AFPR: Fuel System Design - Q's for future sticky

GC89

1J-THIS
Jun 13, 2007
938
3
18
38
Spokane, WA
I am currently designing my fuel system and have some thoughts that I haven't seen discussed in the copious amount of previous FUEL and AFPR threads. This also is the reason I am hoping some type of Sticky can come out of this eventually to prevent future threads and because I know I am lazy and hate digging through subscriptions for info I found previously.......so

There are 3 major EFI fuel system designs

1. Returnless

The Fuel Pump output is controlled (usually by an ecu and pwm) and a sensor at or near the rail provides feedback. Fuel pressure is maintained fully by the fuel pumps duty cycle and adjusted to maintain the desired value observed by the sensor at the rail.

Pros: Simple system in terms of fuel lines/returns and regulator. No excess fuel heating or excess power wasted by the pump: fuel is supplied on Demand. Also less fuel evaporation for same reason. Claims as good or better fuel pressure consistency than a manual regulator.

Cons: Potential for Lag: While the sensor can sample fuel pressure extremely fast there is a theoretical potential for lag between when the PWM's duty cycle % is increased > pump speed reacting > pressure increased through the line to the rail. Corvette's and other OEM performance lines are using this technology but I am unsure of current forced induction applications.


2. Return Regulated

Factory Configuration*
The Fuel pump supplies fuel directly to the rail at one end. At the opposite end is a return line which goes to the regulator which effectively acts as a dam. The regulator maintains the desired pressure and bleeds off any excess via a return line which routes back to the tank. Can be used in conjunction with pump control to minimize the amount of work the pump is doing and consequently the amount of fuel being returned at idle and low load conditions. (Factory system uses a hi low load relay triggered by the ecu which forces the current through a .73 ohm resistor to limit pump speed.)

Pros: Max fuel delivery from the pump is provided straight to the rail. This design has the lowest potential for inadequate fuel supply due to a sudden increase such as low load > WOT & high boost.

Cons: Excessive fuel heating occurs as pump is running at a high output (epecially if supplied constant 12V) and heat generated is transferred directly to fuel in tank. Excess fuel supplied is further heated as it passes through the hot rail just off the top of the engine and back to the tank.

attachment.php


3. Supply Regulated

The Fuel Pump supplies fuel via supply line to the regulator which then bleeds off pressure above the desired setpoint. This set pressure is delivered to the rail. This design can also be used with some type of pump control to limit over supply conditions and excessive fuel being returned at idle and low load situations. Used in some OEM applications and sometimes with the regulator at the rear next to or mounted in the tank for a near returnless system.

Pros: No electronic lag issues like a returnless system due to the pump supplying a higher than needed volume to maintain the desired pressure. Reduced fuel heating compared to a return regulated system as the excess fuel that is returned does not pass through the rail.

Cons: Potentially less reactive design compared to return regulated as fuel is limited before reaching the rail. Rely's on regulator reaction to rapid increases in fuel demand.

attachment.php


Other Considerations

Reaction: Obviously this is very important, if your fuel pressure dips even slightly due to not being able to adjust to demand a lean condition can occur and destroy an engine.

Fuel Heating Something I don't see discussed much on this forum compared to others. Warmer fuel = less dense = leaner for given quantity. This is not so much a concern for a dedicated drag car with short runtimes but for a street car or autocross car which encounter long run periods it can be a serious issue. IJ for one has commented on encountering AFR's shifting leaner after periods of run time and not being able to pinpoint or explain it until he felt his tank after a long drive one day and it was HOT. AFRs shifted lean across the board could also lead to a lean condition under high load and if severe enough be just as dangerous as inadequate reaction.

FPD Fuel Pressure Dampener* This little part is often chucked but is included on virtually all OEM systems and for good reason. The best discussion I have seen can be found HERE

If built out for a sticky I would also include common info I have compiled including: fuel rail fitting sizes, return and supply line sizes, common regulators and controllers ect which comes up quite frequently
 

Attachments

  • return regulated.JPG
    return regulated.JPG
    21.5 KB · Views: 89
  • supply regulated.JPG
    supply regulated.JPG
    24.1 KB · Views: 91

GC89

1J-THIS
Jun 13, 2007
938
3
18
38
Spokane, WA
Please chime in if you have anything to add about my above statements or if I am blatantly wrong on anything please correct me. I am no expert just a SM regular who has built a few cars

My take on the above configurations.

1. Returnless - Optimum system if it could be engineered so that pump lag was not an issue. I'm Not an expert on available components or if this is possible, I have yet to see it implemented on a Supra or any other compatible forced induction application. Please link if you know of something.

2. Return Regulated - Proven setup, used by OEM as well as powerhouses such as Duane, Reign, Dr. Jones and others. Fuel Heating is a real issue as mentioned above and I am building primarily a road car so I would really like to avoid it. This is my last option but if I go with it I would use be using a fuel controller that has a base duty cycle % as well as a linear increase after a set input value base on a 0-5V map sensor signal.

This should help alleviate some heating by running the pump just slightly above what is needed for demand at all times. Should not have the lag issues a return-less system may have due to still having a higher volume supplied and bled off by the regulator. I am also considering leaving the FPD in place but placing it on the return line of the rail and using a banjo > -6AN fitting (credit to supraguy@aol). This should leave the benefits but reduce the risk of it being any type of a restriction at high flow.

3. Supply Regulated - This is the option I am most interested in exploring. I would still use a PWM fuel pump controller so the pump is not generating excess heat and returned fuel. I would also incorporate a dual feed system, with a line at both ends of the rail, one will maintain the FPD using the -6AN fitting. I believe the regulators reaction time concerns are not a real issue especially with dual feeds so that no injector at the end of the line as in the stock system. I believe this could be the best of both worlds being able to realistically keep up with demand and minimize fuel heating.

Similar to this diagram - to carb = to opposed ends of rail.
attachment.php


supraguy@aol's fitting solution

p2051722_1.jpg


Long Posts I know but I am hoping we can get some good discussion on the topic and build a definitive source of info for others.
 

Attachments

  • supply regulated dual.JPG
    supply regulated dual.JPG
    26.8 KB · Views: 85

GC89

1J-THIS
Jun 13, 2007
938
3
18
38
Spokane, WA
Yes, I have one in my oiling system. Didn't know they had them for fuel applications..........looks like many newer OEM applications use them. That could be the ticket to ensuring a return less or supply regulated system is 100% reliable. Thanks JJ
 

SeeUSmile

Member
Nov 22, 2012
249
0
16
Australia,Victoria,Melbourne
Ahhh been a while since i've seen an informative thread :D Cheers for the good info.

Lets see.... well gasoline has a low specific heat capacity compared to water.... so it would gladly take heat easily equally as it cools.

So as far as fuel temp is concerned...
-ambient temp is gonna be whatever nature sees fit which is beyond our control
-Fuel pressure will be constantly changing to whatever loads, vacuum, boost and driving style the engine is going to see and fuel pump switching 9-12V or should you decide to run a different setup.
So that would be the ever so changing factors

Then the fuel's two major sources for thermal conduction would be from obviously the engine itself and the fuel pump. Not sure or gathered any data as to much how much heat a fuel pump can generate if the motor is out of the equation but nevertheless it's still generates heats. So some form of insulating that source could help. I remember IJ sleeved his fuel pump in a foam block, not sure if it was more so for sound damping or both.

I also heard of inline fuel coolers but then the concern for the flow rate to temperature relationship with that added restriction.

Anywho loving the innovation, intertested to see where you'll end up going with your build.
 

GC89

1J-THIS
Jun 13, 2007
938
3
18
38
Spokane, WA
Thanks ^ I don't hear that much from you anymore JJ, I remember when it used to be the norm on SM :(

One hurdle I failed to mention with a return less system is the ability to use a 1:1 rising FP rate as needed in a forced induction application. Adapting any OEM FMU unit and sensors I have seen to date is out because of this. Most NA>T fuel pressure discussions I have found center around this issue and ultimately end up going returnless. The common statement I see across multiple platforms is 'A return system is how all of the big power racing guys do it'. I could see this changing in the future though. Aeromotive's FMU controller is the only one I have seen to date designed for a performance based returnless system including the 1:1 rising rate capability for boosted apps. It runs about 500 though.

Not terrible when you factor in regulator + another simple controllers cost but I have the reg already and having that much brain behind the fp makes me wonder if that money would be better spent jumping to stand alone now.

These are the two FP controllers I am considering if I go with a return setup: Easy Performance and FuelPro

A simple solution might be to go supply regulated and take a ford fp sensor and put it at the rail on a logging setup. Gradually test for pressure drops by increasing revs and then doing increasingly sharper boost transitions. If drops are found I could try putting an OEM bosch fuel accumulator inline between the rail and the regulator to cover split second pressure fluctuations. An accumulator may also eliminate the need for the FPD.

@SeeUSmile - Yes a cooler is another option, likely on the return size to eliminate any restriction, not sure how effective they are as I have no experience with them.

Anyone else have any useful input?
 

GC89

1J-THIS
Jun 13, 2007
938
3
18
38
Spokane, WA
Feed on the rail (fpd) is M14x1.5 return(fpr) is M12x1.25 I haven't figured out the threads for an an adapter off the stock supply line like driftmotion sells yet so if anyone knows....

SeeUSmile;2051875 said:
Sticking with the stock fuel rail for this one gc?

Yep, Duane's running a stock rail. Good enough for me.

I may drill out the return to a larger size and use it as the feed if I go return regulated though.