Background of the invention It is well known that gasoline or diesel fuel is a multi-component fuel with an average set of properties tailored to satisfy a broad range of engine operating conditions including good volatility for cold start, easy ignition at low loads and with lean mixtures, high knock tolerance at high loads, and at the same time, free from problems such as vapour lock and deposits. Many of these are conflicting requirements and a compromise must be found in blending the properties to provide adequate performance all round with no significant defect in any one aspect, but no outstanding characteristic either.

A corollary of the above is that different fuel groups have been selected intentionally to make up the commercial fuel blend based on the unique properties of each fuel group being better suited for a particular operating condition among a wider range of operating conditions. For example, the paraffin group is more volatile and more ignitable, while the aromatic group has higher knock tolerance, these being mixed to produce the averaged general properties of the blend. Thus for a given multi-component fuel, instead of using it directly, it would be more advantageous to separate it again into its component groups before supplying it to the engine, using a higher proportion of one group in preference to another according to operating conditions. In this way the distinct properties of each fuel group may be utilised in a more targeted manner, thus extending the

operating boundaries of the engine in many aspects while overall still using the standard commercial fuel blend.

Conventional fuel separation systems rely on fractional distillation of the fuel by heating the fuel and condensing the vaporised fraction. For use on board a vehicle, in order to cope with a wide range of variable fuel demand from the engine, known systems have been designed either as a batch system in which the separated fuel components are produced in batches and stored in separate fuel tanks as described in GB Patent No. 2,209,796, or as a continuous supply = demand equilibrium flow system in which the production flow rate of the separated fuel components is variable in response to the demand flow rate from the engine as described in GB Patent No. 2,330,179. Other systems, such as that described in US Patent No. 4,220,120, aim to extract fuel vapour from the fuel tank to support a limited number of cold starts but ignores the consequence that the fuel in the tank would eventually be depleted of the light fractions resulting in poorer and poorer starts as the tank fuel is progressively being used up. All these systems suffer many disadvantages in that they are complicated and costly, and are prone to instability and surges due to the time delay inherent in the fractional distillation process, making it unsuitable for use as stable fuel sources to meet all the transient demands from the engine requiring accurate fuel metering.

Summary of the invention In order to mitigate the above disadvantages, there is provided according to a first aspect of the present invention a method for operating a fuel separation and injection system for an internal combustion engine, the method comprising the steps of drawing a continuous surplus- to-demand fuel stream from a fuel reservoir in the fuel separation system, separating the fuel stream into at least

two flow streams containing different fuel components having different fuel properties, connecting the separated flow streams to respective separate fuel rails arranged in parallel with one another with at least one fuel rail supplying at least one fuel injector, combining continuously the unused surplus flow streams back into a single fuel stream, and returning the recombined fuel stream to the fuel reservoir.

According to a second aspect of the present invention, there is provided a fuel separation and injection system for an internal combustion engine, the system comprising means for drawing a continuous surplus-to-demand fuel stream from a fuel reservoir in the fuel separation system, means for separating the fuel stream into at least two flow streams containing different fuel components having different fuel properties, means for connecting the separated flow streams to respective separate fuel rails arranged in parallel with one another with at least one fuel rail supplying at least one fuel injector, means for combining continuously the unused surplus flow streams back into a single fuel stream, and means for returning the recombined fuel stream to the fuel reservoir.

In the invention, the supply of the separated fuel components is completely independent of the fuel demand from the engine. Because the supply flow is in surplus and constant, and the fuel separation takes place along a steady process path with well-defined processing stages at precise locations disposed along the length of the system, the separated fuel components would always be available at predetermined stations along the length where they behave as constant fuel sources, capable of meeting any instantaneous fuel demand from the engine immediately with very little time delay. This is the important distinguishing feature of the invention from the prior art.

Preferably, a fuel pump is provided in the invention for pressurising the fuel stream from the fuel reservoir, and a pressure relief valve is provided for returning the recombined fuel stream back to the fuel reservoir, such that all the fuel rails in the fuel injection system are pressurised to the same fuel pressure regulated by one pressure relief valve.

In a preferred embodiment of the invention, the means for separating the fuel stream into two flow streams containing different fuel components having different fuel properties comprises a fuel separation pipe section connected to the fuel stream, means for inducing cavitation of the fuel within the pipe section by creating regions of low pressure below the critical cavitation pressure thereby causing the lower boiling point fuel components to form vapour bubbles, means for providing a force field within the pipe section such that the vapour bubbles are caused to migrate towards a predetermined region of the pipe section before collapsing back into liquid while leaving the pipe section, and means for separating the fuel leaving the pipe section by drawing independently from the predetermined region and from the remaining region of the pipe section, thereby producing two flow streams the former comprising predominantly the lower boiling point fuel components that have undergone cavitation and the latter comprising predominantly the higher boiling point components.

Preferably, the means for inducing cavitation is a flow constriction through which the fuel is forced to pass at high speed creating a region of low pressure below the critical cavitation pressure near the throat of the flow constriction.

Alternatively, the means for inducing cavitation may be a propeller driven at high speed within the fuel creating regions of low pressure below the critical cavitation

pressure near the edges and the tips of the propeller blades.

As a further alternative, the means for inducing cavitation may be an ultrasonic transducer vibrating within the fuel creating a high energy acoustic field exceeding the cavitation level.

Preferably, the fuel stream is directed to enter tangentially into the fuel separation pipe section creating a strong swirling flow and a centrifugal force field within the pipe section past the cavitation region, causing the cavitation-formed vapour bubbles to migrate towards the swirling flow axis and the remaining liquid to spread away from the swirling flow axis. Examples of similar bubble separating methods may be found in US Patent Nos. 4,390,351, 5,510,019 and 5,749,945.

In the present invention, the separate fuel rails may be arranged to supply respective separate sets of fuel injectors, each set metering fuel to the engine from the respective flow streams.

Alternatively, a first fuel rail may be arranged to supply one set of fuel injectors, and a flow diverting valve may be provided for selectively guiding one of the flow streams to flow through the first fuel rail and the other flow stream to flow through the other fuel rail.

In this case, the flow diverting valve may also serve as a flow mixing valve for partially mixing the flow streams before guiding the streams to flow through the fuel rails.

After passing through the separate fuel rails, the unused surplus flow streams are combined back into a single fuel stream via a flow junction connecting the separate fuel rails to a single pipe leading to the pressure relief valve.

The invention as described above is ideally suited for adaptation into a conventional fuel injection system which already has a fuel pump, a fuel rail supplying fuel injectors, a pressure relief valve and a fuel return pipe.

When incorporated into the conventional fuel injection system, the present invention enables the engine to take advantage of the distinct properties of the separated fuel components rather than the averaged properties of the whole fuel, and this is constantly available with surplus to meet any demand from the engine. For example, the engine may be started and warmed up using predominantly the lower boiling point fuel components, and subsequently operated at high loads and during accelerations using predominantly the higher boiling point fuel components. Provided that, on average, the cumulative usage of the separated fuel components are monitored and kept in balance, the composition of the fuel in the fuel reservoir would remain substantially the same before and after a short trip.

Compared with known fuel separation methods such as fractional distillation, the preferred fuel separation method of the present invention using cavitation has the advantage that it does not rely on external heating or condensing of the fuel, although the temperature of the fuel has an effect on the cavitation pressure and may be used as a controlling parameter. To this effect, a heat exchanger may preferably be provided for heating and/or cooling of the fuel entering the fuel separation pipe section in order to keep the fuel temperature within predetermined limits.

The fuel stream pumped into the fuel separation pipe section is the surplus-to-demand rated flow of the fuel pump supplying both the metered fuel and the return fuel. This flow is substantially constant which provides the stable conditions necessary for the fuel separation process to take place steadily and accurately, independent of any transient demand from the engine. The unused surplus flow streams are

recombined back into a single fuel stream in the fuel return pipe and this fuel is separated again afresh each time it is re-circulated through the fuel separation pipe section.

There is no need to store the separated fuel components since they are produced continuously and are constantly available, not being subjected to the time delays associated with the fuel separation process itself.

In the present invention, the separated fuel components are used preferentially in different proportions according to engine operating conditions for short periods while control means are provided to ensure that the whole fuel is used over a longer averaged period.

In order to do this, a sensor may be provided for measuring or estimating the composition of the fuel in the fuel reservoir, and the signal from the sensor may be used in a feedback control system for varying the relative usage rates of the separated flow streams for short periods such that the estimated composition of the fuel in the fuel reservoir stays within predetermined limits over a longer averaged period.

Various types of fuel sensor may be used. For example, the sensor may be a device for detecting a change in the specific gravity, the vapour pressure, the refractive index, the acoustic properties or the electrical properties of the fuel. An example of a fuel vapour pressure sensor may be found in JP Patent No. 7198710.

Preferably the liquid level in the fuel reservoir is kept constant by topping up continuously or intermittently from a separate fuel storage tank. In this case, because the volume of the fuel in the fuel reservoir is constant, any change detected in the composition of the fuel in the fuel reservoir would be a direct indication of the relative proportions of the fuel components that have been used. For

good sensitivity in indicating a change in fuel composition, the volume of the fuel reservoir should be as small as possible.

Brief description of the drawing The invention will now be described further, by way of example, with reference to the accompanying drawings in which : Figure 1 is a schematic view of a preferred embodiment of a fuel separation and injection system of the invention, and Figure 2 is a schematic view of an alternative embodiment of a fuel separation and injection system of the invention.

Detailed description of the preferred embodiment In Figure 1, separation of the fuel is achieved in a cylindrical or conical fuel separation pipe section 10 with closed ends. The pipe section 10 has a tangential inlet pipe 14 and two outlet pipes 22,24. A device for inducing cavitation 16a, 16b, or 16c is mounted within the pipe section 10 and submerged in the liquid fuel. The device may be a flow constriction 16a, a propeller 16b or a ultrasonic transducer 16c serving to create regions of low pressure below the critical cavitation pressure of the lower boiling point components of the fuel. When cavitation occurs, the lower boiling point liquid components are transformed into vapour bubbles and these are dispersed within the remaining higher boiling point liquid components.

To separate the two-phase components, a centrifugal force field is created inside the pipe section 10. This can be achieved in several ways. First, fuel is forced at high speed tangentially into the pipe section 10 from the inlet pipe 14 imparting a strong swirling motion as the fuel moves

generally along the pipe section 10. Second, in the case of the flow constriction 16a, the position and the direction of the flow constriction are such that the jet emerging from the constriction is tangential to the pipe section 10 creating a strong rotation in the fuel within the pipe section 10. Third, in the case of the propeller 16b, the spinning of the propeller in the liquid once again produces a swirling flow field in the pipe section 10.

Because of the density difference between vapour bubbles and liquid fuel, the centrifugal force field would cause the heavier liquid to migrate outwards away from the swirling flow axis and the lighter bubbles to migrate inwards towards the swirling flow axis, thus separating the two-phase components. As shown in Figure 1, the swirling flow is represented by the rotating arrow, and the lower boiling point fuel components that have undergone cavitation are concentrated within the central core region of the pipe section 10 while the higher boiling point fuel components are spread around the peripheral region, as the fuel moves generally towards the outlet pipes near the exit end of the pipe section 10. A concentric wall 18 is positioned near the exit end of the pipe section 10 to separate the central core region from the peripheral region thus separating the fuel stream into two flow streams connected separately to the two outlet pipes 24,22 respectively.

The preferred fuel separation method described above has advantages over other known fuel separation methods such as fractional distillation in that it does not rely on external heating or condensing of the fuel, although the temperature of the fuel has an effect on the cavitation pressure and may be used as a controlling parameter. To this effect, a heat exchanger (44) is provided in Figure 1 for heating and/or cooling of the fuel entering the fuel separation pipe section (10) in order to keep the fuel temperature within predetermined limits.

In Figure 1, the fuel separation pipe section 10 is integrated with a fuel injection system having a fuel reservoir 38 topped up to a constant liquid level from a fuel supply pipe 40 connected to a separate fuel storage tank (not shown), a fuel pump 12 delivering a surplus-to- demand fuel stream to the fuel separation pipe section 10, two fuel rails separately connected to the two outlet pipes 22,24 of the fuel separation pipe section 10 and supplying respectively two sets of fuel injectors 26,28, a flow junction 30 connecting the two fuel rails downstream of the fuel injectors to a single pipe 32, a pressure relief valve 34 and a fuel return pipe 36. Because of the surplus-to- demand flow rate delivered by the fuel pump 12, at least some of the fuel pumped into the fuel separation pipe section 10 is unused and returned to the fuel reservoir 38.

The pressure relief valve 34 regulates the return fuel so that the pipe section 10 and the fuel rails are collectively maintained at the same elevated fuel pressure. In this system, the engine may be supplied simultaneously with the two separated fuel components in various proportions according to different operating conditions.

Figure 2 is similar to Figure 1 except for the. disposition of the fuel rails connected to the two outlet pipes 22,24. In this case, a flow diverting valve 20 is provided for selectively guiding the flow stream from one of the outlet pipes 22 or 24 to flow through a fuel rail supplying one set of fuel injectors 26 and the flow stream from the remaining outlet pipe to flow through another fuel rail with no fuel injector. By selectively diverting a flow stream to the fuel injectors 26, the engine may be supplied alternately with one or the other of the separated fuel components according to different operating conditions.

In use, an engine management system determines the appropriate flow stream/s to be supplied to the engine and sets the flow diverting valve 20 or triggers the appropriate

fuel injectors 26,28 accordingly. After a short period, the composition of the fuel in the fuel reservoir 38 may gradually change if one of the flow streams is consumed in a greater proportion relative to the other. A sensor 42 in the fuel reservoir 38 monitors the fuel composition and the signal is used by the engine management system to proportionate or alternate the usage of the two flow. streams in order to keep the composition of the fuel in the fuel reservoir 38 in balance over time. For example, the engine may be started and warmed up using predominantly the lower boiling point fuel components, and subsequently operated at high loads and during accelerations using predominantly the higher boiling point fuel components. Provided that, on average, the cumulative usage of the separated fuel components are in balance, the composition of the fuel in the fuel reservoir 38 would remain substantially unchanged.

In Figures 1 and 2, the fuel in the fuel reservoir 38 is topped up to a constant liquid level from a fuel supply pipe 40 connected to a separate fuel storage tank (not shown). In this case, because the volume of the fuel in the fuel reservoir 38 is constant, any change in the composition of the fuel detected by the sensor 42 would be a direct indication of the relative proportions of the fuel components that have been used. For good sensitivity in indicating a change in fuel composition, the volume of the fuel reservoir 38 should be as small as possible.

It would be clear from the Figures that the invention has significant advantages in that the preferred system may be constructed by simply adding a compact fuel separation unit and controller to a re-circulating fuel injection system of substantially conventional design. The preferred fuel separation unit is low cost, robust, well suited for mass production and may be fitted to new vehicles as well as retro-fitted to older vehicles.