Another try to build an engine running with diesel fuel

[Chicken]

Michael, my engine will be a hot bulb type engine with preheated hot bulb.

Progress: Finished the 1:3 gear with new governor shaft. the governor response is way better now.
Things left to do before 1st start attempt:

- fuel tank and fuel piping
- throttle lever which changes the tension of the governor spring
- spring loaded roller chain tensioner
- some lubricators
- hot bulb preheating system
- intake manifold

best regards,
Alex

 

[Chicken]

Hi!

Not too much to report over Christmas. But I've managed to finish the roller chain tensioner. It's not spring loaded, but easily adjustable using a screw.

best regards,
Alex

 

[Mechanicboy]

When start up the semidiesel engine, do not rotate the flywheel in direction of rotation. Bunch starting only, also prime the fuel one or two stroke of fuel pump and rotate the flywheel back until it will ignite and rotate in right direction.

I owned the Sabb semidiesel (hot bulb engine).

Diesel combustion engines can be divided into three main types:

Semidiesel engine with a glowing hot bulb. Compression ratio is 5-7 and fuel pump pressure ca. 80 kp / cm2

Semidiesel engine with a hot bulb. Compression ratio is 7-10 and the fuel pump pressure between 80 -150 kg / cm2.

Dieselengine: Without hot bulb. The appropriate compression ratio depends on the design of the cylinder head. The figure is usually between 14:1 and 16:1 for direct injection engines, and between 18:1 and 23:1 for indirect injection and fuel pump pressure between 105-650 kg / cm2 (Common rail diesel pressure up to 1500 kg/ cm2, to start up engine the fuel pressure is about 250 kg/ cm2)

 

[Mechanicboy]

To adjust amount of fuel and timing.

White smoke: Too late ignition timing (Blue smoke: too much engine oil)

Black smoke/ hard knocking: Too early ignition timing

Glowing hot bulb under load, in case loss of power: Too much fuel

Engine is running before it stopped and white smoke: 1: Not enough heat of the hot bulb, need more heat. 2: Loss of heat of hot bulb due the cylinder head is robbing the heat from hot bulb to example the cylinder head is wellcooled by coolingwater. The solvent: 1: Make long and thin wall body between hot bulb and cylinder head to prevent loss of heat in hot bulb. 2: Increase ratio of compression (If the engine is without knocking under max load and poor/impossible idling means too low ratio of compression, a very weak of knocking under max load and idling is nice = ratio of compression is ok)

The injector in semidiesel is designed to regulate by operator: The atomized fuel is injected into the hotbulb under idling (ignite by hot bulb) and the thin beam of the fuel (unatomized) is injected into the cylinder under high revolution under load (ignite by compression).

 

[Chicken]

Hi!
I recently mounted the engine onto a wooden board. Furthermore I have finished the Diesel tank which rests just below the cylinder. It has a little acrylic glass level indicator. The pressure line between injection pump and injector is also finished. It's made from 3 mm copper pipe.
The cooling water pump is also mostly done. It's an ordinary rotary vane pump and will be belt driven from the camshaft.
If it's finished then all that is left are some lubricators, so I'm getting closer to the first start attempt

 

[Herbiev]

Looking forward to watching it run

 

[Chicken]

Hi!

Ready for first start attempt:

Using a small propane burner I heated my hot bulb to dark red (barely visible).
I flipped the engine several times and got lots of white smoke from the exhaust indicating lots of evaporated diesel. In some cases, I also got some very weak combustion, but no real strong power stroke. I tried several different injection timings, but none was successful.
The injector spray is not the finest, but not too bad in my opinion?

What may be the reason that it never really started?
Some thoughts:

Compression too low. But actually it's a 10:1 compression?
Spray from the injector is too bad?
Bad injection timing, but I've tried many different positions?

Currently, I think the most plausible reason is a bad spray from the injector and thus bad fuel evaporation making it difficult to get a ignitable fuel/air mixture.

Do you have any other thoughs?

best regards,
Alex

 

[Mechanicboy]

When you had the problem to start up the engine, the white smoke out of exhaustpipe (unburned fuel) means not enough compression of ratio or not enough heat by hot bulb. The engine will ignite when fuel is ignited by heat of hot bulb + right compression of ratio.

Try to raise up to 14:1-15:1. The hot bulb will evaporate the bad atomized fuel until it will ignite. Heat up the hot bulb to glowing before start up the engine.

Ignition timing can be varied depending on size of hot bulb and compression of ratio. Set first to 5 degree and for each 5 degree between start test up to 20-25 degree before increasing the compresion of ratio until the engine will run with correct timing and compression of ratio.

Edited due disturb by my childrens..

 

[ZAPJACK]

Hello Eirik,
On two stroke engine, especialy on "hot bulb" engine. there are theoric and real compression ratio.
The first one is mesured between TDC and BDC, the second one start when transfer and exhaus port are closed.
That's perhaps why à Lanz have "only" 1/6 compressio ratio
LeZap

 

[Mechanicboy]

Hello Eirik,
On two stroke engine, especialy on "hot bulb" engine. there are theoric and real compression ratio.
The first one is mesured between TDC and BDC, the second one start when transfer and exhaus port are closed.
That's perhaps why à Lanz have "only" 1/6 compressio ratio
LeZap

In a small engine such as a model semidieselengine is much difficult due loss of heat is big than in big semidieselengine, need more compression of ratio. In a ship dieselengine to example Sulzer, the compresion of ratio is not high as in dieselengine made for automobile due the heat by compression in big engine is higher than small engine.

 

[Chicken]

Hi!

Made some troubleshooting today. Looks like the intake ball valve of my injection pump is leaking a bit, so the whole system has problems working against the compression of the engine causing the injector to just poorly drip.
Also, the intake valve is not 100 % tight and causing some compression loss.
I'll get these things sorted and try again.

 

[Mechanicboy]

Take the ball and give a light knock with the rod to form the valveseat (both inlet and outlet valveseat) inside the injectionpump. As rule the semidieselengine has a extra valve before entering into injection to prevent leakage caused by compression/pressure by combustion. Also 1 ball in outlet side and 1 ball before entering into injection.

 

[Mechanicboy]

Here is the drawings of the injector who is used in Sabb semidiesel, it is not difficult to make the injector. The deflector is a angled disc who is rotated by operator to run engine in idling or under load. Under load with atomized fuel against hot bulb will make problem with engine, then rotate the deflector away from orifice then the fuel is ignited by compression. The disc is cooled by injector body under load. Forgotten to tell in early post, all the outlet valve is closed by spring, both the pump and injector . The injector is cooled by waterjack in Sabb semidieselengine, in modelengine try with less cooling to better evaporating of injected fuel, but no fuel is boiling inside the injectorbody.

 

[Chicken]

Hi!

I'm struggling with the injection pump. Always, one ball or both balls are leaking. I tried to produce a good valve seat by giving them a light knock, but it doesn't seem to work. What's the best material to machine the valve seats? my first try was brass, But I feel it's not the best material.
My ball valves are 2 mm in diameter. Which diameter should the passage be drilled?

I thought the most difficult part of the injection pump would be a good tight fit of the plunger, but that was very easy. The ball valves are very challenging!

best regards,
Alex

 

[Mechanicboy]

See at the picture how the valveseat is best for ball as valve, use brass or bronce, both is good as material. 2 mm ball is maybe too little for a big engine (unlike the Find Hansen engine who is about big as WEW engine when we are talking about bore/stroke) You can use 3-4 mm ball where the valve seat is not difficult to make with ownmade drill as i showed in the drawings. The angle is 90-110 degree in drill. And last work, give a light wack with rod on ball to make good tight between ball and valve seat. Use valvespring (thin wire, light springload on valve) to keep ball on place if your valveseat is in horizontal where the ball is not on place on valveseat due force of gravity. About passage: For a ball on 3 mm will i use 2-2,5 mm passage, 4 mm ball use 3-3,5 mm passage. And the hole size in valvehouse ballsize + 0,25mm-0,5 mm to example 3 mm ball drill 3,25 mm or 3,5mm hole in halvehouse. Fuelpassage no problem due fuelamount is very little for each stroke of injection.

 

[Mechanicboy]

The other alternative is the valve same as you find in the dieselpump in dieselpowered car. The valve is very reliable and easy to make to your injectionpump. The valve is made of stainless steel from M6 bolt, valve seat in brass or bronce, the spring is made of stainless steel wire. After the job is done, lap the valve on seat with Autosol Chrome polishing paste or similar such as fine diamond lapping paste and clean up the parts free for dirt. Do not make groove on valve by lapping! Use all valves both inlet, outlet and before entering into injector.

 

[deverett]

Rather than give the ball a whack, it may be better to burnish the seat.

Get a spare new ball - steel if your valve seat is bronze/brass - and glue it to a rod of a size that will hold the ball exactly central over the valve seat. Turn the rod a few times in each direction by hand. Do not use lapping compound. This will put a minute flat on the valve seat that will be a mirror image of the ball surface. Use a new ball for the valve.

Dave
The Emerald Isle

 

[Chicken]

Hi!

It turned out that it's not the valve seats that are leaking! I'm using springs to retain the balls in place. I found that the ball valves will leak if the spring pressure is not 100 % dead on. Even if the spring is only distorted by < 0,1 mm, the uneven pressure on the ball will immediately cause leakage! The leakage is minimal, but already enough to spoil the whole setup. Consider that the maximum pumped volume is only 0.01 mL (10 µL)!
If I turn my pump 90° and just use gravity without any springs, the pump works well and builds up excellent pressure! Unfortunately in this position the pump does not fit into my setup. I will see if I can change this to use the pump without any springs.

best regards,
Alex

 

[Chicken]

Hi!

Allright, I've played around with the injection pump the whole day and I think I came up with a version that is worth tying. I removed any springs and made the pump setup vertical, so only gravity will move the ball valves.
This pump now creates 100 bars easily and holds them several seconds, which is very good I think. I still use the 2 mm ball valves. While working on the pump, I ran into several more issues and I want to share my experience:

- Don't use springs on the injection pump ball valves. If the spring is not pushing perfectly dead on, there will be enough leakage to spoil the pump
- Don't use threads for the valve bodys or other parts of the injection pump. All parts must be soldered. Tiny bubbles of air can catch in the threads and turn the pump unuseable
- Make all passages as small as possible and avoid any sharp corners. Again, tiny bubbles of air can be trapped in narrow corners and they are very difficult to remove
- A tiny bubble of air anywhere in the pump will make it inoperative, bleeding the system almost took me 1 hour.

Now, I must also make a new injector. The current one works OK for larger amounts of fuel. But in the case of very small volumes, a drop may form at the nozzle tip. If a drop has formed, there is no spray anymore. The drop will just keep growing and eventually drip of the injector. I found that 1 drop is about 50 µL, which are 5 strokes of the injection pump at maximum flow rate.
I will now try to build a spring loaded injector that requires a certian pressure to open. This should solve any dripping problems.
Anyway I still have some hope to get this engine running some time!

best regards,
Alex

 

[Mechanicboy]

I saw at the homepage to Find Hansen, there is not drawings. But i learned out by this homepage, the angle at injection valve was at 14 degree and opening pressure at 8 bar. You need the injector tester with manometer to rig up to adjust opening pressure and test the injector before use. The drawings is scematic only. As it works the small injector to inject the engine with high pressure to atomize the fuel. The fuel in Find Hansen hot bulb engine was turpentine or kerosene due thin viscosity (easier to atomize the fuel and ignite by hot bulb)