20 cc Horizontal 2 Stroke Diesel

[Roger B]

Thank you Brian 

I improved the test setup with a stand and a swinging arm for the follower bearing. The test weights are M16 nuts suspended on a bit of wire coat hanger. This seemed to operate at a bit under 1000rpm which is at the lower end of the expected (hoped for) speed range of the engine. I won't be able to develop this much more until I know what the running speed, required travel and force will be, in other words when and if the engine runs 

I need to look a bit more at possible linkages. I originally planned for the cone to move inwards towards the crankcase but it may be easier if it moves outwards There will still be space to mount a drive gear if it ever runs well enough to become part of a tractor 

<a href="https://www.youtube.com/watch?v=Xw76yqUuhrM" target="_blank">http://www.youtube.com/watch?v=Xw76yqUuhrM</a>

[Brian Rupnow]

Tell me more Roger, please. I see that you machined a 30 degree cone and a 45 degree cone. Why?--Did one work better than the other?---Brian

[Kim]

This is just fascinating, Roger. I'm with Brian - I've never seen a governor like this...
Kim

[Roger B]

Ok, as with everything I do on here I am learning as I go along so there may be errors 

The initial concept for this governor was taken from an illustration of an American Bosch distributor injection pump. I think that it works with one fixed part and one moving part, the balls being there own thrust race. I made a trial assembly for this with a 45° cone as that is what it looked to be in the picture. I made an initial trial just holding the fixed part in my fingers (there is a video clip of this). The force didn't seem to be much so I tried using 6mm balls instead of 5mm ones. This was better. I then thought that a shallower angle would increase the force so I made A 30° cone. I didn't feel much difference.
The problem with this design is that the balls are rotating at half shaft speed and as the force is proportional to the square of the speed the force is reduced to 1/4. Some quick experiments showed that allowing the whole assembly to rotate did increase the force so I made the current test piece based on the Junkers illustration.
Having made the test stand I will go back and make some quantative tests with the 30° and 45° fixed cones.

I hope that this helps but feel free to ask more questions.

[Roger B]

I have also attached below a PDF of the section of the drawing and the complete drawing of the engine so far (it makes a good 'off site' backup).

[Roger B]

I did some more testing of the different types of governor and learnt a few things 

For the fixed cone system there needs to be enough clearance when the cone is closed so the balls only have two point contact.
The hysteresis can be quite high. This is partially due to friction and partially due to the r term in the force equation. As the balls move out r and hence the force increases giving a positive feedback.
The drag on the fixed cone system is quite high. For the first tests I just let the screw rest against the support. This was partially why the first test with 6mm balls would not rise. For the later tests I restrained the screw with my finger. Some of the drag may be due to the surface finish as these parts were made as quick trials.
The difference in speeds between the fixed and rotating cones was a factor of 0.6, less than I would have thought.
The difference between 5mm and 6mm balls was only a factor of 0.9 whereas the difference in mass was a factor of 0.7.

I have summarised the test results in a spreadsheet below.

<a href="https://www.youtube.com/watch?v=rxU6Xi4FAEw" target="_blank">http://www.youtube.com/watch?v=rxU6Xi4FAEw</a>

[Brendon M]

Keep up the good work Roger, this is very interesting to watch

Also, thank you for taking the time to explain your rotary table improvements

[Roger B]

As this engine uses a lot of unions I tried Regner as an alternative supplier to Holzapfel who I have used up to now. The Regner unions have a smaller spanner size and offer nipples for 2mm and 3mm pipe. They are also cheaper if bought in 10s.

After an unsuccessful start attempt (in the fuel injection thread) I decided to confirm that the compression was what I though it was. The design is aiming at a 20-1 compression. I made an adaptor to fit in place of the injector with a ball valve right at the tip. This was tapped 1/8" BSP and I then had a fitting to go to 1/4" BSP although this has a large dead volume.
The first trial with a 1/8" BSP 10 Bar gauge went straight off scale. With a 16 bar gauge from an old pressure regulator I was seeing 18 - 20 Bar. I'm not sure how much the spring behind the ball valve will affect the reading  Obviously if it is too strong it will not open and I will not see any pressure at the gauge 

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

[Admiral_dk]

Hi Roger

Are you sure that the ball valve Works as it should ...?... I ask because my compression tester has to be reset to lower the reading.

Other than that I think you are doing great and will continue to follow 

[Roger B]

Hi Per,

I paid a bit more attention to the sealing of the joints on the 1/8" - 1/4" BSP adaptor and the leak off rate improved somewhat and I rapidly came against the end stop of the gauge  The next step I have is the 300 Bar gauge from the other side of the pressure regulator. This seemed to settle down at around 25 Bar which for a 20-1 compression ratio with some heat loss due to the small size of the engine is not too bad.

[Craig DeShong]

I'm following along Roger    I LOVE the governor 

[Roger B]

Thank you Craig    , I hope the engine runs so I get a chance to try the governor design 

I had another start attempt. I adjusted the injection timing, tried with the weaker injection spring that I use on the petrol engine, tried with no spring at all  The results were all much the same. As I raised the fuel rack I got increasing amounts of blue smoke for the first couple of millimetres. If I raised it further there was no more blue smoke but oil being ejected from the exhaust. I think that this means my injection pressure/penetration is too high and most of the fuel is ending up on the piston crown. The bench tests have all been just rocking the engine over TDC by hand which gives a fuel jet of 10-20 cm. If this is being cranked at ~1000 rpm with drill the jet may be much stronger. When the smell of diesel has reduced I will make some more trials with alcohol in the cellar. If the jet is too strong, I have several options: Increase the size of the hole in the nozzle (easy, a new nozzle is an evenings work but will need cleaning in the ultrasonic bath after lapping). Increase the injection period to reduce the pressure (not so easy, a new cam with curved flanks will require much more design that the current flat flanked one). Get the mushroom injector working (unknown).

[Roger B]

I had another try with the needle injector opened out to 0.5 mm and with the mushroom injector. The results were similar, blue smoke as I started to raise the fuel rack followed by a spray of unburnt diesel from the exhaust as I raised it more  When the smell has gone I think that I will remove the cylinder head and see if I can get any clues as to where the spray is going 

<a href="https://www.youtube.com/watch?v=GRmWe5r2c10" target="_blank">http://www.youtube.com/watch?v=GRmWe5r2c10</a>

[Tennessee Whiskey]

Roger, still quietly following along. If I may: back in the early days of true diesel fueled farm and construction equipment here in the states, it was quite common to start the engine on gas (petrol) and once running, the diesel fuel took over. Might be a way to judge things in the “running state”

Cletus

[zeeprogrammer]

I've also been watching.

I have to admit...a lot going on here that I don't understand...which is fine! Always learning.