Fuel Tank Cracks - have me baffled.

Capt. Geoffrey Thorpe

Capt. Geoffrey Thorpe

Touchdown! Greaser!
Joined
Jun 7, 2008
Messages
16,022
Location
DXO124009
Display Name
Display name:
Light and Sporty Guy
My ride (homebuilt LSA taildragger Merlin GT) hasn't flown since last November. While it was sitting in the hangar over the winter, one fuel tank cracked in two places. The tank is supported on the inboard / outboard ends (short ends) and cracked where the side and top are welded along the inboard / outboard ends. I would think that these are the least stressed joints on the tank - so I don't understand why they would crack there and not along the long unsupported front and back. The tank previously cracked on the bottom of the outboard end - under where it is cracked now. The airplane first flew in 1998 and has about 500 hours on it. Tank holds 8 gallons and has no internal baffles. Vent was clear when I blew through it today. 20160624_093749.jpg 20160625_085347.jpg

I suspect that when it was originally welded there wasn't enough penetration and the weld must be pretty thin - but other than that, I got nothing...

Ideas?
 
My guess is the flat sides of the tank are flexing. cracking the weld material. If they could be slightly curved outward, the joint stress would be lower.
 
That looks like the tank is flexing, I'd check for a restricted vent system. or a missing support stringer. might also be, not properly normalized after welding.
 
poor grade of aluminum .poor welding.
 
Tom-D said:
That looks like the tank is flexing, I'd check for a restricted vent system. or a missing support stringer. might also be, not properly normalized after welding.
Click to expand...
It's aluminum.
Vents are clear.
The sides are supported nearly the full length.
 
I would have thought it was CRES, stainless. Probably a penetration problem, but maybe it should have baffles that would reduce flexing.
 
Last edited:
Is your hangar next to a major highway or railroad track? Looks like a fatigue failure due to vibration and maybe a brittle, thin weld bead.
 
A lot of the comments above refer to steel weld problems and are not applicable.

What is the aluminum alloy? I assume it should be probably 3003, but that info should be given to the welder. A compatible rod should have been used .
 
nrpetersen said:
A lot of the comments above refer to steel weld problems and are not applicable.

What is the aluminum alloy? I assume it should be probably 3003, but that info should be given to the welder. A compatible rod should have been used .
Click to expand...

When you ask what alloy it is how can you assume its 3003? or make the insinuation that the wrong filler rod was used?
 
nrpetersen said:
A lot of the comments above refer to steel weld problems and are not applicable.

What is the aluminum alloy? I assume it should be probably 3003, but that info should be given to the welder. A compatible rod should have been used .
Click to expand...
Thanks, I think everyone figured out it was aluminum when the Capt. told us, above.

It's crap alloy. It's corroded and deep scribe lines.
 
The fuel header tank on the PA46-310P and 350P will crack on the welded seams just like yours. Sometimes they just split for no reason at all. We've had some shops try to clear debris from the drain valve using an air hose. Our tank is made with check valves that don't allow the air to pass then splitting the welded seam.
 
Is there a type club for the Merlin? Maybe this is a known problem (with, or without a solution)? Just wondering if there is any history available for you to study. Perhaps another material for tanks like plastic, fibreglass.
 
GlennAB1 said:
It's crap alloy. It's corroded and deep scribe lines.
Click to expand...
There are no scribe lines anywhere on the tank. There are some fine scratches here and there that show up when the light is right. Corrosion is minimal (unless it is happening inside the overlap between the top and side.)
 
Let'sgoflying! said:
Is there a type club for the Merlin? Maybe this is a known problem (with, or without a solution)? Just wondering if there is any history available for you to study. Perhaps another material for tanks like plastic, fibreglass.
Click to expand...
No type club that I am aware of - but the Avid / Kitfox is similar - and I have heard complaints about the metal tanks in those. Long term, a composite tank may be the answer - I just need to find an E-10 resistant resin.
 
kmead said:
The fuel header tank on the PA46-310P and 350P will crack on the welded seams just like yours. Sometimes they just split for no reason at all. We've had some shops try to clear debris from the drain valve using an air hose. Our tank is made with check valves that don't allow the air to pass then splitting the welded seam.
Click to expand...
Interesting.
My current theory is that the actual penetration was too shallow and oil-canning of the flat top may have flexed it enough to break the weld if it is really shallow. I'm going to try and get into the crack with a fine Deremel cutter and see what I can see before I get it TIG welded.
 
nrpetersen said:
What is the aluminum alloy? I assume it should be probably 3003, but that info should be given to the welder. A compatible rod should have been used .
Click to expand...
I don't know the exact alloy - there are no markings on the sheet that remain. I would assume that when the tank was made new by the kit manufacturer they would have know what it was when they welded it.

The repairs that I have had made are holding just fine.
 
Dan Thomas said:
Is your hangar next to a major highway or railroad track? Looks like a fatigue failure due to vibration and maybe a brittle, thin weld bead.
Click to expand...
That's my thought - inadequate penetration.
The airport (ONZ) is on an island with no trains or major highway or any road with more than 2 lanes. And speed limits are 35 or less.
 
GlennAB1 said:
Water in the fuel froze?
Click to expand...
Possible. But these were top seams above the fuel level - condensation? The fuel is probably E-10 (I don't bother to check for ethanol when I pick up fuel at the local gas station) which would tend to keep small amounts of water mixed in.
 
GlennAB1 said:
Expanding ice might just fracture the weakest weld?
Click to expand...
doesn't happen that way.the ice must be trapped in confined space in order to do damage other wise it just forms a ball of ice in the bottom of the tank. Plus the tank was only half full.
 
Well, I poked a knife into the crack and it was pretty easy to cut it along another half inch or so - which suggests to me that there just wasn't enough penetration in spots. It doesn't look different from the outside from one weld to another, but...

Got it cleaned up, polished off some of the scratches with scotchbright for the fun of it. Got into the crack to clean up the inside surfaces. I'll get it TIG welded (again) and hope that there aren't more thin spots ready to let go.

Given that I can probably find a used TIG rig for what two new tanks would cost me, that just might be my long term solution.
 
curious how the welder inerts the tank, still hearing about boom events from such repairs, let us know what technique is used.
 
Let'sgoflying! said:
curious how the welder inerts the tank, still hearing about boom events from such repairs, let us know what technique is used.
Click to expand...
Fill about half full of water, dump in about half a bottle of Dawn Dish Washing Liquid, put an air hose into the tank (I plug it into the fuel outlet) and set the air so it slowly bubbles. Let it sit for about 18-20 hours. Good to go.
20160611_142730[1].jpg
 
Let'sgoflying! said:
curious how the welder inerts the tank, still hearing about boom events from such repairs, let us know what technique is used.
Click to expand...
I etch the interior of the tank with this (http://www.aircraftspruce.com/catalog/cspages/metalprep.php?clickkey=12033) or equivalent acid cleaner allow to set 1 hours and rinse, that will resolve and hydro-solvents in the tank dry, preheat to see if there is any steam coming out, wait until it quits, then empty a CO2 fire extinguisher into the tank and place the cap on it.
now weld it.
 
The west Texas way is apparently to run hose from a pickup exhaust into the tank for 10mins, then have at it. Having doubts about that being 100% effective
 
Tom-D said:
I etch the interior of the tank with this (http://www.aircraftspruce.com/catalog/cspages/metalprep.php?clickkey=12033) or equivalent acid cleaner allow to set 1 hours and rinse, that will resolve and hydro-solvents in the tank dry, preheat to see if there is any steam coming out, wait until it quits, then empty a CO2 fire extinguisher into the tank and place the cap on it.
now weld it.
Click to expand...
Let'sgoflying! said:
The west Texas way is apparently to run hose from a pickup exhaust into the tank for 10mins, then have at it. Having doubts about that being 100% effective
Click to expand...
The tank should be purged with a shielding gas, which carbon dioxide and carbon monoxide are not.
 
GlennAB1 said:
The tank should be purged with a shielding gas, which carbon dioxide and carbon monoxide are not.
Click to expand...
No O2 - - no boom. Yes Argon is better but who has the test equipment to tell what % Argon is in the tank?
 
GlennAB1 said:
You aren't concerned with the quality of the weld? Why not throw some carbon in it!!!
Click to expand...
Carbon in an aluminum weld ??? don't we try to keep these as clean as possible?
 
Then why would C02 be ok?

Does the FAA have anything published about welding?

Try page 5-24 of the AMT airframe handbook found on the FAA website:

TIG Welding Aluminum

"Argon or helium shielding gas may be used, but argon is preferred because it uses less by volume than helium. Argon is a heavier gas than helium, providing better cover, and it provides a better cleaning action when welding aluminum."

"When welding tanks of any kind, it is a good practice to back-purge the inside of the tank with a shielding gas. This promotes a sound weld with a smooth inner bead profile that can help lessen pinhole leaks and future fatigue failures."
 
Last edited:
You must log in or register to reply here.
Back
Top