Solenoid diesel injection pumps

[minh-thanh]

Interesting idea, but you probably don't need to make the plunger too small - 1mm
Roger B's post data is pretty accurate for my 10cc engine. 10cc needs about 0.5mm3 of fuel

 

[The Ignoble Troll]

I haven't done the full math yet on the fuel consumption but this is going to be a blown two stroke so it would at ~14 psi boost which would mean that it would burn twice what a similar sized 4 stroke engine would burn. The goal of making the plunger such a small diameter is to take advantage of the mechanical force provided by the solenoids, which will likely be a lot less force then provided by the driven cranks in many of the four stroke models.

The other reason for such a small plunger is actually following your advice, minh-thanh, and making sure I can adjust both timing and spray volume.

I also am departing form the forum norms a little in that I plan for the engine to drive a load and not just run(not that there is anything wrong with that!). The single cylinder test engine will likely drive a generator as a dummy load, but I want the five cylinder version to drive its own water pump, fuel pump, injector pump, generator, lubrication pump and blower while still having enough power left over to be used to drive a real world application. To that end, it is going to take a lot more fuel to run usefully. When I work out the back pressures created by the compressed air in the cylinder I may go to a larger pump plunger diameter and bring down the total possible injection pressure.

I am up in the air on what I want the injection to look like. Also up in the air if I want to go preburn chamber or not. I would like to be able to burn diesel OR vegetable oil, so a chamber temperature of 500°C is required to burn off the glycerine deposits.


Five cylinders will give me about a 500cc two stroke engine which is where a cheap AMR500 blower will come in with, now getting really hopeful, a turbocharger rated for a smaller engine, tied in to provide compressed air to the blower, once the engine is up to speed.

For now the plan is to devise a fuel metering system that is as bullet proof as possible and drive the single test cylinder off of compressed shop air. If that works I have a large vaned motor that I want to use as the single cylinders blower. If that works, then it is on to adding more cylinders.

My lathe and mill are very small, both manual Taig, so I will probably end up building a modular engine and add cylinders as I go with a linked but not solid crank.

 

[minh-thanh]

I also am departing form the forum norms a little in that I plan for the engine to drive a load and not just run(not that there is anything wrong with that!). The single cylinder test engine will likely drive a generator as a dummy load, but I want the five cylinder version to drive its own water pump, fuel pump, injector pump, generator, lubrication pump and blower while still having enough power left over to be used to drive a real world application. To that end, it is going to take a lot more fuel to run usefully. When I work out the back pressures created by the compressed air in the cylinder I may go to a larger pump plunger diameter and bring down the total possible injection pressure.

I am up in the air on what I want the injection to look like. Also up in the air if I want to go preburn chamber or not. I would like to be able to burn diesel OR vegetable oil, so a chamber temperature of 500°C is required to burn off the glycerine deposits.


Five cylinders will give me about a 500cc two stroke engine which is where a cheap AMR500 blower will come in with, now getting really hopeful, a turbocharger rated for a smaller engine, tied in to provide compressed air to the blower, once the engine is up to speed.

For now the plan is to devise a fuel metering system that is as bullet proof as possible and drive the single test cylinder off of compressed shop air. If that works I have a large vaned motor that I want to use as the single cylinders blower. If that works, then it is on to adding more cylinders.

My lathe and mill are very small, both manual Taig, so I will probably end up building a modular engine and add cylinders as I go with a linked but not solid crank.

Those are very interesting projects !!
----------
----------
1mm ... Really very very small . I can not imagine how you will do it
I also had the thought of making a 1.5 mm cylinder , tried every way and all failed . and decided to give up - the future is unknown ...
I failed does not mean you can not do it . Anyway you can make it a little bigger , from 1 to 2.5mm maybe 3 mm .

 

[The Ignoble Troll]

Those are very interesting projects !!
----------
----------
1mm ... Really very very small . I can not imagine how you will do it
I also had the thought of making a 1.5 mm cylinder , tried every way and all failed . and decided to give up - the future is unknown ...
I failed does not mean you can not do it . Anyway you can make it a little bigger , from 1 to 2.5mm maybe 3 mm .

I agree, making a 1mm cylinder would be killer. The cylinder will probably be closer to 1cm.

Only the gland will be 1mm.

From https://www.google.com/search?q=plu...MjOwTe4y4UM&vssid=_W4arZ_7xPIqw0PEP39mGiQ0_46

 

[minh-thanh]

Quite interesting .
If you use an inlet valve, the spring force must be very light.
With my engine, the spring only holds the valve in the closed position, with almost no force.
Either way, with the small volume, making the valve + valve seat... completely seal is a challenge.

 

[The Ignoble Troll]

Quite interesting .
If you use an inlet valve, the spring force must be very light.
With my engine, the spring only holds the valve in the closed position, with almost no force.
Either way, with the small volume, making the valve + valve seat... completely seal is a challenge.

Agreed. That's why I intend to use a positive feed fuel pump and why I want a small plunger. It will cause a rapid rise in pressure which should help seat the valve.

Thank you for your insights.

 

[Roger B]

Quite interesting .
If you use an inlet valve, the spring force must be very light.
With my engine, the spring only holds the valve in the closed position, with almost no force.
Either way, with the small volume, making the valve + valve seat... completely seal is a challenge.

There are two options for the fuel inlet in an injection pump, an inlet port in the cylinder or a ball/conical valve.

The inlet port is simple and reliable but requires a lapped bore. The start of injection is fixed by the mechanical position of the pump piston which makes altering the volume injected without changing the start of injection more difficult.

If a ball valve is used for the inlet the start of injection is when the piston start moving up the cylinder and the fuel flow closes the ball valve. This means that the fuel volume can be altered by altering how far the piston is withdrawn which has benefits. Ball valves are however, as Minh says, an interesting problem. The movement of the ball needs to be closely constrained to ensure a quick closure but there also needs to be space for a free fuel flow. The diameter of the seating needs to be appropriate for the size of the ball otherwise it will not seat properly. This is short article (in imperial units) regarding the design of ball valves:

https://www.goldengatels.org/librar... and Feeding of Check Valves by Bob Cohen.pdf

Making a good seating can also be a problem, tapping the ball with a brass drift usually works but is not certain. I now lap the seatings using one of these ball laps:

https://www.precisionballs.com/ball_valve.php

I have drawn up a couple of designs of wedge controlled pumps with ball valves but have not built one yet so I can’t make a direct comparison to my inlet port style pumps.

 

[minh-thanh]

There is one thing I wonder... a 1mm plunger is too small compared to a 10mm cylinder.. for example, if the plunger stroke is 1 or 2mm and fast, will it have enough effect on the valve or will it just expand the volume of oil in the cylinder a little bit?? And will the expanding volume of oil in the cylinder reduce the time it takes to open and close the valve??

 

[Roger B]

There are two options for the fuel inlet in an injection pump, an inlet port in the cylinder or a ball/conical valve.

The inlet port is simple and reliable but requires a lapped bore. The start of injection is fixed by the mechanical position of the pump piston which makes altering the volume injected without changing the start of injection more difficult.

If a ball valve is used for the inlet the start of injection is when the piston start moving up the cylinder and the fuel flow closes the ball valve. This means that the fuel volume can be altered by altering how far the piston is withdrawn which has benefits. Ball valves are however, as Minh says, an interesting problem. The movement of the ball needs to be closely constrained to ensure a quick closure but there also needs to be space for a free fuel flow. The diameter of the seating needs to be appropriate for the size of the ball otherwise it will not seat properly. This is short article (in imperial units) regarding the design of ball valves:

https://www.goldengatels.org/library/Tech articles/Design, Care and Feeding of Check Valves by Bob Cohen.pdf

Making a good seating can also be a problem, tapping the ball with a brass drift usually works but is not certain. I now lap the seatings using one of these ball laps:

https://www.precisionballs.com/ball_valve.php

I have drawn up a couple of designs of wedge controlled pumps with ball valves but have not built one yet so I can’t make a direct comparison to my inlet port style pumps.

These are the two pump designs. I think that the horizontal plunger version will have a better chance as gravity will help seat the ball. I have not yet identified suitable springs for closing the inlet valve. The bodies are aluminium with a hardened silver steel cylinder and seating for the delivery valve shrunk or Loctited in place (the red pieces). Something similar may be required for the inlet valve seating but this is removable so can be modified later.

The plunger is retained by a sleeve held in place by a grubscrew (blue) which has a bronze sleeve for the tappet to run in (yellow).

 

[minh-thanh]

Hi Roger .
I bought 0.15 mm diameter SUS304 wire to make spring for ball valve

 

[The Ignoble Troll]

There are two options for the fuel inlet in an injection pump, an inlet port in the cylinder or a ball/conical valve.

The inlet port is simple and reliable but requires a lapped bore. The start of injection is fixed by the mechanical position of the pump piston which makes altering the volume injected without changing the start of injection more difficult.

If a ball valve is used for the inlet the start of injection is when the piston start moving up the cylinder and the fuel flow closes the ball valve. This means that the fuel volume can be altered by altering how far the piston is withdrawn which has benefits. Ball valves are however, as Minh says, an interesting problem. The movement of the ball needs to be closely constrained to ensure a quick closure but there also needs to be space for a free fuel flow. The diameter of the seating needs to be appropriate for the size of the ball otherwise it will not seat properly. This is short article (in imperial units) regarding the design of ball valves:

https://www.goldengatels.org/library/Tech articles/Design, Care and Feeding of Check Valves by Bob Cohen.pdf

Making a good seating can also be a problem, tapping the ball with a brass drift usually works but is not certain. I now lap the seatings using one of these ball laps:

https://www.precisionballs.com/ball_valve.php

I have drawn up a couple of designs of wedge controlled pumps with ball valves but have not built one yet so I can’t make a direct comparison to my inlet port style pumps.

Thank you for the reading material, I will go and digest it.

Besides a lapped ball valve with a spring retainer, my other thought was a spring loaded poppet valve as is favored by commercial high pressure plunger pump manufacturers.

 

[The Ignoble Troll]

There is one thing I wonder... a 1mm plunger is too small compared to a 10mm cylinder.. for example, if the plunger stroke is 1 or 2mm and fast, will it have enough effect on the valve or will it just expand the volume of oil in the cylinder a little bit?? And will the expanding volume of oil in the cylinder reduce the time it takes to open and close the valve??

So long as there are no air bubbles (it will be vacuum purged) the fluid will be uncompressable and be forced out. Any expansion would be due to tubing or the valve body, both will be over built to ensure no flexing.

 

[The Ignoble Troll]

There is one thing I wonder... a 1mm plunger is too small compared to a 10mm cylinder.. for example, if the plunger stroke is 1 or 2mm and fast, will it have enough effect on the valve or will it just expand the volume of oil in the cylinder a little bit?? And will the expanding volume of oil in the cylinder reduce the time it takes to open and close the valve??

I did the math wrong on injection volume the first time. You are correct, I think, 1mm will be too small.

The engine will be running on 15 psi air so when the exaust valve closes, it will contain 2 bar absolute not 1 bar like a regular four stroke.

This obviously means that it will require twice the amount of fuel of a unpressurized but blown two stroke or four stroke.


All of this is very hypothetical still. I am sadly still months away from testing, got to reno and sell the old house, so it's great to get all of this feedback from you guys BEFORE I start.

 

[The Ignoble Troll]

For fixing much larger manual hydraulic pumps, like on a car jack, it seemed to me that the valve seats where made from iron. Highly malleable and rust prone straight iron.

 

[Roger B]

I also am departing form the forum norms a little in that I plan for the engine to drive a load and not just run(not that there is anything wrong with that!). The single cylinder test engine will likely drive a generator as a dummy load, but I want the five cylinder version to drive its own water pump, fuel pump, injector pump, generator, lubrication pump and blower while still having enough power left over to be used to drive a real world application.

I like to load test my engines, all the spark ignition ones are coupled to a DC motor as a generator (also as a starter). I have a resistor load bank to see what they can deliver. I used my 12 cc fuel injected twin to drive a small electric car I made for my children many years ago:

 

[ajoeiam]

I like to load test my engines, all the spark ignition ones are coupled to a DC motor as a generator (also as a starter). I have a resistor load bank to see what they can deliver. I used my 12 cc fuel injected twin to drive a small electric car I made for my children many years ago:

If you would care too - - - - would love some information on that small electric car - - - please?

 

[Roger B]

If you would care too - - - - would love some information on that small electric car - - - please?

The car was made many year ago, mostly from various scrap bins. It has a Parvelux 24 V DC gear motor and controller from a cancelled work project. The wheels are old wheelchair wheels from an auto jumble. The chain wheels are from an old electric lawn mower. The motor went into a 7 1/4" battery electric loco. The steering and front axel are from various scrap bin bits. The power source is two 12V 20AH sealed lead acid batteries. It is laid out so it can be driven by an adult or carry up to 4 small children. For the test in the video the battery leads were connected to the generator coupled to my 12cc four stroke twin.

This should be OK on the forum as it is a load for a model engine ;-)

 

[ajoeiam]

The car was made many year ago, mostly from various scrap bins. It has a Parvelux 24 V DC gear motor and controller from a cancelled work project. The wheels are old wheelchair wheels from an auto jumble. The chain wheels are from an old electric lawn mower. The motor went into a 7 1/4" battery electric loco. The steering and front axel are from various scrap bin bits. The power source is two 12V 20AH sealed lead acid batteries. It is laid out so it can be driven by an adult or carry up to 4 small children. For the test in the video the battery leads were connected to the generator coupled to my 12cc four stroke twin.

View attachment 164823

Thank You!!!

View attachment 164824

View attachment 164825

View attachment 164826

View attachment 164827

This should be OK on the forum as it is a load for a model engine ;-)

 

[The Ignoble Troll]

Some thoughts:

Moving the plunger will alter the start of injection point. This could be compensated for by adjusting the timing of the solenoid but is an added complication.

The plunger needs to be sealed against full injection pressure yet able to move freely. This suggests lapped components. A polymer seal would probably bring too much friction.

The inertia of the plunger needs to be considered. If it does not move rapidly to the threaded rod the rate of pressure rise at the start of injection will be reduced. This will tend to cause dribbling rather than atomisation.

No worries the injector and plunger pump are separate in my design, unless I totally misunderstood your post?

 

[Roger B]

I was thinking about your fuel volume control using an additional spring loaded plunger set in the side of the injection pump barrel. Without the additional plunger as soon as the solenoid is energized the main plunger will rise displacing fuel and causing the inlet valve to close. The pressure will then build up (rapidly as the fuel is almost incompressible) until the delivery valve and injector open. Fuel will then be delivered at a rate determined by the force of the solenoid and the inertia of the main plunger and fuel.

If you now incorporate the side plunger the main plunger will have to displace enough fuel to move the side plunger to the limit screw before pressure will build up to allow injection to commence. With a cam driven system the delay could be readily calculated and would be the same as if a wedge system was used. With the solenoid driven this would depend on the force of the solenoid and the inertia of the main and side pistons.

This may be just a theoretical issue and me overthinking.