Lucas Mechanical PI System Model

[Roger B]

At first I was not sure whether to write a thread on this as it is quite specialised. It started as a discussion on the Porsche fuel injection thread and then I realised that if it worked it would be a good solution for one of my future engine designs.

The injectors are based on my previous poppet/mushroom design but reduced in size as they will not have to fit my existing diesel engines and don’t need the compression seal.

As this system has a quantity of petrol pressurised to around 7 bar (if I have to use the same pressure as the full size system) some though will need to be given to safety.

The injector nozzles are turned from 6 mm hex brass with the bore drilled 0.8mm and reamed 1mm. there is then a clearance bore for the needle and a larger clearance for the spring.

The needle is made in the same way as the larger ones using 1.5mm silver steel:

Centre drill with a 0.5mm bit.

Turn the portion that will be threaded down to 1mm.

Thread M1.

[Roger B]

Continuing the needle sequence:

Support the needle with a 1mm bore brass bush held in the tailstock chuck and turn the next section down to 1mm.

Turn the section before the cone down to 0.8mm to allow the fuel to pass.

Set the cross slide to 5° and turn the cone.

Part off using a 1.2mm bore brass bush to set the parting tool in the right place on the cone.

[Roger B]

Next the union nipples were turned from 3.5mm brass rod.

All the turning was done using an Applitec 1mm parting and grooving tool.

The injector body and M4 x 0.5 union were turned from 5mm brass hex also using the Applitec parting tool  to avoid changing tools and having to reset the cross slide zero position.

The cones on the needles were carefully polished with a 400 grade diamond lap. I decided to try not lapping the cone onto the nozzle as I have done with previous poppet injectors.

Finally I made up an M4 x 0.5 to M5 x 0.5 adaptor pipe using 2 mm od 1 mm bore copper pipe so I could make some trials.

[Roger B]

The injector was fitted to my test pump and tested with lamp oil (kerosene that doesn’t smell). The spray pattern was quite good but not sufficiently atomised to be ignited with a flame.

<a href="https://www.youtube.com/watch?v=GHcJAl6gp80" target="_blank">http://www.youtube.com/watch?v=GHcJAl6gp80</a>

I made an adaptor from a short length of plastic pipe to fit the inlet manifold on my 12 cc twin. After blowing out the lamp oil and priming with petrol. It ran quite well, but obviously much too rich due to the reduced pressure on the helix pump reducing the leakage.

<a href="https://www.youtube.com/watch?v=yVcLcY32dkY" target="_blank">http://www.youtube.com/watch?v=yVcLcY32dkY</a>

Manually adjusting the fuel rack improved the running, however a different control cam will be required.

[Vixen]

Hello Roger,

What an excellent tutorial of how to make the parts for one of your miniature injectors.   

What can you tell us about the tiny spring, which is obviously key to getting the tiny injectors to perform so well? Are they commercially available, or you need to 'roll your own'? Perhaps the more important parameters are spring rate and free length. What closing force on the needle are you currently trying?

Thanks for posting this detail.

Mike

[Roger B]

Thank you Mike   

The spring is a commercial item, full details below.  I decided at the beginning of my fuel injection experiments to use commercial springs for consistency and to allow others to copy what I have done.

The nuts are made from 2mm brass hex.

For actual use the injectors could be made smaller by soldering the fuel pipe directly to the body without using a union. The nozzle hexagon could also be reduced in size.

The injection pump will follow the two cylinder design below using a 2mm diameter shuttle with a maximum stroke of 2mm. Previous trials have shown that this is more that enough to petrol inject a 25cc cylinder.

To allow the pump sleeve to be fitted to the body without damaging the O rings one of the O ring grooves is cut in the body. For this I need a narrow internal grooving tool that will enter a 12 mm bore. They are available, at a price, so I decided to try and modify a small Proxxon boring bar.

I very rarely try and grind lathe tools, preferring to use inserts for consistency, usually DCGT 070204. I ground something that I thought looked right, checking with an offcut with a 12mm hole that there was some clearance.
After a little experimentation with tool height and angle I got a good cut 

[Vixen]

Hello Roger,

Thank you for the spring information, I hope I have not distracted you too much from your excellent miniature injector thread.

I looked back through your previous posts but could not find much spring info. I see you tried a larger diameter spring which needed changes to the lock nuts to keep it centralised, not sure if the performance was better or worse than the 'standard' spring.
 
Am I correct in thinking you give the spring a 2.0mm preload? which would appear to give about 1.0 N  closing force on the needle.
 
I can find similar (but not identical) springs from the Lee Springs Ltd catalogue. Which way should I bias my choice? Is a higher spring rate preferable to a lower rate?

The more I look at the Lucas shuttle valve metering unit, the more I think it may be an excellent choice to meet our model engine requirements. The shuttle will ensure equal fuel delivery to each cylinder and the movable shuttle end stop looks like it is begging to have it's position set by a simple wedge. The wedge position being set either by manifold vacuum or by the throttle position.

The proposed size of you two cylinder Lucas injection pump is also interesting. You calculate a 2mm diameter shuttle with a 2mm stroke will be more that enough to petrol inject a 25cc cylinder. If we consider methanol requires approx. three time the quantity compared with petrol; then a similar 2mm diameter shuttle with a 2mm stroke will be more that enough to methanol inject an 8.0cc cylinder (917?). I calculated a 2mm diameter shuttle with a 1.0 mm stroke for methanol, so we are well within the same ballpark. Hopefully there should be good design read across.

Thanks again. 

Mike 

[dieselpilot]

The more I look at the Lucas shuttle valve metering unit, the more I think it may be an excellent choice to meet our model engine requirements. The shuttle will ensure equal fuel delivery to each cylinder and the movable shuttle end stop looks like it is begging to have it's position set by a simple wedge. The wedge position being set either by manifold vacuum or by the throttle position.

That's exactly how the system worked. Race versions used a throttle linked fuel cam. Road car systems used a manifold reference. The challenge will be feeding 6 or 12 cylinders.

The previous thread had me thinking I should try to build one for an RC aircraft engine.

[dieselpilot]

Roger, will you be using a diaphragm where the original does?

[Roger B]

I haven’t got into the details of pressures with this system yet. I am going to try the original Lucas pressures so that the injector should open at around 3 bar. That means a lower spring rate may be better. My running trials so far have been made with an injection pump that will reach at least 100 bar.

My 25cc horizontal engine has a 2mm bore pump and the working stroke is around 1mm so there is plenty of capacity to spare on petrol.

My initial trials will use a mechanical connection between the throttle and the shuttle position. Moving on to a diaphragm will require a good vacuum signal which requires a multi cylinder engine with a not too wild cam.

This first version of the shuttle pump is very much a proof of concept so I won’t be hardening and lapping, just using silver steel in reamed bores. The silver steel is typically 0.01mm undersize (in spec.) and the reamed hole are typically 0.01mm oversize (in spec. for H7) which give a running clearance of 0.02mm. This is too much for a 100bar diesel pump but may be ok for a 7 bar petrol pump 

The pump sleeve is made from 12 mm silver steel drilled 5.8 mm and reamed 6mm.

The pump rotor is 6 mm silver steel drilled 1.9mm and reamed 2mm. The ends are threaded m4 x 0.5.

[Roger B]

The sleeves (I made an additional as a spare) were then set up inn the mill using an edge finder and drilled through 1.5 mm.

The outlet ports were opened out with a 4.5 mm end mill for the sealing O ring.

The sleeves were then turned 180° and positioned with a 1.5mm drill to mill the fuel inlet passage on the other side.

[Vixen]

 Following every move 

Mike

[jcge]

Nice work Roger. I'm fascinated to see how this goes.
Regards
John

[steamer]

Watching VERY  closely!

Dave

[Roger B]

Thank you all for your support and interest 

Next up was to drill the ports in the rotor. The two holes have to be at 180° and offset by 10mm. I decided to use my small Proxxon RT as I could get one hole between a pair of jaws and the other outside.

With hindsight (so accurate) It would probably be better to drill the rotor and sleeve to the required depth, drill the required ports in a RT or dividing head whilst still attached to the stock, part off and then ream the bore to size. This would probably be essential for a 4 or 6 cylinder sleeve.

This operation went ok and I could assemble the rotor in the sleeve and put a 1.4mm drill into both ports at the same time 

I am still working on the design of the body. One end will have the cam/wedge to control the shuttle position, the other some form of drive coupling (very small Oldham style?). Both ends will have to have a return line for fuel that leaks past the shuttle or rotor. I will also need to select a suitable way to retain the fixed shuttle stop. For the finished item some form of Loctite would work, but in the experimental phase it needs to be a removable mechanical fixing. I am looking at either a very small grubscrew in the rotor, maybe M1.6 x 2, or a M1.4 screw from the coupling end.

Lots to think about   

[Vixen]

I am still working on the design of the body. One end will have the cam/wedge to control the shuttle position, the other some form of drive coupling (very small Oldham style?). Both ends will have to have a return line for fuel that leaks past the shuttle or rotor. I will also need to select a suitable way to retain the fixed shuttle stop. For the finished item some form of Loctite would work, but in the experimental phase it needs to be a removable mechanical fixing. I am looking at either a very small grubscrew in the rotor, maybe M1.6 x 2, or a M1.4 screw from the coupling end.

Roger,
"When in Rome do, as the Romans do".  When in Switzerland, "make it like a Swiss watch".

Hope that helps. 

Mike

[Roger B]

I turned a pair of nuts for the rotor from 8mm brass hex, one short one for the control end and a long one for the drive coupling. As anticipated I had to cut a 0.5mm run out groove at the end of the threads so the nuts seated on the rotor. I could then assemble the rotor and sleeve using shim washers to centralise the ports.

I will need an electric fuel feed pump for the trials, potentially working up to the prototypical 7 bar. The Modelcraft pump I have been using struggles to reach 2 bar so I wondered if I could modify one, possibly with some reduction gearing.  The gears are definitely designed for flow rather than pressure and I can’t see a way to remove the drive piece from the motor without damaging anything. It seems firmly pressed or bonded in place. I will make my own pump using 24T MOD 0.5 brass gears and a Portescap 28mm motor from the bits box.

The body is a piece of 20mm square aluminium, drilled 11.5mm and reamed 12mm. There are M5x0.5 threaded holes for the two delivery valves and the fuel inlet. I then cut a 1.5 x 0.7mm groove for the O ring that will seal the sleeve to the body. At the other end the O ring groove is in the sleeve to allow it to be assembled without shredding the O ring on the ports.

[Roger B]

Next up was the first attempt at the three metering components. These are all from 2mm silver steel. For this first trial I am not hardening and lapping them.

The fixed stop and the control stop were turned down to 1.5mm for the appropriate lengths and the end of the fixed stop was taped M1.4 for a fixing screw.

[55fairlane]

Watching this with great interest,  I spent a lot of time at Kinsler fuel injection (Troy Michigan), I raced hemi powered cars, and they were motivated by Hilborn mechanical fuel injection, so I needed to learn to meter alcohol.  We were drag racing,  the Cuda made a mere 900hp (big big horsepower in its day) to 3000 plus on nitro methane  for the digger....mechanical fuel injection will build rpm faster then super charger, and doesn't rob near the power off the engine.   

Turned several carburetor intake manifold to mechanical fuel injection manifold no one made a great manifold for small Fords......

...do you need a vapor separator catch can ? Or can you you simply return by passed fuel straight back to the tank? What are you doing for a high speed by-pass ? Same with low speed by-pass? Gonna use the high speed by-pass as a rev limiter?  How are you calculating the leanth of the air horn (bell on top of the injection stack)?

I'm gonna have to dig out all my books and information to brush up on Lucas metering............if memory serves me right in late 70's or early 80's USAC racing , you ran carburetors and alcohol.......someone showed up with a Ford , Lucas metering on gasoline,  and wooped the track! USAC outlawed gasoline and mechanical fuel injection.....

Loving this build!

[Vixen]

Hello Roger.

Still following closely.   

Mike

[steamer]

Hello Roger.

Still following closely.   

Mike

Absolutely!!!

[Roger B]

Thank you all for your support 

The next challenge is a miniature Oldham coupling. The first section screws onto the end of the rotor, butting up to it to help keep the alignment. The small gap confirms this is the case. I used a short piece of 8 mm brass hex with an M4 x 0.5 mm thread (left over from making the nuts) as a holder to drill through 1.5 mm with a 2 mm counterbore for the holding screw for the fixed stop.  The holder was then moved to the Proxxon mill to cut the 2mm slot.

[Roger B]

I turned an offcut of Delrin down to 6mm for the coupling element. This was then drilled through 1.4 mm to allow the element to be held for the second operation. For this step I needed an M1.4 screw with a 2mm diameter head. A commercial socket head screw was held in a 3 mm brass adaptor and carefully turned down to 2 mm diameter. This seemed to work and left sufficient material around the hexagon.

The Delrin blank was set up in the Proxxon RT and the tongue was milled, turning through 180° at each step. The 3 mm length was then parted off.

[Roger B]

The coupling piece was put back in the holder and set at 90° in the RT. It was then removed so that the coupling element could be fixed to it with the reduced M1.4 screw. This assembly was put back in the holder and the second tongue was milled. The finished coupling element was quite difficult to photograph.

The next piece was the drive part of the coupling. This was made from 6mm silver steel turned down to 4mm for the drive shaft. The 2mm slot was cut once again with the Proxxon mill followed by a quick trial assembly.

[Roger B]

Watching this with great interest,  I spent a lot of time at Kinsler fuel injection (Troy Michigan), I raced hemi powered cars, and they were motivated by Hilborn mechanical fuel injection, so I needed to learn to meter alcohol.  We were drag racing,  the Cuda made a mere 900hp (big big horsepower in its day) to 3000 plus on nitro methane  for the digger....mechanical fuel injection will build rpm faster then super charger, and doesn't rob near the power off the engine.   

Turned several carburetor intake manifold to mechanical fuel injection manifold no one made a great manifold for small Fords......

...do you need a vapor separator catch can ? Or can you you simply return by passed fuel straight back to the tank? What are you doing for a high speed by-pass ? Same with low speed by-pass? Gonna use the high speed by-pass as a rev limiter?  How are you calculating the leanth of the air horn (bell on top of the injection stack)?

I'm gonna have to dig out all my books and information to brush up on Lucas metering............if memory serves me right in late 70's or early 80's USAC racing , you ran carburetors and alcohol.......someone showed up with a Ford , Lucas metering on gasoline,  and wooped the track! USAC outlawed gasoline and mechanical fuel injection.....

Loving this build!

Thank you for your interest 

As the Lucas system uses timed injection pulses rather than continuous flow a lot of the complications of the Hilborn system goes away. Engine speed is automatically compensated for by injecting on each induction stroke.

Load compensation is from the throttle position on competition engines and the manifold vacuum on ‘normal’ engines.

[55fairlane]

 Kinsler.com lots of neat mechanical fuel injection

I'm guessing a rotary valve is used to control each pulse of fuel? Is that valve driven off the crank? Belt drive?

I need to look into Lucas much closer

[Roger B]

For various reasons I decided to modify the body design so that each end plate has a spigot to fit into the body and hold it central.

This requires a new, longer, body. The 12mm diameter 1mm long spigots were turned on the end plates and they were parted off from the stock. The spigot could then be held in a 12mm spring collet to allow the other side to be faced. It all seems to fit together ok.

A few weeks ago I ordered some gears for the fuel pressure pump from Ali Express. I received the expected 6, very long, pinions and a bag with grub screws and a hex key. These  are 24T MOD 0.5 with a usable length up to nearly 12mm. I hope the tooth profile is good enough for up to 7 bar with petrol.

[Roger B]

So continuing with the body. The inlet and outlet holes were drilled and tapped M5x0.5 and 2mm fixing holes were drilled in the end covers. These were then spotted through to the body and tapped M2. 1.2mm drain holes were drilled in each end cover to release any fuel that leaks past the metering system.

The control end cover required some thought. For the initial tests a simple screw adjuster is required but this must be convertible to a quick acting lever for engine trials. It is also important that the hole in the end cover is less than 2mm diameter to minimise the chance of the adjuster coming loose and the metering shuttle shooting out followed by a jet of fuel at 7 bar. I finally decided to press in a 4mm diameter bronze bush with a 1.5mm hole for the adjusting pin and an M3 thread for adjusting. Even if it becomes unscrewed the shuttle will only reach a maximum stroke of 2.5mm. For the final design I can press the bush back out and replace it with something else.

The drive end is drilled and reamed 6mm for a 4 mm bore bronze bush. Unfortunately the cover pulled out of the spring collet as the drill broke through. I was able to finish this off in the 4 jaw SC chuck as the concentricity requirement are not great.

For the next version I will develop an alternative machining sequence for the covers.

[Admiral_dk]

Great to see more progress on this very ambitious Project  + good to hear that the latest mishap was easy to cure 

I haven't chimed in earlier this year, as I had a minor work related accident on a eBike January 2nd. 
Didn't give it much thought at the time - only bruised skin on the left knee + it became swollen ....
Ten days later I was admitted to the local Hospital with a serious Infection for 76 hours + several weeks of laying on my back when I got back home .....
I did enjoy reading MEM and this thread very much - but didn't feel up to replying at the time ....

Best wishes

Per   

[steamer]

Sorry to hear that Per.....I know I dont bounce like I used to.

looking good roger!   following along

[55fairlane]

So continuing with the body. The inlet and outlet holes were drilled and tapped M5x0.5 and 2mm fixing holes were drilled in the end covers. These were then spotted through to the body and tapped M2. 1.2mm drain holes were drilled in each end cover to release any fuel that leaks past the metering system.

The control end cover required some thought. For the initial tests a simple screw adjuster is required but this must be convertible to a quick acting lever for engine trials. It is also important that the hole in the end cover is less than 2mm diameter to minimise the chance of the adjuster coming loose and the metering shuttle shooting out followed by a jet of fuel at 7 bar. I finally decided to press in a 4mm diameter bronze bush with a 1.5mm hole for the adjusting pin and an M3 thread for adjusting. Even if it becomes unscrewed the shuttle will only reach a maximum stroke of 2.5mm. For the final design I can press the bush back out and replace it with something else.

The drive end is drilled and reamed 6mm for a 4 mm bore bronze bush. Unfortunately the cover pulled out of the spring collet as the drill broke through. I was able to finish this off in the 4 jaw SC chuck as the concentricity requirement are not great.

For the next version I will develop an alternative machining sequence for the covers.

Man o live look at you go! This will be awesome when you get it all straightened out! Really enjoying the build

[Admiral_dk]

Thank you Dave. I'm doing fine now - but it took some time to get back to a decent speed off walking, biking etc. ....
And I completely agree on the 'Bouncing back time grows with age' thing - rather anoying .... isn't it 

Per       

[Roger B]

Hello Per, I’m glad you feel up to contributing again   

Here are the current drawings including the concept for a 6 cylinder version. I have attached the .DWG if anyone is interested (it is also a useful offsite backup).

[Roger B]

The stroke adjuster bush is turned from a offcut of bearing bronze. I just drilled the 1.5mm hole before pressing it in place as I think it would have collapsed if I had already opened it out to M3.

The drive end bush was turned from the same material and after being pressed in place it was reamed 4mm again. As the alignment was not perfect I used the floating reamer holder to avoid a tapered/bell mouthed bore.

The adjuster bush was then drilled and taped M3 for the stroke adjusting screw.

[Admiral_dk]

Slightly of topic - but does that Depth-Stop (in the second to last picture) work well for you Roger ...?

Every time I try any kind of stop mounted on the Drill-bit itself - I always end up in trouble .... 

I ask as it worries me a bit considering the precision required in this build ....

Per       

[Roger B]

Per, I have used this technique successfully on drills and end mills (generally when cutting the delivery valve seats on the 'normal' style injection pumps).

The bush/collar has to be the same size as the drill/end mill and the screw neeeds to be on a land.

I also use it as a depth guide rather than a stop so there is no real force applied.

[Admiral_dk]

Good to hear Roger .... it was just my immediate thought when I saw the picture 

Per       

[Roger B]

The next step was to machine the O ring grooves for the seal between the sleeve and the body.

The body was put back in the 4 jaw SC chuck in the same position as when it was bored. A quick check showed acceptable run out so I didn’t have to swap to the 4 jaw independent chuck.


<a href="https://www.youtube.com/watch?v=_d1Xz4tdWHs" target="_blank">http://www.youtube.com/watch?v=_d1Xz4tdWHs</a>



The groove was then cut with tool I had made earlier from a Proxxon boring bar. The groove in the sleeve was cut with the Applitec grooving tool and the other end was chamfered to allow entry into the body O ring.

The stroke adjusting hand wheel is a knurled aluminium disc, threaded M3 and tapped M1.6 for a grubscrew.

[Roger B]

The shaft is a piece of M3 allthread turned down to 1.4mm at one end for the stop pin. The adjuster was then assembled and the thread was parted off to give the required 3 mm maximum protrusion.

I would like to have a flat-bottomed hole for the delivery valve seat. For the helix pumps I use a 4.5mm end mill but I don’t have a 3.2mm end mill and so I will try using a 3.2mm drill with the end cut off square. Hopefully this will be adequate for brass.

I decided to make a trial using a piece of 7mm brass hex as will be used for the final version. The first hole was 2mm rather than 1.8mm and then ream to 2mm (as I would normally do for the ‘real’ part). It would have been better to start with the 3.2mm hole as the drill tended to snatch. The modified drill seamed to cut without chatter and the surface didn’t look bad, although difficult to photograph.

[Roger B]

I need to make a few trials with the fuel feed pump. There are various options for materials and construction. The initial attempt will be with brass gears in an aluminium body. I have used Delrin gears in my water pumps and steel gears in the Junkers oil pump. Delrin may be ok in Alkylate fuel but I think I prefer brass over steel as the lubricating properties are not so good.

I cut a couple of bits of aluminium for the body and cover. The main bearing will be an Oilite bush but as the ones I have are a little short I turned the flange off one to act as an extension using a 4mm tapered mandrel.

The gears as supplied have a small chamfer on the outside of the teeth which would cause additional leakage. I faced this off and then parted off 5mm slices.

To fix the driver gear to the shaft I lightly knurled the last 3mm of the shaft and then pushed it into the gear. It seems to be firm and runs reasonably truly.

Schillings suggested either silver soldering, with the potential problem of cleaning off any excess, or using coarse diamond dust to lock the two together.

[Roger B]

Next the body and cover were milled to size and the sealing face of the cover was milled flat (it was just the extruded surface). The bores for the shafts were drilled 3.8mm then reamed 4mm using the floating reamer holder. The separation of 12mm was set using the mill leadscrew.

The mesh of the gears seems a little too loose but I will continue with this for the moment.

[Roger B]

The ports were drilled 5mm with a separation of 11mm. I think they may be better with 10mm separation but once again I will continue with this. The inlet and outlet were drilled 4.5mm and taped M5 x 0.5.

To bore the cavities I mounted a piece of silver steel in a 6mm MT2 collet in the headstock and turned the end down to 4mm to fit the pump body. The body was then clamped to the faceplate and the bores were roughed out with a 12mm end mill.

Problem, the bores were over size     

I had used this technique before without problems but I realised that the previous pump had 6mm bores so the non-centre cutting end mill was not a problem. With 4mm bores it was being pushed off centre, hence the oversize bores     Start again