WO1997045634A1

WO1997045634A1 - Fuel evaporators

Info

Publication number: WO1997045634A1
Application number: PCT/GB1997/001416
Authority: WO
Inventors: John Rendell Conrad Pedersen
Original assignee: Individual
Current assignee: Individual
Priority date: 1996-05-25
Filing date: 1997-05-23
Publication date: 1997-12-04
Anticipated expiration: 1998-11-25
Also published as: AU2910497A; GB9611030D0

Abstract

A carburettor metering system incorporates an evaporator (1) for vaporisation of fuel into an air flow. The evaporator (1) comprises an evaporator matrix (2) having an extended evaporation surface, two spray nozzles (6) for supplying fuel to the evaporator matrix (2) so that fuel droplets are captured at a plurality of locations distributed over the evaporation surface, and an air flow duct (4) for passing an air flow over the evaporation surface to cause the fuel droplets to be evaporated into the air flow. When used in a gasoline engine, such a carburettor promotes evaporation of a sufficient quantity of fuel early enough in the induction process that any remaining spatial variation of fuel concentration in the air/fuel mixture is small and of sufficient spatial scale as not to promote production of nitrogen oxide, at least under higher load operating conditions.

Description

"Fuel Evaporators"

This invention relates to evaporators for vaporisation of fuel into an air flow and

is concerned more particularly, but not exclusively, with carburettor metering systems

incorporating such evaporators.

Experimental data from gasoline engines shows that the production of nitrogen

oxides (NOx) in engine exhausts is increased by the presence of non-uniformities of fuel

concentration in the air/fuel mixture supplied to the engine at the time of flame passage

in the combustion process. Nitrogen oxides are among the unwanted and legally

restricted pollutants emitted from engine exhausts. It can be assumed, by analogy, that

non-uniformities of fuel concentration in other kinds of combustion process, such as in

furnaces for example, will also lead to increased production of unwanted nitrogen

oxides.

Furthermore the experimental data indicates that the non-uniformities of fuel

concentration responsible for production of nitrogen oxides can be on a small spatial

scale, such as might result from the incomplete, or only recently completed, evaporation

of a mist or spray of distributed fine fuel droplets at the time of flame passage. It is

known that a uniform air/fuel mixture at the time of flame passage is advantageous in

reducing emissions of unburnt or incompletely burnt hydrocarbons, in allowing the use

of air/fuel mixtures of reduced mixture strength (weak mixtures), and in increasing the power output for a given weight of air/fuel mixture of given strength

The mechanism responsible for increased production of nitrogen oxides as a

result of variation of non-uniformities of fuel concentration has not been positively

identified. Furthermore it is not known over what range of spatial variation the

mechanism is of significance, although it is known that gross variation, due to a few

large liquid fuel droplets remaining in the mixture at the time of flame passage for

example, has negligible effect on the production of nitrogen oxides It is possible that

the mechanism is dependent on the spatial variation of fuel concentration within the

flame reaction zone, in which case the spatial range over which the mechanism is of

importance would be comparable with the reaction zone thickness

It has been reported in The Sixth Symposium on Combustion, p 106-7,

"Structure of Laminar Flames", Robert Fristom, that the thickness of the reaction zone

in a low pressure propane/air mixture (defined as the region in which propane is being

rapidly consumed) is about 0 8 mm at 0.25 bars At higher pressure, as in an engine

combustion chamber, the reaction zone would be expected to be slightly thicker, since

the laminar flame speed is slightly less An alternative definition of reaction zone

thickness, defined as the region within which the heat release rate is at least a quarter of

the local maximum, would give a thickness of 2 mm This might therefore be considered

to be the range of spatial variation over which the above-mentioned mechanism for

production of nitrogen oxides is likely to be significant. In an engine supplied with a spray of fuel, the volume of air at ambient pressure

is typically about 10,000 times the volume of liquid fuel for ordinary, slightly lean

mixtures prior to compression, whereas, after compression, the air volume is likely to

be about 1,000 times the original fuel volume, so that the average fuel droplet can be

considered to exist in a volume of air having linear dimensions of the order of 10 times

the droplet diameter According to I Mech E , C 81/83, "Laser Video Imaging and

Measurement of Fuel Droplets in a Spark Ignition Engine", D B Peters, measurements

of the fuel spray from an injector has given a Sauter mean diameter (the diameter of a

hypothetical droplet having the same volume/area ratio as that of the whole spray) of

25μm, with the spectrum giving a volume mean diameter (the diameter of a droplet of

such a size that a number equal to the total number would have the same total volume)

of about 20μm Thus the droplet separation in an engine supplied with such a fuel spray

would be of the order of 0 2mm if all the droplets survived the compression process

It will be noted that this droplet spacing is distinctly smaller than the reaction zone

thickness given by either definition above

Furthermore the fuel droplets persist in the induction and compression processes

because the gas temperature in the inlet manifold and the cylinder during the induction

stroke are such that evaporation of fuel is slow Whilst the temperature rises during

compression, such a rise in temperature occurs late in the compression stroke For

example, at a typical compression ratio, only half the final rise in temperature has

occurred when the piston has travelled more than 80% of its stroke Thus the time within which the conditions within the cylinder promote rapid evaporation of fuel is very

short. Furthermore, during rapid evaporation of fuel, the mixture strength in the vicinity

of the droplet is substantially greater than the rich combustion limit so that further time

is required to allow this local very rich mixture to diffuse.

International Published Patent Application No. WO 94/1621 1 discloses an

evaporator consisting of a stack of plates which are spaced apart so as to define narrow

air passages therebetween and so as to provide porous evaporation surfaces along the

sides of the passages to which fuel is supplied from a spreader tube so that the fuel is

distributed over the evaporation surfaces by capillary action. However such an

evaporator suffers from a number of disadvantages in practice. One of the main

disadvantages is that, under high transients, it is possible for droplets of fuel to break

away from the evaporation surfaces into the air flow, thus causing inhomogeneity of the

resulting air/fuel mixture and a consequent increase in nitrogen oxides and hydrocarbons

in the exhaust emissions.

It is an object of the invention to provide an improved evaporator for

vaporisation of fuel into an air flow with a view to decreasing the emission of nitrogen

oxides.

The invention is defined in the accompanying claims. When used in a gasoline engine, such an evaporator should be designed to

promote evaporation of a sufficient quantity of fuel early enough in the induction process

that the remaining spatial variation of fuel concentration in the air/fuel mixture is small

and of sufficient spatial scale as not to promote production of nitrogen oxide, at least

under those operating conditions of higher load when nitrogen oxide production would

otherwise be expected to be substantial

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

Figure 1 is a vertical section through an evaporator in accordance with the

invention,

Figure 2 is a horizontal section through the evaporator of Figure 1;

Figure 3 is a detail of Figure 1,

Figure 4 is a vertical section through part of a further evaporator in accordance

with the invention,

Figures 5, 6 and 7 are explanatory diagrams; and

Figures 8 and 9 are vertical sections through further evaporators in accordance

with the invention

Referring to Figures 1 and 2, the evaporator 1 comprises an evaporator matrix

2 having an extended evaporation surface, fuel metering means 3 for supplying metered fuel to an air flow in the form of fuel droplets, and air flow means in the form of a duct

4 for passing the resultant air/fuel mixture over the evaporation surface of the evaporator

matrix 2 in intimate contact therewith so that fuel droplets are captured at a plurality of

locations distributed over the evaporation surface, and so that the fuel droplets are

subsequently evaporated into the air flow to produce a highly homogenous air/fuel

mixture which is supplied to a combustion zone. The fuel can be delivered to the air flow

by the fuel metering means 3 as a spray by means of two spray nozzles 6, as shown, with

a spray plume designed to provide as uniform wetting of the evaporator matrix 2 as

possible. Existing fuel injection systems may be used for supply of fuel to the nozzles

6. Alternatively the fuel can be boiled to a vapour which is then introduced into the air

flow by the fuel metering means 3. In the latter case some of the fuel will condense to

form a fog of fine droplets on mixing of the rich hot vapour with the cooler air flow. In

both cases fuel droplets will be present in the resultant air/fuel mixture.

As the air/fuel mixture incorporating fuel droplets produced in this manner passes

towards the combustion zone the temperature of the mixture rises so as to cause the

droplets to undergo evaporation. In a spark ignition engine, this evaporation process

would take place in the inlet manifold during induction and compression. In a furnace

the evaporation process would occur as the air/fuel mixture approaches the reaction

zone. However the approach to mixture uniformity occurs in two stages, namely a first

stage in which vapour evaporates from the liquid droplets and diffuses into the air

between the droplets, and a second stage in which there remains a local concentration of fuel in the mixture which decays with time although the droplet is no longer present

in liquid form

In the first stage the gas consists wholly of fuel vapour at the surface of the

droplets whereas, between the droplets, the gas consists of air with a concentration of

fuel vapour which varies with position and the time since the fuel was admitted Both

stages occupy a finite period of time which depends on the temperature and pressure of

the mixture, as well as the distance through which the vapour must diffuse, that is on the

number, and hence the size, of droplets present in a given volume of air

However the evaporator matrix 2 is designed to capture at least a proportion of

the fuel droplets in the mixture as it passes over the evaporation surface The purpose

of the matrix evaporator 2 is to capture those fuel droplets which are too large to

evaporate and diffuse to practical uniformity prior to reaching the combustion zone, so

as to provide a large liquid surface with substantial air flow over the surface, and so as

to shorten the time for the fuel to be evaporated into the air flow with a view to reaching

practical uniformity of the mixture by the time of flame passage Very small fuel

droplets which are present in the mixture can escape capture by the matrix evaporator

2 without having any deleterious affect on uniformity of the mixture supplied to the

combustion zone 5

In order to promote capture within the evaporator matrix 2 of fuel droplets of sizes likely to be present in the mixture, it is advantageous to cause the air flow, in which

the droplets are carried, to undergo an abrupt change, or many abrupt changes, in flow

direction as it passes through the evaporator matrix 2. This has the effect that the

relatively massive fuel droplets fail to follow the diverted direction of air flow and collide

with the solid evaporation surface so as to be captured thereby.

At high flow rates, the changes in direction of the droplets in the evaporator

matrix 2 can be made to cause even quite small droplets to impact on the surface of the

matrix so as to become trapped in the matrix and to subsequently be evaporated into the

passing air flow. Droplets which escape capture are largely evaporated in flight through

the matrix and the inlet manifold so that few droplets remain in the mixture passed to the

engine. The advantage of such capture within the matrix is that the matrix temperature

can be lower, namely at a temperature approaching the fuel dew point if the fuel is

spread onto a large surface with narrow air spaces Such a lower temperature is

important in enabling high output torque to be obtained At lower flow rates, the

capture of droplets within the matrix is less effective since the accelerations of droplets

within the air flow are reduced. As a result more of the smaller droplets escape capture

by the matrix, although such droplets then have more time to evaporate in flight

Furthermore, at low flow rates, there is less need to restrict the temperature of the input

air in the interest of high output torque. Indeed it is advantageous for the air

temperature to be increased at low load since this reduces the pumping losses and allows

the use of leaner mixture by increasing the flame speed. Also there is a general trend to lower nitrogen oxide concentration at low load in engines, so that the need for

uniformity of fuel concentration is less at low load.

In a non-illustrated embodiment of the invention, the evaporator matrix

comprises a filter medium to which the mixture of air and fuel droplets is admitted in

operation. Optionally the filter pores are graded so that the pores are of decreasing size

in the direction of air flow through the filter with the result that larger fuel droplets tend

to be captured in the portion of the filter closer to the inlet of the filter whereas smaller

fuel droplets tend to pass through this portion so as to be caught in the portion of the

filter nearer to the outlet of the filter. The grading of the filter prevents larger fuel

droplets flooding the filter in such a manner as to fill many small pores of the filter whilst

presenting only a relatively small exposed fuel surface to the air flow. The filter can, for

example, consist of a reticulated foam pad which may comprise layers of different pore

size. The reticulated foam filter can be produced from a polyurethane foam by passing

a caustic substance or a flame through the foam so as to remove the intermediate walls

between the cells of the material whilst leaving behind thin strips of material at the

intervening boundary lines between adjoining cells. The resulting open cell structure is

then plated with a metal, such as nickel, so as to produce a reticulated metal foam of low

throughflow resistance but having the ability to capture small fuel droplets passing

therethrough.

In an alternative embodiment of the invention, as illustrated in Figures 1 and 2, the evaporator matrix 2 comprises, within a flow duct 10, a stack 1 1 of generally parallel

metal plates 12 of a typical thickness of 0.25mm spaced apart by spacers (not shown)

which define narrow passages 13 of a typical width of 0.5mm for flow of air/fuel

mixture between the plates 12. As is best seen in the detail of Figure 3 showing the end

parts of three of the plates 12, evaporation surfaces 14 are provided on both sides of the

plates 11 (apart from the outermost plates which have such an evaporation surface on

only one side) so as to provide an extended surface area for evaporation of fuel into the

air flow. Inlet portions 15 of the plates 12 are staggered relative to one another and

serve to divert the air flow passing in a direction 16 transverse to the plates 12 to a

direction substantially parallel to the plates 12 in which the air flow passes along the

narrow passages 13 between the plates 12. The resulting abrupt change in direction of

the air flow on entry into the passages 13 between the plates 12 causes the more massive

fuel droplets in the air flow to impinge either on the ends 17 of the plates 12 or on the

upwind faces of the plates 12 as shown in the figure, as a result of the fact that the

droplets are unable to negotiate the change in direction. The spacing between the plates

12 and the lengths of the plates 12 may be chosen to ensure that substantially all the fuel

droplets which are too large to evaporate to practical uniformity before combustion are

caught within the evaporator matrix.

The evaporator 1 of Figures 1 and 2 is shown incoφorated in a fuel metering

system for a spark ignition engine with the output from the evaporator matrix 2 being

supplied to the engine manifold by way of an outlet duct 5 incorporating a throttle 7. In this system hot and cold air flows are supplied along inlet ducts 8 and 9 to a

temperature controlled gate 18, and the resultant mixture of hot and cold air is supplied

to the duct 4 by way of a pleated air cleaner 19 The evaporator 1 has a drain point 20

to enable liquid fuel to be recirculated to the pump of the fuel metering means on cold

starting

Figure 4 shows a variant of such an evaporator matrix for increasing the

probability of capture of fuel droplets by providing passages 22 through the matrix 23

of zig-zag shape so that the air flow is subjected to a plurality of abrupt direction

changes as it passes along the passages 22 At each location in which the direction of

each passage 22 changes there is an abrupt change in the direction of air flow, and this

causes fuel droplets within the air flow to impinge on the matrix surfaces so as to be

captured thereby Thus most fuel droplets of sufficient size will tend to be captured by

the matrix 23 with the larger droplets tending to be caught closer to the inlet of the

matrix 23 and the smaller droplets tending to be caught closer to the outlet of the matrix

23 In the illustrated embodiment the passages 22 are defined by layers 24 of filter

medium arranged on a support grid 25, with the layers 24 having a pore size which

decreases from layer to layer and with the pores of adjacent layers being arranged in

relation to one another to provide the required changes in direction of the passages 22

In a further variant the evaporator matrix comprises a stack of generally parallel

metal plates defining narrow passages therebetween and having a single bend 30 (Figure 7) or double bend 31 (Figure 6) followed by a straight section 32. Generally, in a flow

of gas along a curved rectangular duct 35, as shown diagrammatically in Figure 5,

secondary vortices 34 are produced in the gas whose axes are streamwise and whose

lateral dimensions are comparable with the duct height, alternate vortices rotating in

opposite directions. These vortices 34, which are analogous to the Gόrtler vortices

found between rotating coaxial cylinders, are set up (at low flows) at a Reynolds number

given by :

Re < 36 (R / d )'^/'

where R is the inner bend radius

d is the plate spacing

Re is v d/v

v is the mean duct velocity

v is the kinematic viscosity

Thus, at a low flow relative to the flow at full load, the passage of the mixture

of air and fuel droplets around the bends 30 and 31 in the variants of Figures 6 and 7 will

give rise to such vortices which will assist in deposition of the fuel droplets on both

surfaces of the plates, and also in distributing the evaporated fuel across the passages

between the plates. The vortices will persist for some distance downstream of the bend

or bends so that deposition and distribution will continue along the straight section 32.

Furthermore the liquid fuel deposited on the plates is distributed over the plates by shear

stress due to the flow of air and fuel droplets. The evaporator arrangement chosen for a particular application depends on the

range of flows involved. If the flow range is modest, as in a speed governed engine, it

is possible to arrange for the matrix to capture sufficient fuel droplets over the whole

flow range without special measures having to be taken However, if the flow range is

large, as in automotive use where both speed and torque vary over a large range, special

measures must be taken to allow for the large range of flows. Such measures may

include an arrangement for increasing the charge temperature at low load, such as will

be described below with reference to Figure 8, or an arrangement for returning some of

the air/fuel mixture from the output of the matrix to the evaporator inlet at low engine

flow, as will be described below with reference to Figure 9

Figure 8 shows a further embodiment of the invention in which the evaporator

1' comprises a rectangular duct 4' which is divided into two air flow conduits 40 and 41

provided with separate throttles 42 and 43 upstream of an air mixing zone 44 which is

itself upstream of the spray nozzles 6' and evaporator matrix 2'. Furthermore a heater

45 is provided within the conduit 41 to heat the air passing through the conduit 41. In

operation of the embodiment of Figure 8 air of a suitable high temperature, which has

been passed through a heat exchanger heated by engine coolant, for example, is passed

through the air cleaner 19' before flowing through the conduit 40 and/or the conduit 41

depending on the positions of the throttles 42 and 43. At low load only the throttle 43

is open so that the air supplied to the air mixing zone 44 has passed along the conduit

41 and has been heated by the heater 45. At higher loads, the throttle 42 is fully open and the throttle 43 is also open so that air is supplied to the mixing zone 44 along both

the conduits 40 and 41 with the result that air which has passed along the conduit 41 and

been heated by the heater 45 is admitted to the air mixing zone 44 together with air

which has passed along the conduit 40. Thus the temperature of the air supplied to the

air mixing zone 44 is high at low loads but is reduced progressively in the upper part of

the load range to a minimum which will allow adequate heating of the matrix 2' to enable

most of the fuel droplets to be captured at the higher loads Such an increase in the

charge temperature upstream of the matrix 2' has the effect of increasing the volume

flow through the matrix to thereby improve droplet capture, increasing the evaporation

rate of droplets which escape capture, reducing engine pumping losses by reducing

charge density, and increasing the flame speed, thus allowing the use of leaner mixtures

However the increase in charge temperature cannot compensate for change in engine

speed

Figure 9 shows a further embodiment of the invention in which the evaporator

1" incorporates a rectangular mixing tube 50 centrally located within the rectangular

duct 4" and having the matrix 2" located within one end so as to provide an annular

return passage 51 surrounding the mixing tube 50 for returning part of the mixture

outputted by the matrix 2" to the inlet of the mixing tube 50 at low load. The control

of the return flow of mixture can be by means of a blower of suitable performance or by

use of an ejector pump arrangement as shown in Figure 9. In this ejector pump

arrangement, the high velocity air flow through a partly closed throttle 52 is used to entrain mixture from the matrix output. The throttle 52 consists of two pressure-

balanced plates 53 and 54 hingedly connected to the end of a rectangular input duct 55

In order to ensure pressure balance and to thereby maintain symmetry and prevent the

high energy flow of air from the throttle from attaching to one side of the mixing tube

50, the plates 53 and 54 are hinged near their centre and form variable chambers with

their ends which are vented to upstream pressure by vent holes 56 in the duct 55. The

variable chamber is closed off by virtue of the fact that the end of each throttle plate 53

or 54 is adjacent an arc-shaped wall 57 or is connected to the wall by flexible bellows

(not shown) When the throttle 52 is fully opened, air passing through the air cleaner

19" and a heater 58 flows at low velocity through the throttle 52, and the throttle plates

53 and 54 restrict the area for return flow of mixture to the inlet of the mixing tube 50.

However, when the throttle 52 is partly closed, the air passes through the throttle 52 at

higher velocity and entrains mixture from the matrix output so as to cause mixture to be

introduced into the inlet of the mixing tube 50 together with the input air. The return

of some of the mixture from the matrix output to the mixing tube inlet can be made to

compensate for both load and speed changes

In either of the embodiments of Figures 8 and 9 the matrix 2¹ or 2" may comprise

parallel plates and/or a pad of reticulated metal foam as described above. Furthermore,

in a variant of the embodiment of Figure 9, the heater 58 may be provided within a

conduit in parallel with a cold air bypass conduit in an arrangement similar to the

arrangement provided in Figure 8. In each of the embodiments described above, the total free volume in the

evaporator matrix is chosen so that the time taken for the air to pass through the matrix

is sufficient to allow vapour to diffuse from the wetted surfaces of the matrix into the

air flow, so as to provide adequate mixture uniformity at the outlet of the matrix. The

average vapour partial pressure at the outlet is determined by the required average

mixture strength, and the vapour pressure at the wetted surfaces is dependent on

temperature. Thus the minimum final charge temperature required depends on the

degree to which the vapour has diffused to produce substantial mixture uniformity at the

outlet. The relevant parameter is the dimensionless group

DI h²

where D is the vapour diffusivity in air (length²/time)

t is the transit time

h is a typical dimension of the matrix transverse to the air flow, such as plate-to-

plate spacing.

The variation in mixture concentration decreases exponentially as this parameter

increases. Thus a finer matrix structure, that is smaller h, can have a smaller transit time.

The time is given by (free volume/volume flow rate). The shape and size of the evaporator matrix for any application is a compromise

between the constraints of cost, volume, pressure loss and charge temperature A finer

evaporator structure allows less volume or lower temperature, but increases the pressure

loss.

Claims

1. An evaporator for vaporisation of fuel into an air flow passing through the

evaporator, the evaporator comprising an evaporation surface (2, 2', 2", 23), fuel

metering means (6, 6', 6") for supplying metered fuel to the evaporation surface wholly

or partially in the form of fuel droplets so that fuel droplets are captured at a plurality

of locations distributed over the evaporation surface, and air flow means (8, 9, 40, 41,

50, 55) for passing an air flow over the evaporation surface to cause the fuel droplets

at said plurality of locations to be evaporated into the air flow

2 An evaporator according to claim 1, wherein the fuel metering means (6, 6', 6")

is adapted to introduce fuel droplets into the air flow produced by the air flow means (8,

9, 40, 41, 50, 55) so that the fuel droplets are carried to the evaporation surface (2, 2',

2", 23) by the air flow

3 An evaporator according to claim 2, wherein air diverting means (15, 22, 30, 31)

is provided for producing a change in direction of the air flow in the vicinity of the

evaporation surface (2, 2', 2", 23) in order to cause a proportion of the fuel droplets in

the air flow to collide with the evaporation surface

4. An evaporator according to claim 2 or 3, wherein the evaporation surface is

defined by a filter matrix (23) having an open cell structure through which the air flow is passed by the air flow means (8, 9; 40, 41; 50, 55) so that fuel droplets within the air

flow are captured at a plurality of locations within the filter matrix.

5. An evaporator according to claim 4, wherein the filter matrix (23) comprises a

reticulated metal foam.

6. An evaporator according to claim 4 or 5, wherein the filter matrix (23) is graded

so as to have pores of decreasing size in the direction of the air flow so that larger fuel

droplets tend to be captured in an upstream portion of the filter matrix and smaller fuel

droplets tend to be captured in a downstream portion of the filter matrix

7. An evaporator according to claim 2 or 3, wherein the evaporation surface is

defined by a series of parallel plates (12) which are spaced apart so as to define narrow

passages (13) therebetween through which the air flow is passed by the air flow means

(8, 9; 40, 41; 50, 55) so that fuel droplets within the air flow are captured at a plurality

of locations distributed over the plates

8. An evaporator according to claims 3 and 7, wherein the air diverting means

includes inlet portions of the parallel plates (12) incorporating at least one bend (30, 31)

followed by straight portions (32) of the plates.

9. An evaporator according to any preceding claim, wherein temperature control means is provided for varying the temperature of the air and fuel supplied to the

evaporation surface so that the temperature is greater at low loads than at high loads

10. An evaporator according to any preceding claim, wherein feedback means is

provided for recirculating a proportion of the outputted mixture of fuel and air to the

evaporation surface at low loads

1 1 A carburettor metering system incorporating an evaporator according to any

preceding claim.