FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1947
1947 - 0956.PDF
556 FLIGHT JUNE 12TH, 1947 GAS TURBINE COMBUSTION a state oi random turbulence. A smallswirler immediately surrounding the burner was retained to provide the neces-sary stabilizing vortex. This arrange- ment was first employed on '' reverseflow" engines of the original Whittle pattern and for obvious reasons is termeda "colander" system. Applied to engines of the straight-through type, the colander was sur- rounded with an additional member, alsopierced with holes and situated a short distance away, the space between form-ing an equalizing zome. A further addi- tion was made in the form of a " snout,"so disposed as to pick air off from the stream over a section where its velocitywas approximately uniform. The size of the opening was chosen to form ametering orifice, although a. second orifice was usually incorporated. This is Fig. 1. Arrangement of typical combustion chamber show- ing "snout," additional orifice and two types of cooling air shoulder. shown in the general arrangement of achamber. Fig. i. The fuel is introduced into the com-bustion chamber as an "atomized" spray of fuel droplets. It would appearthat there may, for any prevailing set of conditions, be an optimum degree ofatomization. Extremely coarse atom- ization which results in unburnt fuelpassing through the combustion zone will, result in a calorific loss. Excessivelyfine atomization, resulting in the whole of the fuel being gasified before combus-tion takes place, may also result in a lowering of the combustion efficiency. Atomizers of the anvil type in which.i compact fuel stream impacts on a rigid object have so far met with little suc-cess. Atomization resulting from impact of the liquid stream upon the surround-ing gas, as in the combustion chamber of a diesel engine, is probably not sowell suited to the lower pressures pre- vailing in a gas turbine. Swirl Atomisers This method, although still used withsuccess on the Metropolitan-Vickers tur- bine, is somewhat unpredictable and theangle of divergence of the spray is in- sufficient to suit most combustion cham-bers Various constructions are possible forthe swirl atomizer used in the vast najority of present-day chambers. Themost common one is Fig. 2(A) where a series o. holes are drilled tangentiallyinto the swirl chamber which is closed at the rear end by an inserted plug. An-other is the plate type, Fig. 2(B), where the entry holes take the form of slotspierced in a thin metal plate. In a third type the entry is through a seriesof helical slots cut in a tapered plug. Fig. 2(C), and in which the enteringfuel has a slight forward component in addition to its tangential one. The tangential velocity increases fromits initial value at the entry slots to pro- gressively higher values as it approachesthe centre. As velocity would become infinite at the centre where the radiusis zero there must be a limit to which the stream can approach the axis andan air core is formed corresponding to that diameter at which the velocity issuch that all the^pressure energy in the fuel has been converted into kineticenergy. The fuel therefore traverses the swirlamber and the exit orifice in the form r >f a swirling column of liquid surround-ing the air core. At the point of emerg- ence from the ori-fice we have the condition that thetangential velocity increases as wemove towards the centre, whereas theaxial velocity in- creases as we movetowards the out- side. Any elemen-tal annulus of fluid issuing from theorifice would therefore spreadout along the sur- face of a cone, theangle of which is defined by theratio of tangen- tial to axial velocities. Considerationwill show that the apex angle of the cone will be greatest for the fluid near-est the centre where the tangential velo- city is greatest and least for that near ORIFICE PLATE ,\\\\\\\N\\\\\\\X the inner wall of the orifice, which meansagain that the inner elemental layers of liquid will tend to penetrate the outerones. An ideal fuel spray would 'consist of alarge number of droplets of uniform dia- meter. Actually it isv a collection ofdroplets whose diameters vary over a wide rangft. In a typical spray it is pos-sible to find a small number of 200 microns (o.oo8in) diameter while thesmaller droplets will range in size to below 10 microns (0.0004m) diameterCombustion will depend not only on thf SWIRL PLUG Three types of swirl atomizer: (A) Tangential hole : (B) Slotted plate ;(C) Tapered plug. Fig. 3. Temperature contour diagram at turbine entry (Deg. C.) drop sizes present but on their size dis-tribution. A normal spray consists of some 5 to 10 million fuel droplets perc.c. of fuel atomized. The droplet dia- meter range of 20 to 1 corresponds to amass range of 8,000 to 1. Unless a very large number of droplets is included inthe count, there is the possibility of large droplets being incorrectly estimated and,although these form only a small propor- tion by number, they form a large pro-portion by mass. The problem of igniting a kerosenespray differs from that of igniting petrol. A high-tension plug of modified form andwith a gap between the electrodes of the order of o.o6oin, must be supplied witha stream of sparks of sufficient duration and energy content to volatilize the fueldroplets before the resulting vapour can * be ignited. It is general practice toemploy a coil with a high-speed trembler to accentuate the inductance rather thanthe capacity component of the spark. This is the reverse of the ordinary prac-tice on piston engines. Torch Ignition On early designs where the plug wasarranged in the main fuel spray it was difficult to find a position where the plugwould give good ignition and yet not overheat and burn out in running. Re-cent practice, however, is to use the torch igniter. In this device the plug is ,,located well away from the combustion * zone and a small auxiliary atomizer isarranged adjacent to the plug points. The spray from this atomizer is pro-jected in the form of a torch into the main chamber, thus igniting the mainfuel spray. The time taken for the gases to traversethe combustion zone of a given engine is nearly independent of engine speedover the whole of its operating range; the increase in air mass flow with speedbeing compensated for by the increased density. Hence the only factors whichcan influence the time are the cross- section of the combustion chambers andtheir effective length. The true criterion is the dimensionless ratio of combustionchamber section area to nozzle guide vane area which, on representativemodern engines varies from 7 to 14. It is desirable that the temperature
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
