Port Injection Strategies Paper: Key Takeaways on Open Valve Injection?

[NewbVetteGuy]

I REALLY enjoyed this University of Brighton (2011) PhD candidate's paper on Port Fuel Injection Strategies that can help drive learner burn mixtures, and am interested in some of the key take-aways that someone should have for getting the most out of a port-injected engine.

Curious what others think: https://www.researchgate.net/profile/Mo ... dacted.pdf

He spent some time looking at droplet size and quoting Meyer and Heywood (1999):

• *Typical (older) PFI injector produces droplets between 70 and 150 micrometer
  *Larger droplets (ability to atomize throughout their travel through the intake tract and cylinder) depend upon "fuel film transient behavior (surface tension?)
  *Intake airflow velocity doesn't really influence the mean droplet past the intake valve below 1,000 RPM (but a strategy of only injecting into an open valve DID result in smaller mean diameters going into the cylinder)

Hindle 2008 quotes that were interesting:

• *multihole injectors produce a finer droplet distribution with narrower angles, which helps a lot at low velocities (obvious...)
  *Single hole injectors have a higher momentum which means better targeting and better ability for fuel droplets to fight reversion and make it into the cylinder

Paper looked at other injector technologies that can improve atomization and mixing: air-assisted injection, swirl tips, nozzle tip heating, ultrasonic. (I found an ANNOUNCED heated tip fuel injector from Delphi that was supposed to help with startup and E85 but appears to have been pulled from the market.)


A GREAT look into closed valve vs. open valve injection- which was largely what I was looking to learn and how I found the paper.

• * Closed-valve: spray angle and trajectory more important than droplet size 60-70 micrometer droplets bouncing off the intake valve are typical at WoT
  *Open-valve: starting @ 2,000-3,000 RPM air flow starts altering the fuel droplet trajectory, but does not significantly impact droplet size yet. Testing say a droplet size of 100 micrometers when open-valve injection started near the valve open; BUT when open valve injection occured late in the compression stroke, fuel droplet size was reduced to 50-80 micrometer (a big accurate injector timed backwards from the intake valve close (Large, accurate Injector Dynamics style injectors timed backwards from valve close) saw an increase in atomization.)
  * SMD values under 25 past the valve can significantly decrease wall wetting with OVI strategies (Holthaus in 1997 theorized that you'd need droplets smaller than 10 micron; this seems to be an upward revision of that number)
  *A high velocity exhaust system is even more important if adopting an OVI strategy (Wish I understood this part...)

--An old Speedtalk thread on the potential benefits of a large injector, late injection strategy saw a quote from Mark MK E (wish I had the link still) stating "I've seen 1/2 dozen dyno charts that show a 3-5% hp increase going from port injection with 80% duty cycle to a 35% duty cycle with Injector Dynamics injectors with narrow 5-10 deg spray patterns." -If a fine atomization into a faster moving air-stream results in more fuel evaporation and more cooling going to the air vs. the walls and intake valve, then I could see such a strategy giving port fuel injection more of the cooling benefits that we see from carbs / throttle body injection. -Is that's what's happening here? (I think so.)

Later in the paper it was mentioned that open valve port injection reduces exhaust gas temperatures by 30-40 degrees C vs. firing upon a closed valve; isn't this just evidence that more cooling is happening in the cylinder and that such a strategy could potentially allow for more knock resistance at a given DCR / cylinder pressure?

It was also mentioned that open valve port injection improved the ability to burn lean mixtures.
A Stone 1999 quote stated that 10% of fuel could be saved if cycle to cycle variations could be eliminated and mentioned that fuel concentration at the spark gap early in the ignition event has a strong coorelation to later max pressure development.
The study's experiment then quoted Zhai and Lai 1995 and Berkmuller 1997 that stated that open valve injection gives you an average of 4.3% richer fuel stratification around the spark plug.
--> My takeaway is that a large spark plug gap combined with a late Open-valve injection can significantly extend the lean-burn limit, which was clearly called out as a key take-away in the study's 2nd Abstract paragraph. (Lean limit extended from 14:1-17.5:1 @ 1,000 rpm and low pressures, and from 17.5:1 to 12:1 at 1,500 RPM and 1.5 bar. (This was using an intense "duel split injection" strategy that isn't possible on a dual valve engine.)


-With direct injected engines ALL the cooling potential of the fuel is going directly into the cylinder and we see higher compression ratios as a result. Feels like a late, open valve port fuel injection strategy is moving port injection closer in that direction. (Especially if combined with an injector that's starting out with a smaller fuel droplet to begin with as it's got so much less TIME / DISTANCE in the intake tract to evaporate.)



I thought that it highlighted some interest potential advantages of a late, open-valve via large, accurate port fuel injectors injection strategy that I hadn't seen testing data or academic references for previously, anyway.

Some of this makes me really wish for a large capacity, 12 hole port fuel injector with their 50um initial droplet size, but large, or even moderate capacity and 12 hole injectors don't seem to really be a combination that exists...

I like the idea that a late open valve port injection strategy has the potential to buy you a little bit of extra breathing room with a high DCR engine on the edge of pump gas AND when combined with a powerful ignition system and a decent spark plug gap, to push lean cruise mixtures even a bit leaner.



We've all seen the port injection vs. carb or throttle body injection results a bunch of times now, but it would be interesting to see port injection maximized for the best possible atomization and cooling in this way vs. carb / throttle body injection, IMHO; maybe someday...

I've also seen a few high velocity header dyno tests that clearly highlighted that fuel was being sent out the exhaust at overlap and a large and late open valve injection strategy seems like it would help with that if the fuel isn't being introduced until after the exhaust valve has already closed.

I know that some folks have tried switching to timing their port injection backwards from ivc (Belgian1979) and reported no real gains, but I'm interested in hearing from others who have tried this.


Adam

[FC-Pilot]

I would have to imagine that the engine platform and port and chamber design would be big contributors to this potential gain. Some of our old antiquated stuff we are all still using probably would be a waste of time on this theory. Only one way to find out.

Paul

[David Redszus]

It's nice to see someone take on a complex subject and provide technical information from qualified sources.

We have built and raced engines with high pressure (450psi) fuel injection system since 1974. For many years they
assumed the cloak of mystery and dark science. Until we were able to apply a bit of mathematics to the beast.

One of the major problems is to deliver an adequate amount of fuel within the time allotted. As the engine runs
faster, the same injection period causes an increased duty cycle and requires more crank degrees. An example follows
based on an injection period of 10ms.

rpm....duty cycle...CS degrees
2000.....16.7%......120
4000.....33.3.......240
6000.....50.0.......360
8000.....66.7.......480
10000....83.3.......600
12000....100........720

Clearly, at 6000 rpm, an injection period of 10ms would require 360 degrees of open inlet valve. Inlet valve durations
rarely exceed 280 degrees, resulting in a substantial amount of fuel being sprayed on a closed intake valve.

If an injection period of 10ms is inadequate to provide the required amount of fuel, either a longer injection
period must be provided or a larger injector selected. But larger injectors with limited injection periods, will
deliver much larger droplet sizes than a smaller injector that is open longer. But a smaller injector being open
longer will spray fuel on a closed valve where it will cook and coke until the valve is opened.

High speed engines often require multiple injectors per cylinder; sometimes sequentially phased.

One solution is to use very high pressure injectors that can deliver the require fuel volume in the available
allotted crank degrees during which the inlet valve is open, and with a very small droplet size.

And that is one feature of direct injection systems. The injection pressures are mind-boggling. While they do
not cool the inlet air charge due to evaporation prior to entering the chamber, their vapors also do not
displace induction air. Now we have a trade-off to consider: charge cooling vs vapor displacement.

At the other end of the spectrum, we have carbs and throttle body injection that will deliver fuel during
valve overlap and deposit rivers of liquid fuel in walls and valves.

If I were to consider building race engines again (I won't), I would strongly consider designing and building turbo charged direct injection giant killers.
engines

[Circlotron]

Something I mentioned on another thread some time ago - Injecting while the valve is open might help to sweep in additional air courtesy of the kinetic energy of the fuel droplets.

[NewbVetteGuy]

One of the major problems is to deliver an adequate amount of fuel within the time allotted. As the engine runs
faster, the same injection period causes an increased duty cycle and requires more crank degrees. An example follows
based on an injection period of 10ms.

Sure, but if you're trying to use the late, open valve injection strategy to improve fuel economy at highway cruise RPM, then you just size the injector to be large enough so that it's spraying after overlap ends and before the intake valve closes at whatever your RPM is at say 75 mph and don't worry about spraying during overlap or on a closed valve at high RPM? (Does that make sense as a sizing strategy to minimize droplet size by not having a gigantic injector with huge holes that make huge droplets and still being able to inject enough fuel between exhaust close and intake close on the intake stroke?)

One solution is to use very high pressure injectors that can deliver the require fuel volume in the available
allotted crank degrees during which the inlet valve is open, and with a very small droplet size.

And that is one feature of direct injection systems. The injection pressures are mind-boggling. While they do
not cool the inlet air charge due to evaporation prior to entering the chamber, their vapors also do not
displace induction air. Now we have a trade-off to consider: charge cooling vs vapor displacement.

-Sure but so many modern engines aren't really that "airflow limited" now with their shallow valve angles, huge valves, and huge intake ports, are they? With heads being so good that intake durations can be shorter, if you don't have as much O2 because you're cooling the chamber more and the intake less, can't you just add an appropriate amount of duration to make up for the loss in airflow from the charge cooling? (Peter's got the extra dough, he can afford to pay some for Paul.)

-Feels kinda like a post Nuclear Age warfare concept: after the nuke we had nearly infinite lethality so the focus of weapons development and military strategy had to significantly change. We now have a TON of airflow, but not a ton of octane, so we trade some airflow, which is plentiful, for more octane tolerance ala combustion chamber cooling and power from the efficiency gain we get from the higher compression ratio that's now possible, because we're more octane-limited than we are airflow limited. (At least the lowly pump gas plebeians like me are octane limited.)


Adam

[NewbVetteGuy]

Can anyone explain why or guess at why the author of the paper says that high velocity exhaust systems are so important with late Open Valve Injection strategies?


High velocity exhaust systems ALWAYS seem good to me, but I can't see why they'd be particularly useful or important with such an injection strategy that the author would've added / stressed that point?


Adam

[NewbVetteGuy]

We have built and raced engines with high pressure (450psi) fuel injection system since 1974. For many years they
assumed the cloak of mystery and dark science. Until we were able to apply a bit of mathematics to the beast.

Wow. This was 450 psi port injection!?? With some sort of super special purpose-built injector that could deal with the pressures and still have good response times for idle / low RPM or was this for an application in which only high RPM operation really mattered?

-I don't understand how all these ultra high fuel pressure engines are not a gigantic explosion risk if something goes wrong... A large quantity of highly atomized fuel getting shot around a hot engine bay.... Terrifying when I stop to think about it.

I have to think that the OEMs doing direct injection have either decided it's not a real risk or they've mitigated it significantly somehow.


Adam

[dannobee]

I have to think that the OEMs doing direct injection have either decided it's not a real risk or they've mitigated it significantly somehow.

Well, they do have some experience with diesels having over 30,000psi of fuel pressure. The 3,000 psi in a gasoline direct injection is a walk in the park.

[David Redszus]

Sure, but if you're trying to use the late, open valve injection strategy to improve fuel economy at highway cruise RPM, then you just size the injector to be large enough so that it's spraying after overlap ends and before the intake valve closes at whatever your RPM is at say 75 mph and don't worry about spraying during overlap or on a closed valve at high RPM?

Before any determination of injector sizing, it is necessary to determine how much fuel will be needed.
Considering a 2.5L, engine running at 6000rpm, we find that:
it will consume 4.98 lbs of AIR per minute per cylinder...
AND 0.362 lbs of FUEL per minute per cylinder.

That comes out to 0.07 cc fuel per revolution or 0.00006 lbs fuel per rev.
Those are very small amounts of fuel that must be injected within a very small amount of time. Injector fuel
delivery properties are very carefully selected and controlled in order to deliver the proper amount of fuel at
the right time under the changing operating conditions. Injector fuel mass delivery is non-linear with duty cycle.

Sure but so many modern engines aren't really that "airflow limited" now with their shallow valve angles, huge valves, and huge intake ports, are they?

Engines are "airflow limited", even modern engines. Consider that the maximum airflow (per stroke) exists at piston speeds well below redline or power peak piston speeds. The actual inflow air mass is not constant across all operating conditions, even steady state conditions. For maximum efficiency, fuel must be carefully metered to exactly meet required conditions.

We now have a TON of airflow, but not a ton of octane, so we trade some airflow, which is plentiful,

We have neither a ton of air flow nor a ton of octane. Some airflow will be traded for turbulence to assist fuel
evaporation; and some is displaced by fuel vapor.

Compression ratio increases produce small returns to performance improvement in terms of combustion efficiency CE.
DCR 7 = 54.09 CE
DCR 8 = 56.67
DCR 9 = 58.55

The best way to improve engine efficiency and performance is to eliminate combustion variations and malfires.
This means adjusting fuel and timing in response to changes in a multiple of variables: density,
temperature, humidity, velocity, purity, EGR, etc, etc. on the fly and for each cylinder individually.

For a really fun exercise, select any engine at random, and draw a fuel injection delivery map considering load
and speed as well as an ignition map, from scratch.

[NewbVetteGuy]

Sure, but if you're trying to use the late, open valve injection strategy to improve fuel economy at highway cruise RPM, then you just size the injector to be large enough so that it's spraying after overlap ends and before the intake valve closes at whatever your RPM is at say 75 mph and don't worry about spraying during overlap or on a closed valve at high RPM?

Before any determination of injector sizing, it is necessary to determine how much fuel will be needed.
Considering a 2.5L, engine running at 6000rpm, we find that:
it will consume 4.98 lbs of AIR per minute per cylinder...
AND 0.362 lbs of FUEL per minute per cylinder.

That comes out to 0.07 cc fuel per revolution or 0.00006 lbs fuel per rev.
Those are very small amounts of fuel that must be injected within a very small amount of time. Injector fuel
delivery properties are very carefully selected and controlled in order to deliver the proper amount of fuel at
the right time under the changing operating conditions. Injector fuel mass delivery is non-linear with duty cycle.

This was all super great, David.

I'm trying to replicate the math that you did with my engine, which I think I've done. I think you provided that to illustrate the absurdly tiny amounts of fuel that need to be introduced and how hard the challenge is.

Still trying to figure out how to size an injector to support the much smaller window that's available if I only want to spray when the intake valve is open, and the even smaller window that's available if I don't want to be spraying during overlap...

-I'm simply going from an initial sequential multiport injection strategy timed to the intake open event to having one instead timed backwards from the intake close event AND I want to ensure that it's not done during my 62 degrees of overlap.

(Of course at higher RPM there is not sequential injection and it's just flooding the ports, but at low cruise RPM I want the benefits of injecting only on an open valve and trying to avoid the overlap period.)

-Should I just need a slightly larger injector than the typical sizing for this strategy?

I was honestly expecting something simple like "if you're looking to size injectors for an OVI strategy simply size your injectors in the normal manner but choose a 50% duty cycle instead of an 85% max duty cycle" or something like that.


I do like calculating the theoretical fuel flow required like this though and then being able to potentially look at how much time I had to inject that much fuel at a specific RPM and get a more accurate view of sizing. (Might try this next.)


Adam

[NewbVetteGuy]

Playing with engine math in a way that makes sense to me to estimate OVI strategy backwards from intake close and avoid the overlap period:

Assuming that I only want to do OVI at up to 3,000 RPM and under for highway cruise efficiency increases (my car cruises @ 2,355 RPM @ 79 MPH (I have 70 mph highways so 79 is a possible highway cruise speed for me), my 5.7L 350 @ 3,000 RPM and a 100% VE (not real, but I'd rather be safe on injector sizing) requires 301.7 CFM of air (21.1 lbs / min or 1,266 lbs/ hr)

If I'm running a rich AFR of 13:1 (but I plan on running very lean at low load while cruising), then I need 1,266/13= 97.4 lbs / hr of fuel @ 3,000 RPM.
97.4 lbs / hr split amongst 8 injectors is only 12 lbs / hr.

But 12 lbs / hour I think assumes that the injectors are constantly spraying that 12 lbs / hour all 720 degrees.
My cam's intake as a seat to seat duration of 280 degrees and an overlap of 62 degrees -so I have only 218 degrees to inject that 12 lbs / hour worth of fuel. (I have to squirt all the fuel in in 30.3% of the available total time.) -I need 3.3x as much fuel flow to get it all in during the only 30.3% of the time that I have.

12lbs / hour *3.3= 39.6 lbs / hour injector required to inject all the fuel that my engine needs @ 3,000 RPM in only 218 degrees.
(My current injectors are 36 lbs /hour @ 3 bar and if I ran them @ 4 bar they'd flow 41.6 so would be totally doable.)
Probably made many wrong assumptions and maybe some math mistakes, but there's my ugly attempt at doing it a way that made sense to me.



David, why isn't this as simple as calculating the fuel requirement for the engine at an RPM safely above the highway cruise in lbs / hr and then multiplying that by the % of "degrees available for injection"? ((intake seat-to-seat duration - overlap / 720))

I get that if I wanted to try and use this injection strategy as a way to better cool combustion at WOT and max RPM, then I'd end up with a requirement for a huge injector compared to typical injection guidance and @ idle you can start running the injector in a non-linear response area if it's not a good injector. (I'll do that math next with my potentially horribly flawed method.)


Adam

[NewbVetteGuy]

Dead Simple (horribly flawed?) Method of Sizing Port Fuel Injectors for OVI while excluding overlap period:

Applied to a 350 estimated @ 440hp at up to 6,000 RPM:

1. Obtain Traditional 85% fuel injector sizing # from online calculator: 440hp, 8 injectors BSFC .5 for NA engine, 85% duty cycle = 32.4 lbs / hr https://www.witchhunter.com/injectorcalc1.php
2. Determine Degrees available for injection: (Intake Seat-to-seat Duration - degrees of overlap) 280-62= 218 degrees
3. Calculate %of total time available to inject fuel: 218 degrees / 720 degrees: 30.3% and injector multiplier (720/218)= 3.3
4. Multiple original injector size by injector multiplier based upon % of total time available: 32.4 lbs * 3.3 = 107 lbs / hr injectors required (1,123.5cc/min)

-A 400 hp engine can get by with an Injector Dynamics ID1050 at 3 bar for this 440 hp a 1050 would work with slightly increased pressure or max duty cycles all the way to 6,000 RPM. (Assuming this math works, which is hugely questionable at this point...)


Adam

[David Redszus]

Dead Simple (horribly flawed?) Method of Sizing Port Fuel Injectors for OVI while excluding overlap period:

Applied to a 350 estimated @ 440hp at up to 6,000 RPM:

1. Obtain Traditional 85% fuel injector sizing # from online calculator: 440hp, 8 injectors BSFC .5 for NA engine, 85% duty cycle = 32.4 lbs / hr https://www.witchhunter.com/injectorcalc1.php
2. Determine Degrees available for injection: (Intake Seat-to-seat Duration - degrees of overlap) 280-62= 218 degrees
3. Calculate %of total time available to inject fuel: 218 degrees / 720 degrees: 30.3% and injector multiplier (720/218)= 3.3
4. Multiple original injector size by injector multiplier based upon % of total time available: 32.4 lbs * 3.3 = 107 lbs / hr injectors required (1,123.5cc/min)

-A 400 hp engine can get by with an Injector Dynamics ID1050 at 3 bar for this 440 hp a 1050 would work with slightly increased pressure or max duty cycles all the way to 6,000 RPM. (Assuming this math works, which is hugely questionable at this point...)

Adam

Fuel pressure actually refers to the differential pressure between the fuel and air into
which the fuel is being injected. If fuel at 50psi is being injected into ambient air, the
differential pressure is 50-14.7=35.3psi. If sprayed into a diesel chamber with a
compression pressure of 22 bar the fuel pressure must exceed 323psi or there will be
no fuel flow. It becomes apparent that the fuel pressure differential will change with
changes in fuel pressure and inlet air pressure.

Fuel sprayed into a low pressure manifold will increase the delivered fuel volume and
fuel mass. Fuel sprayed into a high pressure manifold such as a turbo application will
decrease the delivered fuel volume and fuel mass. Fuel flow is determined by the square
root of the product of fuel density and differential fuel pressure. Doubling the fuel
pressure will increase fuel flow by 41.5%, but a 10% increase in fuel specific gravity
(.72-.79) will increase fuel flow by only 5.0%.

The fuel flow through a specific injector nozzle will depend on the product of orifice
area (A) and the discharge coefficient (CD), called CdA. An injector may have a large
nozzle area with a low flow coefficient. Or the reverse may also be true.

The discharge coefficient (CD) is a multiplier used to adjust for a number of variables
that affect the flow of a liquid through a small hole. Since there are quite a few variables
which have an effect on fuel flow, they are often measured on a test bench using
standardized test nozzles.

The CD of a fuel orifice is determined by:
1. Fuel mass flow rate (velocity)
2. Orifice length/diameter ratio.
3. Orifice/approach area ratio.
4. Orifice surface area.
5. Orifice surface roughness.
6. Orifice inlet and exit edge chamfers.
7. Fluid specific gravity.
8. Fluid viscosity.
9. Fluid surface tension.

Item 1 is determined by the injector nozzle size, design and pressure. Items 2 through
6 are a function of nozzle design. Items 7, 8 & 9 are determined by fuel characteristics
and temperature.

Since exact values for discharge coefficients is rarely available, various methods have
been developed to estimate its value based on Reynolds turbulence number.

Regarding injector dwell time, it is important to recognize that fuel mass flow rates are
not linear with dwell time. A duty cycle consists of an opening period, a fully open
period, and a closing period. The opening and closing ramps flow much less fuel than
during the fully open period. Posted flow numbers often do not consider the fuel density
and are based only on fully open flow numbers.

After you have calculated the probable fuel mass required and have selected a candidate
injector, it is good practice to flow test the injector using the dwell time calculated;
and make the dwell time adjustments as needed.

[Bill Chase]

Sure, but if you're trying to use the late, open valve injection strategy to improve fuel economy at highway cruise RPM, then you just size the injector to be large enough so that it's spraying after overlap ends and before the intake valve closes at whatever your RPM is at say 75 mph and don't worry about spraying during overlap or on a closed valve at high RPM?

Before any determination of injector sizing, it is necessary to determine how much fuel will be needed.
Considering a 2.5L, engine running at 6000rpm, we find that:
it will consume 4.98 lbs of AIR per minute per cylinder...
AND 0.362 lbs of FUEL per minute per cylinder.

That comes out to 0.07 cc fuel per revolution or 0.00006 lbs fuel per rev.
Those are very small amounts of fuel that must be injected within a very small amount of time. Injector fuel
delivery properties are very carefully selected and controlled in order to deliver the proper amount of fuel at
the right time under the changing operating conditions. Injector fuel mass delivery is non-linear with duty cycle.

Sure but so many modern engines aren't really that "airflow limited" now with their shallow valve angles, huge valves, and huge intake ports, are they?

Engines are "airflow limited", even modern engines. Consider that the maximum airflow (per stroke) exists at piston speeds well below redline or power peak piston speeds. The actual inflow air mass is not constant across all operating conditions, even steady state conditions. For maximum efficiency, fuel must be carefully metered to exactly meet required conditions.

We now have a TON of airflow, but not a ton of octane, so we trade some airflow, which is plentiful,

We have neither a ton of air flow nor a ton of octane. Some airflow will be traded for turbulence to assist fuel
evaporation; and some is displaced by fuel vapor.

Compression ratio increases produce small returns to performance improvement in terms of combustion efficiency CE.
DCR 7 = 54.09 CE
DCR 8 = 56.67
DCR 9 = 58.55

The best way to improve engine efficiency and performance is to eliminate combustion variations and malfires.
This means adjusting fuel and timing in response to changes in a multiple of variables: density,
temperature, humidity, velocity, purity, EGR, etc, etc. on the fly and for each cylinder individually.

For a really fun exercise, select any engine at random, and draw a fuel injection delivery map considering load
and speed as well as an ignition map, from scratch.

With all the advances in ecu processing power it makes a guy wonder why they haven't went to a step up transformer for the injector drivers. When you can snap the average injectors open in half the time by using 20 volts vs 12-14 you could do a lot more with a lot less injector in the given time you have to inject the fuel into an open intake valve. See drag race guys with Holley EFI going to 18-20 volt systems for this very reason, With the biggest benefit being quickly getting the injectors into a linear area of operation. yet the oems have not followed suite. What do they know that's not being discussed here?

[David Redszus]

With all the advances in ecu processing power it makes a guy wonder why they haven't went to a step up transformer for the injector drivers. When you can snap the average injectors open in half the time by using 20 volts vs 12-14 you could do a lot more with a lot less injector in the given time you have to inject the fuel into an open intake valve. See drag race guys with Holley EFI going to 18-20 volt systems for this very reason, With the biggest benefit being quickly getting the injectors into a linear area of operation. yet the oems have not followed suite. What do they know that's not being discussed here?

You have opened a real can of worms. Here are just a few of them:

Test fluid
The flow through an injector will vary with the test fluid being used. The volume flow
will vary with the viscosity of the fluid while the mass flow will be affected by the specific
gravity of the fluid. Various fluids are used by the industry to more closely resemble the
characteristics of the fuel fluid that would actually be used. Lighter, less viscous fluids
are used for gasoline while heavier, more viscous fluid are used to test diesel injectors.

Fluid temperature
The test fluid standard temperature is 21oC +-2.0oC, measured at the injector inlet.
If some non-standard temperature is used, injector fuel mass flows must be
recomputed based on the change in density and viscosity due to temperature change.

Injector temperature
The injector temperature should be allowed to stabilize to the same temperature as
the test fluid 21oC+-2.0oC.

Fluid pressure
The standard test pressure is the differential pressure measured at the injector inlet
and in the manifold. The test pressure is determined by the application and is held to
within +-0.5% (kPa). Not all injector test flow data has been obtained at the same
test pressure. It is important to note the test pressure that was actually used and to
make necessary adjustment to compensate for the actual pressure that will be used
in the specific vehicle. Bosch uses a standard test pressure of 300 kPa unless other-
wise noted. Considering allowable pressure variance, a test pressure of 300 kPa
could actually range from 298.5 kPa to 301.5 kPa.

Injector mounting position
The injector shall be mounted vertically spraying downward. Actual engine installation
may utilize a different position angle.

Injector preconditioning
Injectors must be purged of all fluids and air for 10,000 pulses @5ms PW and 10ms P
using fresh fluid at standard test pressure. This comes out to pre-flushing the injector
for 2.5 minutes before testing. Used fluid should be discarded if contaminated.

Flow measurement
The flow rate may be measured in units of volume or mass. Mass is preferred in units
of g/s or mg/pulse. Dynamic flow measurement shall always precede static flow
measurement since it produces less injector solenoid heat build-up.

Injector warm-up
Prior to measurement, injectors shall be warmed-up for not less than 5000 pulses
at 5ms PW and 10ms P. This comes out to be 25 seconds of injector warm-up.

Coil resistance
Measurement of injector solenoid resistance permits the determination of the current
requirement for a given system voltage.

Typically, injector solenoid designs are either high or low resistance coil units.

High resistance coils (10-16 ohms) are used with a saturated circuit driver. The
driver switches battery voltage allowing current to be controlled by coil resistance.

Low resistance coils (1-4 ohms) use a current limiting driver. The driver allows a
peak current for rapid opening responses followed by a lower current level (the
holding current) for the remainder of the pulse.

Resistance measurement of the solenoid is performed at 21oC +-2.0oC with an
accuracy of +-0.01 ohms.

Inductance (mH)
Inductance of an electromagnetic solenoid is a function of the number of turns in the
coil and the permeance of the magnetic circuit. It is an indirect measure of the material
properties and geometry of the flux path.

Inductance, along with resistance, provides the time constant when the voltage is
known and can be used in predicting the initial current rise profile.

Inductance (in milliHenry) is measured with the injector in the closed position (at 21oC)
using a Wheatstone Bridge. The test frequency used is 1.0 kHz + 0.5% with a voltage
of 1.0V rms. The unit of measure is the milliHenry with a resolution of +-0.01mH.

Operating voltage range
With high resistance injectors, the minimum opening voltage is critical and may be
inadequate during cold cranking or low battery voltage. Low resistance injectors are
not as sensitive to lower voltage conditions since they use current limiting drivers.

Static minimum opening voltage (SMOV)
The static minimum opening voltage is the voltage (measured at the electrical
connector) required to open the injector solenoid. An accelerometer is used to
determine pintle motion using 0.1v steps every 20.0ms until the injector opens.

Dynamic minimum opening voltage (DMOV)
Dynamic minimum opening voltage is a measure of system response to low voltage
conditions. The type of circuit driver has a significant effect on injector opening
response.

My Bosch fuel injector data base currently includes 656 different part numbers for
fuel injectors. That does not include direct injection or diesel injectors.

I said over and over again, this dance is gonna be a drag.