632C2

The following is a write-up on some fuel system tests that I performed on my 1967 542” tri-powered stock-appearing Corvette with cast iron exhaust manifolds. Back in November of 2011 I posted a thread about the car making 671 rwhp on a DynoJet chassis dyno. After reading a thread by Alan Rothman that was based on him being concerned about the fuel system on his 10 second 127+ mph ’71 Corvette, I decided to run some tests. I was concerned that my Holley 12-454-25 170 gph pump might not be up to the task.

Just for arguments sake let’s say that my motor produces 800 hp at the flywheel. When you multiply that number by a BSFC number of .45 lbs/hr(estimate) and then divide it by 6.2 lbs/gallon for gasoline you would come up with a fuel requirement of approximately 58 gph at WOT. I have heard that you should select a pump with a free flow rating that is at least twice what the motor can actually consume. This would mean that I would need a fuel pump with at least a free flow rating of 116 gph. The Holley I am using should be more than sufficient based on their published numbers. But I had to try to find out for myself.

The system I had in place at the beginning of these tests is described as follows: the 3/8” pickup at the gas tank is attached to a 3/8” ball valve which connects to the 3/8” fuel line (about 11 feet long). The fuel line connects to the inlet side of a Holley 12-454-25 170 gph mechanical fuel pump with a -8 Fragola Push-Lite fitting. Another -8 Fragola Push-Lite fitting is attached to the fuel pump outlet and from there it goes into the Tri-Power 3/8” hard line (about 3 feet long) up to the first of two fuel blocks. The 3/8" fuel lines go from the fuel blocks to the carburetor bowls. The factory sintered fuel filters are used in each carburetor bowl. No regulator is used with this fuel pump and the fuel pressure is pre-set by Holley to 7.5 psi. All hard lines are 100% stock.

I modified this somewhat to enable the testing while the motor was running – a requirement for a car with a mechanical fuel pump. I disconnected the 3/8” fuel line that leads up from the pump to the first fuel block. I attached one leg of a 3/8” ‘Y’ fitting to this line. Another leg of the ‘Y’ goes to the factory fuel block and the final leg goes via 3/8” rubber hose to a 3/8” ball valve that hangs under the car. I have this car on a 4-post hoist that connects to a DynoJet chassis dyno with all the testing going on underneath. From the ball valve I have a 6” piece of 3/8” ID hose that allows the fuel to be directed into a 1 gallon container. The ball valve gives me the on/off capability.

The first series of tests involved the Barry Grant style 1-gallon test. Essentially the BG 1 gallon test is a measurement of how much time it takes to fill a 1 gallon container with fuel as delivered to your carburetor(s). His site states that a 12 second car should fill the 1 gallon container in 35 seconds, 30 seconds for an 11 second car, and 25 seconds for a 10 second car. I ran the first 2 systems twice and the third was run 3 times. All tests were performed with the Holley 12-454-25 mechanical fuel pump. For #2 and #3 the existing fuel system up to the Holley pump was eliminated with the 12’ of hose/line coming directly out of a 5 gallon container and feeding into the Holley pump. All tests are run at 1000 rpm:

System 1: Original (as described above) fuel system = 38 seconds (95 gph)

System 2: 3/8” ID rubber gas hose appx.12’ long feeding from 5 gallon jug = 38 seconds (95 gph)

System 3: ½” ID rubber gas hose and ½”OD steel tubing from 5 gallon jug = 38 seconds (95 gph)

This was a real eye opener. This test confirmed Rick’s (540 Rat) statement to me about 3/8” OD fuel line not being a restriction. I contacted Rick with regards to this testing because of his engineering background. I would like to thank him for all of the invaluable information he provided me with. Since it took 38 seconds to fill the 1 gallon jug that equates to a flow rate of approximately 95 gph – and that is at idle. That is well over the 58 gph that my motor requires. I then ran this test at 3000 rpm and the times were identical.

I also performed a test similar to System #2 with 3/8" rubber hose and an Aeromotive 140 gph electric fuel pump that operates at 14 psi. With no regulator attached and the pump back by the 5 gallon jug, it took 25 seconds to fill the 1 gallon container which equals 144 gph. With their regulator installed at the front of the 12’ hose and set to 6.5 psi it took 35 seconds to fill the 1 gallon container which equals 103 gph. This test was run just for comparison sakes.

The next series of tests were performed as a “flowing fuel pressure” test. I use this “flow gauge” when I am setting up the nitrous on my 632 car. The device is pictured here: http://www.appliednitroustechnology....flowguage.html. It is used to set fuel pressure based on the real demands of your nitrous jetting. There is a jet that fits into the end by the gauge that is based on the jetting you are currently running. You turn on the fuel pump while the end of this device is flowing fuel into a container and you can accurately set your flowing fuel pressure. I thought this might be a good test of what is happening under WOT demands so I connected it to the end of the ball valve.

I calculated the area of each .110” needle and seat and multiplied that value by 3(Tri-Power, remember) then calculated the needed diameter to match that area as approximately .190”. I placed a .190” jet into the end of the flow gauge. I ran each of these tests multiple times with the results being very close and, unfortunately, very dismal. The fuel pressure before the flow gauge was opened up was around 7.5 psi with all three of the above outlined systems. When the flow gauge was opened up the pressure dropped on all three systems to under 1 psi. I would have to think that these numbers would be even worse when the car is making a quarter mile run.

Again I ran these tests at 3000 rpm and again the results were identical. Clearly this current fuel pump setup was not up to my expectations or requirements.

Rick had been insisting that I needed a fuel pump that provides at least 14 psi instead of the 7.5 psi that mine had. I would then need a regulator to dial it back for the carburetors. I resisted because I absolutely have to have this thing looking stock when you open the hood. A regulator up by the carburetors will not work. After the above testing results I realized I had to make some modifications. I purchased a Holley 12-704 regulator and decided to try to shim the spring in the Holley fuel pump to increase the pressure. The regulator and all the extra plumbing is hidden under the car and can not be seen from the engine compartment. While I am sure that it would be better to have the regulator closer to the carburetors it is simply not going to happen in my case.

In order to verify my next rounds of testing I installed a Racepak Ultra Dash with data acquisition capabilities. I am now able to monitor and record data from the following sensors: rpm, voltage, oil pressure, oil temperature, water temperature, fuel pressure prior to the regulator, and fuel pressure at the carburetor bowl.

I disassembled the pump and machined 3 spacers (.100”, .150”, .200”) that would allow me to shim the pressure spring. Using a valve spring tester I recorded the spring pressure at two different spring heights for the spring with no shims and for the spring with each of the 3 different sizes. I wasn’t sure what thickness of shim I could get away with without risking damage to the pump so I decided to install the .150” first. While the pump was disassembled I also internally ported the inlet and outlet housings hoping to increase the flow a little.

These changes resulted in a fuel pressure prior to the regulator of 10.5 psi with fuel pressure at the carburetor bowl adjusted to 6.0 psi. I wanted at least 14.0 psi prior to the regulator. Since this .150” shim only increased the pressure by 3.0 psi, I didn’t want to spend any more time testing the other shims. At this point I decided to eliminate the idea of using shims and ordered Holley’s high pressure spring assembly.

Once this diaphragm assembly arrived I installed it in the pump and was rewarded with much higher fuel pressure. At 900 rpm the fuel pressure prior to the regulator was 16.0 psi while the fuel pressure at the carburetor bowl was adjusted to 6.5 psi. At 2000 rpm the fuel pressure prior to the regulator jumped to 17.0 psi while the fuel pressure at the carburetor bowl remained at 6.5 psi. Now I felt I was getting somewhere.

I ran the BG 1 gallon tests with the motor at 900 rpm and at 2000 rpm. I also decided to run the BG 1 gallon test with the flowing fuel pressure assembly attached – a test I had never tried before. Attaching that flowing fuel pressure assembly to the end of the hose would cause a significant restriction and should result in a longer time to fill the 1 gallon container. The following is a summary of the results:

........................................ ........................................ .900 rpm……………...2000 rpm

BG test with 6" of -6AN hose after ball valve:.....45sec / 80 gph............34sec / 106 gph

BG test with .190" jet after ball valve:..................50sec / 72 gph...........37sec / 97 gph

Again, I was very happy with these results. At 2000 rpm the time to fill the 1 gallon container was now the same as the electric pump. When the .190” jet was added to the equation it only took 3 seconds more to fill that same 1 gallon container. This time increasing the rpm had a definite and positive impact.

The final set of tests were based on the flowing fuel pressure assembly. The following are the results :

.................................... 900 rpm....... 2000 rpm

Pressure after pump:.....4.5 psi.............5.5 psi

Pressure at carb bowl:...2.0 psi.............3.5 psi

The increase in rpm again had a positive effect especially on the pressure at the carburetor bowl.

Do these tests have any real-world validity? I sure hope so but the final proof will come when I take the car out to the drag strip.

Steve

4 Speed Dave

Yes for a street type of application, however the same system in a hard launching wheels up drag car you will find its limitations being the long suction line and the ability to pull the required fuel to feed the engine under acceleration forces.

Looking at a pure steady state fuel requirement say a dyno room a moderate mechanical pump works fine. However in a real world racing application with outside g forces working against the pump this normally increases the requirements for a larger pump with a short suction line and a pressure feed from the back up to the engine bay.

Neat data though.


Dave

632C2

Thanks Dave. I agree with what you are saying and will be quite interested to see how this works out for my deal. One thing I don't have to worry about is having "a hard launching wheels up drag car". Just a streeter with some drag radials.

Steve

427Hotrod

Uh huh........

Wow Steve...great stuff! When you showed me what you were working on (hey guys....a month or so ago I was standing under this Vette looking at all of this!) I never dreamed you were going to come up with all of this. No..that's a lie....I KNEW you would do like usual and follow it to a good ending through logic and work. Really cool stuff!!

Rick's been telling us this for years. For some reason he's convinced all that math and science and engineering stuff actually works on this Hot Rod stuff too! And it does!

I think Rick's favorite line is something likec...*One test blows away a whole lot of theory*!!

Thanks for posting!

JIM

PS-Now it's time to go get the 1/4 panels nasty at the track and get thrown out!

632C2

Jim, I am getting exhausted responding to all of these questions.

Steve

427Hotrod

Remember...you always need to preface a good thread with something from this list:

And on and on. These topics will guarantee at least 1000 views and 800 posts within 24 hours!

See ya,

JIM

540 RAT

Steve, your work is always “VERY” impressive, among the best I’ve seen. I’ll be looking forward to your write-up after track day.

540 RAT

Yeah, my exact quote is, “A single test can ruin a perfectly good theory”. I’ve seen that happen many times, in Industry and in our Hotrod and Racing hobby. An actual test is the “real deal”, even if it goes against what the common “thinking” is.

Rick