Engine tuning theory thread #2 - Variable fuel pressure

[Extinct]

OK, this is an easy one.

Why did GM vary the fuel pressure instead of just running a single (max) pressure ?

Was it just to reduce the parasitic load on the engine ?

[CHEVY052500HD]

possibly to keep the fuel at a lower temp? Then compressing it to max and making more heat then needed, and more parisitic load.

[IdahoRob]

Go into the scan tool and DVT and raise pressure at idle. Fuel pressure is a major player.

Turn on the garden hose with a varible spray nozzle set on mist. Now vary the pressure via the petcock. What happens to water volume and spray pattern?

[bonescarolfi]

Im no expert at this but having the right pressure helps with proper fuel atomization and two the higher or lower the pressure the less or more time it takes to inject x amount of fuel so having variable fuel pressure is very useful in controlling diesel engines.

[Extinct]

OK, let's go at this from a little more theoretical perspective, ok ?

Regarding the hose post, of course the flow rate and pattern change with an increase in pressure. That isn't the question. The question is why would GM choose to use pressure as a controller rather than pulse width.

Most IC engines do a more efficient job of burning fuel that is atomized in to finer droplets. Lower pressure does not help this, it typically hurts. That being said, some knowledgeable fuel injection specialist and engineers on this board have stated that the nozzles in our engines are designed in a manner to deliver highly atomized sprays in the design presssure range.

If you go back to the water hose scenario, the point has been made in previous threads that our fuel pressure range is such that the pressure doesn't make a material impact on atomization or pattern. Regardless of if that is true or not, the real question is why would GM want to do that ?

We also know that the fuel pressure change will change the flow rate, so a higher pressure will take less time to deliver an equivalent amount of fuel.

So, what are some possible reasons GM would want to do this (looking for answers here from the experts):

1. Spread the fuel burn over a longer time period, piston travel, and crank angle so that:

a. Increase power ?
b. Increase fuel efficiency ?
c. Decrease engine noise ?
d. Decrease emmissions ?

Let's look at each one of these in turn see if this makes any sense ?

a. Increase power.

At max pressure and flow, the burn will take place over a narrower time period, increasing the rate of pressure rise in the cylinder. We can see that GM cranks up the pressure at higher rpm and fuel flow where more torque and power are desired (presumably because there is less time to get the pressure rise in), but why would they lower the pressure at lower flows and rpm ? It would seem to me that a quicker pressure rise would increase efficiency by reducing the resisting pressure of too early a pressure rise and reducing the amount of pressure rise that occurs after optimum crank angle/piston position ? Based on this, it would seem that the goal of varying (lowering from maximum) injection pressure would not be to maximize power per injected amount of fuel ? Perhaps a longer burn time increases power and efficiency from a given amount of fuel by allowing more time for complete combustion and exposing the fuel to more cylinder swirl ?

Discuss...

b. Increase fuel efficiency

One could reasonably argue that by injecting the fuel pulse over a longer time period one would obtain a more complete and efficient burn as discussed in the last line of a, could that be what GM had in mind ?

c. Decrease engine noise

Diesel and gas engines rattle because of pressure shock waves vibrating the pistons and cylinders. Part of the Duramax quiet technology is the pilot injection pulse which tends to help smooth out the pressure waves of combustion. Could GM have desired to smooth out these pressure waves even more by varying fuel pressure so that the burn takes place over a longer time period, creating more of a controlled burn than an explosion ? again, as stated in A however, this would seem counter to max efficiency because some efficiency might be lost to crank angle. Perhaps the efficiency vs noise tradeoff is worth it however ?

d. Decrease emmissions

This last one would seem to make sense, as longer burn times across more crank angles and piston positions would increase the completeness of the burn, reducing unburnt hydrocarbons and CO emmisions.

If we look closely at the stock fuel pressure map, we can see that the pressure curve isn't very linear, in fact if you think about the likely cells that would be used in any acceleration or steady cruise events, you can see that a curve with three distinct regions would occur.

1. A relatively low slope region of lower pressure around idle and low load, low rpm
2. A steep slope (almost straight up) region of increasing load and rpm.
3. Another low slope region near max pressure at high loads and rpms. There are some interesting variations here as past a certain point pressures are actually decreased. Why does this happen ?

Now, hopefully we can have a more productive discussion. I guess this question wasn't so easy after all

[FMOS]

I think you pretty much answered your own question.

Firstly, yes, injectors designed for common rail diesel pickups are designed to operate over varying parameters. Fuel atomization will be sufficient over the range.

I would say that you definetly want lower fuel pressure at low engine speeds and load ratings. The biggest factors that GM... Izuzu wanted to play here was probably clean combustion, environental factors and noise. Keeping the pressure lower with a longer pulse width will allow the fuel to burn over several degrees of crank turn... versus one short shot close to TDC.

You can imagine the noise it would make if it injected the same amount of fuel over a shorter period (how about pilot injection... isn't this situation what it was designed to help with). And think of how much more quickly the combustion reaction needs to take place, complicated by the fact that cylinder pressures are not as high at low speed/load, less cylinder heat... To me it sounds like it would all add up to be pretty inefficient.

Sounds like were talking about 12/24v dodges here!

I dont want to shut down the discussion but I think the answer is much more simple than you might be thinking. Reduce noise, increase efficiency. Hopefully someone can prove me wrong though, I often think things are more simple than they really are.

[Extinct]

I started out thinking the answer was pretty simple also, then the first few responses I got were not really delving in to the theory of things, so I tried to turn the discussion more toward the operating theory.

I don't have a deep background in tuning Diesels, but I do have BS and 20 years exp. in Mech. Eng. plus probably another 7 on top of that tuning engines back when they were controlled with a carburetor and distributor.

I like to really start by understanding the theory behind why things are built they way they are before you go down the path of changing/optimizing them. I know there are a lot people on this board that have a lot more experience and expertise with diesels than me and I am just looking to start a good learning discussion and get confirmation or correction to what might be my crazy ideas and perspective.

Thanks for your post !

[vortecfcar]

Originally Posted by Extinct

OK, let's go at this from a little more theoretical perspective, ok ?

Regarding the hose post, of course the flow rate and pattern change with an increase in pressure. That isn't the question. The question is why would GM choose to use pressure as a controller rather than pulse width.

Most IC engines do a more efficient job of burning fuel that is atomized in to finer droplets. Lower pressure does not help this, it typically hurts. That being said, some knowledgeable fuel injection specialist and engineers on this board have stated that the nozzles in our engines are designed in a manner to deliver highly atomized sprays in the design presssure range.

If you go back to the water hose scenario, the point has been made in previous threads that our fuel pressure range is such that the pressure doesn't make a material impact on atomization or pattern. Regardless of if that is true or not, the real question is why would GM want to do that ?

We also know that the fuel pressure change will change the flow rate, so a higher pressure will take less time to deliver an equivalent amount of fuel.

So, what are some possible reasons GM would want to do this (looking for answers here from the experts):

1. Spread the fuel burn over a longer time period, piston travel, and crank angle so that:

a. Increase power ? Not a major concern at this level

b. Increase fuel efficiency ? Not a major concern

c. Decrease engine noise ? The longest pulse/lowest pressure will offer the lowest engine noise, assuming stockish timing

d. Decrease emmissions ? This is the huge one. NOx emissions are closely related to heat of combustion, which is closely related to cylinder pressure, which (you guessed it) rises with rail pressure.

Let's look at each one of these in turn see if this makes any sense ?

a. Increase power.

At max pressure and flow, the burn will take place over a narrower time period, increasing the rate of pressure rise in the cylinder. We can see that GM cranks up the pressure at higher rpm and fuel flow where more torque and power are desired (presumably because there is less time to get the pressure rise in), but why would they lower the pressure at lower flows and rpm ? It would seem to me that a quicker pressure rise would increase efficiency by reducing the resisting pressure of too early a pressure rise and reducing the amount of pressure rise that occurs after optimum crank angle/piston position ? Based on this, it would seem that the goal of varying (lowering from maximum) injection pressure would not be to maximize power per injected amount of fuel ? Perhaps a longer burn time increases power and efficiency from a given amount of fuel by allowing more time for complete combustion and exposing the fuel to more cylinder swirl ?

Discuss...

b. Increase fuel efficiency

One could reasonably argue that by injecting the fuel pulse over a longer time period one would obtain a more complete and efficient burn as discussed in the last line of a, could that be what GM had in mind ?

c. Decrease engine noise

Diesel and gas engines rattle because of pressure shock waves vibrating the pistons and cylinders. Part of the Duramax quiet technology is the pilot injection pulse which tends to help smooth out the pressure waves of combustion. Could GM have desired to smooth out these pressure waves even more by varying fuel pressure so that the burn takes place over a longer time period, creating more of a controlled burn than an explosion ? again, as stated in A however, this would seem counter to max efficiency because some efficiency might be lost to crank angle. Perhaps the efficiency vs noise tradeoff is worth it however ?

d. Decrease emmissions

This last one would seem to make sense, as longer burn times across more crank angles and piston positions would increase the completeness of the burn, reducing unburnt hydrocarbons and CO emmisions.

If we look closely at the stock fuel pressure map, we can see that the pressure curve isn't very linear, in fact if you think about the likely cells that would be used in any acceleration or steady cruise events, you can see that a curve with three distinct regions would occur.

1. A relatively low slope region of lower pressure around idle and low load, low rpm. This region is at low pressure to allow for smooth idle. Changes in fuel rate as small as a couple mm3 will bounce the idle. High pressure drastically lowers the amount of injector on-time variance tolerated by a smooth idling engine. Engine controller processing rates and hardware precision is less critical at low pressure.
   
2. A steep slope (almost straight up) region of increasing load and rpm Make note of the area unaffected by the steep slope. The <40mm3 and <2000 RPM range of steady cruise. I'll go out on a limb here and speculate that this is the area where cruise NOx emissions required some extra dialing in by the GM calibration engineers.
3. Another low slope region near max pressure at high loads and rpms. There are some interesting variations here as past a certain point pressures are actually decreased. Why does this happen ? Fuel rate is decreased as well <- this is your answer

Now, hopefully we can have a more productive discussion. I guess this question wasn't so easy after all

My thoughts are in blue.

Nick

[scramHD]

Thanks fellas....

[McRat]

Fuel evaporates and draws the heat out of the air. Only evaporated fuel burns, liquids don't burn. Superhigh pressures make smaller droplets, which evaporate faster.

If you spray the fuel in faster than air temp rises, the fuel only partially burns, then fuel sticks to the piston and cylinder, which makes it harder to burn. As the heat climbs from the fuel that did burn, suddenly a bunch of it all ignites at once. You hear a "knock", and the fuel that was stuck on walls only partially burns, and you get black smoke.

What would be nice would be a variable injector, that could increase and decrease it's size as needed, but that tech isn't in our engines. So we reduce the pressure and/or use a pilot shot to mimic a smaller injector.