Small Injector Pumps

[johninsarnia]

My main hobby is making small engines and having made several of the spark ignition variety I am trying to move on to hot bulb and hopefully full diesel types. The main challenge is to make the injector pump and the injector itself to handle minute quantities of fuel and in the case of a diesel engine to develop a high enough pressure to finely atomize the fuel. For the size of engine I am interested in (~30cc) a pump with a bore of 1/16" (~1.5mm) and a stroke of somewhat less than this is about the right size.
Looking at full size practice injector pumps rely on having the piston and cylinder made to close tolerances to minimize leakage back past the piston. With the pressures involved (certainly in a diesel engine) it is not possible to use any type of elasomeric seal. I have experimented with using O ring seals around the piston but they do not seem to last very long.
When you read about manufacturing diesel injector pumps they talk about very tight tolerances on diameters and cylindricities down to hundreds of a thousandth of an inch and all the special (hence expensive) equipment needed - which is beyond the scope of any amateur.
In the model engineering world there is someone in Denmark (Find Hansen) who has succeeded in making small diesel engines that work and he has lots of videos on YouTube showing these. In some he discusses construction techniques but does not get into how he makes the pump. Looking at the videos it appears that he just relied on a close fit and there is no significant leakage when the engine is run. I have no idea how he achieves this and he refuses to pass on his knowledge of how it is done.
I am wondering if anyone out there has had any experience in this area? Taking a simple approach, if you just use a reamer to produce a 1/16" bore, find a piece of 1/6" rod that is a tight fit initially then run the two together using metal polish until the fit seems right, leakage is still excessive so lack of cylindricity is presumably the problem. I have a 1/16" lap on order and my next move is to try to produce a more accurate bore then work on getting a piston to fit properly, probably using an external lap.
Any thoughts would be appreciated.

 

[PeterT]

Not me. I'm just an interested lurker because it is obviously a challenge. I assume you are aware of HMEM & MEM where this discussion is ongoing & some are 'approaching' success.
Are you the OP of this thread?

Request - Injector designs

Hello, I was curious and hopeful that those folk who have successfully built their own diesel injectors, might share their designs and words of wisdom in this thread. There is a bunch of fantastic work spread across the site and for those of us dreaming of compression ignition, it would be...

www.homemodelenginemachinist.com

I have no idea how he achieves this and he refuses to pass on his knowledge of how it is done.

I've seen his videos. While I'm a big fan of his general knowledge sharing & model engineering accomplishments, especially considering modest equipment, I just cant get behind his tight hole mode for the injectors. Hey, its his prerogative. He obviously went through some effort to make it 'work'. Maybe the secret sauce is in his will, or maybe he will take it to the great shop in the sky. Personally, I find it a bit off-putting. There are many model engineers who have arguably accomplished just as much or more in other capacities & share their techniques. For that reason I respect them more. Because even knowing how they did it doesn't mean you can replicate it. That's the highest form of compliment. And many of these same individuals go out of their way to mention they stand on the shoulders of others before them when learning the craft. That just resonates for me personally. I totally understand masking items with commercial or rights implications, but I have no basis to speculate this pertains. We may never know & that's just the way it is. But my guess is - probably like a (really) good magicians trick. Cards really don't vanish into thin air, but your attention is drawn elsewhere during the demo. Maybe there is something 'else' that he is not talking about. That would also explain a general reluctance to elaborate.

 

[DPittman]

Well I am a long way in my machinist capabilities to offer any real advice to someone that is clearly light years ahead of me however, I have rebuilt a diesel injector pump of a very simple and proven design that might be an achievable and repeatable thing for your application. The Cav/Lukas distributor style injection pump is what I'm suggesting. You may already be familiar with them but if not check them out. I can find a link for them if needed.

 

[Bandit]

I have not watched any of his videos. Having done a very small amount of work on diesel injectors for pre electronic engines quite some time ago. I am in no way an expert. These were Cummins injectors, the Detroit injectors are very similar. A supply pump pushs fuel to the injectors, the injector makes the high pressure for injection into the cylinder. Detroit supply pump pressure around 30 psi, Cummins varied the supply pressure. The camshaft pushed the plunger in the injector to make the high pressure. Camshaft timed for injection to inject fuel relative to piston position.
Anyway the metal was "very" hard and very polished, clearances "very" tight.
Other systems used a injection pump, some individual pumps on a block driven by engine, some, one pump using a distributor system to supply high pressure fuel to the injectors. In all cases the materials are/were very hard and very polished and tight clearance that pumped the fuel.
The one set of Cummins injectors I worked on were final polished using a wood dowel, by hand, these were special for this job! Had 2 injectors that had plungers not quite free moving enough, some time to get that rite! Injectors were flowed on a machine for the purpose, pumped a certain amount of fuel in a certain number of pulses. Yes, measured in cc's/pulses/time.
Not sure how I would tackle this problem for a very small engine. Very early diesels did not use an injection pump for fuel delivery, they also ran at lower compression ratios. I believe reading that some of them used coal dust for fuel, and some used peanut oil. Later ones did use injection systems but operated at lower injection pressures and also at lower compression ratios which would be able to use the lower fuel injection pressures. Some diesels engines required preheating engine head/heads before they would start! Once running, would maintain heat to keep running.
I think I would look into a 10 to 12 compression engine with a slightly more volatile fuel. A injection pump at maybe 600 to 1000 psi.
The lower pressures would allow a bit more clearances in the pump and possibly the use of hi-speed steel material for the injection pump.
I guess it could also depend on if the engine was a straight compression ignited engine, or an assisted compression ignited engine.
Please accept my apologys for a long story, with perhaps no info that usable.

 

[Bandit]

Typed while@DPitman was posting.

 

[PeterT]

Roger B on the MEM forum has been making great progress using what I consider quite modest machines & tooling. I don't follow the posts & progress it as closely as I should, but I'm quite sure he has a running proof of concept engine. Now onto more tweaks & improvements.

Fuel Injection Systems Part 2

Fuel Injection Systems Part 2

www.modelenginemaker.com

 

[johninsarnia]

I posted the item while having coffee this morning and by tea time this afternoon there have been several very interesting responses. Firstly; I have looked at some of the discussion on HMEM but some time ago so I will get back to seeing what is new on that forum.
Just addressing what is done full size: They do have the advantage that volumes are normally a minimum of 10 x larger which makes things somewhat easier. On the other hand they need to design for thousands of hours operation which means parts have to be much harder than if one is designing for tens of hours. One other aspect is that there is quite a lot out there about tolerances, hardness etc. but I have never seen anything about the machinery used in the manufacture of pumps. (I assume very small, specially designed, precision cylindrical grinders?) Does anyone have any information on this?
Concerning the comments about Find Hansen I too have wondered whether about whether everything is real. Aside from making the tiny injector and pump he talks about having completed an engine during one winter. He must work extremely diligently to achieve this. The main parts OK but all those fiddly bits with a flyball governor operating a wedge to control pump stroke? Cutting bevel and skew gears just using a lathe must also be quite a challenge.
Having said this it now seems that the Chinese are now making a small diesel engine looking very much like the ones made by Find.

There is also one of Find critiquing and improving one of these engines so it does look as if the Chinese have managed to set up some form of production facility to make miniature injection equipment.

 

[PeterT]

Yes, I saw that video too. The Chinese tabletop engines draw some comments both positive & negative. In some cases they look 'remarkably similar' to well known home shop designs with no credit to the designer, so there's that. Others designs are unique & even innovative in some respects. Some ME's feel these are bad or the hobby because they are 'assembly' vs scratch built. The flip side is very few are fortunate to have a $ machine shop & construct an engine, so what's wrong with shelling out cash for a model engine that runs. I'm cool with this. Nobody is pretending to have built it. It might even spark interest to get into the hobby. But from what I've read & seen, some of the construction methods & quality control & post-sale assistance is apparently an issue, so buyer beware.

 

[Mike R]

Hi John,

I have been studying this for about a year now and I'm actually close to making chips on a test pump and injector that is close in size to what you describe.
1.5mm (0.059") bore with up to several mm stroke (stroke is a variable).

I have chosen a few options to try and make it, we'll see how I succeed but I'm hesitant to start as I have too many other unfinished projects ahead of this one. I will say that this is the one project that has me most engaged right now and I keep telling myself that is just a small project....

I purchased pin gauges in 1.50mm as a starting point for the plunger. I also bought some reamers in 1.48mm and 1.49mm, and finally a needle lap to hopefully bring the bore to final size. I actually need to look up what drill size I should be using with the reamers...

I think the pumps used by Find Hansen and others get by with a small amount of leakage, the leakage is compensated for by the governor so in the end it doesn't matter too much if there's a bit of leakage, as long as the pressure is built up sufficiently to have good injection & atomization. Also, look at the model engines this way - the bar for them is that they operate, where as the bar for a full size engine is to make good power, not pollute, be efficient and not leak everywhere. So we can take a few manufacturing shortcuts on the models we build.

I need to also be ready to make an injector, as without it the pump is a bit useless and I won't know how effective the pump is. The injector will be very much a close copy of what Find Hansen has designed (mushroom type).

 

[Susquatch]

Hi John, welcome to CHMW.

Please consider introducing yourself, your interests, and your machines in the new member introductions sub forum.

If you are who I think you are, you already know that I spent a chunk of my career designing full size engines for mass production. I can tell you that designing diesel pumps and injectors is no easy task. I can certainly appreciate how difficult it would be at small model engine scales. I've never done it, nor have I even thought about it.

On the other hand, the challenges have to be quite different. Large vs small, tight emissions control vs none, 20 year life vs 20 hours, hot & cold weather vs STP, wide fuel quality vs a standard test fuel, max power/torque vs good enough, performance vs good enough, vibration not an issue, etc etc.

I guess that what I'm trying to say is that you have some room to be less than perfect yet plenty satisfactory even if the scale itself is a very major problem.

Since you are asking about various designs, I assume that you are not locked into the same design concept as the original engine that you are modelling. Although I'm again not familiar with them, it might be worth looking at what one of the many diesel lawn tractors use. While not as small as yours will be, they are still a LOT smaller than what a truck or tractor uses. I'm thinking Deere Yanmar or Kubota engines. Chatham Injection may even have a broken or damaged pump that they would let you look at or even keep. The Kubota engine in my Utility gator is this little tiny toy diesel engine. Apparently, they make a little 276cc engine. That's still 10x what you are trying to do.

Regardless, starting with something that is already quite small vs a big truck or car engine might be a good way to figure out what might work in your scenario.

 

[johninsarnia]

Hi John,

I have been studying this for about a year now and I'm actually close to making chips on a test pump and injector that is close in size to what you describe.
1.5mm (0.059") bore with up to several mm stroke (stroke is a variable).

I have chosen a few options to try and make it, we'll see how I succeed but I'm hesitant to start as I have too many other unfinished projects ahead of this one. I will say that this is the one project that has me most engaged right now and I keep telling myself that is just a small project....

I purchased pin gauges in 1.50mm as a starting point for the plunger. I also bought some reamers in 1.48mm and 1.49mm, and finally a needle lap to hopefully bring the bore to final size. I actually need to look up what drill size I should be using with the reamers...

I think the pumps used by Find Hansen and others get by with a small amount of leakage, the leakage is compensated for by the governor so in the end it doesn't matter too much if there's a bit of leakage, as long as the pressure is built up sufficiently to have good injection & atomization. Also, look at the model engines this way - the bar for them is that they operate, where as the bar for a full size engine is to make good power, not pollute, be efficient and not leak everywhere. So we can take a few manufacturing shortcuts on the models we build.

I need to also be ready to make an injector, as without it the pump is a bit useless and I won't know how effective the pump is. The injector will be very much a close copy of what Find Hansen has designed (mushroom type).

Yes, in some respects what we are trying to do is less onerous than making diesel engines and injectors for the real world. On the other hand there is no way any of us in the hobby world can afford the equipment people use in the commercial manufacture of diesel injection systems - even the old tech type before common rail and solenoid injectors hit the scene.
Looking at machining techniques we are more familiar with, one can turn to maybe 0.0005" grind to less than 0.0001" but to make these pumps one is looking at 1 or 2 orders of magnitude better than this so that lapping to final size is really the only option. I have a 1/16" lap on order and plan to just try making a component with a bore that has good cylindricity, then try to make a piston that will fit, grinding initially then using an external lap (and a lot of patience.)
I did have a set-up earlier using a O ring seal around the piston. It worked OK initially but not for very long.

 

[johninsarnia]

Hi John,

I have been studying this for about a year now and I'm actually close to making chips on a test pump and injector that is close in size to what you describe.
1.5mm (0.059") bore with up to several mm stroke (stroke is a variable).

I have chosen a few options to try and make it, we'll see how I succeed but I'm hesitant to start as I have too many other unfinished projects ahead of this one. I will say that this is the one project that has me most engaged right now and I keep telling myself that is just a small project....

I purchased pin gauges in 1.50mm as a starting point for the plunger. I also bought some reamers in 1.48mm and 1.49mm, and finally a needle lap to hopefully bring the bore to final size. I actually need to look up what drill size I should be using with the reamers...

I think the pumps used by Find Hansen and others get by with a small amount of leakage, the leakage is compensated for by the governor so in the end it doesn't matter too much if there's a bit of leakage, as long as the pressure is built up sufficiently to have good injection & atomization. Also, look at the model engines this way - the bar for them is that they operate, where as the bar for a full size engine is to make good power, not pollute, be efficient and not leak everywhere. So we can take a few manufacturing shortcuts on the models we build.

I need to also be ready to make an injector, as without it the pump is a bit useless and I won't know how effective the pump is. The injector will be very much a close copy of what Find Hansen has designed (mushroom type).

Somewhere Find mentions 3mm for the pump bore and he also mentions 0.3 mm stroke with the engine running. It would seem better to use a smaller (1.5mm, 1/16") bore so the stroke is ~4x larger and easier to control. Because he says nothing about how he makes these pump parts I start to wonder whether he has found some commercially made items that can be modified for what he needs. He seems to put some emphasis on how simple his workshop is so you wonder how he can achieve the precision needed to make these pump parts from scratch.
One has to believe there must be some leakage but in the videos I have not observed any drips. I have been cranking the theory to estimate what clearances are tolerable, just to get some feel for what one should be aiming at.

 

[PeterT]

0.3mm is a tiny stroke.
What sort of maximum pressure do you figure the pump would see?

 

[Mike R]

At very high pressures even a tiny clearance can cause leakage, at much lower pressures the same clearance may be absolutely fine. Also, an elastomer may be fine for low pressure sealing but incapable of handling high pressure.
A pump does not need to leak externally to have leakage past the piston.
I'm planning on dealing with the leakage issue by having the high pressure leakage path route back into an area of lower pressure, i.e. think of it as 3 stages: high pressure fuel, low pressure fuel and no fuel. The back up plan is to add a gland seal at the top of the plunger shaft if leakage is an issue (i.e. seal between low pressure fuel and no fuel areas).

As for Find H. and others pumps, I think they just deal with leakage as a fact of life, as I said the governor automatically will increase the required stroke if the delivered fuel is insufficient (i.e. reduced speed), so leakage is not as big an issue in the pump as long as it can deliver the required fuel volume and pressure. The theoretical stroke may be 0.3mm but the actual operational stroke may be more to achieve the pressure and volume delivery.

For pressures, the pumps are probably generating peak pressures of ~1500psi (100bar). It will depend on the injector and its opening pressure setting.

 

[johninsarnia]

0.3mm is a tiny stroke.
What sort of maximum pressure do you figure the pump would see?

It is hard to know what sort of pressure his pumps develop. He does mention compression ratios of 20 or 22:1 for his diesel engines so the compression pressure must be at least 500psi and the injection pressure will have to be significantly greater than that. One sees pressures of 3000psi for old type diesel engines injection and nowadays they are 10x greater than that.