Model Engine Maker
Supporting => Engine Ancillaries => Topic started by: Roger B on November 14, 2021, 08:36:12 AM
-
My first fuel injection thread was getting a bit unwieldy so with some new thoughts I decided to start a new thread. The original is here https://www.modelenginemaker.com/index.php/topic,4906.0.html
The first piece will be a new, smaller, helix controlled injection pump. The last run of the 2 stroke diesel showed significant damage due to overtravel/overshoot of the cam follower so I decided to reduce the size and weight of the various components. The first step was a Mod 0.4 rack and pinion in place of the earlier Mod 0.5 one. The rack is wide enough that the sliding motion can take place there so I can reduce the width (and weight) of the pinion.
The pump element is made much as before, the guide for the pinion is reduced from 5 to 4mm. The cylinder was drilled 1.85mm and reamed 1.95mm using a floating reamer holder. A 1.97mm pin gauge would just enter the start of the bore. The delivery valve seating is finished with a 4.5mm end mill rather than a 4mm one. This makes it easier to cut the M5 x 0.5 thread for the delivery valve body.
-
Well I will still be following your progress, as I'm sure will a lot more here :cheers:
Per
-
I will be with you also Roger!
Don
-
Watching along!!! Hope you have a 12 plunger unit coming along! 8)
-
I will still be following along Roger. This is great stuff.
-Bob
-
Thank you all for your support and interest :) This is, as I have said before, a long term project but I hope to be able to offer some solutions. A 12 cylinder version is just a lot of single cylinder versions :) The Porsche 917 pump was actually quite interesting. It had two rows of pumps with the rack down the middle so you needed plungers with left and right hand helixes.
Having finished the pinion and the end of the pump element I tried it with the return spring. A 1.2kg mass compressed the spring around halfway. Is that too much or too little :thinking: no idea, it will have to wait for some running trials. Next up I drilled the inlet port in the pump element and milled the locating flat. I did this using my home made fine feed for the little Proxxon drill. I could also have moved the vice over to the Proxxon mill which would also have kept the alignment.
The pump body is again made much as before using the 4 jaw independent chuck to give the offset. The final stage was a quick sanity check that the recess to locate the element was of the correct depth.
-
Still following along! :ThumbsUp:
Dave
-
Thank you Dave :)
Next up I moved to the small Proxxon tools and milled the cavity in the pinion for the tappet. The inlet hole in the body was finished with a 4.5mm endmill to give a flat bottom and the most threads for the inlet union. I used the small Proxxon drill to tap the M2 threads to attach the flange to the body and then moved onto the tappet. This is made from a short piece of 3mm keysteel so it can be hardened. The ends were shaped by filling rather than some complicated mill set up with rounding cutters and the centre was spotted from the pinion using a 4mm diameter centre drill.
The next steps are to harden and lap the element and to grind the helix in the plunger.
-
Next up is the pump mounting flange. There were several possible ways to make this however I decided to do most of the work in the lathe. It is fixed to the body with M2 countersunk screws. The final operations will be completed the next time I'm in milling mode.
-
The flange was milled to width and then profiled with some filing buttons. Next the slot for the control rack was milled with a 5mm endmill and then opened up to 6mm to be a closed fit on the rack.
A 30mm piece of the rack was cut to approximate size with a hacksaw and finished by milling.
-
The rack is held in place with a small 2mm thick brass plate. This was cut to size with the Proxxon table saw so didn't need any finishing. It's held in place with 2 M2 capheads.
Due to various design changes the spigot on the flange was 3mm instead of 4mm. The was a suitable piece of aluminium with the right diameters in the odds and ends box so I just parted off a 1mm slice.
The pump element was hardened by heating to red for a few minutes and then quenching in water. When it was made the 1.96mm pin gauge would pass straight through. After hardening a 1.95mm pin gauge would not enter. The bore was lapped using Acro needle laps starting with 5 micron diamond paste, finishing from 1.96mm to 1.97mm with 1 micron paste.
The first needle lap is already worn after maybe 6 pumps. That's why the are sold in packs of 10 I guess.
-
Still crossing all my fingers etc. and praying that you succeed and may it not take too many more tries before you are happy with the result.
:popcorn: :cheers:
-
A little bit more on this one.
The delivery valve seating was lapped with a 3mm diamond charged ball as before and then I needed to cut the metering helix in the plunger.
This is another thing that tests my lathe's abilities. I have worked out a change gear set up that gives the required 4mm pitch. As the cutting is controlled by the leadscrew handle at low speed I can set the gears with zero clearance to reduce the backlash. The original Hobbymat milling angle plate was modified to allow me to set the Proxxon hand drill/grinder at the required 32°. The cut was made with a grey Proxxon disc at 20 000rpm with an infeed of 0.05mm and a slow drive from the leadscrew handle. The slide was withdrawn after each cut to further reduce the backlash effects.
As this set up takes some time I cut a second helix on the other end of the piece of 1.97mm pin gauge. After a light deburr with a fine diamond one the plunger passed smoothly through the pump element.
-
I fitted the new smaller helix pump onto my two cylinder engine, the details are here:
https://www.modelenginemaker.com/index.php/topic,7883.msg248312/topicseen.html
There were various problems with this design:
1) The O ring seal between the pump element and pump body requires close control of tolerances and careful smoothing of almost hidden edges to avoid damaging the O rings.
I conceived this design to allow for hardened pump elements and a modular system to make a multi cylinder pump. I’m now not so sure that this is an optimal solution, maybe a simple hardened sleeve shrunk/Loctited into an aluminium body might be better :thinking: Care would be needed to ensure that the inlet port in the sleeve lined up with the port in the body, easy with Loctite, harder with a shrink fit.
2) Fuel leaks past the plunger.
I accept that there will always be leakage past a metal pump plunger in a metal barrel. This could possibly be reduced by using some form of seal around the end of the plunger but I think that would be another level of technology with a 2mm diameter seal :headscratch:
The other option is to manage the leakage. With the current vertical layout the fuel leakage flows down onto the operating mechanism and washes the lubricant away resulting in wear and seizing. This would be less critical with diesel which has some lubricating properties. My first two injection systems have/had horizontal pumps so that leakage past the plunger does not get into the mechanism.
Maybe a new design with the pump horizontal would be better as it would allow for the fuel leakage to by led away from the mechanism.
3) Stroke/volume control and the inlet port.
My first pump designs were based on full sized diesel pumps where the inlet valve is a port in the side of the pump barrel. This removes the need for an actual inlet valve and any sealing problems. It also allows for a precise control of start of injection which is necessary for true diesel operation. For operation on petrol (gas, benzine) it creates a problem as the cylinder is under vacuum when the port opens causing the petrol to vaporise. This can be solved, as in full size practise, by pressurising the fuel above it’s vapor pressure ~0.8 bar/10psi.
Controlling the volume by controlling the pump stroke works and is quite simple but does not give constant timing, which is not a problem for manifold petrol injection. It is however very sensitive to wear in the operating mechanism. Even if the actual working stoke remains the same increased clearances in pivots etc. reduce the delivered volume. When the working stroke is around 0.2mm this is a problem.
Controlling the volume with a helix ground in the plunger is more complicated. It gives a constant start of injection point and seems fairly linear. This seemed to work ok with diesel, but with petrol the leakage past the plunger washed the lubricant away from the rack and gear make the movement stiff and jerky.
Next steps.
I think I need to split the development into two streams:
a) The helix controlled pump for diesel, continuing on the current path but using a simple hardened sleeve.
b) A horizontal pump with separate inlet and delivery valves for petrol. This will reduce the sensitivity to wear in the operating mechanism, hopefully remove the need for a separate pressure feed pump and allow separation of the leakage past the plunger.
The picture are taken from the two cylinder engine thread so the numbers are out of sequence.
-
I can think of one more option - an up-side-down pump - that way leaking fuel do not wash the moving parts .... Sad to say that is all I can come up with right now :-[
Per
-
Hi Per, I did look at that design, it does solve a number of problems but the killer is how to get the air out of an inverted plunger ::) The optimum seems to be a horizontal pump with vertical valves as was used on many of the hot bulb/semi diesel tractors. The example from Lanz uses a spring loaded cone valve and a gravity ball valve in tandem, no idea why :headscratch:
-
I looked at the horizontal petrol engine whilst it was running and the was no obvious leakage from the plunger. This engine has the first injection pump I made and quite a few running hours so I decided to dismantle and check the pump I made for the vertical engine, not good ::)
The plunger measured at around 1.965 so about 5 microns undersize. A 1.98 mm pin gauge would enter the barrel for 5 - 6 mm. With the 1.98 mm pin gauge in the barrel there was no detectable leakage and the fuel feed pump held pressure for 10s of seconds. A 1.99 mm pin gauge would enter the barrel for 1 - 2 mm.
I evidently got something wrong during the making of the plunger and barrel and will probably have to remake both :( maybe I can save the pump element but the 1.965 plunger is not useable.
The leaking fuel washing the lubricant away fron the excentric and plunger caused significant wear in less than 5 mins running. I haven't measured it but I think a new excentric strap will be required :toilet_claw:
https://youtube.com/shorts/wEaZnRayJUs?feature=share
-
Hi Roger,
I glanced back in this and the previous thread but didn't see a drawing of the general arrangement of the pump, or injector. Do you have one you can point me to? I'm not familiar with this type of pump but would like to know more.
Regards
Mike
-
Hi Roger,
Sorry for the problems :( But isn't that always the way it goes? :-\
But that's why we love the hobby, right?
Your video doesn't seem to work for me. Is it public?
Thanks,
Kim
-
Thank you both :)
YouTube seems to have uploaded this as a video short (????) and I can't see it on MEM. I have tried to upload it again:
https://www.youtube.com/watch?v=qhWlly2ebUA
https://www.youtube.com/watch?v=qhWlly2ebUA
Mike, This is a bit made up as I go along, my drawings are electronic back of an envelope sketches. Here are a couple of exploded views, I will try to do a couple of decent layouts. The pump is based on a full sized jerk pump with helix metering. The injector is based on a GM pintle design.
-
Yup, I can see the video in your latest post!
Thanks Roger,
Kim
-
:ThumbsUp:
Here are layouts of the pump and injector. Hopefully they will make sense together with the exploded parts.
-
Thanks Roger,
That was enough information to enable me to give myself a youtube education in jerk pump with helix metering. Looking forward to your progress with these miniature examples.
Mike
-
I lapped the element out to 1.98 mm and ground two new helixes in a 1.98 mm pin gauge. These were then cut to length and the non helix end was annealed then threaded M2. This was all much as before except I used an ER11 chuck rather than an ER25 to get better access.
The pump was rebuilt, fitted to the engine, primed and tried.
https://www.youtube.com/watch?v=I96e8rIJkbs
It ran quite well with no obvious leakage from the pump, but a seep from the injector that looks to be due to a crack in the soldered joint at the union.
Now the lubricant is not washed away from the rack mechanism it tends to move on it's own and has to be held. Progress in one way but I only have two hands so cannot measure RPM etc.
Next up is to repair or replace the injection pipe. I have always hard soldered the unions on the diesel versions. Then I need to sort out a control mechanism for the rack.
-
Good job getting that tough nut to crack, Roger! :cartwheel: :wine1:
-
Thank you Ron :)
This fuel injection thread is a bit of a lonely path ::) I know some people have got it to work but there are no real details available. If I can even list what doesn't work and why I will hopefully add to the knowledge base. The goal is to give a way to build your own injection system :)
-
It's' a runnah Roger!!! That is a definite success!! I've seen others runninn, but none of them were running at that speed!!!
Good job Man!!!!!! :ThumbsUp: :ThumbsUp: :ThumbsUp: :ThumbsUp: :jumpingsmileys:
Dave
-
Hi Roger, It might seem like a lonely path, but I am sure that I am not the only one that looks forward to and reads every post. I like your detailed record of the process with the descriptively titled pictures.
I am in awe of the accuracy your are achieving, and your patient approach to developing solutions to the problems one by one.
Great to see the current stage running so well in that video.
MJM460
-
I really appreciate your detailed explanations - that is a heck of a good result. Superb running engine!
-
Engine runs great , You have succeeded :ThumbsUp: :ThumbsUp: :ThumbsUp: :ThumbsUp:.
Congratulations !
Thanks for sharing !!!
-
I can atest to reading everyone of your post about this subject - hoping for you to succed and me to be a bit viser on this field :ThumbsUp:
Agree with Dave - nice fast run - even and steady :praise2:
Per
-
Thank you all for your interest and support :) :)
I am happy in a strange way that the problems were due to my manufacturing errors rather than fundamental problems with the design ::)
I am moving this pump back to the two cylinder engine thread to work out a control linkage.
The updated design for petrol injection will remain on here. I have produced a couple of layouts for a horizontal and vertical version. They are based on an aluminium body with the hardened and lapped pump cylinder and delivery valve seating (red) Loctited in place. The pump tappet housing/ return stroke limit is made as a separate piece (blue with yellow bronze bush). There is a drain hole for any fuel leakage (green). I am not sure if the inlet valve will need a spring, obviously it will for the vertical version, and the union/seating may need to be something harder than brass. This has a conventional style camshaft rather than an excentric and stroke adjustment is via a wedge with a single pivot point. This will not be as linear as the earlier versions with a floating wedge, but equally the wedge can't keep popping out.
-
Hi Roger 1
I think, with the vertical version, inlet valve: you need a spring in the ball valve - but the pressure is light, it just holds the ball valve in the closed position.
For the horizontal version: no springs needed - you can make springs if you like
-
Thank you Thanh :)
My thoughts are much the same. I am not sure if I can avoid a spring in the horizontal version due to vibration :thinking: The current 360° twin is not that well balanced and I guess it is reaching around 8 000 rpm.
-
As mentioned in the four stoke diesel thread I need to do some further work to improve the atomisation and spray angle of the injectors. I will keep to the current design but will experiment further with the needle angle and finish and with the springs.
The needles are turned from 2mm silver steel and like last time I will make two blanks, one on each end of a length of stock. The ends are centre drilled with a 0.5mm drill, turned down to 1.4mm and threaded M1.4. The shank is then supported with a brass bush while the section before the cone is turned down to 1mm. The cone is then turned, in this case 10° included angle, and parted off to give the required 1.6mm diameter at the end.
-
The two needles have different lengths to allow some experiments with springs. I may have to make a slightly longer body as well. The next stage was to mill some feed slots in the guide are of the needle. In the first experiments I thought I could allow the needle to float but trials showed that this did not work. The most successful injector of that series was where I had drilled the spring seat 2.4mm instead of 2.5 mm so it centred the needle. The area of the slots is a little low for the fuel injection time and volume but seem to work.
The cones were then hardened and polished with a fine abrasive stone. I am hoping to do away with the grinding and polishing operations and just move onto lapping. Last time it didn't work but I had also moved from 1 Micron to 5 Micron lapping paste (don't change two things at the same time ::)
Next up I made some M1.4 nuts from 2mm brass hexagon. Luckily the swarf gnomes didn't get any :)
-
It must be very hard to keep Deflection to a minimum @ those tiny sizes :thinking:
(don't change two things at the same time ::)
Always a prudent advise (that we always keep) if you need to know what is the reason you get a certain result :ThumbsUp:
Per
-
Hi Roger - I've thoroughly enjoyed reading about your methods and results in developing this injection system.
Viz. fuel leakage past the plunger (clearance to bore) - I understand that Bosch introduced (for racing Porsche's using PED6KL series pumps), a tiny oil passage to the side of the bore to create an "oil seal" due to high relative viscosity of the oil vs fuel in the tiny clearance between the plunger and bore. Also addressing lubrication of the plunger beyond the intrinsic lubricity of the fuel.
You can see it as a small diagonal passage in this sectioned plunger body assembly
Hope this info is useful to you
John
-
That probably also means you don't need as tight tolerances! (Or rather, it can handle a bit of wear over time), such a simple and elegant idea
-
Thank you John, an interesting idea :ThumbsUp: I had thought about grinding a couple of oil retention grooves in the plunger as a next step. At the moment I am happy if it works at all before I start looking at extending the life span. Do you know how big the Bosch plunger is? I would guess around 6mm.
-
The next step was to make some more nozzles. These were turned from 10mm stainless steel. 8mm Hex would possibly be better but I couldn't find any in sensible lengths and it doesn't take long to mill the hexagon at the end.
The bore is carefully reamed 1.4mm as the quality of the seat depends on the bore being as round as possible. The nozzles and needles are then identified with punch marks and fine grooves to keep them in pairs after lapping. They are lapped with 1 micron diamond paste thinned with light oil using the little Proxxon lathe and then cleaned up in the ultrasonic bath.
https://www.youtube.com/watch?v=p7-P_VkiGbw
https://www.youtube.com/watch?v=p7-P_VkiGbw
-
Hi Roger - the plungers for normally aspirated Porsches were 6mm and up to 7mm for 917 turbo.
Your success in running an engine at this scale is extrordinary in my opinion !!
Watching with great interest - thank you for posting your work for the world to see.
Regards
John
-
Thank you John :) A bit more on this.
I made a couple more injector bodies from 6mm brass hex. Now I have the tailstock die holder it can all be done on the little Proxxon lathe. The injectors I have made to this design have been very inconsistent ::) There is one good one which will run on the diesel and several that won't give consistent atomisation. Going back to the previous generation without the guided needle the one that worked best had the spring retained in a 2.4mm hole. As the bore of the spring was a close fit on the 1.4mm needle this acted as a guide. Now I have the guided needles a 'double' guide may be causing problems. I bored the spring seats out to 2.5 mm without obvious success :headscratch: The springs are quite a close fit on the needles so maybe that is causing some restrictions :thinking: I have ordered some new springs with a 1.5mm bore 2.5mm OD, neither easy to find nor cheap ::)
I also needed to modify my test pump for higher pressures so I drilled a second hole twice the distance from the pivot. This will allow 5kg/50N on the 2mm plunger.
Finally an attempt to show my injector test method and a more spectacular way of checking the atomisation.
https://www.youtube.com/watch?v=KhNYT6AAsCI
https://www.youtube.com/watch?v=PcREpDLM64Q
-
Switching from model engines to model flamethrowers?
-
That does look a little dangerous, Roger! :D
Though I'm quite sure you keep flammable things far clear of your testing grounds!
Kim
-
Hi Roger
...... and several that won't give consistent atomisation. Going back to the previous generation without the guided needle the one that worked best had the spring retained in a 2.4mm hole. As the bore of the spring was a close fit on the 1.4mm needle this acted as a guide. Now I have the guided needles a 'double' guide may be causing problems. I bored the spring seats out to 2.5 mm without obvious success :headscratch: The springs are quite a close fit on the needles so maybe that is causing some restrictions .......
Rather than clutter up your thread, I've added a link here that you might find useful ...https://pbase.com/9146gt/the9146gt_28rsr_engine1_bosch_rsr_fuel_injectors_floating_needle (https://pbase.com/9146gt/the9146gt_28rsr_engine1_bosch_rsr_fuel_injectors_floating_needle)
These PETROL injectors open at 15-18 bar.
Regards
John
-
Thank you all :)
The flame thrower test isn’t that risky. I move flammables out of range, the total fuel volume burnt is a couple of mm3, I have a damp towel to hand and if all else fails there is a fire extinguisher by the door.
My current injector design was loosely based on the Lucas petrol injector which is similar to the Bosch petrol one. This was fine for manifold injection spark ignition engines. I found that nozzle guidance was better for higher pressures which I based on the GM poppet diesel injector.
Manifold injection for a spark ignition engine is much easier as there is nearly a full crankshaft revolution for the fuel to vaporise before it has to ignite. Cylinder injection for compression ignition has only a few crankshaft degrees to atomise and ignite.
I also carried out some output trials on the diesel injection system, spinning the engine at various speeds with different electric drill and measuring the output from the injector with a measuring cylinder. The effect of leakage past the plunger is quite noticeable. Testing that injector with the hand pump suggest an opening pressure of around 150 bar.
-
Hello Roger,
Those conclusions make interesting reading, especially the lower time dependence of manifold injection v direct cylinder injection. Obvious perhaps, when you think about it, but most of use don't think that far.
Oh! next time you do that exciting 'flammenwerfer' test/ experiment, why not try a dark as well as the light coloured background. It may make the flame front, even more visible and exciting to watch.
Just a thought
Mike
-
Oh! next time you do that exciting 'flammenwerfer' test/ experiment, why not try a dark as well as the light coloured background. It may make the flame front, even more visible and exciting to watch.
Or better yet, turn the lights off!
-
Hi Roger - I wasn't aware of the Lucas petrol injection system - thanks for enlightening me. And I agree completely re evaporation of the petrol in the inlet tract vs atomisation requirements for injection into the cylinder for diesel.I really posted that link as it gives a pretty good visual around component proportions (spring and pintle needle) for reference
I really do applaud your work on this - great stuff !!
John
-
ou ou ou! 12 cylinder manifold injection? 8)
-
It's been done, just lots of little bits :)
-
By golly, does that thing actually run!?
-
By golly, does that thing actually run!?
Word has it, it did and it sold at auction for $40K...
Though honestly....it's a poor representation of a 917 engine.....
-
This is the only information I have on it:
https://rmsothebys.com/en/auctions/NY15/New-York---Driven-By-Disruption/lots/r1029/182098
Apparently it sold for $1000 per cc.
-
As I am still waiting for the new injector springs here is an aside about another obscure petrol injection system.
I found the Simms MkII petrol injection system in a German Language Technical Book from 1966. Internet searches found a couple of references in Commercial Motor:
https://archive.commercialmotor.com/article/21st-october-1960/40/petrol-injection-by-simms
Petrol Injection By Simms
21st October 1960, Page 40
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
DETAILS of a petrol injection system which is simple and relatively cheap were announced on Wednesday by Simms Motor Units, Ltd. Whereas previous equipment of this type has been reserved for expensive high-performance cars because of its complexity and price, the Simms system, when supplied in large quantities, is to be offered to manufacturers at some £7 more than the cost of conventional carburettors. This, it could be applied equally to cars or petrol engined commercial vehicles.
It consists of a cam-operated plunger type pump driven from the camshaft gearing and a low-pressure injector which delivers fuel into the inlet manifold. Throttle pedal position provides quantitive control of the fuel injected, and output is modified for engine speed by a centrifugal governor and for load by a diaphragm actuated by manifold depression. A thermostatic control incorporated in the cooling jacket of the engine provides a form of automatic choke for cold starting.
The one pump will suffice for engines of any size, fuel delivery being altered by modification of the pump cam profiles.
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
https://archive.commercialmotor.com/article/29th-october-1965/40/simms-to-make-and-market-petrol-injection-system
The only picture and technical description are from the German book. I have translated it here:
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Petrol injection from Simms
This British pump is built for small engines and correspondingly small injection quantities (5 mm2) and for very high speeds. With this new design, a major step has certainly been taken towards the further introduction of Petrol injection in small engines, especially with regard to the lower production costs compared to a conventional injection system.
The fuel is injected into the intake manifold. Only one pump element and only one injection nozzle is required for all cylinders. Injection takes place in time with the suction strokes of the individual cylinders. In contrast to a conventional injection pump with pistons and cylinders, the Simms pump uses a kind of rubber ball, which is compressed by a ram and relaxes again through its own elasticity. Both a spring-loaded intake valve and a check valve protrude into this rubber ball. The quantity regulation of the fuel brought in by a feed pump is achieved by longer or shorter opening of the intake valve, The ram, which presses on the rubber ball, is lifted over a roller by a camshaft, The cam elevations are only a few tenths of a millimetre. The fuel metering according to speed and load is carried out according to the vacuum generated by a throttle valve in the intake duct. Facilities for cold start and fuel shutoff are planned.
Also interesting is the associated injection nozzle, which differs from the others in that the path of the spring-loaded nozzle needle is limited by a soft ring made of synthetic rubber. The interaction of rubber disc and spring means that a very good atomization of the fuel at the nozzle opening is achieved at large and small flow rates, regardless of the prevailing injection pressure.
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
I guess that the rubber ball avoids any problems with sealing the plunger and has fairly low inertia which will be better for high speed operation. I can’t yet work out exactly what the rubber disc/ring in the injector does.
-
Hello Roger,
Simms fuel injection systems are still manufactured with lots of stuff on the internet. There may be stuff about the Simms MK2 petrol injection system, with the squezzy rubber balls, but I have not had time to discover it yet.
Another high pressure pump with a squeezy rubber pump element is the Harben high pressure pump. They employ a small rubber tube for the pump. The rubber tube is squeezed by a hydraulic cylinder. Have a look at https://www.harbencalifornia.com/pumps/p-pump/ There is a description and animation in the bottom right hand corner.
I have often thought that a rubber tube diaphragm pump, similar to the Harben, could make the basis for a simpler petrol injection pump; without the need for all those precision fits. You would need to devise some variable throw into a petrol injection pump
More stuff for you to think about, if you think it has any merit.
Cheers :cheers:
Mike
-
This thread along with my own engine build have got me started reading up on mechanical FI more than I ever have before, starting from a near vacuum. The kind of info offered by you, Roger, and Vixen in the last couple of posts above are great and fall right into my hands!
I've looked at what I can find on the net about the SPICA system used by Alfa Romeo (not much), and the more well known Lucas system (quite a lot more). The Lucas "shuttle metering" looked well suited to model engine use and home shop building until I realized it relied on high fuel pressure (100 psi). Nope, not doing that.
Anyway, Thank You both for sparking and feeding my curiosity. Whether I will ever take on actually building an FI unit is an unknown at this time, but I no longer dismiss it out of hand!
-
Hello Mike,
The second Commercial Motor article suggests that the Simms system was derived from the Marvel-Schebler petrol-injection system about which I can find even less. The Simms write up suggests that the pump stroke was a few tenths of a mm for a full size engine. I'm not sure how it could be reduced for a model sized system. The hardness of the rubber would also be critical to getting consistent results.
Hello Ron,
SPICA, Alfa Romeo, Kugelfisher, Peugeot and Bosch, Mercedes Benz and Porsche were all derived from standard diesel pumps with different control systems. They all offered a fairly direct control of the injection quantity using a helix spill system as long as the fuel was kept above it's vapour pressure. The injection pressure was generated in the pump plunger. The Lucas system used the pump shuttle for metering but the injection pressure was generated by the feed pump which meant there was a lot more pipework with petrol under moderate pressure. The metering shuttle should be fairly easy to make but the distribution ports would be more difficult. It also will only work with an even number of cylinders.
-
Hello
The "few tenths" is probably because the size of the rubber ball is large.
What if we considered this fuel injection similar to the low pressure mechanical fuel pump that feeds a carby, that uses a lever pressing on a rubber diaphragm to achieve pumping action?
Would be easier to scale that design down? And a wedge can be used to vary the amount of stroke for the ram that would press against the rubber.
Edit: gee whiz I thought I read through the last posts correctly, I see Vixen has proposed a similar design as well.
In all cases, the basis of these pumps is to remove high tolerance fits from the equation. Actually I now have an appreciation for the mechanical pump in my Toyota - for me the rubber diaphragm represents an obvious failure point, but so long as the rubber is OK it won't ever leak unlike a piston setup. Only problem is I probably can't buy new diaphragms for it.
-
Hi Roger
Also interesting is the associated injection nozzle, which differs from the others in that the path of the spring-loaded nozzle needle is limited by a soft ring made of synthetic rubber. The interaction of rubber disc and spring means that a very good atomization of the fuel at the nozzle opening is achieved at large and small flow rates, regardless of the prevailing injection pressure.
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
...edit....
I can’t yet work out exactly what the rubber disc/ring in the injector does.
Roger - my guess is that the needle is limited in its lift ("path" above) under varying pressure and flow rate, much akin to the "lift stop" of the cutaway poppet nozzle of your post #46 above
Regards
John
-
I don't know anything about fuel injectors, other than what's in the name, but I am reading this Roger, and greatly enjoyed your 2-stroke diesel thread. :ThumbsUp: Took me a few days to read through it all, but definitely worth it! :popcorn: :cheers:
-
Thank you :) I have had a lot of fun with all this development stuff ::) :wine1: Sometimes I feel a bit like the odd kid playing by himself in the corner so it is good to know others are enjoying it as well.
-
Roger I admire your persistence!
:cheers:
Don
-
Thank you Don :)
Following some discussions I have decided to try and make a needle type injector for the four stroke diesel. This requires moving the fuel inlet a long way up the body to clear the head with some complicated deep drilling and silver soldering. I decided to try a few of the difficult stages first. The body will require a 20mm deep 1mm diameter hole which will then have to be plugged. The first 6mm depth was drilled 1.2mm to allow the silver solder to flow round the plug. The rest drilled 1mm with small pecks and frequent cleaning of the drill. That worked :) I then started to drill the 1mm crosshole that will feed the needle. This can't be finished until the main bore is drilled otherwise it may deflect the bore. Next I turned a 1mm mild steel plug from some 5mm black bar, the smallest I had. This was supported on a piece of 1mm piano wire in the crosshole (I hoped the silver solder wouldn't wet the piano wire if it got that far) and silver soldered in place.
-
OK - so now you take us on the next leg of the Journey :ThumbsUp:
If not to the destination - then at least closer to it :cheers:
Best wishes and a Happy New Year :cheers:
Per
-
That's some pretty fine turning! :popcorn:
-
Nice work getting a 1mm drill down 20mm Roger !!
-
:popcorn: :popcorn: :popcorn: :ThumbsUp:
Getting it done with what you got! Love it!
Dave
-
That's some amazingly fine (in both senses of the word) work there, Roger! :popcorn:
Kim
-
Thank you all for your interest :ThumbsUp: :ThumbsUp:
I know that the Proxxon machines don't tend to get good comments but 4800 rpm for drilling the deep 1mm holes and 2800 rpm for turning the 1mm plug are useful ( I could have swapped the belt pulleys and gone to 4000 or 5000 rpm for the final cuts but it didn't seem necessary).
The next steps were to drill and tap the ends of the body. The bore was made as before, drilled 1.3mm and then reamed 1.45mm. It appears that the silver solder flowed down the plug.
The next step was to tap the first section of the cross bore M1.4 and screw a grub screw 2mm deep. The silver soldering didn't seem to flow well but hopefully it will hold the plug in place.
-
The grub screw was filed flat and the end of the body was turned down to 8mm. The plug seemed to stay in place :)
The next piece was the fuel inlet which will have to be set a 45° to clear the head studs and be drilled through to the passage in the body. The 1mm bore was only partially drilled to align the final hole through.
-
The fuel inlet mating face was milled 45° and a 1mm hole for a locating pin was drilled. This hole was then spotted through to the body and the rest of the 1mm plug was used to hold the parts in place for silver soldering. With the inlet in place I then continued drilling the 1mm bore until it broke through into the 1mm bore in the body. Luckily the drill didn't break and the passages are well enough aligned that I can blow though from the inlet to the bore for the needle :)
-
Nice fabrication job on that Roger. I particularly like the most elegant vise clamp straps. Tastefully done!
I have found that small diameters don't need to have uber high rpm but they do need feed control That said, the Proxxons seem to be performing well!
Dave
-
Always a pleasure to watch you work, Roger. :wine1:
-
Thank you both :)
The fixing flange was cut from 2mm thick steel and silver soldered in place. It all seemed to fit so the next step was to mill scallops either side of the body to allow my box spanner to reach the fixing nuts. This design is full of compromises, I am trying to use parts from my first injectors which I know work in a confined space and I reduced the separation between the fixing studs from 16mm to 14mm to gain a bit more space in the cylinder head.
I used an 8mm round ended mill with 0.2mm depth of cut which seemed to work. The last details were finished with a file, often quicker than a milling setup. It fitted and i could tighten the clamping nuts :)
Next up is the nozzle. I know I can drill down to 0.2mm in brass, I need to see what I can do in stainless steel ::) The first trial will be 0.35mm
-
Interesting new design :ThumbsUp: and I see what you mean about lack of space / planning ahead :)
Per
-
That is just impressive work, Roger! :popcorn:
Kim
-
Thank you both :ThumbsUp:
The nozzle was turned to length from some 5mm stainless steel, with a 0.5 x 0.5 mm step to clear the partially threads in the body. This was threaded M5 x 0.5 and checked for fit in the body and then in the cylinder head. This was then spotted and drilled 1.35mm then reamed 1.45 mm ready to drill the nozzle hole. I had decided to start with 0.5mm to reduce the risk. It appeared to be ok, there was a little resistance and then it appeared to break through. The drill looked ok :)
The next step was to cut the seating with a D bit I made for the previous experiments. What I didn’t realise was that the tip of the 0.5mm drill had broken off and was still in the hole in the nozzle. This piece then jammed and wrecked the D bit :toilet_claw:
-
I made a few more unsuccessful experiments with 0.5mm and 0.35mm drills at various speeds up to 5000 rpm on the Proxxon lathe. Back to brass.
I repaired both nozzle drills, the bit are Loctited into the holders so a bit of heat releases them.
Repeating the procedure on a brass nozzle resulted in a problem free 0.35mm hole :)
Now I need to make new D bit. This starts as a length of 1.5mm silver steel with a 90° point. The needle then has a 60° point to give close to line contact. Due to historic parts of the design I now need to lap it down to 1.45mm before filling the flat.
-
More fine and precise and teensy work. :praise2: :ThumbsUp:
-
Bummer about the broken bits. :(
Glad you were able to recover. Sorry the stainless didn't work out either! But now you've learned. What kind of Stainless are you using? 303?
Kim
-
I like your die holder lap! :ThumbsUp: :popcorn: :cheers:
-
Thank you all :) :) The material was a piece from the scrap bin, probably similar to 303. I am not sure if the European materials are a direct equivalent :headscratch:
When the blank was a good fit in the nozzle I used the previously made filling fixture the make a flat to just under half the diameter. The D bit was then hardened and polished in the polishing fixture with a fine diamond file.
I then used the D bit to finish the seating in the nozzle. The next stage was to cut a small step to clear the unfinished thread in the body and thread it M5 x 0.5.
It all fitted together but unfortunately I didn’t allow for the step and cut one thread too many ::) I don’t think it will affect the sealing of the injector to the cylinder head.
-
Bummer about the broken Drill-Bit -> D-Bit :hammerbash:
Smart Jig for filing and polishing the D-Bit :ThumbsUp:
Is it hardned like a filing button ?
Per :cheers:
-
I like the D-bit jigs :ThumbsUp: :popcorn: :cheers:
-
Thank you both :)
The jigs are made from 7.98mm diameter silver steel (nominal 8mm), the one for filling was milled down to 4.10mm and the one for polishing to 4.03mm so that the flat on the D bit will not quite reach the centre. Whilst still in the vice both were drilled and tapped for an M2 grubscrew. They were both then hardened.
-
Some impressively fine work, Roger! I still can't get over 0.35mm drills, it seems ridiculous that they actually work. :o
-
Thank you :ThumbsUp:
Next up was the needle. A 60° cone was turned on the end of a length of 1.5mm silver steel and the next 15mm was turned down to 1mm in 5mm steps. I have previously done this using a tool post grinder but I wanted to see if it could be simply turned.
The injector body was then lapped with 1 micron diamond paste until a 1.48mm pin gauge just passed through. The needle was then lapped to fit the body and finally the needle tip was lapped into the nozzle.
-
I'm impressed that you were able to turn such a small a diameter with that much unsupported length hanging out. :praise2: :wine1:
-
Thank you :) A sharp insert, high speed (2800rpm) and shallow cuts (0.05mm).
I assembled it and tested it without the needle spring. The 0.35mm hole gave a nice jet but no atomisation.
https://www.youtube.com/watch?v=JAZUb5ams2E
When I applied pressure to the needle there was a significant leak from the nozzle fixing thread. I decide to reduce the size of the anulus on the nozzle to increase the clamping pressure and to recut the seating in the body. This worked but I then had a leak from the fuel inlet union. After some inspection and thought I realised I had cut the seat 90° not 60° :toilet_claw: This was easily fixed.
The final step was to cut the needle to length and assemble it with the spring and spring cap from an existing injector.
The spray pattern is somewhat offset, probably due to a lack of concentricity between the nozzle and the needle. The atomisation looks good and the opening pressure is adjustable in the correct range.
https://www.youtube.com/watch?v=hv5i8zO_tx8
https://www.youtube.com/watch?v=pQytbUvBtBE
Finally everything that had been in contact with the diamond paste went in the ultrasonic bath
-
Incredibly small and meticulous work, Roger! I find it just amazing! :popcorn:
Kim
-
Roger - that is looking great !
Whilst the spray pattern may not be as concentric as you'd like, there appears to be good atomisation of the fuel stream.
And the video of the 0.35mm nozzle stream looks viable for non diesel manifold injection as well
Bravo !
Regards
John
-
Thank you both :)
I fitted it to the engine for a trial. It runs but there is still a lot of un atomised fuel coming out of the exhaust. You can also see some leakage past the needle coming out of the centre of the spring adjuster.
https://www.youtube.com/watch?v=NncOlT-Vfxw
This design has some problems for manufacture. The blind passages are the main one, followed by the sealing and alignment surface between the nozzle and the nut. I have been playing around with various designs and have come up with a version based on full size designs. The fuel passages are open at the end and the sealing faces are exposed and can be easily lapped, rather then one being buried. The nozzle and the end of the body will be 7mm diameter to keep the parts aligned and the clamping thread will be M8x1.
-
That is sweet Roger. I admire the effort here. Very few have accomplished what you have done and are doing! Well done sir! :NotWorthy:
-
Looks like a good plan. If development was easy... :ThumbsUp: :wine1:
-
Great to see it running, Roger, still following. Looking forward to your latest design, and hoping it will do exactly what you want it to! :popcorn: :popcorn: :cheers:
-
Sound like a happy Runner :ThumbsUp: and good to see it run again :cheers:
I like the latest draft - but I have one small sugestion, to help with leakage .... (just thinking out loud).
Make the Blue Collar slightly longer so it goes up over the solid above the thread + remove the top end of the thread with a grooving tool and put a small O-Ring there.
Perhaps a bit premature - as you haven't disclosed where the leak is yet :embarassed:
Per :cheers: :popcorn:
-
Thank you all for your interest and support :)
Per, I have attached a similar drawing of the current injector. The centring and sealing of the nozzle are via an M5 x 0.5 thread and a narrow ring as in my delivery valves. The new version moves the centring to a closely machined 7mm diameter on the nozzle and the body. The sealing is then by a 1mm wide ring between the nozzle and the body.
Looking at this design there was limited space for M8 x 1 threads so I purchased an M8 x 0.5 set.
The nozzle is made from 8mm brass rod. I started in the Hobbymat as the headstock bore of the Proxxon is only 7mm. I have kept the 4 jaw SC chuck on for making the body. The flange was turned to 6.99mm and the nozzle bore was completed much as before. I decided to go for a 0.2mm orifice this time.
-
The final stage was to cut the fuel recess with a 5mm end mill. There was a little chatter at the start but this didn't affect anything important.
-
Wauw - great finish down in this small size :praise2:
So this absolute latest, looks like you fairly easy can change the Nozzle - if you want to test different Orifice versions ...?
Per
-
Very nice! You make some REALLY tiny holes, Roger! No small feat there! :popcorn:
Kim
-
OK, I like metric and I've happily worked in metric at times, but I don't think in metric. So 0.2/25.4 = 0.0078"! Wow. Not only is that a tiny drill to use, but I have to wonder just how they make those in the first place!
-
I have a small book called "Making Tiny Drills" where it's done with a grinding wheel on a watchmaker's lathe. But not a twist drill! Amazing work Roger! :ThumbsUp: :ThumbsUp: :popcorn: :popcorn: :cheers:
-
Roger - 0.2mm is a pretty small hole !! well done.
Does the nozzle hole, cleaned up with the D bit, fit the shank of the drill to hold it concentric for drilling the final 0.2mm dia hole ? Or is there some other witchcraft going on ? Or did I just miss the explanation ?
Needless to say, I'm very impressed.
John
-
amazing dedication, very few are able to make a working injection system for model engine...that works !
as you are drilling such a tiny hole (0.2mm !) from the opposite end of an already drilled part, the concentricity has to be the highest...
-
Thank you all :) This design is planned to allow me to experiment with different seating and needle angles as well as different nozzle sizes. The drill holder helps to guide the nozzle drill however the actual centring comes from the D bit.
I have thought some more about the manufacturing sequence for the body. In the last version the silver soldering was carried out near the end which caused scale on some of the machines surfaces. This time the minimum will be machined before fixing the fuel inlet and the clamping flange. This required a threaded stub mandrel to hold the body as the fuel inlet interferes with the four jaw chuck.
This time I remembered to cut the seating for the fuel inlet union with a 60° centre drill rather than a 45° countersink.
-
More fine watchmaking skills on display.
Mike
-
Thank you Mike :)
After a quick check the fuel inlet was parted off. The body was then mounted on the mandrel to be turned down to 8 mm for the clamping flange. The clamping flange was made with drilling and filling from some 2mm bright bar. Next the fuel inlet was set up to be milled at 45° and to drill the 1 mm locating pin hole.
-
The locating pin hole was then spotted through to the body. The locating pin is another piece of the 1mm mild steel I turned for the last version. It looked a bit sad after silver soldering :( There was good flow around the fuel inlet but poor penetration around the flange. I used the same set up as before so I don't know quite why. It was well fluxed.
It will work so I will keep going, I still have several opportunities to scrap the body ::) The body was cleaned up and mounted back on the mandrel to turn and thread for the fixing nut and then to drill and ream for the needle.
-
The intricate work just never ends, does it Roger! Looking good! I don't see anything wrong with that solder joint :ThumbsUp: :popcorn:
Kim
-
Thank you :) I will remake the body but will keep using it for the moment.
The next steps were to drill the two intersecting 1mm holes for the fuel feed. Unfortunately they didn't meet and I ended up breaking through into the main bore ::) I then moved on to the clamping nut using a 5mm end mill as a boring tool for the seating which meant I couldn't see what was going on ::) It fitted together but the nozzle didn't seat properly.
-
Sorry it didn't quite work out .... All that nice work :wallbang: - or can you save it ?
I'm a bit perplexed - why the seperate feed bore - I mean, the other channel ends up in the main bore anyway (as far as I can see) :headscratch:
Per :cheers:
-
:(
Kim
-
Thank you both :) The problems were not unexpected, If you look at the coloured picture of this design breaking through into the main bore has to problems, firstly it reduces the sealing length from around 20mm to around 6mm and secondly it gives a trapped volume that is difficult to bleed the air out of.
Looking at the pieces more closely there is a small ridge in the clamping nut that I removed using a 3mm end mill as a boring tool and the fuel inlet is not square, confirming that I will need to make a new body. I'm not sure what was different with the silver soldering but the flow on the two versions is totally different :thinking: :headscratch:
-
Onwards with the new body much as before. I decided only to silver solder the fuel inlet in place, the fixing flange will be screwed onto the same thread as the nozzle clamping nut. The silver soldering looks much better this time, like the first attempt. I now know what the difference is. I bought two sizes of silver solder, 1mm and 1.5mm, both are the same material. The first time and this time I used the 1.5mm for the last one I used the 1mm so maybe they are not actually the same ::)
The various drillings went ok and meet up where they are meant to meet without breaking through anywhere else.
-
Good to see that you had more success this time :cheers:
Per
-
The first time and this time I used the 1.5mm for the last one I used the 1mm so maybe they are not actually the same ::)
Better keep that 1.5mm solder somewhere separate and safe, away from the other. And labeled, "Use Me"! Glad you're back on course. :wine1:
-
Looks really good Roger. :ThumbsUp: :popcorn: :popcorn: :cheers:
-
Thank you all :)
The end of the body was threaded M8 x 0.5 for the clamping plate and nut. I them milled some flats on the nut to aid fitting and removal as well as making a threaded clamping plate. This means there is only one silver soldered joint on the injector and it can be made before I drill the small bores so they won't be affected by scaling.
-
Continuing with the needle injector. Clearance slots for the clamping nuts were cut with an 8 mm ball end mill. This was easier than on the mark 1 as the clamping plate can be unscrewed. The bore was then lapped to a good fit on a 1.48mm pin gauge. The needle was then started from a piece of 1.5 mm silver steel. The tip was turned to 60°. Is a 60° cone in a 90° seating sensible? I am hoping for a very narrow seat so time will tell. The end of the needle was then turned down to 1mm to allow fuel flow into the nozzle. The tip and the first part of the shank were hardened and then lapped to 1.48 mm to be a close fit in the body. The needle tip and the nozzle seating were then lapped together. After that it all went in the ultrasonic bath to remove the residue of the diamond paste.
-
The needle was then parted off from the rest of the rod, which makes further lapping of the seat more difficult. I would need to make some sort of extension holder.
When the injector was assembled for the first trials the needle showed a tendency to stick open. Careful investigation showed the end of the needle was around 5 microns oversize. This was lapped down to size which seems to have resolved the problem :)
I had also received the new springs for the mushroom/poppet style injectors which should allow the needles to seat freely. Time for some more Rorschach inkblot tests.
This is the new injector with the spring holding the needle shut, looks possible.
https://www.youtube.com/watch?v=DzFSroSf4nc
This is with the spring removed and just the 0.2mm hole. Not enough atomisation.
https://www.youtube.com/watch?v=6Ds0Bn7oF7c
The mushroom injector with the new spring. Looks good.
https://www.youtube.com/watch?v=bCXCwJt1Ci4
Now I need to find time for some running trials.
-
My guess is that the Injector in the first video is the best, followed by the last - and that the middle has a too concentrated cone ....
Test runs might prove me very wrong ;D
Looking forward to see the Test Results :ThumbsUp: :cheers:
Per
-
Such fascinating, detailed work, Roger. I love the careful, methodical way you go about building and testing your injectors!
I learn a lot from following your posts!
Kim
-
Hi Roger - I echo Kim's comments above.
Taking such time to develop your ideas, methods and to share your results is greatly appreciated
John
-
Thank you all for your comments and support :ThumbsUp: :ThumbsUp:
A couple of the bodies had 2.4mm spring seats which I bored out to 2.5mm using a 2mm endmill as a boring bar.
I had previously made two nominal 10° needles. One works quite well and the other tends to jam. 10° is quite close to a self-locking taper so I needed to check what I actually had. A bit of experimenting with my protractor allowed reasonable measurements. The one that jammed was around 8°, the one the worked and the one from the last running trials were 10°.
I decided to try a 12° needle as well. I made a pair as before and selected the one that was the best fit in the body. The initial test had a lot of big droplets but this improved with some more lapping.
10° needle:
https://www.youtube.com/watch?v=FDyp44PP63o
12° needle first test:
https://www.youtube.com/watch?v=SB3FpXEC6t8
12° needle after more lapping:
https://www.youtube.com/watch?v=DFYyD5ss6KE
Time for some running trials :)
-
I had some interesting but not very successful trials ::)
The new needle injector started and ran immediately but was quite inconsistent. None of the 4 mushroom/pintle injectors ran but one produced some thick black clag from the exhaust rather than the usual unburnt fuel :headscratch:
I then went back to the needle injector. It ran, but not stably, as before then there was a loud pop and the engine stopped and would not restart. The exhaust pushrod had dropped out due to an excessive clearance :toilet_claw:
As there was no adjustment left on the screw I called it a day and started to pack up. When I pulled the fuel pipe off the injection pump to drain the fuel back into the container the pipe had become unsoldered :( Maybe this was causing an air leak and some of the inconsistent running :thinking:
More work to do.
-
The lab tests continue! :popcorn:
Sorry that they weren't as successful as you'd hoped, Roger. Hopefully resoldering the pipe fixes the issue and you can continue with your fascinating learning process.
Kim
-
Thinking out Loud - I wonder if that very long and unsupported Copper Fuel-Line between Pump and Injector, vibrations + posible resonance => breakage @ the ends :noidea:
Some of your spray patterns are good (close to very good, if it wasn't for the bigger droplets present too).
Still hoping and wishing for your complete succes Roger :ThumbsUp: :cheers:
Per
-
Even a bad result is a good one! sounds like a "reset" and try it again .
Dave
-
Thank you all :)
I resoldered the fuel inlet pipe and then started to analyse the injectors.
The spray pattern from the mushroom injector that gave the thick black clag was tight with minimal big droplets.
The result from the needle injector was poor and I felt that the needle wasn’t sealing. When I took the nozzle off the needle was stuck, suggesting the engine was running (badly) with a 0.2mm open hole. Maybe a slightly smaller hole would run, avoiding most of the sealing problems. 0.15mm should be possible, maybe 0.1mm.
I felt that the needle was sticking/jamming due to the long (20mm) bore that the it runs in. This was reduced to 10mm by drilling first 10mm out to 2mm.
The pattern was still not improved so I will make a new needle.
-
I really wish that I had the knowledge to tell you what to do .... but if I did - I would have done it a long time ago.
I have only see the spray from an old 'Oil-Burner' (Oil Boiler according to Google) and from the 'Carburator' on a 'Velo Solex' (French Moped) - this is actually a low pressure Fuel-Injection System and not a Carburator.
In both cases a very fine 'Mist' (I couldn't see any droplets with a preteen eyevision) - coming out as a cone ....
I have seen a number of press releases from introductions of new Motorcycles (or upgraded versions of a model) - 'More holes of a smaler diameter than previous xxxx, results in a better atomization of the fuel -> more complete combustion -> more power and better fuel economy'.
In your case more holes are not needed - but a smaler diameter might help .... as you think about doing.
Per :ThumbsUp: :cheers:
-
Thank you Per :ThumbsUp: As you know I am just making it up as I go along ::)
I Turned up a new needle and lapped it to fit the bore and into the nozzle seating. The pattern looks somewhat better :)
I had another look at the first new generation needle injector and the bore has significantly drifted off centre so to be recycled :(
-
As the diesel appeared to run with the needle stuck open and just a 0.2mm orifice I decided to make a simple open injector with a non return valve.
The nozzle was turned from brass, which may have a limited life but will show if it will work or not. The non return valve is based on the injection pump delivery valve. Setting all the depths for drilling to gat a suitable length orifice is a little tricky but I seemed to end up with around 0.5mm.
-
Not much special about the body, the bore was reamed 2mm and the ball seating was lapped with a 3mm diamond ball lap.
The injection seems to be more of a squirt than an atomisation with the hand test pump:
https://www.youtube.com/watch?v=S9XOfphk89w
https://www.youtube.com/watch?v=fS5owr5DIJQ
A mushroom injector appears better:
https://www.youtube.com/watch?v=9D-_l4zTTv0
As does the latest needle injector:
https://www.youtube.com/watch?v=V5M0mXQIXQ8
Hopefully I can do some running trials in the next couple of weeks :)
-
I see that you turn to KISS - 'Keep It Simple Stupid' ....
.... and it has saved my bacon on a number of times, too ...! ;D
Come to think about it - your new aproch might work even better with an 'Indirect Injection System' .... if you got the room for it .... :noidea:
Per :cheers:
-
Not much special about the body, the bore was reamed 2mm and the ball seating was lapped with a 3mm diamond ball lap.
The injection seems to be more of a squirt than an atomisation with the hand test pump:
https://www.youtube.com/watch?v=S9XOfphk89w
https://www.youtube.com/watch?v=fS5owr5DIJQ
A mushroom injector appears better:
https://www.youtube.com/watch?v=9D-_l4zTTv0
As does the latest needle injector:
https://www.youtube.com/watch?v=V5M0mXQIXQ8
Hopefully I can do some running trials in the next couple of weeks :)
Hey Roger,
Just an observation, the nozzle in your test video is hanging at the end of a copper tube....and I'm noticing it's vibrating. That vibration is going to have a fairly significant effect on you paper spray pattern. If you're counting on that....just sayin....
Dave
-
Thank you both :ThumbsUp:
Per, I have thought about indirect injection, it makes the injection theoretically easier but increases the heat loss due to a larger surface area combustion chamber. The current four stroke diesel combustion chamber is about as small as it can be. There's a poster on HMEM i am following who is looking at building an indirect injection diesel.
Dave, I'm not too worried about the actual pattern at the moment, just the atomisation. Previously while I was trying to work out how to keep everything concentric it was important and someone (possibly Per?) pointed out the vibration of the pipe. The spray from the needle and mushroom injectors lights easily, the jet from the open nozzle just makes my lighter oily. With a much faster pressure rise in an engine turning at 2000 rpm rather that me pushing the handle on the test pump things may be different :thinking: :headscratch: :wine1:
-
I can only agree on the heat loss - was just thinking out loud ....
Best wishes :cheers:
Per
-
I had a quick trial with these injectors today. The workshop I use has been reorganised so the area I can use has a parquet floor rather than concrete so my 'workmate' is rather more lively.
A cold start with the needle injector was as good as last time, supporting the airlock theory.
https://www.youtube.com/watch?v=nHYDfQaNw98
The open injector would not fire properly at all, there was some light blue smoke and a few pops but nothing useful. Most of the fuel came straight out of the exhaust :(
https://www.youtube.com/watch?v=f6O0ms1kg7c
As a confirmation I refitted the needle injector which performed as before.
https://www.youtube.com/watch?v=nynKOU7mE0E
If you look carefully you can see the fuel leaking past the needle coming out of the hole in the spring adjusting screw. Maybe I need to add a return pipe.
-
Easy starting with the old model Injectors :ThumbsUp:
If you look carefully you can see the fuel leaking past the needle coming out of the hole in the spring adjusting screw. Maybe I need to add a return pipe.
Considering that this engine isn't using much fuel (or most likely not doing so) - it is certainly a percentage lost that way. It appears to contain some polutants too - so if you decide to do so, run it through a filter ...
I guess that you already have a seal where it leaks .... :noidea:
Per :cheers:
-
I have been discussing fuel injection with another member on here and started thinking about my last diesel trials.
The engine ran with the injection needle jammed open so I assumed that the fuel was being atomised via the 0.2mm orifice. I made an open nozzle with a 0.2mm orifice and tried that without any success. I now think that the needle jammed partially open and the atomisation was due to the small opening, but how small?
I decided to try and measure the needle lift for both the needle and mushroom injectors. The test pump was set up and primed with lamp oil (like diesel but not so smelly) and I set up a DTI to attempt to measure the lift. There was nothing measurable :thinking: Any movement of the indicator needle seemed due to the lack of rigidity of the simple set up. This suggests that the lift is just a few microns and the fuel is atomised by this gap. Lots more to think about :headscratch:
https://www.youtube.com/watch?v=4ylfLu_BJJY
https://www.youtube.com/watch?v=x84gqNyPM6w
-
Hello Roger,
Could you instead make a "micrometer" style setup where you can gradually decrease the gap to the point where it starts running?
-
Hello Brendon,
Technically yes, but I would run out of hands holding the drill, adjusting the fuel rack and adjusting the injector. Maybe it would be easier when I fit a governor the look after the fuel rack :thinking: I also suspect the gap may be speed dependant so adjusting the spring may be more practical :headscratch:
As I have said many times I am making this up as I go along, all thoughts and ideas are most welcome :) :wine1: