BauerR

Thanks for the help, guys! Looks like PC-7 may be the strongest option, though.

GregBBRD

That's an interesting conclusion to take away from this thread....

BauerR

After looking at the options, the PC-11 offers the option to harden under water... which is not something I need. The PC-7 is a stronger version without the water hardening option.

GregBBRD

This particular thread is filled with reasons why epoxy doesn't work, with solutions using heat.

That's the interesting part about your conclusion....

I don't repair fuel tank inserts....at all....ever!

Of course, because I own a shop that repairs 928's, my legal exposure/moral issues of doing a repair that fails in the middle of the night in someone's garage, burns the house down, and kills the entire family, is completely different than that of people doing crazy repairs at home.

Therefore, I throw fuel tanks with spun inserts in the trash.....every single one of them. My attitude is simple....fuel is, by far, the most damgerous thing present in a vehicle (besides the driver.) If you can't afford to properly repair the fuel system (tanks, hoses, etc.) you can't afford the car.

I hunt down used tanks to replace ones that have failed....and I keep brand new tanks, in inventory. Both my insurance agent and my conscience feel much better than trusting an epoxy that will not actually stick to the original plastic.

BauerR

Well, if it makes you feel better, I've already purchased another tank from Kuhn 2 days ago. I still want to have a go at fixing this one.
So because I purchased another tank and fixing my current one, does that mean I can afford the car now?

dr bob

I know that there are more pics someplace, but here's what I found in a quick search on my travelling laptop:

-- We used a length of 3/8 threaded rod to fashion a puller to get the sleeve installed again correctly. The first pic shows the tool as we were ready to start drawing the sleeve into place. What you don't see is what's inside, guiding and pulling the sleeve. From the rear inside towards the end that you see here, there's a 3/8" nut, a 3/8" fender washer, and a 1/2" USS flat washer, and a nut. These are sandwiched so the washers are concentric, with the OD of the fender washer matching the OD of the sleeve, and the OD of the 1/2" flat washer matching the ID of the sleeve. In effect, the smaller OD washer inside the sleeve keeps it all lined up in the middle of the sleeve.

-- Continuing inside on the threaded rod, there's a nut, another 1/2" USS washer, and a nut to hold the washer concentric to the rod inside the sleeve. Now the rod is pulling straight and square to the sleeve. That last nut towards the front needs to be far enough back that the outer face is still a little behind the front end of the sleeve.

-- We tie-wrapped some mechanics wire to both ends to facilitate install and removal.

-- A thin layer of JBWeld was applied to the outside of the sleeve and to the inside of the plastic tank nozzle, after thoroughly cleaning both.

-- On the outside, I used a bearing washer with ID that fit around the lip on the tank nozzle. You can see that lip clearly in the last two pictures. So that washer is installed, then a 1/2" USS washer and finally a regular 3/8" washer and a nut (the one the wrench is sitting on in the first picture).

-- A little further out on the rod, there are two nuts tightened against each other so they don't turn on the rod.

-----

-- The second picture shows my gloved hands holding the rod from turning with a wrench on one of the locked-together nuts, as I use the other wrench to tighten the nut towards the washer stack to gently draw the sleeve into place. With everything cleaned, and lubricated by the JBWeld, I turned the nuts until the sleeve was drawn in. I wanted the end of the sleeve firmly against the inside end of the tank nozzle, right behind the sealing face. The third picture shows a progress pic, where you can see the outer of the two alignment washers inside the sleeve and captive by a nut, and the end of the sleeve not quite all the way to the end of the nozzle. The wrench is sitting around the threaded rod just inside the washer stack there for the picture; there is no nut there. The nuts were snugged back up, and the sleeve was drawn in the rest of the way. I could easily tell the sleeve was all the way in, as the tension on the nuts suddenly increased.

-- The last two pics show the sleeve in place, after the JBWeld set up for a day with the tool still holding it all together. One shows the squeeze-out from drawing it in (removed before it set up...), the other the cleaned nozzle end and the sleeve threads where the sleeve is tight against the end of the nozzle.

-----

The sealing face and the lip at the end of the nozzle, best seen in the last picture, is where the sealing happens. With the threaded sleeve supporting that face from behind, the fitting for the screen (or in-tank pump) holds a soft round ring gasket in a groove in the fitting, securing it so it can meet the plastic tank squarely. Snug the fitting to compress the gasket some against the plastic to effect that sealing.

I used a thin film of Teflon thread sealing paste on the threads inside of the threaded sleeve and the outside of the fitting. This lubricates the threads, and hopefully will keep water that falls out of the fuel from corroding the metal parts together again. On the ring gasket, a very thin film of Hylomar was applied to all contact surfaces to allow the fitting to go in without distorting the ring gasket. The actual tightening force on that fitting to clamp the ring gasket was minimal, like contact plus less than half a turn. Kinda like the way you fit an oil filter to avoid distorting the ring gasket on that.

So--- The epoxy you choose is there to keep the threaded insert from turning as you install the hose and filter fitting. That's all. It doesn't really participate in the "sealing" function at all. Hopefully it will still be working if/when you need to remove that fitting again sometime in the future. Remember, It Does No Sealing.

All of the sealing is done by the ring gasket, squeezed gently between sealing face and lip on the end of the plastic tank nozzle, and the hose-and-filter insert that threads into the sleeve in the tank. To that end, it's imperative that you avoid damaging that plastic lip as you grapple with getting the threaded sleeve out of the tank after it "spins". That means you may need to --carefully-- pick out the old ring gasket as a first step on removal, and avoid using any tooling on that sealing face that might damage it. So no prying with big screwdrivers against it.

-----

The pictures:

davek9

Welding temp for HDPE is 300 deg F, from the plastic welding chart I found.

just an FYI,

Dave K

GregBBRD

I'm hoping you are smart enough to understand what I'm saying...

Some of the procedures that people perform on this Forum are absolutely crazy....using some sort of a glue, which does not actually bond to the plastic tank (on the very bottom of a fuel tank to hold in a main insert) is just one of these things.

People on this Forum do these crazy things because they are too cheap to do the proper repair.....be careful of what you read.

The basic rule of the Internet that everyone should follow:

Just because you read that someone did it, you need to be smart enough to do the basic logic of why you shouldn't do it.

GregBBRD

Dr. Bob:

Although you are correct that the insert does not do any of the sealing....it is all that keeps that seal from leaking. When the fuel breaks down the epoxy (which is not bonded to the plastic.....nor recommended for submersion in fuel), the insert will slip and there will be a leak.

I don't keep track of every single stupid prior attempt at repairing these cars that I see...that would take up every hour of my entire life.

But I can guarantee you that I've replaced over 20 tanks that have been "repaired" by some prior idiot....which come into my shop either leaking or empty from the insert coming loose.

Nothing. Repeat nothing you smear around on that insert is going to bond that insert to that original plastic!

polecat702

What's wrong with using an A/ N bulkhead fitting? They're available for aircraft applications. I can reach into the tank with the sender removed, to counter hold the fitting for tightening.

BauerR

I thought about making something like that. You would have to figure out an external filter.

polecat702

You could probably thread the ID of the bulkhead fitting to accept the screen, or make a new screen. The boss in the tank looks like there's enough material to come up with a replacement set up. Isn't there a replaceable external filter on all the cars? Some have the internal pump, and some have the screen.

BauerR

I haven't dropped my tank yet to see exactly what I'm working with. But i am thinking of moving in that direction.

All fuel systems, even the 928s with the internal pump setup, have some type of pre-pump filter or screen.

dr bob

The internal pump was installed because the boiling in the suction end of the primary pump became critical in high fuel temp situations. Putting a relatively restrictive filter between the tank and the pump would make this problem even worse. Find a suction sock that will fit to whatever you are installing as a tank fitting, and install filters after the pump discharge only.

---

While a custom or bulkhead fitting through the tank wall seems like a Good Idea, you still have that small matter of figuring out how to seal the end of the factory outlet nozzle.

---

Were I doing this "from scratch", a full-sized pump would get dropped into the tank from above, and the discharge would come through a second fitting at the top of the tank, next to the return line. Electrics would pass through the gauge sender top plate too. That way there would be no penetrations in the tank below normal liquid level.

---

I seldom disagree with Greg. Having the adhesive lose grip between the threaded sleeve and the plastic tank nozzle will not cause a leak. As stated and with Greg's agreement, all the sealing is done by the ring gasket between the threaded hose fitting and the lip at the end of the tank nozzle. Having the sleeve "loose" to possibly turn inside that nozzle ignores the somewhat snug fit there even with no adhesive. Again, the purpose of the bond between the threaded sleeve and the tank is to allow you to tighten (and loosen) the hose fitting to maintain correct pressure on the ring gasket. With the threaded sleeve brought tight against the end of the nozzle, as shown with the little tool we made, pressure on that ring gasket is maintained even if the adhesive fails.

I'm not suggesting that Greg's advice to avoid cheezey half-fast fixes isn't sound; it is. Use all my shared experience at your own risk. I have the experience of analyzing and fixing exactly one tank this way, while he's replaced a lot more than that. I am merely sharing/documenting that one experience. Greg's liability is very different from mine and that of most folks here. Our fix was done in a controlled environment, thought through carefully, and particular to our own situation. Your mileage may vary. Your conditions, technique, preparation and results may vary. Use your best judgement, knowing your own capabilities. Unlike Greg, no warranty is expressed or implied as to the suitability of this or any other repair experience shared with readers here to their situations and problems.

And yes, knowing what we know now, we should have stopped pulling on the original fitting before the threaded sleeve "let loose" from the plastic. The tank should have been drained and removed completely, and the aluminum hose fitting should have been sectored and removed in place. Perhaps the next person who walks this path will share torque wrench settings used to find the critical breakaway limit for the sleeve in the plastic tank, so others can make an educated decision and avoid spinning the sleeve at all. New tanks are $1400 last I looked. Used ones offer exactly the same parts you have now, with the same well-aged ring gasket doing the work of sealing the bottom of the tank. All we needed to do was replace that little ring gasket. Instead, like many 928 projects, it quickly blossomed into a larger ordeal.

Good news: The subject car has been leak-free for a half a year so far, and will likely see use on a PCA driving tour tomorrow. We will inspect the area again soon to verify the status of the seal, and will report back on what we find.