Lift pumps and EFI surge tanks

[CDMBill]

At the risk go being repetitive I'm looking for the collective BS wisdom on surge tank lift pumps. I'm fabricating a tank within a stock tank setup, see my project cars thread under GIRTST, and I've ordered a Carter OEM'S replacement pump to pull gas from the main tank into the Aeromotive sump/surge tank which will feed the A1000.

Is there a better alternative? I'm not running a true fuel cell due to cost, nor a true surge tank as prefer to keep the fuel tanks isolated from the cabin as this is in a convertible and I have a sold steel trunk floor which the fuel tank rides under unlike earlier Mustangs.

[hauen]

I think with a minimum of baffling you could just run an inline pump, that sump seems plenty deep. At the risk of sounding like a total 'tard I think you could make a louvered lid/cap/baffle with the louver openings facing forward that just sits over the top of the sump so fuel flows down and in so when you pick the front end up the fuel has to really work to climb out. With the troubles I've been having lately, I'm probably the last guy who should be giving fuel system advice though.

[OldMachinist]

I just did this and the instructions say drill a 3/8 hole on each side of the sump at the bottom of the stock tank. I actually drilled 2 holes on each side about 1 1/2 " from the ends. Don't you think this will supply the sump adequately? Or am I not understanding something here?

[TheSilverBuick]

Am I reading this right, you are setting a pump up to pull from the main tank and pumping into the sump area in the main tank? Sounds overly complicated and an extra strain on your charging system to me. Baffle and those one way flaps I would think would be adequate with enough fuel in the tank. Do you plan on being less than a 1/4 tank at WOT? I know it's extra weight, but I'd figure at at around 1/2 tank with a sump like the one I've seen in your pictures that uncovering the pickup point would be near impossible when accelerating forward. Like hausen though, I may not be the best source of fueling advice

[CDMBill]

So Randall you are right in that I'm trying to run with the least amount of fuel on board as possible. With holes drilled more or less as Oldmachinist describes, which reflect the standard recomendation from Aeromotive I'd still have to have 1/4 to 1/3 of the tank full at least to avoid starving the pump inlet on acceleration. It's deceleration that is trickier with low fuel levels even with the deeper, much bigger sump than I have now. Deccel at the drags happens to a minor degree at the end of a burnout and to a very major degree at the end of a run and the turn out. I experience this now at any level below about 2/3's of a tank so I have been running it full (20 gal.)

The other issue is for auto-x and road course which was why they recomended the one way valves on the front and sides, and which has led me to the lift pump idea so I can have the sump be quite full even with the main tank almost empty. In addition to the one way valves they suggested adding baffles which would divide the tank into thirds with a tapered baffle extending from each of the corners of the sump out towards the front corners of the tank. Part of the issue there is that the side walls of the sump lean in so the one way valves would be slightly open under nuetral conditions and I would end up with identical fuel levels in and out of the tank unless I reconfigure the sides of the baffles. The front side is angled out and presentes no issue other than deciding on the height of the location

I was planning to add top baffles to the sump to better contain slosh, but still allow the fuel to be added through the orginal filler neck and to spill over filling the rest of tank when its time to hit the road. I haven't started cutting yet so all options are open.

The power required by the little lift pump is tiny compared to what the main pump draws and that is dwarfed by the electric fan. All of which is fed by a 200 amp alternator, so I am good on power.

Lastly, I want to be able to mix 100 octane unleaded into a small amount of the pump premium in the event we start getting E10 or worse 'summer' blends at some of the gas stations as I am at the edge of the pump gas window with 11.86:1 compression. So far so good with Cali Chevron 91, but on the road you never know what you may get. So using the sump as a kind of surge tank for a particluar run makes that easier.

This is how I think myself in circles.

[TheSilverBuick]

IMO, if you are going to a surge tank set up, then go all out on a surge tank set up. Box in the whole sump area except a small hole on the very top that might be 1.5-2.0" in diameter for filling up at the gas station. No side or top baffling or flaps, just adds complication and ways for things to go wrong. Let the Carter pump (it's an intank pump right?) pump into it full time and an A1000 it from there. The Carter pump doesn't have to keep up with the A1000 on volume either, just has to be close enough to your expected fuel consumption rate while auto-crossing. Works best if your EFI return line goes into your surge tank as well.

Mine is a little under 1 quart in volume I figure, and the mechanical pump rate feeding it is a bit less than the electric pump feeding the injectors. By my math it would take nearly 12 minutes of WOT at maximum electric pump capacity for my surge tank to run dry with the mechanical pump pumping into it at it's peak rate. Now based on my actual fuel consumption (injector flow rates and duty cycle) the mechanical pump just barely lags and may go forever without starving or like an hour, but in either case, not even a Bonneville 10 mile course run would run the <1qt surge tank dry. If your's is a few gallons you'd be in great shape I'm sure.

[CDMBill]

Randall, that's the conclusion I'd reached as well, but I wanted to bounce it off the folks here to see what I was missing. The lift pump adds another point of failure which is whay I was equivocating on the baffles and one way valves which are pretty simple. I have a spare A-1000 that is for the N20 setup I'm not running at Drag Week and the little semi-generic Carters (yes its an in tank unit) are available everywhere. The EFI return is to the sump/surge tank so that is covered.

[TheSilverBuick]

I had repeated failures of cheapy universal electric pumps(12psi TBI versions), so went to a mechanical fuel lift pump. Might of been my installation (not convinced of it, but could be) or from using bottom shelf parts, I don't know, but about once a year I was getting stranded because it would quit pumping, and popping a new one on there would work every time.

You run a fuel pressure gauge so should be able to tell pretty easily when there is a hiccup, plus you could put a couple gallons in the fill spout (if it goes right into the surge part) to go several miles on the surge tank if the lift pump fails. I'm contemplating installing one that feeds into the MSIII.

**Maybe a small ball style check valve at the bottom of the surge tank that allows tank fuel into the sump when the fuel level goes lower than the tank level for driving purposes and if the carter pump fails. But my K.I.S.S. feelings is not to use flaps and such that would be more prone to seeping/leaking back into the main tank when parked and pumps off. Keep the surge tank always topped off.

I've ran my car down to within a gallon of the specified capacity of the tank (and where I have ran out years ago) with no hiccups or mis-firing of the engine.

[Matt Cramer]

I don't like the idea of using an OEM Carter fuel pump to feed an Aeromotive A1000 unless you're using three or four of the Carters - it seems to me like the Carter pump is going to be a restriction instead of helping. And if you're using multiple Carters, you can just skip the A1000. Or depending on your horsepower level, you may be able to use a single in-tank Walbro.

[TheSilverBuick]

I don't like the idea of using an OEM Carter fuel pump to feed an Aeromotive A1000 unless you're using three or four of the Carters - it seems to me like the Carter pump is going to be a restriction instead of helping. And if you're using multiple Carters, you can just skip the A1000. Or depending on your horsepower level, you may be able to use a single in-tank Walbro.

Have you seen the size of surge tank he is running? The dang thing looks like a 5-gallon gas can! Well maybe 2 gallons anyways. The Carter just has to keep up at normal cruising since the A1000 EFI return line goes back into the surge portion. At WOT and the Carter could be completely turned off and Bill will have what appears to be a couple gallons of fuel for the A1000 to pull from before there is trouble. I know he isn't at WOT very long on the Autocrossing courses, not with 1,000 ponies on tap. Some simple flow rate math would be able to tell for certain though.

[BKBridges]

My personal experience is that check valves leak, fuel pumps fail, and overly complex systems create maintenance issues. In-tank pumps require a lot of grief to service in most cases. An external surge tank with the return dumping into it will accelerate the fuel heating and subsequent hp reduction (-1 hp/+1degree?) If youve got to run a surge tank, IMO just buy a fully engineered one new and cob it into the fuel tank. Two Walbro pumps mounted side by side in the tank will supply what ever you need. The 98 Viper fuel pump/surge tanks work well and can fit two 255lph pumps (or the new same package size larger capacity Aeromotive in-tank pumps). In tank pumps MUST be covered with fuel all the time as they have no way to cool themselves without liquid contact, necessitating a particular minimum fuel level. I run two in-line stageable Walbros in parralel mounted to the floor pan at the bottom level of my tank. The tank itself has some baffling around the draw tube that reduces sloshing during 3 axis acceleration. I use the largest "sock" filter on the "draw tube" I can find as they also work as a one way fuel reservoir. A large volume fuel filter is also a good concept for reducing the chance of a "hiccup" due to acceleration forces. Slalom and drag racing with an 1/8th tank has been no problem, although with only 300hp my launch is a little soft at the strip. Just saying...
BKB

[Bob Holmes]

Matt, would the fact that the Carter is supplying fuel with no head pressure to push against change your opinion?

Our roadrace car runs an ATL drop in fuel trap; we've not had any issues with fuel pickup: http://www.google.com/url?sa=t&rct=j...k08dsKk6-RknIg

I find this company's solutions very interesting: http://www.intengineering.com/integr...ith-pumps.html

[CDMBill]

Matt the engine makes ~1000 hp. We just updated it to an MS3X BTW. The car will get a real fuel cell set later this year, but in the mean time the stock tank with the much larger Aeromtive sump is the plan. In drag race mode I think the 110 lph Carter is enough to feed the Aeromotive sump/surge tank. If not the form factor is similar with a 255 Walbro. I've decided to add some fuel cell foam and baffles to keep the little pump wet and control slosh a bit better. I'm also adding a fuel temp sensor which I'll be logging so I can test the fuel beating/heating issue.

I experimented last year with a small separate 5 gallon fuel cell, which was inconclusive as we had screwed the tune and the chassis setup badly and couldn't tell what caused what, but we lost hook and 6 MPH before the DNF.

That setup is now a dedicated N2O system which isn't on the car at the moment.

[Matt Cramer]

It should work with a 5 gallon surge tank in a drag application. I don't particularly like it, but there's a difference between doing things right and making it work, and each one has its place and time.

[CDMBill]

So between the last time I updated this and now we went to Pomona for the NMCA race and saw some fuel pressure anomolies once I was able to datalog the 0-5v output from the AutoMeter electric gauges sensor.

The top four traces show RPM (white), staged at about 2900 then when the throttle is mashed (green) the RPM rises, the AFR's (yellow and red) briefly jump and then level out except for a small bump at the shift.

The next are a repeat of RPM (red) Pulsewidth (tan) coolant temp (yellow) and battery voltage (green) Of note at the hit pulsewidth rises and gets wiggly as the along with the RPM trace as the converter engages at around 5700.

The thrid four are Manifold Absolute Pressure MAP (purple) which jumps from around 60 KPA staged to 96 with all eight barrels open; Mnaifold Air Temp MAT which gradually falls as cooleer air enters the fornt facing scoop down track (ambient was around 100*) and Cankcase KPA which rises form 12 to 49.

Last is the Spark Advance (yellow) which is a flat 26 form 5500 to peak, two MPH traces front tire (white) and rear (red) and the star of our show the the fule pressure trace (green) which falls away form the staged rail pressure of 44 to the crazy squiggle that floats down to a range of 25-38 until the run ends and then it sanps back to 44 (not shown as this snap shot only show first and part of second gear)

Before this when we saw the fuel pressure trace from prior testing in Tucson I had re routed the inlet plumping to eliminate a .5" valve and a slightly kinked line, lowered the filter and pump assembly to assure gravity feed, replaced the A1000 Marine spec pump with a new standard A1000, and rewired the fuel pressure sensor to ensure it was on the clean power and grounds as the other sensors. Nothing changed in the fuel pressure trace. The cars performance responded to fuel table changes in MPH, and launch RPM for ET as there has been a persistent stumble on the launch as I try to get the accel tables working as well as they had with the EMS-Pro MS1 tune. Power braking helped get around the stumble but the 60 ft. times weren't great as it killed the chassis reaction.

Ideas, thoughts? No carb swaps please.