Getting an electronic fuel injector to only open partially

[ELS122]

How do I make a gasoline fuel injector open partially? I tried just undervolting one but no matter what I tried it would just be either fully open or fully closed.
Maybe using high frequency AC with a DC offset? Maybe there's a specific type of injector that allows this?

The reason I want this is to be able to adjust the fuel particle size on-the-go. To maximize engine efficiency over the entire rpm range.
I'm not very concerned about the impact of fuel flow when the injector is partially open, since I can compensate for that later and I just want something that gives me an adjustable particle size first.

[bdunham7]

The reason I want this is to be able to adjust the fuel particle size on-the-go. To maximize engine efficiency over the entire rpm range.
I'm not very concerned about the impact of fuel flow when the injector is partially open, since I can compensate for that later and I just want something that gives me an adjustable particle size first.

What effect on fuel droplet size do you anticipate from various pintle positions?  I'm just curious...

As for hanging a solenoid valve in an intermediate position, I think it would be pretty tricky.  You would have to do something like using a PWM signal and then somehow measuring the pintle position (via a change in inductance) in between the pulses using some sort of test signal.  Or something like that.  I don't think typical port injector changes inductance much during the intermediate phase that you are looking to stabilize, so...

[ELS122]

The reason I want this is to be able to adjust the fuel particle size on-the-go. To maximize engine efficiency over the entire rpm range.
I'm not very concerned about the impact of fuel flow when the injector is partially open, since I can compensate for that later and I just want something that gives me an adjustable particle size first.

What effect on fuel droplet size do you anticipate from various pintle positions?  I'm just curious...

As for hanging a solenoid valve in an intermediate position, I think it would be pretty tricky.  You would have to do something like using a PWM signal and then somehow measuring the pintle position (via a change in inductance) in between the pulses using some sort of test signal.  Or something like that.  I don't think typical port injector changes inductance much during the intermediate phase that you are looking to stabilize, so...

Well a big difference, smaller outlet orifice = higher velocity = more breakup, atomization.

The problem is that fuel injectors are already meant for PWM, they are PWM driven to control the fuel flow by changing the duty cycle, and at 8000rpm, the whole intake cycle happens in around 2 miliseconds, and in that time span they're PWM controlled... who knows how extremely fast you'd have to drive them to control the position... 10khz? 100khz?

Oh and btw they are normally triggered at a much higher current, triggered around 11 amps then held with 500mA. presumably to make them open faster, and when they're open lower the current so they dont melt

[AndyBeez]

What kind of fuel injection are you trying to modify - direct cylinder or manifold are common 'modes' of injection.

https://en.m.wikipedia.org/wiki/Manifold_injection
https://en.m.wikipedia.org/wiki/Petrol_direct_injection
https://en.m.wikipedia.org/wiki/Common_rail

Lots of options. EFIs are often optimised to the mechanics, at great expense to the engine manufacturers.

[ELS122]

What kind of fuel injection are you trying to modify - direct cylinder or manifold are common 'modes' of injection.

https://en.m.wikipedia.org/wiki/Manifold_injection
https://en.m.wikipedia.org/wiki/Petrol_direct_injection
https://en.m.wikipedia.org/wiki/Common_rail

Lots of options. EFIs are often optimised to the mechanics, at great expense to the engine manufacturers.

Port injection, Direct injection wouldnt benefit from this.

[Alti]

You could control fuel pressure according to rpm-s.
This should be fast and easy, just put an electronically controlled bypass valve.

[Coordonnée_chromatique]

This should be fast and easy, just put an electronically controlled bypass valve.

I remember the bill of a german study office (with a black horse in the center of the crest) that was around 60 gigabucks in order to implement a Bosch injection system on an french motor.
It was just a injection upgrade on a existing motor.

[ELS122]

You could control fuel pressure according to rpm-s.
This should be fast and easy, just put an electronically controlled bypass valve.

That will have very slow response, no good.
Also there's no electronically controlled fuel pressure regulators when I checked.

[JPortici]

You could control fuel pressure according to rpm-s.
This should be fast and easy, just put an electronically controlled bypass valve.

That will have very slow response, no good.
Also there's no electronically controlled fuel pressure regulators when I checked.

Denso has as system in which the fuel pressure sensor is directly at the injector, one for each injector. They do it to calculate the quantity of fuel going in the chamber with greater precision. It's for diesel engines, marketed with the name iART.
Also for diesels, mantaining open/ partial opening is achieved via PWM. Injector is first fully opened as fast as possible (high current) then the current to the valve is modulated. But i don't know about gasoline as we usually just ignore the fuel injectors

[ELS122]

You could control fuel pressure according to rpm-s.
This should be fast and easy, just put an electronically controlled bypass valve.

That will have very slow response, no good.
Also there's no electronically controlled fuel pressure regulators when I checked.

Denso has as system in which the fuel pressure sensor is directly at the injector, one for each injector. They do it to calculate the quantity of fuel going in the chamber with greater precision. It's for diesel engines, marketed with the name iART.
Also for diesels, mantaining open/ partial opening is achieved via PWM. Injector is first fully opened as fast as possible (high current) then the current to the valve is modulated. But i don't know about gasoline as we usually just ignore the fuel injectors

Diesel fuel injectors will most likely wear out quite quickly since gasoline is a much worse lubricant than diesel.
Also diesel injectors have a return line... I think it's for bleeding air out of the injectors, not sure.

Do you mean they modulate the average current by the PWM signal duty cycle? or that they PWM drive the injectors at a constant duty cycle but modulate the max current?

[langwadt]

The reason I want this is to be able to adjust the fuel particle size on-the-go. To maximize engine efficiency over the entire rpm range.
I'm not very concerned about the impact of fuel flow when the injector is partially open, since I can compensate for that later and I just want something that gives me an adjustable particle size first.

What effect on fuel droplet size do you anticipate from various pintle positions?  I'm just curious...

As for hanging a solenoid valve in an intermediate position, I think it would be pretty tricky.  You would have to do something like using a PWM signal and then somehow measuring the pintle position (via a change in inductance) in between the pulses using some sort of test signal.  Or something like that.  I don't think typical port injector changes inductance much during the intermediate phase that you are looking to stabilize, so...

Well a big difference, smaller outlet orifice = higher velocity = more breakup, atomization.

The problem is that fuel injectors are already meant for PWM, they are PWM driven to control the fuel flow by changing the duty cycle, and at 8000rpm, the whole intake cycle happens in around 2 miliseconds, and in that time span they're PWM controlled... who knows how extremely fast you'd have to drive them to control the position... 10khz? 100khz?

Oh and btw they are normally triggered at a much higher current, triggered around 11 amps then held with 500mA. presumably to make them open faster, and when they're open lower the current so they dont melt

what type of injectors is that? direct injection gasoline?, they are surely not the usual injectors used in port injection gasoline engines

[mendip_discovery]

I dont have my usual contact to ask about this (ford fuel injection nerd).

But my 2p, it takes a given time to open completely could you only pulse long enough for it to open a small amount. Even dropping the current so it opens slower.

Alternative idea is to add some extra smaller injector for the lower speeds and swapping over to larger ones.

[bdunham7]

Well a big difference, smaller outlet orifice = higher velocity = more breakup, atomization.

The problem is that fuel injectors are already meant for PWM, they are PWM driven to control the fuel flow by changing the duty cycle, and at 8000rpm, the whole intake cycle happens in around 2 miliseconds, and in that time span they're PWM controlled... who knows how extremely fast you'd have to drive them to control the position... 10khz? 100khz?

Oh and btw they are normally triggered at a much higher current, triggered around 11 amps then held with 500mA. presumably to make them open faster, and when they're open lower the current so they dont melt

You're making a lot of assumptions that you should consider more carefully.  There are a lot of different specific types of injectors and they all have different designs and drive requirements.  I'm pretty sure that there are many that will behave pretty much the opposite of your stated expectation.  Also, not that it matters to your original question, but the injector on-time is typically not correlated with the 'intake cycle'; the specifics of that are the subject of quite a bit of research and testing when an engine is designed.

[Gregg]

It certainly is difficult to out engineer large companies that specialize in trying to gain fuel economy in order to achieve government mandated regulations. 
But maybe you could consider adding compressed air ahead of the injector to make an air-fuel emulsion similar to the way air bleed orifices worked on carburetors.

[Miyuki]

Most fuel injectors are not spraying from the needle but from an array of tiny precisely machined orifices in a plate on its tip
Opening the valve just partially will decrease the spray pressure, so it will just make big droplets
Such ancient injector construction was used on throttle body injector with very low pressure spraying at the throttle plate and just evaporating the fuel in the long intake manifold

[joeqsmith]

I wonder how your other valve worked out?  I figured for sure you would have got a kick out of the old man's history on it.  Hard to argue with his success.

https://www.eevblog.com/forum/chat/200hz-air-valve/

At one point, we had a quartz window and laser that we could use to take photos of the patterns.  We did whole studies on how the orifice effected it.    That's been decades ago and before the piezo injectors.  Anything I have worked on was always an on/off type valve.   We did work on a multi-pulse system.  Basically firing the injectors several times for one combustion cycle.   Think of integrating the area under the curve.  We could then change the shape of it.  It also allowed us to trim individual cylinder.   Seems like a whole different life time......

[ELS122]

Most fuel injectors are not spraying from the needle but from an array of tiny precisely machined orifices in a plate on its tip
Opening the valve just partially will decrease the spray pressure, so it will just make big droplets
Such ancient injector construction was used on throttle body injector with very low pressure spraying at the throttle plate and just evaporating the fuel in the long intake manifold

That's the fucking idea.
It's amazing how everyone here is unable to understand the question I asked.

[bdunham7]

That's the fucking idea.
It's amazing how everyone here is unable to understand the question I asked.

Perhaps you don't understand the question you asked since you seem to be stating it backwards.

Well a big difference, smaller outlet orifice = higher velocity = more breakup, atomization.

[TERRA Operative]

Injectors are optimised to spray correctly at a certain fuel pressure range.
Altering the pintle position when open will affect the pressure of the fuel exiting the valve portion of the injector before it hits the director plate that actually determines spray pattern and droplet size. Less open pintle = less flow = less pressure at the plate and sub-optimal spray pattern.
Also, the injectors in a manifold injection system don't actually spray directly into the open combustion chamber anyway, they spray onto the back of the intake valves to help keep them cool and to evaporate the fuel for better atomization.

My thinking to increase atomization would be to increase fuel pressure then decrease the injector on-time to compensate for correct air/fuel ratios.

But before diving into injector droplet size, I'd be doing more 'tried and true' methods of increasing efficiency and power, like forced induction, porting and polishing the intake system, flow matching the intake and exhaust ports, adjusting intake and exhaust runner lengths to optimise the powerband for the application at hand and designing the exhaust system to scavenge at the same RPM range.

Then use an aftermarket ECU to tune the car to a better/leaner tune on a dyno instead of relying on an overly conservative OEM tune.

[Miyuki]

I wonder how your other valve worked out?  I figured for sure you would have got a kick out of the old man's history on it.  Hard to argue with his success.

https://www.eevblog.com/forum/chat/200hz-air-valve/

At one point, we had a quartz window and laser that we could use to take photos of the patterns.  We did whole studies on how the orifice effected it.    That's been decades ago and before the piezo injectors.  Anything I have worked on was always an on/off type valve.   We did work on a multi-pulse system.  Basically firing the injectors several times for one combustion cycle.   Think of integrating the area under the curve.  We could then change the shape of it.  It also allowed us to trim individual cylinder.   Seems like a whole different life time......

Thar must be crazy work to do. I know how annoying it can be to just adjust the injection for different types of injectors with different on and off response times.
And now with modern piezo-style injectors, multi-injection is something all engines do.

That's the fucking idea.
It's amazing how everyone here is unable to understand the question I asked.

Big droplets will always have worse efficiency

[sokoloff]

at 8000rpm, the whole intake cycle happens in around 2 miliseconds,

At 8000 RPM, each cylinder in a four-stroke engine is doing 4000 intake cycles per minute or 67 per second. You also have some valve overlap and intake air flow inertia, so you have nearly an order of magnitude longer than you’ve calculated.

[HighVoltage]

1. You need a PWM signal to the injectors, that will give you better control over the injector timing, opening lift....
2. Best to have a fuel injector test stand (can be DIY)
3. Even better to have a high speed camera to know what the fuel injector droplets look like.

There is little linearity in this and big changes in droplet size, when moving away from the optimum fuel injector signals. I used to have a BOSCH paper on this, just can not find it right now.

 

[ELS122]

It certainly is difficult to out engineer large companies that specialize in trying to gain fuel economy in order to achieve government mandated regulations. 
But maybe you could consider adding compressed air ahead of the injector to make an air-fuel emulsion similar to the way air bleed orifices worked on carburetors.

A carburator actually does change the fuel particle size depending on the flow by it's nature.
Only problem is that you cannot set the minimal particle size, there's nothing to limit the fuel droplets from beading up and splattering all over the intake at low flow velocities.
That could be fixed by making the carburator slightly pressurized, and controlling the fuel going to the boosters with a solenoid, sort of like an EFI-Carburator hybrid. But at this point it would be far more practical and versatile to design a new fuel injector that's meant to have varying droplet size. This would also allow you to force a small droplet size for cold starts and/or not require a heated intake manifold because droplet size could just be controlled by a temp sensor right on the injector.

[ELS122]

at 8000rpm, the whole intake cycle happens in around 2 miliseconds,

At 8000 RPM, each cylinder in a four-stroke engine is doing 4000 intake cycles per minute or 67 per second. You also have some valve overlap and intake air flow inertia, so you have nearly an order of magnitude longer than you’ve calculated.

The intake cycle doesnt take 720 degrees of crankshaft rotation to complete, instead it's more around 1/4 of that, 180 degrees. furthermore the injector is PWM'd to control the volume of fuel injected. so it's way faster than 2ms

[ELS122]

I wonder how your other valve worked out?  I figured for sure you would have got a kick out of the old man's history on it.  Hard to argue with his success.

https://www.eevblog.com/forum/chat/200hz-air-valve/

At one point, we had a quartz window and laser that we could use to take photos of the patterns.  We did whole studies on how the orifice effected it.    That's been decades ago and before the piezo injectors.  Anything I have worked on was always an on/off type valve.   We did work on a multi-pulse system.  Basically firing the injectors several times for one combustion cycle.   Think of integrating the area under the curve.  We could then change the shape of it.  It also allowed us to trim individual cylinder.   Seems like a whole different life time......

Thar must be crazy work to do. I know how annoying it can be to just adjust the injection for different types of injectors with different on and off response times.
And now with modern piezo-style injectors, multi-injection is something all engines do.

That's the fucking idea.
It's amazing how everyone here is unable to understand the question I asked.

Big droplets will always have worse efficiency

>Multi-injection is something all engines do
I wonder why when you say:
>Big droplets will always have worse efficiency.

Must be an oversight of the non-out-engineerable engineers   

I'll just go and convert my engine to gas vapors because liquid fuel is lower efficiency... you know.. smaller droplets that evaporate faster always mean higher efficiency right