Fuel pressure.

[fetuhoe]

The principal is sound, but I'd really hope the margins on the actual numbers are a bit wider than that!

Also does the pump offer any restriction when it's off?

The only head that would matter is the height difference between the fuel level and the level of the needle valve in the carb,

This assumes a gravity based system as on an old motorcycle.

Fuel pumps tend to have inlet and outlet valves which prevent flow when they are not driven.

[Bruce M]

If the fuel is below the level of the needle valve, there should be zero head. If the fuel level is above the needle valve there is some positive head (as above).

If the needle valves are not sealing due to a tiny scratch in the valve seat, any amount of head will result in a small drip due to a siphon effect when the tank level is above the valve. This would be unlikely to have a noticeable effect on the road as the fuel is getting used at a greater rate.

However, if there are valves in the pump maybe that should prevent it. Perhaps the valves are not effective against this?

[mycar]

I popped over to the Registry to ask a couple of questions. There was a certain amount of topic creep but I've learned a couple of things.

http://forum.porsche356registry.org/vie ... =1&t=42674

The fuel pump in a 356 doesn't appear to incorporate any sort of valve to prevent flow when the engine is not running. I eventualy did find the diagram I was looking for which in itself is probably not that useful but the text that came with it says that any more fuel than 1/3rd of a tank is above the carbs level.



My 356 fuel tank is round 7' high internally so any static head created whilst the car is parked up level is about 4 3/4". What sort of pressure does that create ? To me that's more seepage than pressure and certainly doesn't equate with the idea of 30kg of petrol forcing itself on the poor delicate needle valve.

[fetuhoe]

The Model AH Pump which is shown in the diagram I previously posted is a 356 Pump and the internal check valves are clearly shown.

The simple way to check if they restrict flow when they are not working is to disconnect the inlet and the outlet ports, attach a clean length of hose and simply blow though the line and see if you get an output.

A Bendix/Facet pump will allow fuel flow due to head but I haven't seen a diaphragm pump that will.

The 'Carb Height' Line isn't clear and it is the position of the needle valve when closed that matters.

Even of this static head (4.75") did have an influence in would be around 0.127 psi so going up steep hill could be more of a problem

I don't have a 356 mechanical pump to test but a Hillman Imp Mechanical pump isn't affected by a static head as we just checked.

I would be interested to know more.

[Bruce M]

Further material:

http://www.sparkingplugs.com/resources/ ... p+39-2.pdf

[Bruce M]

The A pump appears to have 2 one way valves so that the diaphragm draws fuel from the tank in the down stroke and towards the carb on the upstroke. Will this prevent flow from tank to carb?
Maybe when the diaphragm is forced by the spring it blocks the ports? But that would depend on what position the pump stopped at or if it is removed from the engine?

(Just thinking out loud so probably talking s@&7)

[mycar]

I would have thought that was unlikely.

David Jones over on the Registry is our man here, he has an extraordinary breadth of knowledge and says... "The valves are one way only to allow fuel through to the carbs. They will stop fuel flowing back to the tank when the engine is off but if you have a full tank and a bad float needle valve you will fill the sump with petrol."

Shame Porsche didn't think to fit a proper pump to their cars.

As well as the solenoid valve to cut the fuel, David posts details of a valve to control leakage. This looks to me to be a simple and effective piece of kit.... if only I could spec it correctly for 356 use.

The options are many and varied.

http://straval.com/products/relief-valv ... oc4nhw_wcb

[fetuhoe]

Is it likely that the inlet check valve from the pump opens when the head is 0.127psi? and is this also enough pressure to open the outlet check valve.

Any valve will have a 'cut off' pressure because there must be some mechanism holding it onto the seat or it doesn't function correctly so there will always be some small spring force present.

I can understand that this pressure will be low but not this low.

We now also need to deal with the pressure drop of the needle valve which also needs to fail when the pressure difference is only 0.127psi.

This would simply mean that the needle valve had failed so substantially that the engine would be overfuelling all of the time as the pressure of the pump would simply force fuel out of the main jet which would create a small pressure drop.

The other consideration is the pressure drop needed to force the fuel along the fuel line between the pump and the carb.

I would be surprised if a pressure drop of 0.127psi would result in any meaningful flow along a 6mm ID fuel line.

If you do some basic calculations using an average Renolds Number for Gasolene and a low surface roughness index you come up with a typical pressure drop of around 0.15psi for 30cm of hose.

I can accept that this calculation is subject to error but it shows that filling a sump due to the head created by a full tank does seem unlikely if you take into account all the barriers.

The fuel then needs to find its way down the intake into the combustion chamber past the piston rings and into the sump, quite a tortuous path.

I believe that main reason for engine filling with fuel when using this type of pump is not due to the fuel tank head forcing fuel down the intake but a failed diaphragm.

I don't think the basic engineering supports the 'fuel head' argument so without measurements I would be sceptical.

It would be really straightforward to make the measurements.

We have a carburettor test system all set up with the correct test fluids and could easily set up a test to measure flow though a static 356 pump versus head.

All I need is the use of a pump.

[mycar]

There are reports of damaged fuel diaphragms resulting in hydrolocked engines.

The other possibilities are :-

1) The floats sinks check for leakage into them.
2) The float level was incorrectly adjusted.
3) The float valve is faulty.
4) The float valve has dirt in it, preventing it from closing.
5) Shim washer/valve seating surface has a scratch.

That's plenty of suspects, the last two being very tricky characters. It seems that fuel will find its way through the smallest passage given the smallest pressure. It's just easier to tell folk to turn off the fuel tap.

How does that valve that I linked to above look to you Chris ?

[fetuhoe]

Have you tried rebuilding the pump?

It seems as if the internal check valve wafers in the pump are not seating which is common on these old pumps and I believe that some of the early VW pump repair kits are supplied with new wafers.

The simple way is to put a loop into the fuel line which is above the head in the tank is this is feasible and wouldn't spoil the appearance.

The problem with the 'one way' valve that you have shown is that the 'cracking' pressure may well be low and it will have no effect. There are some ball valves that work the same way but again they are not reliable at low pressure.

You need a valve that will only break with a suction pressure which is higher than the pressure head and they are tricky to find. They tend to be a gate style valve and when they fail they fail closed.

A simple solenoid valve operated by a relay would also work - or a tap.

[Ian 2.2S]

- or a tap.

Exactly . Which is why every old car/motorcycle/lawnmower manufacturer fitted a petcock when you have a situation with a fuel level being above a carburettor. You are constantly fighting gravity and a fluid that likes to find the smallest gap.
My brief encounter with a certain '65 car (911) leaves me in not doubt that Porsche had some shortcomings where fuel pumps and carburettors are concerned
But I do like Chris's suggestion for a solenoid operated valve - simple elegant solution.

[mycar]

- or a tap.

Never mind the tap, we can do better than that !

This reply from the valve people in New Jersey :-

Hello Michael,

Thank you for the inquiry! I believe you are on the correct track to resolve the issue as explained.
However, we do not manufacture a valve sensitive enough in a compact envelope for this type of application.
Perhaps an automotive low pressure check valve or ignition driven solenoid valve might be better choices.

I am not certain, but you may also need a product approved for automotive use on public roads as well.

Please keep us in mind with your future industrial valve needs.

(P.S. That is a fantastic automobile and best of luck with your search!)

Best Regards,

Wilson Diaz
Project Engineer
Straval Machine Co. Inc.
1-973-340-9955 X116
--------------------------

So that's a dead end.

My pump along with the carbs and engine have all recently been rebuilt Chris, in my case I don't think it will be the diaphragm that's failing. (it bloody better not be)

The solenoid valve was also suggested by David Jones on the Registry, how would that work ? Are suitable solenoid valves readily available in 6 volt ? Would it be reliable ?

The loop sounds like an elegant solution, sounds like simplicity itself. If a loop is plumbed into the fuel supply, higher than the fuel tank then any fuel that reaches the carbs must be pumped ?

I ran a laser round my car to check the levels, starting with the top of the fuel tank there seems to be a fair bit of room in the engine bay for such a loop.





Interesting that from the top of the tank to the carb top is nearly 9". Add to that the needle valve length and the static head is about twice what I thought it was. Pretty much any fuel in the fuel tank is sitting above the needle valve.



And with that I had to rush about and didn't have time to work out exactly where the loop would be plumbed in. I'll have a look tomorrow.

Edited to refine content,

[Bruce M]

Hmm... a simple loop can be as high as you like, but the siphon effect will draw the fuel up and back down (the weight of the falling liquid creates a vacuum that draws up more liquid). Unless the is an air break at the top to release the vacuum but that not going to help when you want fuel to flow.

A non-return valve might work if it has a cracking pressure of 1psi or so but that's not what they are designed for.

I see why Porsche fitted the tap now.