I had an interesting flight at Chehalis yesterday -- the engine was really winding up, and it flamed out with an empty tank right after the triangles.  My launcher said he felt a "cold sprinkle" as I flew by on the first lap.

On inspection, the tank has a stress crack right where one of the tubes exits the tank.

It's pretty obvious that I over-stressed the metal.  That tube goes in fairly close to the top outside edge of the tank, and the way I installed it I tightened up the radius of the bend at that point, and bent it further than 90 degrees -- so it's pretty stressed right there.  Moreover, for a while I just had the tank shoved in there, and it was that tube that was holding it in position (I've since started shoving in bits of foam to space it away from that wall of the tank compartment).  So it was stressed and sustaining some vibration.

It got me to wondering, though -- I'm using a cheapo metal brake that bends the metal over a sharp edge -- should I be intentionally radiusing the bends?  If so, by how much?  1/16", 1/8"?  The inner radius is probably 1/32" or so currently.  I'm using K&S tinplate.

This is the first time this has happened to me.  It's maybe only the fifth or sixth tank I've built, and by no means the one with the most miles on it, but it's the first one that's in a full-fuselage model.

What do you experienced tank home-builders do?

  This is a pretty rare issue with the mild steel "tin" tank material I use (K&S) and I do nothing in particular to reduce the tendency.

     Brett

What do you experienced tank home-builders do?

Buy plastic clunk tanks

Joking aside, I've only "built" tanks for speed planes and they were kits. Bends were already there. I just had to solder them together.

That's specifically the reason I buy plastic tanks. I'm "getting it" now, and I've been successful at helping others set up theirs, and get their engines running. But I'm still scared to use metal tanks myself.

There's been times I've struggled to find a tank that fits. I've wondered if it would be worth figuring out a couple tank sizes to mold. But it would be a healthy investment in time and equipment to make some tanks for me and a couple dudes.

Lauri:

I really like the accuracy I can get from my bender.  But it occurs to me that the one bit that the metal bears against when bending could be replaced with wood!  (It's a really cheap bender -- you bend against a bar that you clamp on).

I'll probably try bending against wood as you suggest, and if I don't fall in love with it, try a wood bit in my bender.

Tim,

You can safely bend annealed mild steel with a bend radius of only 1/2t or so. If you used K&S #254 material, which is only .008” thick, that gives you a radius of only .004”. That is small.

If you used metal tools to make the bend, they probably were deburred if made by machining. That is likely going to be a softer edge than your requirement. If you used a Lee bender, these appear to be made of cold rolled steel and the edges come from the steel mill softer than that. On the one that I bought I left the sharpest edge so I could use it, and then softened a couple other edges to use on thicker material and material with higher temper.

For making tanks in general, I actually prefer a set of wood bending tools (hand seamers) that I made. These are not form blocks that look like a tank and you hammer your metal parts overtop that. I detest that method….

I do the layout work with a Sharpie pen. The use of a scriber is prohibited in aerospace work. It should be for models, too. A scribed line is a great place for cracks to initiate. I assume you didn’t do that.

Out of curiosity, tonight I made up two hemmed samples. I first formed a 90 degree bend using the Lee tool, on the sharpest edge. Then I closed the bend solid in a vise. The vise has aluminum pads on the jaws. The result is a hard 180 degree bend with essentially zero bend radius. Ok, it is not really zero, but is not measurable without sectioning it and using a microscope. I made one sample with the bend across the material grain (recommended practice for minimum radius bends) and one along the grain (fine for standard radius bends and larger). Neither sample cracked.

One way to cause early failures is to form and reform bends during fabrication. This causes so-called “low cycle fatigue.”  If you didn’t do that, then my thought is that your problem was vibration. This can be compounded by installing the tank with stress on it, and then superimposing cyclic loads from vibration.  So I think floating the tank is a good idea. I don’t like tank vents attached to the fuselage. I hard-mount a separate fitting to the side of the fuse and then connect to the tank with flexible line. Control the stress and you control the failures….

I’ve actually had more plastic tanks fail than metal ones. I think the newer tanks are better, and the use of the cap to help control the rubber stopper forces is what made plastic tanks reliable enough to be worthwhile.

Good luck with your tanks,

Dave

Phil,
Any luck repairing split-neck Sullivans with an iron? Or should I box all of mine up and hope the guarantee is good?

"Divot" McSlow

Ouch!

No washers are necessary when you don’t drill the hole for tubing but rather poke it with sonething like sharpened piano wire with correct diameter.
It keaves a nice radiused ”fillet” to tin plate. L