This invention relates to carburettor metering systems and wicks therefor for absorbing liquid fuel for vaporisation into an air flow.

Automotive engines rarely operate at full power, so the part load condition is of greatest importance. This requires accurate metering of the fuel over a very wide range of flow rates. It has been said that the only reason a conventional carburettor can survive is because engines are very tolerant of rich mixture. Thus the conventional system, which cannot provide accurate metering over a wide flow rate range, is designed to provide richer than ideal mixture at operating points - such as low load - when it is not accurate. This approach is not adequate for today's conditions when emissions and fuel economy are subject to legislative control.

It is known that substantial advantages are to be obtained, in terms of part load fuel economy and decrease of exhaust pollution, by operating a spark ignition engine with a fuel/air mixture having excess air over that required for just complete combustion of the fuel, that is with a lean mixture of fuel in air. As mixture is weakened carbon monoxide emission rate falls rapidly to a low level and then remains low. NOx production is a maximum for air fuel ratios of about 17:1 (14.7:1 is chemically correct) after which it falls

progressively. Unburned hydrocarbons fall progressively as mixture is weakened, down to a minimum after which they increase. This increase is caused either by very slow burning leading to flame extinction before completion or by occasional misfiring. Experiments show that improved ignition can hasten the whole combustion process to a degree and so postpone the increase of the leaner mixture. Improved ignition also minimises the risk of misfire. Conventional engines can readily tolerate excess fuel in the fuel/air mixture to a considerable degree. However, lean mixture operation requires precise control of mixture strength to ensure reliable operation without misfiring. Thus conventional carburettor systems are generally unsuitable for supplying engines operating at lean mixture strengths.

A carburettor system suitable for supplying lean mixtures of fuel in air is disclosed in British Patent Specification No. 1,595,315. This carburettor comprises an evaporator for evaporating the fuel into a stream of air and a closed-loop control arrangement for maintaining the mixture strength at a required value in dependence on the temperature drop measured across the evaporator. While such a carburettor is capable of operating adequately in a lean burn system, it has a fairly slow response time, typically of the order of 1/4 second, which can render the engine sluggish in operation. It would therefore be advantageous to

provide a carburettor metering system which provides accurate fuel/air mixtures, particularly lean mixtures, and adjusts the mixture quickly in dependence on load changes over a wide range of air flows. The present invention is also directed to an improved wick for a vaporiser usable in a carburettor metering system. One such wick vaporiser is disclosed in U.S. Patent No. 4,290,401, in the name of the present applicant, and which is incorporated by reference herein. Such a wick vaporiser comprises a plurality of suspended wick elements having bottom ends of unequal length suspended above a liquid fuel reservoir, with the number of wicks that are wet at any one time being dependent on the level of fuel in the reservoir. Temperature measuring means are provided in an air stream both before and after the air stream passes through the vaporiser. The amount of measured temperature drop of the air stream across the vaporiser indicates the latent heat of the liquid fuel and thus the amount of liquid fuel being introduced into the air stream by evaporation. The level of fuel in the reservoir is adjusted in a controlled fashion in response to the temperature drop. This wick vaporiser arrangement provides good closed loop control. However an improved wick construction which provides an improved vaporisation rate would be desirable.

In view of the foregoing it is an object of the present invention to provide a carburettor metering

system capable of accurately controlling fuel flow over a wide range of air flows and mixture strengths.

It is another object of the present invention to provide an improved wick construction which provides an improved vaporisation rate.

These objects are achieved by features as set forth in the claims.

In order that the invention may be more fully understood, reference will now be made, by way of example, to the accompanying drawings, in which:

Fig. 1 is an axial cross-sectional view of a fuel valve of a metering system according to the present invention;

Fig. 2 is a schematic diagram of a metering system for fuel and air in accordance with the present invention;

Fig. 3 is an alternative arrangement of a metering system for fuel and air in accordance with the present invention; Fig. 4 is a schematic diagram of an overall arrangement for a carburettor metering system according to the present invention, showing in particular an arrangement for adding hot dilution air;

Fig. 5 is a cross-sectional view of the wick chamber of Fig. 4 in greater detail;

Fig. 6 is a perspective view of the wick material of the wick chamber of Fig. 5.

Fig. 7 is a perspective view of the vortex

chamber of Fig. 4 in greater detail;

Fig. 8 is a perspective view of the operative end of a spark plug for use with the invention;

Fig. 9 is an enlarged perspective view of the electrodes of the spark plug of Fig. 8; and

Fig. 10 is a side view of the electrodes of the spark plug of Fig. 8.

Referring to Figure 1, the fuel valve 10 comprises a needle valve member 12 movable along a displacement axis 13 with respect to a valve seat 14. The valve 10 has a circular orifice 15 by way of which liquid fuel is injected into a stream of air. The valve member 12 has an outer cylindrical surface 16 and a tapered end defining a frustoconical surface 17. The valve seat 14 has an inner cylindrical surface 18 and a frustoconical surface 19 surrounding the frustoconical surface 17 of the valve member 12. The valve seat 14 is provided with a fuel inlet 20.

The conical angle of the frustoconical surface 17 matches the conical angle of the frustoconical surface 19, and the two surfaces 17 and 19 overlap one another so as to define therebetween an annular passage for the flow of fuel. It will be appreciated that the length of overlap L of the surfaces 17 and 19 will vary substantially in proportion to the width W of the passage as the valve member 12 is moved along the displacement axis 13 with respect to the valve seat 14. Thus the fuel flow through the orifice 15 for a given

pressure difference is proportional to the square of the width W which is in turn proportional to the degree to which the valve member 12 is lifted.

Figure 2 shows the fuel valve 10 connected to a conventional float chamber 21 for supplying liquid fuel to the valve 10. The fuel orifice 15 opens into an air duct 22 provided with a throttle 23. The valve member 12 is connected to a movable diaphragm 24 capable of being deflected in the direction of the arrows 25 to move the valve member 12 with respect to the valve seat 14 so as to vary the fuel flow through the valve 10. Also connected to the diaphragm 24 is a gate member 26 of an air control valve 27. The gate member 26 has a cylindrical wall having a plurality of triangular cut- outs 28 along its edge. Furthermore the gate member 26 fits within the cylindrical end of the duct 22 which defines a seat member 29 of the air control valve 27.

It will be appreciated that, as the diaphragm 24 is deflected in the direction of the arrows 25, the gate member 26 will be moved within the seat member 29 so as to vary the throughflow cross-section of the triangular cut-outs 28. Thus the throughflow cross- section of the air control valve 27 varies in proportion to the square of the degree of deflecting of the diaphragm 24, and hence the degree of displacement of the valve member 12.

The geometries of the orifices of the fuel valve 10 and the air control valve 27 are chosen so as

to ensure that, with a constant pressure difference across the air control valve, the mixture strength is independent of flow over a range limited only by manufacturing inaccuracy, and so that adjustment of the pressure difference can be used to adjust the mixture strength. In this regard the pressure difference across the air control valve 27 is used to effect lifting of the gate member 26.

With this arrangement the mixture strength is proportional to the square root of the pressure difference. If required the mixture strength can be adjusted by arranging for only an adjustable fraction of the whole pressure difference across the air control valve 27 to be used to lift the gate member 26 against its dead weight or a return spring. Typically it is necessary to adjust the pressure difference to provide variable mixture strength and compensate for changes in fuel viscosity and air density.

It will be appreciated that the particular geometries of the valve member 12 and the valve seat 14 of ^* the fuel valve 10 are advantageous because they ensure that pressure differences related to the viscosity of the fuel are substantially greater than pressure differences due to momentum changes of the fuel, and since the essential geometry of the fuel passage is maintained as the throughflow cross-section is varied.

The described carburettor metering system is

capable of accurately controlling the fuel flow over a wide range of air flows and mixture strengths, and is therefore particularly applicable to lean mixture operation. The system typically has a response time of the order of a tenth of a second.

Fig. 3 shows another embodiment of a carburettor metering system according to the invention. In this system, the fuel metering valve is mechanically connected to the air valve. In Fig. 3, a carburettor metering system 30 is shown wherein liquid fuel from a fuel pump (not shown) enters a fuel inlet 31 of a fuel valve 32. The fuel valve 32 has an outlet 33 connected to a fuel chamber 34 arranged as a pressure differential chamber as will be described. The fuel valve 32 may be arranged in the geometry according to Fig. 1 except that the exit is not exposed to the air. The fuel valve 32 has a valve member 35 mechanically connected and operatively coupled to an air valve member 36. The air valve member 36, shown schematically, is arranged as an inverted cup and has triangular cut¬ outs 37 arranged around its periphery. The inverted cup is received in a seat 38 having an annular chamber 39 connected to an air outlet 40. A pressure link 41 connects the air outlet 40 and the upper interior region 42 of the seat 38 to maintain the region 42 at equal pressure with air outlet 40. As is apparent from the air valve shown and described, the valve member moves

upwardly when inlet air pressure exceeds air outlet pressure present in air outlet 40 and upper region 42. As the air valve member 36 moves vertically upwardly, a greater portion of the triangular regions 37 will intersect the annular chamber 39, thus increasing air flow through the air valve. When outlet pressure decreases relative to inlet pressure, valve member 36 drops, thus decreasing air flow through the air valve. Movement of the valve member 36 vertically will cause the fuel valve member to move likewise due to mechanical coupling 43.

The fuel chamber 34 has an upper subchamber 34a and a lower subchamber 34b, separated by a diaphragm 44. The diaphragm 44 is flexibly mounted in the vertical center of the chamber as shown. Mounted on the diaphragm 44 is a fuel valve 45 which is received in fuel outlet seat 46 of chamber 34. A spring 47 is also provided between diaphragm 44 and the top of the chamber 34. The lower subchamber 34b is connected to the junction point of two adjustable restrictors 48 and 49 connected in series across the fuel inlet 31 and fuel outlet 33 of fuel valve 32. The pressure difference between subchamber 34a and 34b is a fraction of the total pressure drop across the fuel valve 32, the fraction being determined by the setting of "the adjustable restrictors 48 and 49.

In operation, excess fuel flow through fuel

valve 32 for a given position of fuel valve member 35 (and in turn a given position of air valve member 36 and hence given air flow) causes fuel valve 45 to move vertically upwards towards a more closed position, due to drop in pressure in subchamber 34a relative to subchamber 34b, and hence corrects fuel flow excess. The spring force from spring 47 is substantially constant since air pressure drop to lift air valve member 36 is constant. A fuel drain 50 disposes of leakage fuel, which, since it is relatively small in quantity, can be added to air flow since fuel has already passed metering fuel valve 35.

Restrictors 48 and 49 control the air/fuel mixture strength. Either one of these restrictors can be fixed. Both of these restrictors allow only a relatively small flow compared to fuel valve 32. Any flow through restrictors bypasses metering fuel valve 35 and no flow enters or leaves subchamber 34b except in transients. The best arrangement for metering and mixture monitoring is to operate the basic system at a constant mixture strength, because thermometers which are used to measure temperature drop (across an evaporator for example) do not respond quickly enough to changes. However, duty requires a lean fuel mixture at part load, and a richer full strength mixture at full load.

At part load it is advantageous to reduce the weight of charge taken by engine. Reducing the charge

by using throttling is not desirable because it lowers intake manifold pressure, needing more work from the engine and causing back flow of exhaust. One can advantageously heat the air/fuel charge with waste heat to reduce density and hence the weight without change (or substantial change) of pressure. Heating has to be rapid to cover transients.

Fig. 4 is a schematic showing an overall arrangement for a carburettor metering system according to the invention in the environment of a vehicle, for example, and shows in particular an arrangement according to the invention for adding hot dilution air. The schematic layout shows an air cleaner 55 providing air to an air valve 56 connected in turn to a mixture throttle 57. The mixture throttle 57 is connected to a fuel wick chamber 58 having a wick 59, which will be described in more detail below. The wick chamber 58 functions as an evaporator for the liquid fuel. The vaporised fuel/air mixture then goes to a vortex chamber 60, to be described in more detail below, and then to an intake manifold 61.

Also shown in Fig. 4 is an air path on the left, wherein air from air cleaner 55 is provided to a diluent throttle 62. The diluent throttle 62 is controlled by cam 63 whereas the mixture throttle 57 is controlled by cam 64, and both cams are controlled by the same camshaft turned by the accelerator pedal of a vehicle. The shapes of the cams would be determined by

engine testing.

The output from the diluent throttle 62 is connected to a heat exchanger 65, which may be physically located adjacent an existing heat source such as an exhaust manifold or exhaust pipe. The heat exchanger 65 is connected to a thermostatic valve 66 and then to the vortex chamber as will be described below.

The arrangement shown in Fig. 4 according to the invention provides dilution air downstream of an evaporator wick system. In this arrangement, since wick exit flow and diluent are both gas, the proper proportion is not difficult to arrange or maintain. The present invention provides for using hot dilution air with the added advantage that raising inlet temperature extends lean burn range. The precise temperature can be determined experimentally and is achievable. Dilution from a reservoir of hot air provides very fast transient response.

In Fig. 4, diluent air is extracted from the air cleaner 55, passed through its own throttle 62, through a heat exchanger and finally to a second valve 66 to a vortex chamber 60. The diluent throttle 62 has the same temperature at entry as does the main mixture throttle 57, to maintain flow balance. The second (thermostatic) valve 66 is controlled by a temperature sensor and is closed when the air is cold, so that no dilution is provided when the air is cold.

While Fig. 4 shows the mixture throttle 57

upwind of the wick chamber 58 it may be downwind. Experimental results will very well dictate which is the preferred arrangement if any.

Fig. 5 shows the wick chamber 58 of Fig. 4 in greater detail. The wick chamber 58 may take many forms, including those shown in U.S. Patent No. 4,290,401 for example, which is incorporated by reference herein. In Fig. 5 the wick chamber includes an upper conical container portion 70a and a lower conical container portion 70b and having an air inlet 71 and a vapour mix outlet 72. A wick support grid 73, which may be in the form of a wire screen or mesh of a coarser size than the wick, is provided to support a wick 59 in a pyramid fashion as shown. Liquid fuel is shown schematically injected into the wick chamber 58, and can be injected by sprinkling or spraying means to spread the fuel out over the wick.

The wick 59 may be in the form of a closely woven cloth of spun fibers as shown in Fig. 6. The strands in the weave of cloth contain 6 (six) fibers which are spun or twisted in a conventional manner. Liquid fuel injected into the wick chamber 58 will spread by capillary action and lodge in the fine spaces between the fibers and will evaporate into the air as the air streams through the cloth as shown in Fig. 6. The fiber diameter should preferably be as small as possible within the constraints of material strength and availability. A drain 74 is provided for the run-off of

unevaporated liquid fuel. Since fuel injected into the wick chamber has already been metered, the drain 74 is not necessary, and liquid fuel can be allowed to run out of the vapour mix outlet 72. The drain 74 may be connected further downstream of the wick chamber 58.

The small size of the fiber gives maximum surface area for a given weight of fuel resident and lodged in the structure. The small size also gives maximum evaporation effect for given pressure drop in airways. While it is possible to use more than one layer of wick of coarser size, a single fine layer is preferred.

The fibers of the wick may be composed of spun long fibers of wire, glass or natural material. As stated above, an arrangement of six fibers to comprise one strand is the preferred construction in terms of surface area, fuel spreading along the spaces between the fibers within the strands, minimum resident fuel, i.e. quantity of fuel needed to wet the wick surface, and exposure to air flowing in the fine air spaces between the fibers. Of the three materials of wire, glass or natural material, the wire can act as a flame trap and has minimal risk of shedding fragments into the airstream. It is therefore preferred and the effects of heat spread through the metal is probably, on balance, an advantage.

Referring again to Fig. 5, under cold conditions, a fraction of the fuel will likely fail to

evaporate in the cold air passing the wick. The simplest means of disposal is to use a small additional wick in the diluent air stream. The run off is passed to this wick where it joins the main flow as vapour. Referring now to Fig. 7, the vortex chamber 60 of Fig. 4 is shown in greater detail. The vortex chamber 60 has an inlet 80 which receives air/fuel mixture from the wick chamber and an outlet 81 which is connected to the intake manifold for the engine. Hot diluent air from thermostatic valve 66 is provided at side inlet 82 into an enlarged annular vortex mixing chamber 83. The hot diluent air is provided tangentially to the annular chamber and transverse to the air/fuel mixture flow, where the hot diluent air follows a circular helical path. Mixing of the hot diluent air with the cold air in the generally downward vertical direction is encouraged by the fact that the more dense cold air mixture will try to move radially outwards while the hot, less dense diluent air will try to move inwardly.

A spark plug according to one aspect of the present invention will now be described with reference to Figs. 8-10. The illustrated spark plug is of generally conventional construction except for the electrodes, as will be described. The spark plug as shown in Fig. 8 has a body 85 of conventional form including an externally screw-threaded portion 86 and a hexagonal or other portion by means of which it can be

engaged by a spanner or socket to fit or remove it from an engine cylinder head or equivalent part.

Within the body 85 . is a ceramic insulator 12 which is positioned within the screw-threaded hollow cylindrical end of the body 85. There is a gap around the insulator 87 and its shape may be determined in a known manner dictated by cooling or other factors.

Passing through the center of the insulator 87 is a central electrode 88. This extends right through the insulator and terminates in a screw-threaded portion or other portion to which a lead can be connected.

Fixed, as by welding, to the annular end of the body 85 is a side or earth electrode 89. This is of L-shape having one limb secured to the body 85 and the other limb extending over the central electrode 88.

As seen particularly in Figure 9 the central electrode 88 is of cylindrical form but its end is shaped to provide a flat narrow rectangular rail-like surface 90 flanked by two inclined surfaces 91, 92. The side electrode 89 has a surface presented towards the central electrode, this surface being shaped to provide a narrow rail-like surface 93 aligned with the surface 90 on the central electrode and of generally the same proportions. The surface 93 on the side electrode 89 is also flanked by ' inclined surfaces 94, 95. The portion of the surface 90 on the central electrode is parallel sided and of similar width and length to the surface 90 on the central electrode but is

then flared so as to merge with the surface of the side electrode as shown.

The inclined surfaces 91, 92, 94, 95 provide substantial clearance between the electrodes other than in the regions of the opposing surfaces 90 and 93.

As seen in Figure 10 the surfaces 90 and 93 lie at a small acute angle relative to one another with the narrowest gap between them being at the end at which the side electrode 89 is secured to the body 85 of the spark plug.

In use the spark plug is fitted into a combustion chamber of an internal combustion engine in conventional manner and means are provided for applying a high voltage current across the gap between the two electrodes to create a spark. Spark initiation occurs at the point at which the two surfaces 90 and 93 are closest together, that is, at the end towards the junction of the side electrode with the spark plug body. The magnetic forces produced by the current in the spark and the side electrode causes the spark to travel along the" rail-like surfaces 90, 93 towards the free end of the side electrode.

The translation of the arc along the rail-like surfaces will have the effect of supplying heat to and hence igniting a larger quantity of mixture than if the spark were static. This cools the thread of gas which is conducting the electrical current. This increases its electrical resistance and since external resistance.

due to the leads and the coil, or its equivalent, is largely fixed, there will tend to be an increase in electrical energy at the arc.

Furthermore, the substantial clearance at either side of the rail-like surfaces 90, 93 enables the flame front to grow in area as it advances, thus rapidly reaching a larger volume of fresh combustible mixture. Cooling effects of the bodies of metal represented by the electrodes are also minimised. The translation of the arc, extra arc energy, better exposure of combustible mixture and minimised cooling all contribute to a reduction of the duration of the first stage of ignition, the greater reduction occurring under the more difficult case of leaner mixture.

It might be supposed that the large area of the conventional electrodes has the effect of prolonging the life of the spark plug. However, erosion of the electrodes is governed by the amount of heat reaching any particular volume of metal. When, as here described, the arc is deliberately caused to traverse the rail-like surfaces, the amount of heat to any particular volume is minimised with ensuing prolongation of the electrode life. The spark plug as described therefore provides a high efficiency spark initiation and flame generation pattern giving good combustion characteristics and also enabling lean mixtures to be effectively used, thus

enhancing the fuel economy of the engine. Good combustion characteristics also ensure maximum burning of the mixture in the combustion chamber so that exhaust pollution is minimised. The spark plug according to the invention can also have different types of electrodes. For example, the side electrode may extend laterally directly towards the side of the central electrode from the adjacent wall of the body, either at a level with the end of the body or above it.