20 cc Horizontal 2 Stroke Diesel

[Roger B]

Thank you both 

Eric/Cletus, I initially tried the engine on petrol with a different cylinder head with a carb and spark ignition. I know that the crankcase compression and porting were ok for that although after some advice from Per (Admiral DK) I modified the exhaust pipe. I am moving into somewhat uncharted waters here. I have seen pictures and videos of working engines but the constructors give very little detail. Reviewing Find Hansens website it appears that his injector, of the mushroom type, is much smaller than mine being based on a 1mm shaft instead of 2mm. I could make one that small but I hope I don't have to  M1 threads are a bit scarey (as is the cost of the taps and dies)

Carl, don't worry I don't understand it all either 

[fumopuc]

Hi Roger, I am still watching quietly. A great experiment you have started there.

[Roger B]

After it's quarantine period in the garage I bought the engine back into the cellar and removed the cylinder head. Unfortunately there are no real clues, just a film of diesel oil over everything  Maybe I need to try some more settings for the start of injection  Once I can get some form of combustion it is easy to see what makes improvements but until then it is difficult to know what to change next. The joys of experimentation   

[Roger B]

As a next step I bought a small bottle of Diethyl Ether from the local chemist and tried a few drops in the air intake. This certainly produced combustion and a cloud of smoke but it wouldn't sustain. It was also very 'kicky' due to the ether. Luckily my drill is heavy enough and powerful enough to absorb the shocks. I still tend to think that the problem is poor atomisation so I will order some 0.2mm drills and make a smaller nozzle as well as some stronger springs for the mushroom injector.

<a href="https://www.youtube.com/watch?v=WGJdHk97Pnc" target="_blank">http://www.youtube.com/watch?v=WGJdHk97Pnc</a>

[Roger B]

Another attempt  I tried a stronger spring for the injector which gives a theoretical opening pressure of around 120 Bar. This is still using a 0.5 mm nozzle, I haven't lapped the 0.2 mm nozzle yet.  With 0.2-0.3 ml of ether dropped onto the reed valve there were short runs   

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

Getting there 

[AlexS]

Nice progression! 120 bar theoretical pressure looks like a lot! 

May I ask how much heavier the injector spring is now, compared to the old one? In terms of spring constant (N/m). And suspended length of the spring (pre load)/ surface of opening?

I do not know if you've been involved, but I'm thinking.
It would be nice to measure the 'practical' injection pressure. But I think that it would be difficult (or expensive), seen that the pressure increases / decreases rapidly. A simple reading of an arduino + printing or something similar would not be accurate enough.
But how about to measure the amount of injected fuel of a multiply test cycle?

The amount of fuel injected is partly dependent on pressure. Pressure, opening size, and injection time. Of course opening size and pressure changed during the injection.
Or is it an idea to measure the height of the injected fuel? Based on Bernoulli's law, an idea could be given about the injection pressure. Link: https://www.natuurkunde.nl/vraagbaak/7724 (Dutch site).
The same principle is used to calculate, for example, the maximum of a water pump (Hydraulic power” means an expression of the arithmetic product of the flow (Q), Head (H)) . Of course there is a somewhat more constant flow of liquid here.

This ideas could give an idea of what the average injection pressure is. More as a method to 'measure' the differences of injector adjustments.
But I understand that you want to improve the to evaporate of the injection. It is more of an idea.

Alex

[Roger B]

Hi Alex,

The previous spring had a rate of 2.35 N/mm and was compressed to 6 mm giving approx. 14N. The new spring has a rate of 6.7 N/mm and is compressed to about 5mm giving approx. 30N. The needle is 1.5mm diameter and I am ignoring the nozzle diameter to calculate the opening pressure. I am expecting to inject less than 2mm3 per stroke at full output.

[AlexS]

Thank you for the information Roger

[Roger B]

The previous trial runs resulted in a fountain of dirty oil from the exhaust so I decided to drill a 1.5mm drain hole in the bottom of the exhaust bend.

Having lapped the 0.2mm nozzle I fitted it for another trial. Several problems  The flywheel had worked loose in the previous trials but luckily just tightening the grubs screws on the taperloc was enough. I may need to increase the screw size from M3 to M4. The second problem was fuel leak from the injector union which I couldn't fix. I need to study the design again (in the fuel injection thread). The runs I did get were no better than before, but the fuel leak may have been responsible. The drain in the exhaust did work 

[Roger B]

I had another attempt today having fixed the injector union but with no significant improvement. I adjusted the timing in stages from so retarded that nothing happened to so advanced that the kicking shook the starter dog loose in the drill chuck but the best results were where I had been trying.
Trying the normal spring, stronger spring and no spring in the injector generally made things worse. Finally I got no smoke at all and found that the cam follower for the injection pump had fallen off  More thinking required 

[Roger B]

After some thought I decided to make a new higher compression head, reducing the length of the chamber from 10mm to 7mm. As this is to be an experiment I won't be making any water cooling passages. If it works I can shorten the original head.

The blank was made from a rather stringy piece of 50mm diameter aluminium out of the bits box. It was drilled through 5mm, the combustion chamber was opened out to 10mm and then finished off with a 10mm end mill to the correct depth. The cone was turned using my rather dubious spare compound slide. I set out the fixing holes using the small Proxxon RT and drilled and tapped the injector clamping holes M3. The 4.2mm dia. fixing holes were drilled with the bench drill. The Proxxon is not much good above 3mm.

[Roger B]

I then bought the engine in from quarantine in the garage and removed the existing head. There has certainly been some combustion. With the new head fitted I was getting a compression pressure of over 30 Bar rather than 25 Bar with the original head. Maybe it will be enough 
Finally I refitted the injector cam follower that fell off last time ready for the next trials. Outside temperature 1°C so not today.

[Admiral_dk]

We can only hope for more success this time 

    

[Art K]

Roger,
Hopefully you will have a warm day soon and can try it out.
Art

[AlexS]

Good to see you on this project.  Hopefully this gives an improvement!

Is it an idea to insert a glow plug of an rc engine into the cylinder head? To pre-heat cylinder or to initiate ignition? I do not know if this would fit in your cylinder head.

https://www.mecoa.biz/shopdisplayproducts.asp?id=81

Alex