Leaking Fuel Tank Sump....again!

[ALAN COCHRANE]

Hi Everyone

Well this old chestnut has reared it's head again. I noticed that there was a small stain on the garage floor and found that the sump had sprung another leak.
I had already fitted my own cork gasket since the SNG rubber seal deformed when the sump was fully tightened. When I removed it I found a bit of the rubber gasket was still on the mating face of the tank spigot. Bingo the source of the leak-well no.
After failing to cure it with two new cork gaskets the leak has revealed itself to be the braised joint between the middle and upper sections of the sump. This has already been repaired since it was the source of the original leak a few months back. I think it's given way because of the numerous times it's been tightened up. How tight should the sump actually be tightened?
Here's a photo of the latest gasket which I deliberately made to fit tight into the rear under-tray opening so that the sump won't deform it:-
https://www.dropbox.com/sc/ar6ytu4vsxds ... LpP-dVzYga

Alan

[jag68]

Alan - I had a similar problem a few years ago - changed gaskets many times etc. Finally sand blasted the sump and found a few very tiny pin holes had rusted through from the inside allowing fuel seepage. The sand blasting opened these up enough to see them. It's possible that the leak you see from the joint is actually corrosion of the steel tube. I'm really doubtful that you could torque the sump up to a level that would cause a breakage or crack in the braised joint - it is after all quite large, and the nut you are tightening the sump with is itself brass and weak. Over tightening these types of gaskets is, I've found, overdoing it and has often the opposite effect of you are looking for. I'd start with a new sump from SNG (C19302) and go from there.

[ALAN COCHRANE]

I've had the sump joint re-braised as it was the source of the leak. I took it back to the same guy who originally did it and he was a bit mystified as to why it had failed. On the plus side he did it for nothing.
I have to agree about the gasket they are easily squashed out of shape if the sump is overtightened. The design is very poor with the gasket basically a thin rubber ring. It would be much better if the sump mating surface had a recess in it for the seal to sit in.
I made my own gasket from impregnated cork.
To overcome the deforming issue I made the outside diameter fit the bodywork opening, effectively using it as a brace. Interestingly the gasket that was originally fitted was of a similar design and probably for the same reason.
I bought one of the SNG sumps only to find it was the wrong size. The internal thread was 1mm too narrow and the body was 20mm shorter than the original item. They are currently checking the rest of their stock to see if it's a one off.

Alan

[PeterCrespin]

I have to agree about the gasket they are easily squashed out of shape if the sump is overtightened. The design is very poor with the gasket basically a thin rubber ring.

We do this a lot on these 40-50 year old cars. Not you personally especially, lots of us do it. We criticise things that are someone else's cock up, or that have already performed way beyond their intended design life (AKA fulfilled the design brief). Since a rubber ring is not what Jaguar fitted, it's a bit rum to criticise their designers.

It would be much better if the sump mating surface had a recess in it for the seal to sit in.

True, if they hadn't sent tens of thousands of cars out with the simpler alternative shown in the parts book.

I made my own gasket from impregnated cork. To overcome the deforming issue I made the outside diameter fit the bodywork opening, effectively using it as a brace.

Impregnated with what? Good luck but I bet you'll be draining your tank again shortly. Making a thin cork disk a couple of centimetres too wide on OD, in the hope of the ID edge not distorting under clamping pressure, sounds like a triumph of hope over experience. Besides which, there is supposed to be a large foam ring around the sump so I can't see how it would work as the boot floor is lower than the tank base and the sump joint (or should be).

Interestingly the gasket that was originally fitted was of a similar design and probably for the same reason.

When you say 'the gasket that was originally fitted', you mean the PO's bodge I presume? The correct part is not a rubber ring - for the reasons you know only too well. Nor is it cork, which is likewise never used on a narrow, tight, unrestrained threaded circumference seal, for similar predictable reasons of instability and deformation. It's flat ring of a dense fibrous waxy composite, a bit like the 4.2 head gasket stuff, that never pops out and is called a 'washer' in the parts books. If that doesn't work for you, THEN feel free to describe it as Jaguar's 'very poor design'. Ask for C1617, which later changed to C26310 for the 2+2 & S2 cars. The later cars also had a shorter sump which I think was to account for the extra depth of the removable drain plug. You'd be OK to just nudge the pick-up tube up a bit provided you didn't kink it. If vendors keep sending you o-rings (it's not what I was sent) then try a liberal helping of PTFE tape around the threads before assembly and good luck.

Of course, if your guy is really braising [sic] the sump then no wonder you're in a stew.

[ALAN COCHRANE]

Pete

First of all the cork I've used is designed specifically for sealing against mineral based fluids and has been used successfully by the Company I work for, for many decades often in environments far more hostile than a cars fuel tank. It's 2mm thick and not prone to distortion. I'd already fitted a cork gasket when the the weld gave way again. So it was good for 4-5months worth without any degradation.

Secondly the part you mention C1617 was the one I originally ordered but I couldn't get it to fit onto the tank spigot. I was then advised to get part no C26310 and it deformed very badly when I refitted the sump. Both these were from a leading Jaguar parts supplier.

The guy who repaired the sump runs a very well respected radiator repair company and pressure tested the sump to confirm that this was the cause of the leak.
I'm not a metallurgist but the top threaded section looked very much like brass and I assumed that braising would be the weld of choice. I stand corrected if this was the wrong assumption. Perhaps he just waved his wand and it was magically sealed.
The sump had already had a leak from this joint and a PO had used epoxy resin on both the inside and outside to seal it. This eventually gave way restarting the leak. I've no idea what provenance the "original" sealing ring had but it was still performing it's job at the time.

Perhaps I've just been unlucky with the sump and it's not a common problem, but making observations about possible improvements isn't a crime.

Alan

[PeterCrespin]

Sorry, the 'braising' (meat/stew) was a joke on 'brazing' (metal/sump).

You're right the threaded bits are brass. It's part of the design. The bottom of the car would be exposed to lots of water and often salt, so rather than rip the base off the tank or crush a stuck sump, they took the trouble to use a non-corroding metal for the male and female threads, even though it cost extra and took more assembly/fabrication work. Your guy will be a much better craftsman than me, I'm sure, and doubtless did a good job, but something went pear-shaped....

The sumps get bashed around and crushed because they are often very tight to get off. They are only strong at each end where the brass reinforcements are. They come off OK if you grip them at the end with a strap wrench and apply steady leverage - assuming the tank base isn't o thin it ripples like a copper water tank with a stuck immersion heater.

You asked how tight should they be and I didn't answer that. They don't need to be much tighter than an oil filter IMO - tight enough to not fall off or leak, however tight that is. With the o-ring and groove you mentioned, which is probably how they'd do it today, I'd probably settle for hand tight like a filter. However, they don't use o-rings they use a smooth washer made of quite tough stuff and I snug it up with a strap wrench to be sure of a seal. PTFE wouldn't hurt either I guess. Better than RTV silicone for sure, which is not fuel proof.

There is no use of cork that I have ever come across where it is asked to seal a narrow annular joint between tightly-assembled parts without support. The closest seal on an E-type I can think of is the ring of cork around the dizzy base, but this sits in a recess and cannot bulge out and is not clamped all that tightly axially, just held under the clamp plate with a 7/16 set screw on one side. The other places are areas like the tank flanges or the back of the engine sump on your 61, where the cork forms a wide flat seal in a sandwich between even wider clamping surfaces. Cork is fine in these applications where it seals over a wide area.

An o-ring shaped non-captive ring of cork between components like the tank and sump, that are supposed to be tightly screwed together, has no structural integrity to stop it breaking up or popping out like the o-ring did, unless you use some kind of outer sleeve or a serious sealant/adhesive which holds everything together without the hard tightening. I'm sure the company cork is fantastic against lots of fluids but that's not the point - I never said it wouldn't resist petrol. I assume your company uses that cork in a flange-type application, not unsupported narrow o-ring shapes. I'm quite happy to be shown that is not the case, with a photo of something that blows my theory.

Well done for stopping it leaking for 4-5 months or whenever you first noticed the existing drip, but that doesn't make it the right part or let you assemble the car with a big foam ring and strong flat washer as per spec if you want the boot floor hole to act as a restraint to your floppy 2mm thick sheet.

Still, it's your car and the fact many/most people using standard bits don't have leaks, but as your thread says you have recurring leaks with non-spec bits could be something else entirely. Hope it all works out - petrol is expensive stuff.

Pete

[ALAN COCHRANE]

Pete

Sorry a bit slow in the uptake on the humour I must have been thinking about tonight's dinner!
I have been thinking about what you said in respect of the fibre washer. As I have already said I couldn't get the thing to fit around the tank spigot and by the very nature of it's design it can't be bent. If the cork does start to leak I could always enlarge the hole on the fibre washer to allow it to be screwed onto the tank, which is apparently what you would have to do if it fitted in the first place.
I'm beginning to think that the "original" seal was in fact just an O ring coupled with the foam outer ring you mentioned. Unfortunately when I first unscrewed the sump the whole lot just fell off in many pieces. Does this foam ring have a part number? This might explain the complete failure of the rubber O ring I tried.
I did take the precaution of drawing a ring on the cork gasket using the sump as a template and deliberately made the hole too small so that it was a super tight fit. This allowed me to see if any deformation had occurred once the sump was tightened up. It doesn't look as though there was any or at the very least so little that it's not detectable.
Despite this I may yet have to eat humble pie on this one.


Cheers

Alan

[PeterCrespin]

If the ID was snug and you carefully fitted it with some good sealant (and didn't tighten to the point of distortion/popping out), you might be fine in terms of leakage. There's more than one way to seal a cat...

If you already have petrol in the tank and don't do much wet weather driving you might as well leave it if it is dry now. Just get yourself a tuppenny correct part, ready for if the leak happens later.

Not sure of part number but every seller will know what you mean if you say the foam ring around the sump. It's almost exactly the size and shape of a doughnut or fat pineapple ring and can't be mistaken for any other bit of the car (except maybe the heater intake seal which is much bigger diameter but skinnier).

[tinworm]

Hello Alan ,I cant help you with the gasket material , but I can tell you that the later sumps ( which are different than the early multi part type) are silver soldered together - not brazed - between the brass threaded end the steel centre tube (which rots out) and the hexagonal drive lower cap. You will have a job to braze these parts successfully. The correct repair would be to carefully heat and melt out the silver solder turn up a new tube on the lathe and then re-solder back together again - or just buy a new repro unit -which you may have to do now of course.

best regards Barrie

[ALAN COCHRANE]

Barrie

Thanks for the info. Yes I think from what you've said the repair was silver soldered. The outside shows evidence of the solder. I've had a look at the SNG catalogue and found that the foam pad Pete mentioned is part no C18906. I've ordered this along with the fibre washer as insurance in case the cork gasket fails-time will tell.
The repro sumps are a bit of a minefield- I ordered one from SNG and it was the wrong thread and 2cm shorter. So I'm a bit wary of using that route. I'm a great believer in repair rather than replace wherever possible.

Cheers

Alan

[Duckham]

Alan - I recently removed the sump on my tank and used this rubber washer from Ken Jenkins with no issues when I replaced it. http://tinyurl.com/mhj2my7 . It is likely the same one that you got from other suppliers but worked fine for me.

I found it difficult to judge how tight it was as there is a lot of friction between the threads. As the sump thread is orginal and the tank an older repro I assume the threads on the latter aren't quite right. I was aiming for oil-filter tightness as Pete describes it and that seems to have worked OK - I think there is a real danger of overtightening and squeezing the rubber out.

When I first put the car on the road the sump had the same pinholes as in yours. I couldn't get a replacement at the time so my solution was to remove the metal tube on a lathe and replace it with thick wall brass tube that I already had. I pinned this to the brass ends to take mechanical load and then just soldered with plumbers solder.
I was unsure if it would be successful but it has been ( including surviving the use of stillsons).
Joe