The springs he uses look very light. Like the springs from a mechanical pencil.
Request - Injector designs
- Thread starter The Ignoble Troll
- Start date
Help Support Home Model Engine Machinist Forum:
I have looked at Mr Hansen's design very closely. Its not the injector but the fuel delivery system and how the fuel is controlled by the injector pump with the throttle. He showed how he tested the spray pattern and the amount. However, I am of the opinion that controlling the flow is more important then the injector because I dont believe that the pattern has that big of an impact in such a small cylinder volume.
He has a combustion chamber, his spray pattern looks wide enough that it might mostly hit the walls of his burn chamber?
Maybe you are on to something and it's the precise way he can adjust timing and spray volume more then magic at the nozzle end...
...buuuut that tiny simple injector is wonderfully sized and functional.
minh-thanh constantly states that timing and volume control are critical.
I have played around with oil burner nozzles, and they come is a variety of spray pattern widths, and even in solid and hollow cone sprays.
My foundry furnace does not seem to be that sensitive to spray angle, but the interior of the furnace remains red hot, and that is basically acting like a glowplug.
Foundry oil burners can be sensitive to droplet size, and a white paper says that a finer droplet does not necessarily burner hotter/better than a very fine mist. It is a matter of finding the droplet size that produces the largest amount of the hottest flame temperature possible.
I would think that the simpler an injector/pump system is, the better chance of it working reliably/consistently.
I think the Detroit Diesel 671 injectors are cam actuated, but I don't know how they vary the amount injected if they are cam actuated.
The Find-style injector has the advantage of being very small, and allows the larger parts (the pump mechanism) to be located remotely.
Minh's statement about spring tension controlling the release timing is most interesting, and not something that was readily obvious to me.
This is a critical point though for timing of a diesel engine.
.
My foundry furnace does not seem to be that sensitive to spray angle, but the interior of the furnace remains red hot, and that is basically acting like a glowplug.
Foundry oil burners can be sensitive to droplet size, and a white paper says that a finer droplet does not necessarily burner hotter/better than a very fine mist. It is a matter of finding the droplet size that produces the largest amount of the hottest flame temperature possible.
I would think that the simpler an injector/pump system is, the better chance of it working reliably/consistently.
I think the Detroit Diesel 671 injectors are cam actuated, but I don't know how they vary the amount injected if they are cam actuated.
The Find-style injector has the advantage of being very small, and allows the larger parts (the pump mechanism) to be located remotely.
Minh's statement about spring tension controlling the release timing is most interesting, and not something that was readily obvious to me.
This is a critical point though for timing of a diesel engine.
.
I think that the rod through the sides rotates the interior which has the effect of controlling when the internal passages vent extra flow via a helix.
But don't quote me on that.
I will speak from my experience :
Designing and machining an injector is not too difficult for many members here. It's really important to make sure it's 100% sealed - otherwise it will upset the fuel pump system.
Spray pattern : It's very nice - But does it matter?? It depends on your wishes. But if you want a really small injector - smallest in the world
- you have to trade off spring force and then don't expect it to look good - unless experimenting with a larger amount of fuel than the engine needs. That's the problem I often see !!!! - If you don't limit the plunger stroke to the amount of fuel the engine needs, don't expect me to trust that result.
Sealed : Machine an injector and test it with compressed air (1 or 2 bar) and dip the injector tip in water or oil and it will tell you.
Designing and machining an injector is not too difficult for many members here. It's really important to make sure it's 100% sealed - otherwise it will upset the fuel pump system.
Spray pattern : It's very nice - But does it matter?? It depends on your wishes. But if you want a really small injector - smallest in the world
- you have to trade off spring force and then don't expect it to look good - unless experimenting with a larger amount of fuel than the engine needs. That's the problem I often see !!!! - If you don't limit the plunger stroke to the amount of fuel the engine needs, don't expect me to trust that result.Sealed : Machine an injector and test it with compressed air (1 or 2 bar) and dip the injector tip in water or oil and it will tell you.
Roger B
Well-Known Member
To get a small diesel to run the fuel pump, the injector and the combustion chamber are all important. Sufficiently atomized fuel must be injected into high temperature and pressure air in such a way that combustion takes place rapidly but in a controlled way.
As Minh says control of volume and timing are important. The volume can be changed by altering the stroke however this also tends to alter the timing of the start of injection. I think Find Hansen has controls to adjust the volume and timing separately. The alternative is to copy full size practice and fix the timing of the start injection whilst controlling the volume with the end of injection using a helix in the pump plunger. The components will be quite small, a 20 cc cylinder will require around 2mm3 of fuel at full output, say a 2mm plunger and ~1.5mm stroke.
The injector needs to sufficiently atomise the fuel so the droplets are small enough to reach ignition temperature without much delay. This typically achieved using a very small anulus between the nozzle and the needle, either external with a poppet design such as Find Hansen’s or internal with a conventional needle design. Assuming a compression pressure of around 35 bar to reach ignition temperature the injection pressure will need to be 100 bar plus. With a 1.5mm diameter needle this would require a spring load of around 18N. For a poppet design the compression pressure will tend to keep the injector closed so the spring force can be less, for a needle design the compression pressure will tend to force the injector open so more force will be required.
The combustion chamber needs to be compact to minimize heat loss and needs to ensure the air is moving rapidly to allow good mixing with the fuel droplets.
As practical example this is my 20cc four stroke diesel with two different injectors. It has a helix controlled injection pump. The first is just a 0.2 mm diameter nozzle backed with a non-return valve. The engine won’t fire and unburnt fuel flows out of the exhaust.
The second is with a needle type injector, again with a 0.2mm nozzle. The engine starts as soon as I move the fuel rack to start the injection (this was a cold start).
There are various construction pictures of this and other engines here:
https://www.flickr.com/photos/153503473@N05/with/50539849686
Here are the two injector designs:
Roger B
Well-Known Member
To give an idea of the size of my injectors: ( I didn't have any matches )
)
You have done some excellent work. It is impressive because its difficult to go from a design to a working model. What is impressive you have the numbers to go with it although I tend to use imperial because I grew up with it. One question I have and its purely theoretical has to do with the timing of the fuel injector. It would seem that given a pressure and a nozzle area and the required mass of fuel there would be a time difference based on the nozzle velocity. This would mean that the fuel mass could vary based on the delivery pressure as well as nozzle diameter. I am going to assume that the injector opening time is based on the decay of the injector pressure. So just out of curiosity do you consider that in the build of the injector. I also understand that sometimes the best way to get to a answer is trial and error. Just curious.
Curious too, how do you idealize timing on a model engine, EGT or just by looking and listening?
A dynamometer would be the best way.
Roger B
Well-Known Member
Some interesting questions.
Firstly, I am an engineer on the Autistic spectrum so I need to understand how things work. I have some experience with small full-size engines via motorsport and preserved narrow gauge locomotives.
Before I start an engine design I will see what I can research and calculate. This the comes against the limitations of what I can actually make based on my limited skills and limited machine tools. Finally I have to see what I can test, again within my limitations:
Compression pressure; make a one way valve and attach a pressure gauge.
Injection pressure; make a test pump and use a spring balance to measure the force required to open the injector. Calculate the pressure.
Fuel flow; Point the injector into a measuring cylinder and record the flow for a given number of revolutions at different speed and rack settings.
If we assume that the injection is solid, no trapped gas, the injection will start as soon as the inlet port on the pump is closed and will stop when the inlet port is opened again by the helix. The spring in the injector will control the pressure in the system and hence the penetration and spread of the spray. The mechanics of the injector will also affect the spray pattern.
If you just want to optimise the engine a DC motor attached to the engine with a resistor load bank will give you a guide. If you want actual numbers you will need to build a proper dynamometer to compensate for friction and electrical losses.
This is my 12 cc four stroke twin petrol engine undergoing some load trials:
Firstly, I am an engineer on the Autistic spectrum so I need to understand how things work. I have some experience with small full-size engines via motorsport and preserved narrow gauge locomotives.
Before I start an engine design I will see what I can research and calculate. This the comes against the limitations of what I can actually make based on my limited skills and limited machine tools. Finally I have to see what I can test, again within my limitations:
Compression pressure; make a one way valve and attach a pressure gauge.
Injection pressure; make a test pump and use a spring balance to measure the force required to open the injector. Calculate the pressure.
Fuel flow; Point the injector into a measuring cylinder and record the flow for a given number of revolutions at different speed and rack settings.
If we assume that the injection is solid, no trapped gas, the injection will start as soon as the inlet port on the pump is closed and will stop when the inlet port is opened again by the helix. The spring in the injector will control the pressure in the system and hence the penetration and spread of the spray. The mechanics of the injector will also affect the spray pattern.
If you just want to optimise the engine a DC motor attached to the engine with a resistor load bank will give you a guide. If you want actual numbers you will need to build a proper dynamometer to compensate for friction and electrical losses.
This is my 12 cc four stroke twin petrol engine undergoing some load trials:
ninefinger
Well-Known Member
What is ideal? Low emissions (NOx), low particulates, max power, max efficiency, or an engine that starts and runs easily (i.e. fully controllable)?
As a model engine maker, I'm going to aim for the latter, and maybe play in the other sandboxes later...not sure how many of them can be met at the same time - look at full size practice - they are still chasing "ideal".
That's an excellent question, I would say runs smoothly, starts easily in this context.
Then it can power a scrubber to remove NOx and particulate
Similar threads
