inches of mercury. A further 100°C temperature drop to the G.E.C. maximum inlet temperature of 50°C is achieved by evaporative cooling in a number of water-washed flame traps which form the final stage of the cooling system. Treated water is used in the cooling circuit, the maximum flow being 2,500,000 gal/h on a closed circuit using five cooling towers working between 27°C and 57°C as described in Section 5.

Fig. 3.20 Cell 3 gas cooler

If desired, the G.E.C. sets may draw air from the Ceca air drying plant so that Cell 3 tests can be conducted with dry air. Alternatively, an injector in the cell air supply system enables up to 100 lb/s of dry air to be directly induced from the Ceca air dryer

Three fuel systems are available for use in Cell 3. Two of these supply fuel at ambient temperature but the third, described in para 3.7, supplies fuel at temperatures up to 250°C. The capacities of the two ambient systems are 6,000 and 12,000 gal/h at 15 to 100 lb/in2. Both systems operate in a spill arrangement which enables rapid acceleration tests to be carried out at susbstantially steady supply pressure.

Cell 4 is supplied with fuel from branches in the Cell 3 6,000 and 12,000 gal/h systems.

Engine operation during transient acceleration cycles can be measured in addition to the usual performance observations under steady state conditions and for this purpose engien test information is normally fed through an on-line ICL 1904S* computer. Further details of the data gathering systems are included in Section 4.

3.4 Cell 3 Air Heating Plant

The Cell 3 heater is shown in Figure 3.21 and consists of a vertical oil fired cylindrical furnace, 80 ft hight and 20 ft diameter, with the walls lined with tubes through which the air supply is passed. Heat is tranferred to the air tubes by radiation from the 12 oil burners which are fitted on the floor of the furnace and arranged to fire upwards. The air inlet mainfold is mounted at the top and the outlet manifold is situated under the furnace floor, thus the air flow direction is downwards. The heater is arranged to be manually operated during the starting cycle and switched to automatic control once the desired operating temperature has been reached. The heater performance data is given in Table 3.3.

Fig. 3.21 Cell 3 air heater

Table 3.3
Air Heater Performance Data

The air is routed from the heater to the Cell, as shown in Figure 3.17, by way of stainless steel ducting. Air leaving the heater can be controlled over the range 300°C to 600°C as dictated by the test in progress and final temperature control is obtained by mixing heated and unheated G.E.C. air in the mixing sphere.

3.5 Cell 3 Cold Air Plant

The cold air plant works in conjunction with a precooler and pressure dryer to reduce ice formation; temperatures down to -70°C can be obtained by use of a cold air expansion turbine which is designed to give a maximum flow of 100 lb/s. Temperatures between ambient and -70°C can be achieved by mixing warm air from the G.E.C. machines with that from the cold air turbine in the air mixer which has special large valves to meter and blend the supplies.

When necessary, ambient temperature air can be drawn into the air mixer and blended with the cold supply; the circuits which make this possible are shown in Figure 3.17.

Figure 3.22 shows the temperature/air weight flow relationship obtained from the cold air plant and the cold air expansion turbine alternator is illustrated in Figure 3.23. The power generated by the air expansion turbine is absorbed by an alternator of five megawatt output and the electrical current is returned to the N.G.T.E. grid supply.

Fig. 3.22 Cell 3 cold air plant temperature characteristics

Fig. 3.23 Cell 3 cold air turbine and alternator

3.6 Cell 3 Ice-making Facilities

Ice particles in air-breathing engine intakes can have a serious effect on performance. Consequently, two special ice-making plants are installed in association with Cell 3 so that all altitude effects can be investigated. The two plants are:

  1. The ice crystal plant, and
  2. The super-cooled water droplet plant.

The former has the capacity to produce ice blocks which are cut into small particles and injected into the engine under test. Ice blocks can be produced at teh rate of one ton every twenty-four hours and facilities exist for storing four to five tons of ice for an indefinite period. The plant produces particles between 500 and 4,000 microns in size. The particle injection rate is variable between 5.5 and 152 lb/min.

Alternatively, super-cooled water droplet plant can produce a spray of super-cooled water droplets within the engine inlet duct. These will freeze out on contact with the cold engine inlet surfaces and so build up ice formations. The basic air and distilled water system can deal with a water injection rate of up to 600 gal/h. A typical spray rake can pass over 200 gal/h with a median droplet diameter of 20 micron.

3.7 Cell 3 Hot Fuel Supply

To simulate engine operating conditions at high speed in test plant at ground level it has become necessary to preheat engine fuels to correspond with flight conditions where elevated fuel temperature occur when fuel is used to cool the aircraft structure and services.

A plant, located near Cell 3, has been installed to provide a service to heat fuel up to 250°C and pipelines have been laid to service Cell 3.

Mineral oil, heated in two oil fired heater units to temperatures up to 300°C, is passed through coils in a 9,000 gallon capacity dwell tank to heat the engine fuel up to 100°C. The engine fuel is pumped from this dwell tank to Cell 3 via heat exchangers in which the fuel is given a second stage of heating by the mineral oil to raise the fuel temperature to the required level. The system supplies an engine under test with up to 40,000 lb/h of fuel at temperatures up to 250°C and pressures up to 160 lb/in2. Higher fuel flows up to 60,000 lb/h are possible but at correspondingly lower temperatures. A spill system, incorporating a dump cooler, is used and transient flow conditions are possible with substantially steady inlet pressures.

The circuits used for fuel at temperatures above 100°C, including the heat exchangers, are constructed from 18/8 stainless steel.

This system is also used with the heater units unfired to supply Cell 3 with ambient temperature fuel.

3.8 Cell 3 West

Cell 3 West was constructed in 1969 at the West end of Cell 3 suction manifold to provide the large diameter engine chamber required for simulated flight altitude testing of fan engines such as the Rolls-Royce RB 211. The triple shaft design and high by-pass ratio of these engines create a test demand which absorbs the whole of the present N.G.T.E. exhauster resources, namely all eight G.E.C. and the Parsons No. 9 and No 10 machines. (See Figure 2.2).

Cell 3 West assists the commerical exploitation of large fan engine design by enabling full flow cold air tests to be undertaken at flight operating altitudes and under icing conditions. The performance of Cell 3 West operating with the RB211 engine is shown in Figure 3.24 while Figure 3.25 shows a Rolls-Royce RB211 engine being installed.

Fig. 3.24 Cell 3 West typical test envelope with Rolls Royce RB211 engine

Fig. 3.25 Installing a Rolls-Royce RB211 engine in Cell 3 West

A cold air plant capable of cooling an air capactiy corresponding to the maximum RB211 engine requirement has been built as part of the Cell 3 West scheme. This cooler, described in paragraph 3.9 uses a 30% aqueous ammonia solution precooled to -50°C in two cold stores which may be used separately or in series or in parallel operation. In series operation one takn is brought on line to supply the air cooler, followed by the second tank when the "cold" in the first tank has been expended. In parallel operation both tanks are on line together to supply the air cooler. At maximum design conditions there is sufficient "cold" stored to permit the RB211 engine to run for up to 60 minutes but with less arduous air flow and temperature demands the test duration can be proportionately longer. The engine chamber can be run both with and without cold air; in the latter case air is drawn either direct from atmosphere through an intake silencer or through the cooler with no coolant circulation. In all cases the engine is directly coupled to the intake so that all engine tests are of the connected type.

The large diameter engine chambre has also been used extensively to accommodate simulated wet icing testing of full scale helicopter rotors. The Rolls-Royce Olympus 593 engine with a Concorde intake has been tested under simulated wet icing conditions. All tests were conducted in the free jet mode.

Provision has been made on the engine chamber for a pressure connection from the G.E.C. machines in order than the chamber can be converted to a blown test bed if required.

Figure 3.26 shows a general view of Cell 3 West whilst Figure 3.27 shows the front of the Cell with the end dome and inlet ducting removed to expose the inside compartment and a RB211 engine installation. The removal of the end dome permits easy access for installation of the engine.

Fig. 3.26 General view of Cell 3 West

Fig. 3.27 View of Cell 3 West engine chamber with the front dome removed

Figure 3.28 shows the cell end dome in the testing position with the inlet duct connected to the inlet control valve and air cooler. The Cell 3 West engine chamber is 25 ft diameter and 40 ft long and an extension of the Cell 3 exhaust manifold, which is 15 ft diameter, is directly coupled to the exhaust end of the cell. Cell 3 West has been built at ground level unlike Cell 3 and consequently the exhaust manifold passes through two cascaded right angle bends in order to accommodate the height difference. The Cell 3 West exhaust manifold is fitted with bulkhead doors which are shut in non-testing periods when the suction mains may be in use for Cell 3 and Cell 4. The atmospheric to exhaust mainfold pressure difference helps to keep a pressure tight seal. Figure 3.29 shows the arrangement of the engine chamber relative to the eight G.E.C. suction mains and it should be noted that a similar isolating bulkhead is located on the Cell 3 end of the manifold so that installation and modification work can proceed in Cell 3 whilst Cell 3 West is operational.

Fig. 3.28 View of Cell 3 West engine chamber with the front dome in the testing position

Fig. 3.29 Cell 3 West Test Area

To maximise exhauster performance, an internal exhaust gas diffuser is sited in the parallel section of the exhaust duct immediately downstream of the engine exhaust nozzle. The vertical exhaust duct from Cell 3 West which joins the exhaust manifold is fitted with direct injection water cooling sprays and an inbleed valve for altitude trimming purposes. A further inbleed valve is fitted in the exhaust manifold downstream of the isolating bulkhead.

It is possible to measure engine thrust in Cell 3 West as the engine mounting includes special features to permit the movements required for thrust and drag measurements. The engine is clamped to a support frame which is itself supported from flexible rods attached to the roof of the cell. The mounting permits the small deflections which are needed to measure thrust and draft on Davey United Ltd. load cells*. The engine installation includes an automatic-connect bulkhead arrangement so that a large proportion of engine instrumentation and other service supplies to the engine are automatically made when the engine is lifted into its finally installed position. This arrangement is identical to sea level installations at Rolls-Royce, so that it is possible to make speedy interchange of engines between sea level and altitude test facilities.

* The thrust measuring system will be changed in 1982 to an improved arrangement, using Bofors shear force load cells.

The inlet ducting which connects with the cooler is 77 ft long, of which 43.5 ft is external to the cell and the rest internal. This ducting includes an air flow measuring section as well as air straightening gauzes which improve the air distribution at the engine entry plane. A louvre type air inlet control valve is fitted between the air cooler and the cell external ducting.

Both steady state and transient instrumentation is available and test information is normally fed through the on-line ICL1904S* computer. Although the instrument installation is able to provide 700 pressures and 800 temperature channels for steady state conditions although only 1000 can be addressed at one test point. Thirty-six channels of U.V. and 32 channels of magnetic tape are available for transient conditions. A control room adjacent to the cell is used to operate the test plant, the engine and its associated auxiliaries. Further details of the instrumentation services is given in Section 4.

The fuel system is tapped off from that used in Cell 3 and is of similar design. A spill return line is situated close to the engine supply tee-off and fuel in excess of that required by the engine is returned to the plant fuel supply tank. This arrangement eliminates the need to accelerate large quantities of fuel through the long supply pipework during engine transients.

The spray cooling water system operates on a recirculatory basis with a make up supply to replace water evaporated into the air stream. Two banks of spray nozzles are provided and each is fitted with on-off and flow control valves. The quantity of water injected is carefully controlled to ensure than just enouhg is sprayed into the gas stream to maintain an acceptable gas temperature at inlet to the exhausters.

3.9 Cell 3 Cold Air Plant

The air intake cooler consists of 33 modules built in three rows of eleven modules. Each module has 18,500 ft run of 1 in o.d. x 16 B.G. mild steel pipe of plain section. The tubes are galvanised externally and the module headers are interconnected toegether in such a way as would allow the cooler modules to be removed for emergency repairs. Altogether, when all modules are interconnected, the cooler has a total length of 61 ft 5 in with frontal dimensions of 27 ft x 29 ft 6in. The assembly is mounted on its own wheeled carriage and the cooler can be wheeled out of position along its own track if required. Figure 3.26 shows the air cooler in front of the cell, the two cold storage tanks, the smaller defrost tank and the refrigeration plant. The cold storage tanks are nominally 43 ft high x 25 ft diameter.

Fig. 3.30 View of Cell 3 West cooler assembly

Each 500 tonne cold store of aqueious ammonia can be reactivated in twenty-four hours using the refrigeration plant which is sited adjacent to the cold store tanks. This refrigeration plant includes the necessary condensers, evaporators, pumps, etc., and is of conventional design rated at 2.25 million B.t.u./h. A simplified diagram of the cold air system is shown in Figure 3.32

Fig. 3.32 Diagram of Cell 3 West cold air system

Coolant from the pre-cooled store of aqueous ammonia is circulated through the cooler which is divided into three stages for this purpose. Each stage of the cooler has an independent flow control system and coolant is recirculated wherever possible to ensure maximum use of the "cold". Eventually the coolant is returned to the cold store where an interface, formed between the "warm" and "cold" coolant, travels down the tank as testing proceeds. The duration of a run is dependent on the rate at which coolant is drawn from the tank(s) and the position of the interface. Under some circumstances the length of test run may be limited by the falling off in heat transfer coefficient due to ice building up on the outside of the cooler tubes. The second cold store tank was installed in 1980. This tank has been designed to be used independently, or in series or parallel operation with the original tank.

The cooler has been designed to give thirty minutes running time at -37°C with an air throughput of 800 lb/s assuming air inlet conditions of 7.3°C dry bulb temperature and 100% relative humidity using one storage tank. The estimated performance of the cooler using one cold store tank at various air inlet conditions is shown in Figure 3.31. When using two cold store tanks in series or parallel operation the running time will be limited by pressure losses through the air cooler due to ice formation on the tubes. The rate of ice formation will be determined by the atmospheric temperature and humidity conditions prevailing at the time of the test. Optimum cooler performance is achieved by use of a computer program which predicts the cooler performance over a wide range of operating conditions.

Fig. 3.31 Cell 3 West estimated performance of air inlet cooler using one cold store tank

3.10 Cell 3 West Wet and Dry Ice Facilities

Wet icing conditions may be simulated in Cell 3 West by injecting supercooled de-mineralised water spray droplets into the low temperature airstream entering the engine. The droplets remain in the liquid state until striking a solid surface, whereupon they freeze. The resulting ice build-up is recorded photographically and by means of closed circuit television monitors, with video tape-recording facilities. Stroboscopic lighting allows the ice formation on rotating blades to be studied.

The de-mineralised water is injected through 242 spray nozzles fitted in a 93 inch diameter spray rake located in the engine approach ductwork. The spray nozzles are of an air blast atomisation type, manufactured to a Rolls-Royce design, with interchangeable water jets. The atomisation air is heated by a stream heat exchanger and the air temperature is regulated to a minimum temperature of approximately 5&dec;C at the spray nozzle exit, to prevent freezing of the nozzles and water droplets. The water droplet size (volumetric mean diameter) is variable between 20 and 40 micron.

A dry ice facility is available to supplement the wet icing system. The dry ice equipment cuts pre-formed conditioned ice blocks into small crystal particles and injects these particles into the engine inlet airflow from a station in the cell external ductwork, downstream of the air cooler. The mean particle size range is from 500 micron to 2,000 micron with an injection rate of 30 lb/min to 125 lb/min, based on a mean particle size of 1000 micron. Mixed icing conditions may be simulated by combined operation of the wet and dry icing facilities.

In addition to engine testing in the nromal connected mode, with or without icing facilities, Cell 3 West has also the capacity to be used as a free jet facility for simulated icing tests. Figure 3.33 shows a Sea King helicopter fuselage being installed in Cell 3 West in preparation for icing tests whilst Figure 3.34 shows the resulting build-up of ice in the region of the engine air intakes. Figure 3.35 shows the ice formation on the fan blades of a Rolls-Royce RB211 engine during a wet icing test.

Fig. 3.33 Sea King installation in Cell 3 West

Fig. 3.34 Sea King icing test in Cell 3 West

Fig. 3.35 Ice-accretion on the fan blades of a Rolls-Royce RB211 engine

3.11 Cell 4 Test Plant

When gas turbine engines are chosen to operate in supersonic aircraft, there is a need to test the engine in close association with the aircraft air intake system which is designed to reduce the air speed at the compressor face to an acceptance velocity thus converting the forward ram velocity into high engine intake pressure with a resultant bonus in power plant performance.

Although the engine an intake problems arising in supersonic flight are inter-related, a true simulation of all relevant factors is only achieved if full-scale free jets are possible; Cell 4 provides this capability for engines such as the Rolls-Royce Olympus 593 engine for Concorde and it is possible to test this engine and similar power plants with their intakes and control systems under high speed flight conditions. The Cell provides a means of observing on the ground the interaction of the intake and engine at changing altitude, Mach number and incidence. The performance range of Cell 4 is given in Figure 3.36. Figures 3.37 and 3.38 show actual views of the Cell 4 plant and associated machinery and some idea is obtained of the size and engineering complexity of full-scale free jet tests. In addition to testing in the free jet supersonic mode, the cell has also been used for subsonic free jet testing of the engine and intake. Also, by removing the blowing nozzle, connected testing of engines with reheat has been successfully undertaken with an experimental thrust measurement system installed in the engine mounting frame. Figure 3.39 shows the Rolls-Royce Olympus 593 engine installed in the Cell 4 engine chamber; this view is taken from the engine tailpipe end with the propelling nozzle in the immediate foreground, looking forward to the bulkhead which divides the engine chamber from teh intake working section. Figure 3.40 shows the air circuit diagram for Cell 4 which is designed to handle a maximum flow equivalent to two G.E.C. sets, i.e. 400 lb/s, under pressure conditions, with the capability of increasing this flow to 600 lb/s at lower pressures by the use of an injector system.

Fig. 3.36 Performance envelope for Cell 4 plant

Fig. 3.37 General view of Cell 4

Fig. 3.38 Cell 4 plenum chamber

Fig. 3.39 Rolls-Royce Olympus 593 for Concorde installed in Cell 4 engine capsule

Fig. 3.40 Cell 4 process air diagram

The total air mass flow can be provided as dry air from the Ceca air dryer which is described in para 2.8. Normally, the Cell air supply maximum inlet pressure does not exceed three atmospheres absolute and the temperature ranges from 70°C to 470°C using the air heater described in para 3.4. The Cell is built parallel to Cell 3; the relative position of these plants is shown in Figure 2.3 which also shows the inter-connecting ducting with the G.E.C. and No. 9 and No. 10 exhausters.

Flight speed is simulated by a supersonic blowing nozzle which has an adjustable throat and wall contour to provide variable Mach number operation. This nozzle is mounted on a universal carriage which enables aircraft pitch and yaw conditions to be simulated up to angles of +/- 10° at rates of up to 20° and 10° per second respectively. The total weight of the nozzle and moving carriage assembly is approximately 75 tons. Figure 3.41 shows the carriage assembly with the 25 square foot nozzle in position viewed through the plenum chamber with the 20 foot diameter inlet cone removed.

Fig. 3.41 Cell 4 carriage assembly and supersonic blowing nozzle

To match engine performance more closely two variable nozzles are available; the first is 12 sq ft in area, giving a Mach number ange of 1.5 to 3.5 while the second is 25 sq ft in area giving a Mach number range of 1.7 to 2.5.

For subsonic testing the variable supersonic blowing nozzle is replaced by a non-variable sheet steel blowing nozzle which is designed to meet the requirement of the particular cell installation.

A general cross sectional view of Cell 4 is shown in Figure 3.42. The diagram illustrates how the aircraft intake is mounted in the working section immediately downstream of the blowing nozzle with the engine coupled behind.

Fig. 3.42 Cross section of Cell 4 altitude test plant

Spill diffusers, which are shown more clearly in Figure 3.43, are mounted above and below the intake and are used to obtain altitude conditions in the working section by converting low pressure, high velocity air from the nozzle into higher pressure, low velocity air. As the downstream end of the spill diffusers are connected to exhausters it can be seen that the diffusers are a means of aiding the exhausters to obtain altitude. Changes in altitude are obtained by varying the geometry of the diffuser.

Better use of the installed exhauster capacity is obtained by sucking away the low energy air issuing from the supersonic blowing nozzle, normally about four per cent of the nozzle mass flow. This process is termed "working section bleed" and is performed by the No. 9 machine referred to in para 2.3.

Figure 3.43 shows diagrammatically the flow paths in the supersonic nozzle, spill diffuser and engine intake. The geometry of the Concorde intake makes it necessary to provide additional ductwork to accommodate the "ramp bleed" and "dump door" flows which are an essential feature of the Concorde aircraft power plant. The ramp bleed flow is piped away from the upper part of the intake and returned to the working section to be sucked away by No. 9 exhauster machine. The dump door flow is sucked away through one of the two permanent plant blow-off ducts to join the main exhaust system further downstream.

Fig. 3.43 Cross section of Cell 4 intake assembly, and flow ducts for Rolls-Royce Olympus 593

As in all altitude cells, the heat exchanger needed to cool the engine exhaust is a major plant item. The Cell 4 gas cooler has two stages, the first stage cooler is shown pictorially in Figure 3.44 which also shows the six flame torches that ignite any unburnt fuel in the engine exhaust so that explosion risks are eliminated. The exhaust gases leave the water-cooled diffuser duct at temperatures up to 1,700°C to enter the first stage cooler where they are cooled to 1,000°C by a novel gasover-tube matrix; a water flow of 620,000 gal/h is circulated through the matrix. In the second stage cooler, where the gases pass through the tubes in a conventional manner, the temperature of the gases is reduced to 150°C. This second stage also has a water circulation of 620,000 gal/h. Additionally, the flow from direct injection water sprays is used as evaporative cooling to reduce the gas temperature further to 50°C, this latter temperature being the maximum for the inlet flow to the G.E.C. exahusters. The excess water from all the direct injection sprays drains from the air ductwork into a 40 ft deep barometric well. The well is divided into two sections. Clean water is drained into one section from which it is pumped back to the water storage ponds, and water that has been contaminated by the exhaust gases is drained into the second section from which it is pumped into the site sooty water plant described in Section 5. The total recirculating flow of all cooling water systems in Cell 4 under hot running conditions is 2.5x106 gal/h. Losses up to 170,000 gal/h through evaporations, leakages and contamination are incurred.

Fig. 3.44 Cell 4 first stage gas cooler matrix

The exhaust cooling system is manufactured entirely from conventional ferrous materials. During non-running periods corrosion is arrested by filling the system with a solution of sodium suplhite and sodium hydroxide.

Altogether, Cell 4 has a total length of approximately 400 ft of which the 30 ft diameter inlet plenum chamber with its blowing nozzle contributes 36 ft and the working section 10.5 ft. The engine chamber itself has a diameter of 10 ft and a length of 16 ft and the exhaust diffuser 55 ft; the first and second stage coolers have diameters of 26 ft and 18 ft respectively, the total length of the two stages is 160 ft, whilst 105 ft of ducting at the rear end of the cell accommodates the seven G.E.C. exhauser connections.

The fuel system is teed off that serving Cell 3 and described in para 3.3. As both cells do not run at the same time, no inconvenience is suffered by having a common system. The fuel flow rates are 100 gal/min and 200 gal/min for engine and reheat systems respectively. Normally, the engine flight fuel pump is used on engines installed in Cell 4, therefore, the plant fuel system provides fuel to the engine low pressure pump at a set pressure of 25 lb/in2; however, if needed, fuel at pressures up to 100 lb/in2 can be supplied.

Fig. 3.45 Cell 4 control room with test in progress

3.112 Sea-Level Engine Testing Plant (The Glen Test House)

The facilities required for testing gas turbine engines under sea-level conditions are not so complicated as those needed for simulated altitude flight conditions which have already been described; however, userful and much less expensive development work can be conducted on sea-level static beds. Figures 3.46 and 3.47 show views of the New Site sea-level test bed which is known as the Glen Test House (Ground Level Engine Test House).

Fig. 3.46 External view of the 'Glen' test house

Fig. 3.47 Aircraft gas turbine installed in the 'Glen' test bed

The test bed was originally designed for engines of up to 28,000 lbf of thrust, but shear force load cells are now installed permitting thrusts of up to 44,000 lbf to be measured. Inleft air to the test bed is drawn through intake acoustic splitters which are designed to handle 1000 lb/s. Exhaust gases from the engine under test together with induced air are vented to atmosphere through a detuner. The detuner is fitted with water spray cooling, the water being supplied at a maximum flowrate of 42,000 gal/h at 60 lb/in2.

There are various fuel systems supplied to the test bed and fuel can be drawn from either the Tank Farm or from a bowser located at the Glen Test House. The normal low pressure fuel system is capable of supplying 2000 gal/h at 25 lb/in2, the reheat low pressure system can supply 21,900 gal/h at 35 lb/in2 and the reheat high pressure system can supply 10,000 gal/h at 1200 lb/in2.

Figure 3.48 shows a plan view of the Glen Test House and it will be noted that it includes a Fuel System Test Facility (FSTF). This facility consists of a laboratory and motor room housing a 225 h.p. variable speed electric motor coupled to a two-stage gearbox set and also a low pressure fuel supply system. The motor gearbox set is capable of driving a wide range of engine fuel pumps at speeds up to 50,000 rev/min. In addition to the above, the facility also comprises a control room and a clean room for the stripping/servicing of hydraulic and fuel system components. The basic purpose of the FSTF is to allow testing to be carried out on engine fuel pumps and fuel systems in isolation from the engine itself.

Fig. 3.48 Plan layout of the 'Glen' test house

© Procurement Executive, Ministry Of Defence