Model Diesel: 32mm bore, 38mm stroke, indirect injection

[ruben]

Here is the plan of the piston and combustion chamber. Can you read .DWG files?

Hola de qué material es el pistónHola de qué material es el pistón

 

[ruben]

Hola de qué material es el pistónHola de qué material es el pistón

Wat material are made the pistón

 

[Nerd1000]

Wat material are made the pistón

Cast iron in Roger's engine. Mine, aluminium.

 

[ruben]

Hierro fundido en el motor de Roger. El mío, aluminio

el tuyo de aluminio

 

[Roger B]

Heh, would have been nice to have more direct evidence on that front. I've seen some papers where they painted the piston with something fuel soluble and after running a few cycles inspect to see where the paint washed off. Might be one way to attack that question.

I have this picture series from a pre war copy of the 'Modern Diesel'. A quick internet search suggests that the first picture is under vacuum, the middle picture is at atmospheric pressure and the final picture is around 200psi (~13 bar).

 

[Nerd1000]

I have this picture series from a pre war copy of the 'Modern Diesel'. A quick internet search suggests that the first picture is under vacuum, the middle picture is at atmospheric pressure and the final picture is around 200psi (~13 bar).

4000 psi is a fair bit more injection pressure than we are running, I think. At least I'm aiming for 100 bar, about 1500 psi. I'm fairly sure that I've read that the lower pressure reduces 'penetration' of the spray. So presumably our sprays don't go so far. Nevertheless, what we see in that picture makes me think that the spray probably doesn't have much difficulty getting across the small sizes of our combustion chambers and impinging on the piston or (in my case) the swirl chamber wall.

 

[Steamchick]

I don't know much about spray droplets, size, pressures, etc. But I'm sure someone does?
The curious thing I think I remember is the surface tension - is constant - except has a radius part of the function, so a smaller radius is a higher tension...? Then there is Surface energy: Never got my head around that... except I think there is more surface energy "per surface area" with a small radius than a larger one....
So those things (combined?) suggest to me that for smaller droplets, we need higher "pump energy = pressure - to get some shear conditions to create Aerosols instead of Droplets.
Then there is heat: A smaller droplet needs less heat to vaporise than a larger one, and has a higher ratio of surface area to mass, so smaller seems better than larger in fuel droplets, so the droplets vaporise quicker and the gas can then burn as more heat ionises it.
I guess that droplets are formed when the liquid is expelled from the nozzle/jet and the pressure gradient shears the jet of fluid? - I.E. It doesn't take any distance from the jet to convert the liquid at pressure at the jet into droplets without pressure but with velocity? - In which case, I can appreciate how the "spray" does not wash the piston, even though distances are very small.
There must be some science in it that I don't understand, so any advice is of interest. - I just "know nowt about owt" on this subject!
Thanks,
K2

 

[ajoeiam]

4000 psi is a fair bit more injection pressure than we are running, I think. At least I'm aiming for 100 bar, about 1500 psi. I'm fairly sure that I've read that the lower pressure reduces 'penetration' of the spray. So presumably our sprays don't go so far. Nevertheless, what we see in that picture makes me think that the spray probably doesn't have much difficulty getting across the small sizes of our combustion chambers and impinging on the piston or (in my case) the swirl chamber wall.

IIRC some of the modern diesels are running in the 300 bar range for injection pressures.
HTH

 

[Nerd1000]

Last couple of days I made up some mounting brackets, fuel fittings, crush washers and so forth. Timed the injection pump. Broke the high pressure line to the injector and had to braze on a new one. The engine is now complete enough to begin testing for fuel injection system issues, I'm planning to get it running on manual control of the fuel helix before I attempt to build the centrifugal governor. I had an idea to correct my issues with the driver dog without making a new one: I filed the cup point off the setscrew and placed a small soft copper slug in the hole beneath it, then tightened it against the round portion of the shaft. Drive is by friction only now, and it hasn't slipped at all. I guess the copper crushes and conforms to the shaft surface to give a really nice grip. And so, with the sump filled with 5W-30 it was ready for the first test run...

The engine fired and ran for a few revolutions on starter fluid, spraying burnt oil and smoke all over the workbench. It also spat the fuel in the yellow supply line back out with a big bubble of gas, so I know that my injector spring is too weak to hold the needle down and prevent gas pushing backwards into the injection system. The needle has also become much stiffer in the injector since I re-brazed, further investigation needed there.

Further testing was prevented by a blown head gasket. I just can't get the copper head gasket to seal consistently (perhaps my head surface finish is too rough?) so I'm thinking I will try a composite gasket material.

 

[ninefinger]

Is it possible you had hydraulic lock or near hydraulic lock causing the head gasket to fail that quickly? If your injector got stuck open it would not take long to put too much liquid in the cylinder and get too much pressure. Any room left on the head for an adjustable pressure relief valve? It may be worth doing for a development engine where overpressure is a real possibility. Could save the head gasket, or the crank or con rod if you get the head to hold too much pressure. (Find Hansen did one in one of his videos...)

 

[Nerd1000]

Is it possible you had hydraulic lock or near hydraulic lock causing the head gasket to fail that quickly? If your injector got stuck open it would not take long to put too much liquid in the cylinder and get too much pressure. Any room left on the head for an adjustable pressure relief valve? It may be worth doing for a development engine where overpressure is a real possibility. Could save the head gasket, or the crank or con rod if you get the head to hold too much pressure. (Find Hansen did one in one of his videos...)

I did consider this, but the head is pretty cramped, the only space available is between the inlet and exhaust ports which is very narrow.

I'm confident there was no hydrolock, in fact when I removed the head afterwards it was fairly dry with only a thin film of fuel/oil. The injector not sealing meant that fuel wasn't getting delivered regardless, the whole high pressure side of the system got filled with pressurised exhaust the first time the engine fired...

To tell the truth the gasket wasn't sealing properly to begin with. I applied a thin layer of permatex gasket maker to both sides which resulted in a seal, but it would appear the heat and pressure was too much and it blew out the sealant. Not a shock in retrospect, that sealant worked well for sealing the sump etc but clearly no good for head gaskets. When I pulled the head I found that the sealant at the edges of the combustion chamber had been burnt.

 

[minh-thanh]

Hi @Nerd1000
About the gasket
This is what I did with my 4 cylinder diesel engine, it was perfect
Because my engine's cylinder head and water jacket are aluminum, I had to test the hardness of the aluminum gasket against the aluminum cylinder head, gasket needs to be softer.
If the materials of that part on your engine are different: it depends
If you use teflon instead of paper then all thickness is the same, it will be fine .

 

[Nerd1000]

Hi @Nerd1000
About the gasket
This is what I did with my 4 cylinder diesel engine, it was perfect
Because my engine's cylinder head and water jacket are aluminum, I had to test the hardness of the aluminum gasket against the aluminum cylinder head, gasket needs to be softer.
If the materials of that part on your engine are different: it depends
If you use teflon instead of paper then all thickness is the same, it will be fine .

View attachment 157721

Thanks, I like this design. I ordered some fiber reinforced graphite material from eBay already, do you think it would need the metal fire rings? I also thought I might try forming fire rings from shim stock and wrapping them around the soft gasket material.

 

[minh-thanh]

do you think it would need the metal fire rings? I also thought I might try forming fire rings from shim stock and wrapping them around the soft gasket material.

"do you think it would need the metal fire rings?"
Hi
Personally I would do it with a high compression, high temperature engine
I have tried many different ways and failed - maybe I did it wrong
Maybe you will succeed in your own way
Hard to know

Edit :
About the 0.3mm number: just close. (0.23 or... 0.25 or 0.27 )
It's hard to find exactly two different materials just to achieve that 0.3 mm number

 

[Roger B]

Did you aneal the copper head gasket? How thick is it? Can you measure the compression pressure?

 

[Nerd1000]

Did you aneal the copper head gasket? How thick is it? Can you measure the compression pressure?

Yes, 0.6mm, not without making a dummy injector to adapt my compression tester to the cylinder head (I also need to check if the gauge goes high enough).

The compression is certainly high enough to ignite starter fluid (it ran on it) and burn oil (I can smell it even cranking without fuel).

 

[Roger B]

Ok, I was wondering if the compression might be too high.

Next question how flat are the sealing faces of the head and the block? If they are concave you won't get it to seal.

 

[Nerd1000]

Ok, I was wondering if the compression might be too high.

Next question how flat are the sealing faces of the head and the block? If they are concave you won't get it to seal.

As flat as I could get them given how well my mill was machined at the factory (the spindle tram cannot be changed without scraping).

The surface finish is not ideal because I fed a bit fast when milling. There's scope to lap or scrape them in (and with scraping I could probably get a seal with no gasket at all) but I'd rather avoid going to such lengths haha.

 

[Roger B]

Laping or scraping may be a bit extreme, but some fine abrasive paper/cloth fixed to a flat surface may be sufficient. This also has the advantage that if you are not so good with the figure of eight action the surface will tend towards convex which will help the sealing.
I guess you would have to remove the valves to work on the head and drain the oil to invert the cylinder block.

 

[minh-thanh]

Laping or scraping may be a bit extreme, but some fine abrasive paper/cloth fixed to a flat surface may be sufficient. This also has the advantage that if you are not so good with the figure of eight action the surface will tend towards convex which will help the sealing.

Hi Roger !
I read most of your comments, almost exactly what I did as well as the results....
I really appreciate your comments
!