Corvair DiagramCorvair Photo
Corvair Center
home forum corvairs calendar links Corvair Podcast
California Corvairs
Clarks Corvair
Clarks Corvair
“CORSA"



Chevy Corvair License Plate
Chevy Corvair Chrome Wheel
Corvair Center Forum :  Corvair Center Phorum The fastest message board... ever.
Corvair Center 
Pages: 12345Next
Current Page: 1 of 5
Mechanical Fuel Pump Observations...
Posted by: kenzen ()
Date: February 28, 2018 06:23PM

...because no horse is too dead to beat, and the thought of spending time and money to convert to an electric pump to replace something that is inherently simple and should be reliable just pisses me off...

So I've read just about every thread on mechanical fuel pump failures on this forum, and now I'm compelled to kind of consolidate them, and start some analysis and experimentation on my 2 failed pumps, to hopefully "get us well" and not necessarily jump to the electric pump conversions out of fear of the mechanical ones.

Warning: Slightly OCD multidisciplinary engineer with hyperactive imagination - continue at your own risk.

Based on my reading and initial thoughts:
1. Although I have gaskets that are pretty dried up, I don't believe wide-spread pump failures are from fuel quality or content problems. If it was, its an easy fix with a change in materials.

2. Many are observing significantly high discharge pressures. This is not normally attributed to reliability problems in the forum threads, and rather it is discussed in context of stressing the carburetor needle valves, causing the float and valve to become ineffective in controlling the bowl level, witnessed by venturi clusters dripping fuel while running. Recommendations to cut coils, or "get an old spring" are frequently cited on the forum to correct this problem. The discharge pressure is controlled by the spring strength, making this a quality issue of out-of-spec springs.

3. The quality of the gasket surfaces on the pump body components is poor since they are not machine surfaced - they are used "right out of the cast". This lack of flatness is an inherent source of internal leaking, to include getting access to the screw holes that provide porting across the pump layers.

4. The screws holding the body of the pump together tend to loosen, resulting in inadequate pressure for proper gasket sealing. Many have discussed work-arounds for this, including using locktite, using longer screws and putting nuts on the now-exposed screw ends, and I believe someone went so far as to put lock wire through holes drilled in the screw heads. This isn't necessarily a design flaw, but the lack of a more positive approach during assembly to address this common problem should be addressed with a "TSB" to do something more than rely on just friction.

5. Separate from the modern fuel effects, the reliability of the diaphragm and gasket materials with respect to mechanical wear are also discussed in many forum threads. It has been identified that cheaper materials that did not have fabric reinforcement have been used in the past. Clarks page 65 briefly discusses this. If we can assume that the currently used fabric reinforced materials is "the standard" and adequate for acceptable reliability, and therefore disregard the run of pumps from the early 90's that didn't have them, we can consider this choice of material no longer a reliability problem.

6. The failure of the check valves to stay in place, or failure of the valve flap material is also a cited source of pump failure. The failure of the valve to stay in place is simply an assembly quality problem due to inadequate staking of the valves into the pump body. This is easily detected by inspection, and I'd recommend should be inspected before a new pump is installed.

Taking all of the above into consideration, I took some time this evening to do a little analysis of one of my pumps. This pump was leaking out of the top of the cap through the hole.

First I wanted to address the spring strength, and ideally come to some guidance on what an in-spec spring tension should be. Unfortunately I don't have a gauge that reads accurately at low fuel pressures at this time, but I'm ordering one to follow-up this analysis later. I did have a second spring on hand from another pump, so I decided to compare the two to see if there's a big variance. When the spring is installed, it is compressed to a length of about 7/8", which I used as the measurement point for the "static" spring force. Using a kitchen scale I compressed both springs to the 7/8" height. One measured 8.5#, and the other was off the full scale of 11# and I'll estimate it had around 13# of force. I'll also note that I was recently running on the 8.5# spring and was witnessing dripping from my right carb venturi cluster. The carb was just rebuilt and I don't suspect that the needle valve is dirty or the float improperly adjusted.

Noting the spring tensions measured above, I measured the diameter of the pump diaphragm at 2", which makes its surface area ~3.14 square inches. If we ignore the stiffness of the diaphragm, this means the internal pressure in the pump chamber is up to and exceeding 30# possibly 40#(!!!), which also means the dynamic pressures are higher. Since the force of the pushrod on the diaphragm (which draws the fuel into the pump) is practically infinite and easily pushes the diaphragm to its full "inhale" position, we can also assert that the diaphragm relaxing is contributing to the spring force as well on "exhale".

I'm assuming that both the springs measured above where producing grossly over the design discharge pressure spec (2-4.5# I believe). Taking these calculated chamber pressures as "real" (which assumes the check valves can hold this kind of pressure), we now need to consider the quality of the materials, the gasket seating surfaces and screw loosening problems noted above. I'll assert here that these pressures are possibly an order of magnitude above the design pressure, and greatly increase the probability that fuel is being pushed (or squeezed) past the sealing surfaces, stressing the gaskets to failure, and causing leaks (to include dumping fuel down into the oilpan).

From here I'll make this hypothesis: Of all the forum threads I've read (and relying on my memory of this reading), there has no been significant discussion that correlated the high pressure problems from excessive spring strength, to pump leaking and reliability problems. I'll assert the root cause of the reliability problems is the excessive spring tensions, and this can happen with brand new pumps due to the poor seating surfaces and inadequate screw tensions. Poor quality gasket materials only make the matter worse.

For those of you who made it this far into this rant, thanks for your time. I hope this is useful. I'm going to do some spring measurements compared to fuel pressures to determine what an "in spec" spring tension is when I get a gauge and time.

kenzen
66 Monza Coupe 110/PG
Bel Air, MD

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: Frank DuVal ()
Date: February 28, 2018 07:17PM

I am interested in these findings!thumbs up

Frank DuVal

Fredericksburg, VA

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: American Mel ()
Date: February 28, 2018 08:04PM

An EPIC post to be rivaled only by J3M !
Only this one was something of interest to many of us.
Keep reporting your findings.
I will be reading them.
Unlike certain drivel about E.F.I..

.
-----------------------------------------------
WA. state, 1 mile south of the Canadian border,
I am not at the end of the world, but you can see it from here.
'66 Monza Coupe - 4spd, 140 Daily driver beater
'67 Monza Vert. - PG, 140 Daily driver beater

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: krfjkm ()
Date: February 28, 2018 08:08PM

I like where you are going with this analysis--- could someone extrapolate from your theory that if someone replaced the new spring with an old spring their fuel pump "should" be problem free?

KRFJKM
Charlotte NC
1965 Corvair Corsa

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: MattNall ()
Date: February 28, 2018 08:12PM

NO!!! But pressure would be lower!! "possibly" solving a problem PERMANENTLY!! gg





MODERATOR
Sea Mountain, between Charleston Harbor and Coos Bay! SW Oregon Coast
Click HERE for My Website...Click HERE for My TechPages!
..............................110-PG.................................................Webered-Turbo

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: kenzen ()
Date: February 28, 2018 08:28PM

krfjkm Wrote:
-------------------------------------------------------
> I like where you are going with this analysis---
> could someone extrapolate from your theory that if
> someone replaced the new spring with an old spring
> their fuel pump "should" be problem free?


Assuming the "old spring" was the right tension. The golden question is "what is the correct spring tension to achieve in-spec fuel pressure?" And as Matt stated above, if all these new pumps have excessive spring tensions and are eating themselves due to excessive internal pressures, getting a correct spring may put the high pump pressure AND premature failure discussion to bed.

kenzen
66 Monza Coupe 110/PG
Bel Air, MD

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: REM654DPG ()
Date: February 28, 2018 08:47PM

Kenzen, I really appreciate your time to research this issue and for sharing all the details. It was a good read too - Nice work. Hopefully someday we can get this fuel pump problem solved and then move on to the next one for you to figure out for us.

Thanks again,

Regan Metcalf
Portland Oregon

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: kenzen ()
Date: February 28, 2018 08:52PM

(Since I'm not really tired at the moment), here's where I think this will resolve as guidance for achieving reliable mechanical fuel pumps:

1. We will identify mechanical pump suppliers that have (at least) good gasket and diaphragm materials that resist modern fuel and support normal pump operation for "a long time". Or find a rebuild kit supplier for these materials. There's a supplier somewhere in the northeast that allegedly has rebuild kits - this might be cost effective.

2. We will have the specs on what a good spring is, and hopefully have a 3rd party supplier of these springs so we can retrofit them into pumps if needed. This eliminates the "add a fuel pressure regulator" need and (if my hypothesis is correct) ends the pumps eating themselves due to excessive internal pressure. I found one supplier who's willing to make the springs custom, and it doesn't seem like it would be too costly or require large quantity orders. More to come on this.

3. I highly recommend taking a new pump apart before installation to flatten the gasket surfaces and assure the check valves are properly staked. Flattening the surfaces can be quickly done with adhesive 220 sandpaper on a flat surface, and run the surfaces against it. BTW, I did this on a pump base (that slides into the engine) this evening, and noted the center areas of the surface were significantly higher than the edges, which would promote leakage toward the shaft and into the oil pan: This should be checked by everyone running mechanical pumps.

4. Applying locktite or some other technique to keep the body screws tight.

If anyone else has wisdom on this, please jump in.

Thanks for you patience. smiling smiley

kenzen
66 Monza Coupe 110/PG
Bel Air, MD

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: toms73novass ()
Date: March 01, 2018 03:07AM

You are doing a great job!

I have a few old pumps that I believe are original and have not yet gotten to taking them apart. I was going to compare the spring pressures to see what they show. I will use your 7/8" height as a comparison and see what I get.

It would be awesome if we can find all the correct parts to build a bullet proof fuel pump. I am going to be using mine as a boost referenced modified pump, so I don't have the luxury of just doing electrical, with my blow through carbs.

-Tom


63' Monza Spyder Convertable (in process) MY Build Thread
73' NovaSS 454 Big block
86' BMW 325es
98' Dodge 2500 12v Cummins Diesel with 1200 lb torque!
98' VW Jetta TDI, for daughter
01' Audi Allroad Stage 3 twin turbo

NFCC
Grand Island, NY

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: Wagon Master ()
Date: March 01, 2018 05:15AM

I would also suggest that during reassembly, you compress the diaphragm push rod .250 or so while tightening the pump body screws.

'64 Monza Coupe 140 4 speed
'61 700 Lakewood 95 4 speed
'69 Monza Convert. 140 Auto.

Been aircooled since 1973
Northwest Ohio

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: Layton Curtis ()
Date: March 01, 2018 07:55AM

If the surfaces are irregular, why not put the new gaskets and valve body in an old failed known good pump body. Most of us have a stash of old pumps.

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: kenzen ()
Date: March 01, 2018 06:20PM

Layton Curtis Wrote:
-------------------------------------------------------
> If the surfaces are irregular, why not put the new
> gaskets and valve body in an old failed known good
> pump body. Most of us have a stash of old pumps.

That's a solution if you have the old pumps that are good. The goal here is to identify what pumps to buy, and be able to tweak them to the state of "bullet proof" through inspection, correcting manufacturing flaws, and "blueprinting" them to be in spec with what the original GM engineers designed them to be.

kenzen
66 Monza Coupe 110/PG
Bel Air, MD

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: kenzen ()
Date: March 01, 2018 06:33PM

Just a note: I purchased a new pump from Clarks last night as well as the brass fitting for the pump output that feeds the carbs as well as provides the 3rd outlet, where I'll connect a gauge for testing. I'll use this new pump as another set of data points to the experiment. I also bought a 0-15# gauge to support the testing. I have in my head a way to mount an adjustable spring length jig onto a pump by hacking up and mod'ing a pump cap, so that I can adjust the spring tension as the engine runs and dial in the desired discharge pressure. (Maybe this is another mod to consider...hmmmm)

Once I'm able to I plan to measure the low spec 2# tension, the high spec 4.5# tension, and a nominal 3# tension which is where I think a custom spring should be designed to.

Just some data points: The stock springs I have are 1.25" OD at the small end, 1.875" OD at the large end, and the free length is 2". Given a compressed length of 7/8" (or 0.875), this is basically the 50% compression length most springs seem to be spec'ed to. Also, since the stroke of the push rod is relatively short, I'm not concerned about a "constant rate" spring - the linearity of the spring in this short stroke I'll consider effectively constant.

Ken

kenzen
66 Monza Coupe 110/PG
Bel Air, MD

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: MattNall ()
Date: March 01, 2018 06:47PM

I hope Lon Wall watches this!! gg





MODERATOR
Sea Mountain, between Charleston Harbor and Coos Bay! SW Oregon Coast
Click HERE for My Website...Click HERE for My TechPages!
..............................110-PG.................................................Webered-Turbo

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: Ratt643 ()
Date: March 02, 2018 06:58AM

This is an intersting topic. I biult an inline pressure gauge, mounted between the fuel pump tee and the drivers side carburetor. I ha e measured many new pumps fuel pressure and found 9-10 psi. I cjt a speing on one to try to get tbe pressure doen, but that didnt work, the pressure fell to 1 psi. I installed an old spring crom a bad pump and the pressure came down to five psi.
I know there is nothing wrong wkth running a mecbanical pump, but I bave seen several bad pumps putting gas into the oil. Ive got two Corvairs to convert to electric so this problem doesnt happen again. Please keep up the good work!


1964 Monza Coupe

Maryville, TN
U.S. 129 The"Tail Of The Dragon"

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: JimBrandberg ()
Date: March 02, 2018 07:40AM

My buddy Fran Schmit looked into mechanical pumps a few different times. One thing he found last year when examining multiple pumps from various manufacturing generations is that the main body regarding the locating hole is inconsistent. Some folks talk about shortening the pushrod coming out the bottom of the pump if it's longer by measuring its protrusion but that isn't the whole story if the locating hole isn't consistent.

It would seem the distance the pushrod pushes the diaphragm material would affect fuel pressure and fabric longevity, if it's opened too far and stretched it just won't last. Part of the equation is the locating hole and how much the diaphragm is opened with some preload when installed.

Something I haven't done is make a point of installing the pump to the engine with the pushrod on the engine in the down position. Perhaps a fellow could develop a feel for the preload. It's hard to get much feel with two springs and all but maybe if doing it a certain way consistently...

It seems like tightening the screws may be more of a function of compensating for poor metal surfaces. A lot of times they don't really seem loose and sometimes the cheap Grade 1 screws aren't up to much torque. A change to a quality screw would be a good thing.

Fran Scmit worked a little with Dave Herron 15 years ago or whenever that effort at a good pump was afoot. I remember discussion then that the attachment of the pushrod to the fabric was difficult to achieve, that the original AC tooling was difficult to duplicate.

I appreciate your efforts and would love to see a good mechanical pump achieved. It doesn't help that there hasn't been consistency. It seems like the big problem in 1999 was gas in the oil; the big problem in 2005 was valves not staked, the big problem in 2014 is too much pressure... Of course that's way too general of a statement.

I never have enough original springs and shortening and reforming seems sort of crude, maybe I'm not that good at it. Just having a good spring would help my efforts. I should start sanding the surfaces too.

Jim Brandberg
Isanti, MN
CorvairRepair.com

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: JimBrandberg ()
Date: March 03, 2018 07:13AM

I get collections of parts from people. I have a new in the box pump from Clarks with a 2003 postmark. It has slotted screw heads. I'm starting an engine on the cart and installed it temporarily. It leaks a lot externally from the sides, tightened the screws to no avail. Nothing against Clarks, I love them dearly, just a situation report while we're talking about it. Excessive pressure too.

I recently installed a new pump from O'Reillys on a different engine, purchased last fall. In a "Precision" box and I believe it's an Airtex. I took it apart to cut the spring and noticed it has reinforced fabric. No leakage and I didn't check the valve staking because I believe they addressed that a few years ago. I replaced the pump due to a crankcase full of gas with the old pump. The bellhousing seal may have suffered from that "cleansing".

Just a few of my latest adventures. Call me fickle but I would really like to see this mechanical pump thing through. Perhaps the new 2003 would be a good candidate for disassembly and checking the mating surfaces. If it was new new rather than new old I would probably just exchange it.

Jim Brandberg
Isanti, MN
CorvairRepair.com

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: MonzaJPD61 ()
Date: March 03, 2018 11:06AM

Kenzen, this may save your some work...mechanical fuel pump info from many years ago:

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: kenzen ()
Date: March 04, 2018 05:41PM

Wagon Master Wrote:
-------------------------------------------------------
> I would also suggest that during reassembly, you
> compress the diaphragm push rod .250 or so while
> tightening the pump body screws.


Is this because the diaphragm under tension will pull from the pump walls? This is a brilliant observation. Gold Star for Wagon Master!

kenzen
66 Monza Coupe 110/PG
Bel Air, MD

Options: ReplyQuote
Re: Mechanical Fuel Pump Observations...
Posted by: kenzen ()
Date: March 04, 2018 05:45PM

MonzaJPD61 Wrote:
-------------------------------------------------------
> Kenzen, this may save your some work...mechanical
> fuel pump info from many years ago:
> [i63.tinypic.com]

This is awesome! I was looking all over for something like this. If anyone else knows of similar work, please point it to me.

Ken

kenzen
66 Monza Coupe 110/PG
Bel Air, MD

Options: ReplyQuote
Pages: 12345Next
Current Page: 1 of 5


Sorry, only registered users may post in this forum.
This forum powered by Phorum.