More Lucas PI info - some pearls of wisdom here :-)

Over the years I've collected loads of info and books about the PI system, some better than others. I have always meant to scan and post up some of the more interesting stuff. I had my scanner working recently (but not on my spanking new Vista equipped PC - incompatible!) so I thought I'd scan a few things when I came across an old trio of A4 sheets held together with a rusty staple. They appear to be pages from a magazine article but from where I don't know.

The last page has a few small adds on it but no clue as to where it first appeared. The opening paragraph suggests and owners club and the adverts are 1990 vintage - if you know where this came from please let me know and I'll give them the credit.

Anyway, I scanned it and OCR'd it and here it is. I've been through it to correct the OCR errors but if anything doesn't make sense shout up and I'll check the original.

THE LUCAS PETROL INJECTION SYSTEM - MARK II

We are very fortunate to be able to publish these notes on the development of the Lucas PI system. These notes by M. H. Evans were sent to Dr. Michael Bingley. M. H. Evans who is chairman of Rolls Royce has kindly given his permission for the notes to be published. They are sure to be of interest to all PI owners.

1. INTRODUCTION

Origin

The Lucas fuel injection system was originally designed for Rolls Royce. Around the end of the war Lucas designed a fuel injection system for the Merlin aero engine (Probably its tank version, the Meteor and later tanks e.g. the Centurion, had Lucas fuel injection instead of a large Zenith carburettor.) It was not a direct fuel injection such as was fitted to German Daimler Benz aero engines of World War 2 as this would have had lost the effect of charge cooling. Lucas continued to design and build fuel systems for Rolls Royce gas turbo aero engines and they still do.

The Lucas petrol injection system however was designed for a particular application of one of the "B" range military and commercial petrol engines produced by Crewe.

These engines, of 4, 6 and 8 cylinders were designed in the late 1930s for motor cars which might have replaced the Phantom Ill, Wraith and Bentley Mk V if war had not intervened. The post-war cars were powered by versions of the B60 and B80, but a major application of these engines was, and is, powering military combat vehicles.

One particular application of a B range engine was a post-war German light tank. A tight spec. was put on this engine. This led to developing a special cylinder head with individual instead of twinned exhaust ports, and also to Lucas developing the fuel injection system in place of the normal carburettors. In finality the tank was not proceeded with, but the fuel system was employed in racing and found its way into certain production Triumph models.

The System

This system, which is fitted to the Triumph 2.5 PI Mks 1 & 2 and Triumph TR5 & 6, is described in the relevant workshop manuals.

The manuals give adequate information for stripping and rebuilding the system, trouble shooting and setting up both throttle butterflies and the pressure relief valve. They do not, however, give sufficient quantitative information on the "correct" settings for the metering unit camtrack and setting rings.

A great deal of useful quantitative information is, however, given in the Lucas Service Training Centre manual entitled "Petrol Injection Mk ll." But even this does not give a lot of background information that would be useful in keeping the system operative in future years when factory replacement became short.

Further information has thus been sought from key personnel within the Lucas organization. The system was designed by Harry Bottoms, who normally designs aero engine fuel systems. It was developed by Jim Littlehales of the Engine Fuelling and Controls Systems Development, commonly referred to in the past as the Injection Lab. Both have contributed information, but the vast majority has come from Jim Littlehales, whose patience with enthusiasts for the system is saintly' it would have been impossible for him to have helped more.

2. THE INLET MANIFOLDS

Original Design

When I first spoke to Harry Bottoms in 1973 he was horrified to hear that 6 separate inlet pipes and butterflies had been fitted. His original design called for only one or at most two. Fitting six brings obvious synchronization problems.

Early Pattern

Inlets fitted to the Mk 1, and possibly early Mk 2's had individual tickover stop adjusters. Throttles were operated by separate rods which operated the crank of each inlet pair. These rods were actuated by a master shaft below the inlets. These rods were adjustable for length with the result that tickover and pickup were both easy to set up.

Later Pattern

Later systems, for some reason, work on the basis of the first inlet pair butterfly spindle picking up the second, and the second picking up the third through crank levers similar to the systems with multiple SUs. Tickover stops are not fitted and it is much more difficult to achieve a smooth pickup from tickover than with the earlier system. This is because the first pair open before the second and the second before the third once there is any wear on the spindles. For reasons I have been unable to find, this later system, is fitted with a double interlinking balance pipe between the three pairs of inlet castings. Earlier systems only had one balance pipe and seemed to work perfectly well.

3. INJECTORS

These are simple in design compared with diesel injectors. They do not have to contend with the pressures and temperatures of the combustion chamber as they are mounted in the inlet manifold.

First Type

The earliest pattern of at least three designs fitted was designed by Jim Littlehales. It is illustrated in the injection manual. This pattern has a nylon collar which is screw threaded onto the injector barrel. The inner bore is tapered and honed. Thus, when the injector itself is fitted, its "O" ring seal is progressively compressed to form a seal. As far as Jim can remember, no spring circlip was fitted on the nose to retain the injector insert. It was unnecessary as the "O" ring effectively held the insert in place. CAV made these early injectors.

This pattern was dropped once the value engineers got their hands on the system because the screw thread taper bores were costly to make.

Second Type

A further pattern was introduced of which Jim has little clear memory. In this, the injector insert appears to have been inserted from the front end of the injection barrel, the barrel then being swaged over to retain it. They appear to have had no "O" ring seal and to have relied on metal to metal contact of ground faces for sealing. The actual injector was identical to that in the earlier unit except that there was no pip on the valve. The pip was the original grinding centre.

Last and Most Common Type

The third kind, and by far the most common, looks on first sight to be identical to the second, except in that the valve has the pip once more, like the original. These are built, like the original, with an "C" ring seal, but in place of the taper bore there is a stepped bore of two parallel diameters. When replacing the internal "O" ring seal therefore it is necessary to use a thick oil to prevent damage to the "O" ring when it meets the sharp edge at the change in internal diameter. The injector insert is retained in this design by a circular spring ring on the nose of the insert. As with the second pattern, the nylon block is a press fit on the injector barrel.

Dribbling

Whilst Jim Littlehales never experienced leakage problems with the inner 'C' ring seals, this has not been the experience of others who have run cars with this system for many miles. The injector inserts tend to shuffle a little in the injector barrels. Carbon gets between the barrel and the 'C' ring and wears the ring until it is a 'D' section instead of an 'O'. It then starts dribbling, but can be cured with a new 'O' ring once the bore has been carefully cleaned.

Some of the third pattern were manufactured with too short a thread engagement in the inner nylon adjuster nut. They unthreaded themselves and the injector valve then damaged below the valve seat. The nylon adjuster is an interference fit on the thread and is adjusted with concentric Allen Keys on all three types.

Injector Performance

Injectors should blow at between 47.5 and 50 psi. They are all right at anything above 45 up to 55. They should give an even cone spray. There should be no leakage at all at a pressure of 5 psi below their blow off setting. This is what Jim said when I last saw him but I believe there was a production standard of several drops per minute allowable leakage.

Rough running can be experienced if injectors dribble although this has little effect on fuel consumption, if any.

4. INJECTOR LEAD NON-RETURN VALVES

These are located at the junction of each pipe with the metering unit. The valve looks to be of neoprene although Jim describes it as hard rubber. They are designed with plenty of clearance between the valve and the bore in which they are located. In theory this allows the valve to process around on its seating. In practice, however, they tend to stay on the same spot and continuous operation tends to produce a circular groove on the valve face. They then leak, allowing the residual pressure caused by the elasticity of the fuel line to leak back into the metering unit. The next injection stroke of the metering unit therefore spends much of its energy re-expanding the pipe instead of forcing the correct quantity of fuel out of the injector. Also, after running, heat from the engine can vaporize the fuel in the pipe and it then takes a long time to reprime. That cylinder runs "dead" until the line refills. The neoprene valve face needs carefully rubbing flat again if leakage occurs.

5. METERING UNITS

Return Pipe

Early metering units had a "push on" rubber connector to the fuel return pipe. Later ones had a screw on connection. If the return pipe gets blocked, and this is not an uncommon fault, pressure rises in the unit and forces the diaphragm between the metering barrel and cambox towards the latter. This tends to prevent the roller climbing to the maximum depression the fuel is in weak condition up the camtrack. in consequence fuel consumption is increased. The pipe can usually be unblocked either by using a foot pump or airline on the return pipe, or by carefully pushing soft iron wire down it.

Golden Units

One cause of blockage can be copper particles in the return pipe. This happens if the lead rich copper end thrust washer on the metering barrel wears, as the swarf returns by the pipe to the fuel tank. Bill Phelps came across this problem, which in its extreme reduces the alignment of the barrel ports with the injector lines. This reduces fuel delivery to the injectors. When I reported this to Jim Littlehales he said that this problem did occur and that units so affected were referred to as "golden units". Petrol has a low lubricity and, he said, if the end washer did not sit exactly at right angles to the bore (i.e. parallel to the face on which it sits) the petrol escapes unevenly and wear sets in on the "high spot". Once this happens, the process of wear is very rapid. It also has the effect of increasing the clearance between the rollers, camtrack and piston stop, thus making the unit go "fuel rich". A few thou can absolutely ruin fuel consumption. Clearance between the metering barrel and its sleeve, incidentally is a fraction of a thou, although subject to variation in manufacturing tolerance . Tolerances are unusually fine for car components.

Lubrication

The shaft on which the rollers run should be lubricated with a little moly green or moly additive. The piston stop should not need lubricating. Jim Littlehales says they should run in an oilite bush. However, I have seen signs of "pick up" on the piston stop. Jim says that if lubricated at all, it should be with a thin oil like 3 in 1.

Springs

The springs in the diaphragm capsule are critical and must not be changed unless the capsule nuts and camtrack are recalibrated.

A critical fault occurred in some units. The high rate "second" springs were feather ended. In other words the end coil, ground flat and therefore thin to give the spring a flat end, did not touch the first full section coil. As a result the high rate spring did not have sufficient rate until the feather end had deflected enough to touch the next coil. The first part of the spring's compression happened too easily. This has the result of putting a fuel rich kink in the depression/fuel delivery working line right where you don't want it - in the very middle of the normal driving range. A possible cure would be a blob of Araldite to support the feather end on the next coil. This fault is always worth looking for. Feather ends were sporadic throughout production - not just an occasional batch problem.

Wear And Roller Pin & Cam Setting

Frettage corrosion can occur on the pin if run dry. When this happens it increases the clearance between the rollers and the piston stop. The cure is to readjust the camtrack to .002" and .058" clearance at either end of its working range, using feeler gauges. This is easily done with the capsule springs removed using the mouth to suck on a pipe to raise the rollers to their to position. Measurements should be made with the unit in its normal operating attitude and the feeler gauges should be stroked downwards, not upwards, to retain the correct positions given by gravity when there is any "play" either in the nylon ball-joint or roller pin.

Camplates

These are hardened steel and do not wear. However, the constant hammering of the metering shuttle can move the camplate away from the rollers by rotating it about its fulcrum. This sends the system fuel rich. Two setscrews hold the camplate in position on the fulcrum arm. In some cases these are cross slotted screws. Others are Phillips or Posidrive. The former can be adequately tightened whereas the latter have proved unsatisfactory in my experience. The centre tends to trepan out before the screw is tight enough. When tightening these screws I apply Loctite then leave the unit for 24 hours before using it again. On my Mk1 PI my fuel consumption was around 19.7 on overall running - much of which was my 17 mile drives to and from work - until I reset the camtrack. After resetting this went up to 27.4.

Faults of the System

The final demise of the system was the difficulty of meeting the emission control regulations especially at the very low fuel delivery quantities at tickover and low throttle openings. The system had by then already got itself a bad reputation. In part this was because garages did not understand the system and could not set it up correctly. It was also partly due to manufacturing problems with the system. As developed the system was excellent but the value engineers cheapened important bits of it. Furthermore, when manufacture transferred from Lucas Aerospace to Lucas Automotive, the criticality of manufacture was lost. It has been said that the people building the all important fuel pump motors "thought they were building windscreen wiper motors". Oil got under the commutator segments during motor manufacture, with the result that the segments lifted and then the motors failed. Pump shaft seals wore and allowed fuel to leak into the motor. All these problems were overcome in due course, but rather late in the day.

Emission Control

There is obviously a small (but variable in manufacturing terms) clearance between the metering barrel and its sleeve. As a result "inter port" leakage takes place, thus increasing fuel consumption. The greater the clearance, the greater the leakage. It can be as much as 20% more than the quantity metered at small shuttle movements. In 1973 Lucas therefore introduced three production standards - A, B and C. These are marked on the metering unit if built after about July of that year. The unit was run at 2.50 rpm rotor speed and a 0.010 shuttle stroke. Using petrol, average measured quantities per 1000 injections (1000 revolutions) had to be as follows:
A - 8.4 cc to 8.9 cc

B - 7.8 cc to 8.3 cc

C - 7.2 cc to 7.8 cc

All plus or minus 10% with petrol.

Green Top Units

At the end of the production period a three spring capsule was introduced to attempt to get a little nearer to the theoretical depression/delivery curve requirement. These units had a green plastic cap on the adjuster nuts and were called "green top" units. They were only a marginal improvement in attempting to meet emission regulations and few were fitted to cars.

Springs

I have a metering unit with square section springs. Jim Littlehales cannot recall any of these and is suspicious that the springs are not original. He believes all springs were manufactured from round section wire. He told me to check the cam clearances at various depressions against the data in the manual to see if the unit was properly set up.

Pressure Relief Valve Design

The pressure relief valve is a very neat piece of design and was the work of Harry Bottoms. It is very well finished, as are many parts of the injection system. The quality being above normal motor practice and reminiscent of aero engine components prior to World War 2. It does not, however, approach modern practice.

Function

The function of the valve is to maintain fuel pressure at between 100 and 110 psi when the car is running. However it is also designed to achieve one other thing. A concern from the outset of the system design was the case of a "dry tank". In this situation the system had to be capable of repriming itself and the relief valve is designed to facilitate this. Early systems (about 70) had a "well" or small tank alongside the pump. This filled by gravity from the main fuel tank. If the main tank ran dry the first fuel put in it would run down the well. The well was fitted with an air vent pipe which connected to the top of the tank. It could have vented the atmosphere above the fuel tank "full" level except that it is both safer to close circuit back to the tank and also ensures that any air passing down the pipe is filtered (see "filters"). This return pipe allowed the air in the well to be displaced when the fuel flowed in, the well thereby filling completely.

Its Operation

Supposing the system to be dry, the pump will deliver air at a pressure of about 20 psi. At 20 psi the relief valve begins to open. In doing so small bleed holes are uncovered. These allow the air to escape through the valve and back to the fuel tank. Once the air is displaced and fuel flows, fuel begins to flow through the bleed holes. However, as it is denser than air, it cannot escape as fast as the pump delivers it. The pressure then builds up further and at 60 psi the valve moves further, cutting off the bleed holes. No more bleed occurs until the full operating pressure of 100 - 110 psi is reached when the main valve opens. PRESSURE SHOULD NEVER BE BELOW 100 psi at "full chat" on the road according to Jim Littlehales or misfiring will occur.

FILTRATION

Cleanliness of fuel is essential with the system. There is a full flow main filter between the fuel tank and the fuel pump. There are two small coarse filters in the pump inlet and outlet elbows. There is a small nylon filter in each injector barrel and I think, one in the metering unit. Air entering the fuel tank to replace burned petrol is also filtered. There is only one place that a filter is needed but not fitted and that is on the fuel pump breather pipe (see FUEL PUMP).

Early systems had the well referred to in the previous section. Petrol entering it was filtered through a very tiny unit which according to the handbook, was to be changed at something like 12,000 mile intervals. It was totally inadequate. A small quantity of water in the fuel was all it needed to block the filter partially. If the fuel level was low or even with plenty of fuel if you turned left suddenly, the fuel pump found it easier to suck air down the well's breather pipe than from the tank. The filter assumed gravity would always force fuel through the filter down into the well. My engine used to cut out on my 2.5 Mk1 on left hand corners for 2 - 3 seconds while the pump screamed. Colleagues at Hucknell nearly rammed me behind on several occasions when we were hurrying over from Derby to get to work on time. I asked Jim Littlehales what to do. He told me to throw away the well and fit a CAV diesel oil filter (which was similar to or the same as fitted to 2.5 Mk 2 cars). He said it would pass nothing above 2 microns. I did this, but retained the well although clipping the vent pipe so that its maximum rate of passing air was greatly reduced. I never had any more problems. Recently Jim has warned me that these CAV elements are not flushed on production and that fibres can get into the pump and cause damage. He suggests giving a new filter element a really good wash out before fitting it I do, not believe any of the other filters are important, except perhaps the air filter on the fuel tank, and that is doubtful as the breather pipe is very long, goes downhill all the way, and fuel does not burn quickly enough for air movement up the pipe to be speedy enough to carry dust with it.

The Pump

Reliability

The pump is the only part of the system that can let you down completely and leave you stranded. They can do so with absolutely no warning at all. I have had many fail on me.

As originally designed, it was a good piece of equipment but the value engineers got their hands on it and the troubles started. Slowly, their integrity was restored and late pumps were generally very good.

Regardless of date of manufacture, pumps vary greatly. Some have lasted me just over a year. One especially selected for its low current consumption for me by Jim Littlehales ran from the summer of 1975 to the autumn of 1978 when it aired. Cleaned, it resumed work in the summer of 1979 and, at the time of writing, is going as beautifully as ever. The one originally fitted to my 2.5 PI MK 2 ran from 1 January 1975 till the summer of 1979, and the pump on Jim Littlehales own car ran from something like 1972 or 3 until just after my original MK 2 failed in the summer of 1979. Even then it didn't really fail it occasionally skipped a beat. On examination it was found that the carbon brushes had worn out completely and the copper rat tail brush heads were lying on the commutator and supplying the current.

Early Failures

My first failure was of a shaft seal on my MK 1. The pump was then 3½ years old. It did not leak fuel down the vent but entrained air through the seal in some quantities. It thus pumped an aerated mixture, especially when hot. This caused restarting problems and the best solution was not to switch the engine off if avoidable in hot weather.

My second failure followed about a year later. The bearing pins of the pump gearwheels wore until the gears started to cut into the pump casing centre (figure 8 aperture) plate. Since then I have never had any real problem with the actual pump part of the unit.

Later Failures

Recent failures have been of the motor. In almost every case, stripping and carefully cleaning the commutator restores the unit to continued good service. I have done this many times and the pump can run, trouble free, for as long again.

The Motor Bearings

The rotating assembly is supported by two spherically mounted oilite bushes, and positioned at one end by a thin steel thrust washer bearing on one face against a bush and on the other against a circlip on the motor shaft. The other end is positioned by a nylon buffer on the end of the adjustable setscrew projecting from the motor casing's end. I have never experienced pump bearing problems either with bushes or the shaft.

Motor Rotating Assembly

These vary in current consumption. Good ones absorb less than 5.8 A and poorer ones well over 6 A. Jim always prefers 6 A max. (5.8 for hot countries as heat causes cavitation). Lucas manuals are a little ambitious in their claims. Windings can burn out. It has happened to me once.

Commutators at one stage gave mechanical trouble. Oil got below the segments during manufacture with the result that sooner or later they came apart. This problem did not last long. With time however, on all pumps the commutator gets dirty and a certain amount of burning of the copper takes place, causing the pump to stop working. I clean them by mounting the rotating assembly in an electric drill to spin it then gently rubbing fine sandpaper on the commutators. Once this shows as smooth and bright copper all the way round again, I use fine, blunted sandpaper to polish the commutator and finally Duraglit wrapped in a piece of old handkerchief. I then clean out the gaps with a needle and finally clean with a hanky moistened with "tric" or "carbontet". Wrap sellotape on the shaft before "chucking" in the drill.

Brushes

Brushes give no trouble other than in that they wear very slowly away. When reassembling a motor be careful not to let the rat tail brush heads get caught behind the brush holders as, with wear, the brushes will be restrained from moving for- ward and maintaining contact with the commutator.

Leads

Heat and/or petrol vapour can harden the plastic covering. New leads can be made and soldered on to the brush holders although the originals are welded on. Jim Littlehales told me this.

Replacing Brushes

Someone once told me that the paxolin brush, deck and brushes from a Lucas windscreen wiper motor will fit, although only two of the three brushes are needed. I have never tried it has not been necessary.

Heating

According to their current consumption and state of wear, motors can get very hot. They normally run too hot to touch comfortably. After a long run on a summer's day the heat will sink from the motor into the pump on a car that has stopped. Fuel then evaporates and the pump screams. The car often will not start until the whole thing has cooled down. During running the fuel cools the pump so no problem arises. The problem is accentuated in a hot climate.

Best cure courtesy Jim Littlehales is to install a pump cooler. Wrap ¼" bundy pipe in a spiral round a mandrel of slightly smaller diameter than the motor. Then spring the coil open a bit and work it over the pump housing. Connect the bleed valve return fuel flow to the copper pipe. It then provides fuel cooling, the warmed fuel returning to the tank. This works well. A cruder cure is to slap a rag soaked in water onto the motor in hot weather. It does work. Bill Phelps suggested that mounting the pump vertically should help, motor up and pump down, as heat rises.

Magnets

Their power varies but Jim Littlehales does not see this affecting current consumption. They can however crack, so if removing them to clean the motor housing, slide them out with care. Replace them the right way round afterwards!

Noise

Some are quiet, some are noisy, some start noisy and run quieter as they warm up. The noise dips and rises with the flashes of the indicator, or application of brakes. This is only the effect of variation in the current supply available. It is not a cause for concern. Minimum noise requires ½ turn of end float on the endfloat adjusting setscrew.

Vent Pipe

Between the motor housing and the shaft seal which isolates the motor from the pump there is a vent pipe which exhausts through the car floor onto the ground. It is intended to dump any petrol that leaks past the shaft seal and prevent its entering the motor. With a horizontally mounted motor as in the 2.5 PI saloon one wonders if it picks up all the leaking fuel or whether some reaches the motor. In the estate the pump is mounted vertically and this should ensure that any leaking fuel is dumped, quite apart from being a better way of keeping the motor's heat away from the pump.

The vent pipe is the one point in the system where a filter should be fitted, but there is none. This is because the pipe exhausts at one of the dirtiest and dustiest points possible - right behind the rear wheel of the car. This would be fine if no air ever moved up the pipe, but does. When the car is run, the motor heats. The air inside expands and some is expelled. When the car stops, the motor cools and air is sucked up inside the pipe. Quite apart from this the shaft seal usually leaks a little - fuel out when stationary and sometimes when running - air in when the engine is running in many cases. Minute differences in the pump determine whether the seal is subjected to pressure or a depression. Many pumps actually suck air in past the seal and feed it into the system under pressure with the fuel. This causes erosion when the air is humid.

My solution is to push a bit of cellular plastic foam up the vent pipe and, from time to time, moisten it with a squirt of WD40. It is easy to do and at least gives a measure of filtration to the air.

Shaft Seal

The shaft seal as originally designed was in Viton but value engineers changed this for a cheaper and less satisfactory material. Pumps today again have the seal made in Viton and it is very satisfactory. It is a standard proprietary seal made by George Angur.

The seal runs with less interference on the shaft than such a seal normally would. The pressures are not very great and too much friction would cause undue heating. As it is, the Viton seal is rather hard when cold and tends to leak a little until it warms up. It warms quickly once the motor starts however, through friction.

As stated elsewhere in this document, the seal can allow air to pass into the pump or fuel to leak out. If damp air is sucked in, corrosion takes place on the motor shaft and the rust particles stick to the seal lip. They act as an abrasive and a groove is worn in the shaft by the seal. The seal then becomes less effective in its functioning. 1 have seen a number of pumps so affected.

The seal is to some degree lubricated - by petrol splashed at it by the helices on the plastic drive coupling between the motor shaft and pump drive gear. Provided the seal lip is at least wet, it is lubricated.

The Pump

The pump itself consists of three brass plates. The rear one incorporates the inlet and outlet unions and carries the bearing pins for the two pump gear-wheels. The centre plate incorporates a figure of eight aperture to accommodate the two pump gears. The front plate - thin on very early pumps but beefed up on later ones as they distorted under the working pressure inside the pump - abuts to the motor and has one aperture through Which the drive gear passes to its coupling to the meter.

The pump is made to exceedingly tight tolerances and these are critical.

Clearance Gear Faces

The centre plate is slightly thicker than the gear wheels so that there is a running clearance between the gearwheels and the top and bottom plates. The production tolerance on total clearance for the gears was 0.0002" - 0.0008" (2/10 to 8/10 thou). In practice Lucas aimed with great care to achieve 0.0004" (4110 thou) on production build. Above 8/10 thou clearance the pump rapidly loses flow and becomes useless. That is why the tolerance is so critical. (1/10 thou extra clearance loses you 1 gallon/hour at 100 psi approx.)

Clearance Gear Teeth

The clearance peripherally between the gear teeth and the figure of eight aperture is nominally 0.004" (4 thou). Again this is critical. Above that figure you lose one gallon per hour pumping capacity at 100 psi for every extra thou of clearance. Several thou, therefore, and you have zero flow.

There is considerable clearance between the tips of the teeth on the gearwheels and the roots of the teeth on the mating wheel. This is to prevent hydraulic locking occurring by fuel being trapped between the teeth when they mesh.

Bearings

The gearwheels are bushed and run on hardened and ground steel pins mounted in the bottom plate. The driven gear pin is not provided with special means of lubrication but the drive gear pin is. It is hollow and has a hole bored at right angles into it. Fuel flows down the inside of the pin and lubricates the bush through the side drilling.

The bushes in the gearwheels are of a special carbon which has a copper content. The copper is to conduct the heat away which carbon on its own would not. As far as Jim Littlehales can remember, this is Morganite MY3D. The bush is fitted into the gearwheel with virtually zero clearance. The clearance in fact is 0.00005 (½ a tenth of a thou). Inserting the bush normally gives the slightest shaving to the bush.

Bearing clearance between the bush and bearing pin is nominally 0.001 " (one thou) but this is not as critical as the surface finish in the bush bore. There is only one way to get this finish, and it can be done by hand. Use a soft mild steel bar and have syntox (aluminium oxide) sprayed onto it. Then grind the syntox parallel on a diamond wheel. Place the rod in the carbon bush and polish it. It will give a mirror finish.

Bearing Wear

Bushes can pick up metal particles and then they will cut up the bearing pins rapidly. I have only seen this in one (my second ever) pump. Strangely, Jim Littlehales once saw a pump in which the bush to bearing clearances had become massive. The gears had cut their way into the centre plate creating a "pregnant" figure of 8 aperture. It still worked and still pumped fuel.

Gear Corrosion

They don't.

Gear Manufacture

Gears are ground and through hardened. Early gears were hobbed but this left striations on the teeth. Grinding was found more satisfactory.

Wear Between Gear Facings and Plates

In theory there is high pressure on one side of the gears (c 100 psi) and low on the other so a slight leakage occurs across the gear faces and lubricates them. In practice the flow tends to set up on one face rather than both faces of the gear. The other face can therefore run dry and cause friction, heat and wear. In the early days of developing the Mk 1 system Jim Littlehales tried just about everything to achieve lubrication of the gear faces. Finally, in desperation, he inscribed a helical groove on a gear face with a scriber. To his surprise the current consumption of the motor went down, showing that lubrication had been effected. It became known as "Jim's Groove" although its precise functioning was not fully understood. Any loose fibres from the filter used to work their way down the groove and form a blockage. In effect the blockage became a disc brake pad. Mk 2 systems do not use Jim's Groove.

Gears can cause grooving of the top or bottom plate. 1 have seen both although Jim says they are common on the pin plates below the drive gear. As Jim says, they are not very important as they tend to look a lot worse than they really are.

Restoring Side Plate Clearance

If it is desired to remove the grooves, the top and bottom plates can be restored but the centre figure of 8 plate must not be touched as its thickness is so critical. use brand new wet and dry paper on a completely flat surface - a measuring table or piece of optically ground glass of size than the brass plate. The wet and dry must have no kinks or creases.

After sanding you must rub the surface with your fingers with paraffin and keep doing this until your fingers remain clean. This will remove the embedded particles. Lastly, place clean paper on your flat surface and pour Brasso on it rub the plate on the paper