Fuel Tank Fittings

[Stevetin]

It seems that fuel tank fitting leaks are one of the more common issues that I've read about since starting my project. The Sonex Fuel Tank Installation and Warranty guide states, "Do not over-tighten the fittings as they can break an insert loose in the tank". "It only takes a snug, not tight, fit to achieve a leakproof installation". I have a problem with subjective terms like "snug" and "tight", they're not measurable. What's snug to me might be tight to someone else.

I decided to make the torquing of the fittings measurable and ran a series of tests with different Loctite products. I chose Loctite products because I was introduced to a Loctite pipe sealing product called PST back in the 1980's and was extremely impressed. For my test I used Loctite 545, 565, 567 and 5452; Loctite 565 is similar to the Loctite PST of the 80's and Loctite 5452 is designed for hydraulic fittings and is pretty expensive. To better replicate actual conditions I used several 1/8" NPTF raw aluminum couplings with brass 1/8" NPTM fittings.

After applying the four different Loctite products to the brass fittings I screwed them into the aluminum female fittings. I torqued the first set of fittings to 10 inch pounds using a torque wrench (10 inch pounds is basically hand tight). I then torqued the next set of fittings to 15 inch pounds the the last set to 20 inch pounds. Once finished I let all of the fittings set-up for 5 days. After 5-days I connected the fittings to a fuel line and to a lawn mower gasoline tank and filled the tank with gas. I let the fittings sit for a week while making several checks a day looking for leaks.

THE RESULTS: After one week had passed I was surprised to see that NONE of the fittings had leaked, not even the hand-tight 10-inch pound fittings! I decided to use the Loctite 545 for my Sonex fuel tank since that's what this product was basically designed for. I then proceeded to install the brass fittings in the Sonex tank and decided on 15-inch pounds as my target torque setting. This was easy to do with the straight fittings (such as the finger strainer and petcock) but was a little more difficult to achieve with the 90 degree fittings. It seems that when the 90 degree fittings were torqued to 10-15 inch pounds the fittings were facing the wrong direction. When they were facing the right direction the torque wrench was reading 20-30 inch pounds.

Bottom line. I filled my tank with gas and had zero leaks, not one. My take away from this test is that keeping the fittings torqued to very low torque settings, such as 10-inch pounds, does a great job of sealing. A little more torque is still OK but I found no need to torque past 15-inch pounds to insure a no-leak situation. I hope this helps someone.
Steve

[peter anson]

That's good information Steve. Wish I had known about the Loctite sealant when I was building my Sonex. Does the sealant also help prevent the joint from coming loose? You could check that by measuring the torque to undo the fitting.

Peter

[GraemeSmith]

I like this stuff. All the advantages of PTFE without the risk of PTFE tape

https://www.amazon.com/Gasoila-Seal-Thread-Sealant-Degree/dp/B008HPVEP0

[Stevetin]

Good Morning. Peter, I still have the fittings intact so I can easily check the torque of breaking them apart. I'll do that and report my findings. With that said, I'd only be concerned with the lower torqued fittings (10-15 inch pounds) as the higher torqued fittings seem to be fairly tight.

Graeme, yes, I've found these Loctite products to be very good. Years ago Loctite bought Permatex and because of that Permatex makes the same products with slightly different part numbers. I can't say that there aren't better products or that there's not a better way of installing the fittings in the tank but this worked for me. Peter asked a good question and I'm curious to find the answer.

[Stevetin]

Peter, thank you again for asking about the removal torque of my test fittings, it was a real eye opener! Below are the results:
Loctite #545 removal (from 15 inch pounds installation) took 20 inch pounds to remove. Loctite #545 removal (from 20 inch pounds installation) took 20 inch pounds to remove. Loctite #565 removal (from 10 inch pounds installation) took an amazing 60 inch pounds to remove! Loctite #565 removal (from 15 inch pounds installation) took 30 pounds to remove??? Loctite #565 removal (from 20 inch pounds installation) took, again, an amazing 60 inch pounds to remove! Loctite #567 removal (from 15 inch pounds installation) took 45 inch pounds to remove. Loctite #567 removal (from 20 inch pounds installation) took 60 inch pounds to remove! Loctite #5452 removal (from 15 inch pounds installation) took 20 inch pounds to remove. Loctite #5452 removal (from 20 inch pounds installation) took 25 inch pounds to remove.

A few points. First, I'm glad that I chose to go with Loctite #565 for my fuel tank fittings. Secondly, Loctite #565 and #567 are basically the same product but #567 is the higher temperature version. Also, when I originally performed the testing I cleaned all of the fittings with MEK prior to joining them. In addition, I let the fittings set-up for 5-days per the recommendation of Loctite. I can't explain why my 15 inch pound test of #565 removed at only 30 inch pounds; maybe I didn't clean the fittings well enough??? Also, all of the torque results that I've shown were rounded to the nearest 5 inch pounds.

I'm sure that there are other variables that others might learn from performing their own tests but I thought that it was worth the $35 in fittings to perform these tests.
Steve

[DCASonex]

Hope that works for you, mine did not start to leak until the following year. Plane alternated between heated and unheated hangar in Western NY winter.and hot sun in summer. My suspicion is that the differential expansion between the plastic tank and the metal fittings is what eventually loosened them enough to cause very slow leak.

David A.

[Stevetin]

David, I guess I'll have to see if the tank leaks sometime down the road. If that happens I can always go to the Oops fittings, or maybe fabricate on out of aluminum.
Steve

[peter anson]

Steve, you may never have problems with leaking fittings. It is common but not universal. I had no problems until I fitted a capacitive fuel level probe. The original probe was faulty and a couple of ins and outs was too much for the little brass fitting. Even then it only leaked when I flew. I think the fuel sloshing against the probe wiggled it around causing the slight leak. I fixed it by gooping it with PR1422 fuel tank sealant plus a bit of external support. It's lasted 7 years so far.

Peter

[Stevetin]

Thank you Peter, good to know. If I make it a year without the tank leaking I'll report it to this group.
Steve

[WaiexN143NM]

hi all,
just wanted to say thanks to jeff schultz, for compiling the list of sealants. ive posted this before, as we had a very small leak in one of our legacy molded in tank fittings. we used a glue from locktite, called a plastics bonding system. has a primer stick, and tube of glue. looks exactly like a super glue tube. red/blue package. no part number, but bar code on back is 7934082565. for all plastics , poly e and poly p. our fitting on the lower rt sump in tank on verticle side started to barely drip. we drained tank, let sit for a few days. took out brass plug. cleaned outside with scotch brite, wiped w acetone, primer stick tank and fiber washer. glue applied on tank, fiber washer metal washer screw. tightened up. never leaked since. locktite sells this. mfgr is henkel corp, rocky hill ct. this fitting was never used, go figure. avbl from home depot, lowes, ace hardware, true value.

WaiexN143NM
Michael