Success!!!

A couple of weeks ago I developed a small leak from my fuel strainer to gas tank coupling. After reading through all of the threads, I was determined not to break the seal on the embedded sleeve. As was expected, mine was frozen solid. So my quest began.

Supplies needed-

10 lbs. patience
2 plastic drain pans
2 qts. ammonia
2 qts. EvapoRust
2 lbs. dry ice
1/2" id fuel hose
1/2" bolt to plug fuel hose ( I used a spark plug)
impact wrench (highly recommend variable torque electric)

Here is what I did.

1.Drain fuel tank. I opened the top of fuel tank and removed the fuel sender then siphoned almost all fuel from the tank. Then I removed the hose that led to fuel pump and let the rest drip into a pan and evaporate for a couple of days.

2. Place hose clamp around neck of fuel tank and tighten snug.

3. Take a chunk of the dry ice and using gloves hold it on the fuel strainer outlet until it gets nice and cold. (see pic). Because of the difference in the Coefficient of Thermal Expansion (CTE) the aluminum strainer will contract ever so slightly more than the steel sleeve. This will make room for the important next step.

4. Put the 1/2 hose on the fuel strainer outlet and plug it up. (see pic)

5. Pour 2 qts. of ammonia into the fuel tank. The ammonia will break down the aluminum oxide that was formed on the threads of the strainer from oxidation. This is partly the reason that it is frozen in place. Let sit overnight.
At this point, I had a steady drip, drip drip into my plastic drain pan. I checked a couple of hours later and it was a much slower drip. The next morning it was completely dry. At that point I was a little frustrated, because I thought that the drip should have increase in volume from the ammonia eating away the aluminum oxide and clearing a path for the leak to get bigger. Hmmm.

6. Drain the ammonia from the fuel tank and return to its original container.
After looking at the contents in the bottom of the pan, it made more sense to me. There was a fair amount of debris that when dried appeared to be aluminum oxide. (see pic) I surmised that the degraded aluminum oxide filled the gaps in the threads and stopped the leak. But what I want is the leak because that means that the ammonia is getting by the threads and eating away more aluminum oxide.

7. Hit the strainer with the dry ice again. The reason that you must drain the ammonia from the tank to use the dry ice is that the aluminum is a good conductor of temperature and all of the cold is dispersed into the liquid and the strainer never gets cold enough to contract. It took me a couple of times to realize this.

8. Pour the ammonia back into the tank and wait overnight. (this is where the patience comes in)

9. Repeat the above 3 steps one more time.

10. Now drain all of the ammonia from the tank, ice the strainer then fill with EvapoRust. Let sit overnight. The reason I did this was I have an inspection camera and when I put it into my tank I saw some rust on the threads of the sleeve. The ammonia will react with the aluminum and aluminum oxide but it wont do squat to the rust. I used EvapoRust because it is neither a base or acid. After using the ammonia I wanted to make sure that I would not be using another chemical which would cause toxic fumes. Besides EvapoRust works great.

11. The next morning I drained the tank and hit the strainer with dry ice one more time. Then with the impact wrench on the lowest setting I just barely hit the strainer with a bump to tighten. Then I did the same to loosen. I did this about 5 times each way. I slowly got the strainer to move a bit loose. I was worried that it may have been the sleeve moving in the plastic so I used my inspection camera to look at a spot on the sleeve in relation to the tank and tried it again. It moved a little more so I checked my reference mark inside the tank and it had not moved. So this was a good sign. I went ahead and tightened the strainer again then reversed and loosened a bit more each time working it back and forth. each time it moved a little more but was definitely easier as I went back and forth until finally it came out.

Big thanks to everyone who has posted their experiences with this repair on this forum. I would have probably messed this up as others have had I not been informed on the potential pitfalls. Also my father, a chemical engineer had useful input in the chemicals and reactions used.

Good Luck!

 

Wow, very innovative.

However, scrolling down to the photos before reading, seeing that spark plug in the end had me thinking you were going to add some gasoline, 12V to the plug and blow the strainer out

 

After 3rd round of ammonia, I can't say it didn't cross my mind ��

 

We found a significant amount of rust in the bottom of the sleeve in Jim C's '78, likely from water that had dropped out of the fuel. Of course we didn't find it until after we had broken the sleeve free from the tank and prior to getting the aluminum fitting out.

Kudos for the dry ice trick. We talked some about that option but it seemed too far-fetched at the time. I was concerned about freezing the plastic and making it too brittle as we applied torque to the fitting.

Congrats on the success!

 

I found that the transfer of the cold from the dry ice to the plastic was negligible. Kinda like an Ice chest.

 

Great job. Can't say enough about using thermal expansion/contraction to loosen items. Do it virtually daily at work, but I have it easy because most items I deal with are small enough to place directly into our super freezer. (-45 F) Also, EvapoRust is a fantastic product. I've been using it for about 5 years now at work, and it's one of the few chemicals out there in the automotive world that isn't snake oil. If you need to remove rust from anything, and haven't tried EvapoRust, go get some.

 

GJB928, thanks for sharing this idea. I am totally stealing it from you.
Subscribed.


Seth K. Pyle

 

Just one more little 'bonus' provided by Ethanol laced fuel.

 

Amen my brother!

I've managed to pass only a few tanks of fuel through the car each of the last couple years. The last "storage" tank gets a bottle of marine Sta-Bil at the max recommended dosage. Thinking about it, the car can sit with any given tank of fuel for a month or more during the summer, so it probably deserves at least a normal dose of the standard Sta-Bil product in those 'tweener tanks.

----

I wonder what the shelf life of Sta-Bil is...

 

I believe that the strainer and the threaded insert are both aluminum. If that is correct, then dry ice will do nothing-- they will shrink together.

The weak link is the bond (or lack of same) between the aluminum insert and the plastic tank. The only thing securing the insert is the knurling, and plastic is soft enough to allow the insert to spin with only modest torque.

The key is to reinforce the plastic with a metal collar, so that the plastic cannot expand and allow the insert to move. This is a common technique to reinforce metal-plastic connections.

See this post:
https://rennlist.com/forums/928-foru...l#post12771087

Which points here:
https://rennlist.com/forums/928-forum...l#post11611542
(See next post for pictures)

When re-installing, WSM says 25 NM (18 ft-lbs). New gasket, always. And I would advise silver anti-seize on the threads, there was some left on the threads years later in spite of the fuel-soaked environment.

The corroded threads have nothing to do with the seal-- that is a gasket between the strainer fitting and the lip on the "boss" on the plastic tank. If it is leaking, then the gasket is toast.

 

Jim,

On [the other...] Jim C's '78 car here in Bend, the threaded insert is definitely steel. We gave some thought to dry ice and the procedure GJB928 shares, but for various reasons didn't go with that. I'll share what we ended up doing in Jim's thread, once I get the pictures back from the drugstore.

 

You are correct that a metal collar will help hold the sleeve in place. That is why I used a 2" hose clamp.

Actually the sleeve is held in by a bit more than the knurling. There are 4 holes in the sleeve and when the plastic is injected into the mold it creates 4 posts that help hold the sleeve in place along with the knurling.

 

We decided that a 62M24 Ideal hose clamp is the right size. It's 5/16" (8mm) wide, so fits without doing any surgery on the "ribs" that surround the boss around the sleeve. If you don't mind sacrificing the 4 nibs, any #24 screw clamp will do the job. OD of the boss is right at 2".

 

Don't the 4 nibs likely correspond to the location of the holes in the sleeve which create anchor points during the molding process (as mentioned in the post previous to yours)? If so, then it seems you would not want to shave them down losing whatever extra integrity they might provide. Assuming your sleeve hasn't already spun, that is.

 

We didn't trim the little ribs at all. I'm not sure that clamping or not clamping, with or without those ribs, would make s difference with the way the tank is formed-in-place around the sleeve. Checking diagrams I made, the holes in the sleeve are about back well beyond the ~~3/4" penetration. That would put them behind those ribs too.

We talked about using a wider normal-width clamp, but felt that tightening around the ribs would distort the round nozzle. If you've soldered copper pipe or glued PVC sprinkler pipe, you know that you need to have even contact, no distortion if you want things to seal. Ig a pipe is even slightly collapsed so that it's bigger one way than the other, the joint will leak. Carrying that knowledge forward to our insert fit-up with epoxy in the nozzle, we were quite happy to leave the plastic still naturally round.

I'm sure there are a brazillion (that's a lot...) opinions on clamps and such. We decided that it would not be a good idea to use a hose clamp during assembly.