Welding The New Gas Tank

 The first step was to nibble away the flange on the end plates between the Cleco tabs.  My first plan was to nibble with the outside of the flange up so I could easily see what I was doing.  The nibbler would stop against the plate to control how much of the flange was left for welding.

I cut little notches and then cleaned them up.  The Clecos act as stops to assure I don't cut off the tabs.

 It worked fine and left just the amount of flange I wanted for welding, just to the edge of the radius.  The idea is to make an easy fusion weld and the penetration behind the weld would fill the corner making it easy to clean up any flux inside the tank to prevent future corrosion.  A corner weld should also make the bending loads on the weld joint less than at the end of the 1/2" flange.

The problem I found was that the chips come out the bottom of the nibbler and scratched the end plate as it rubbed against the nibbler, bad plan.

 I made a new end plate in about 45 minutes.  I'm getting too good at this.

I'll nibble this with the outside of the flange down this time so the chips are free to fall.  I also put strips of Gorilla tape along the radius to protect it and to control the depth of the nibble.

Again the Clecos act as stops for the tabs.

 It was a little more awkward to do this but it worked.  I first did the perpendicular cuts as I did previously.  It worked but it was slow.  I have a lot of notches on the flanges and then on the skin.

 Next I tried cutting notches at each end and then driving the nibbler along the plate.  I couldn't make it work well.

 Next I tried the same end cuts and then just a sideways motion using the back of the nibbler as a guide along the edge of the aluminum.  It took 3 passes and it was done, and they went quick.

It did take some practice to hold the edge touching the nibbler housing to make a nice straight cut.

In the end this was how I did the remaining cuts.

 Once the notches were cut I cleaned up the corners around the Clecos to reduce the chances of cutting myself on a sharp corner.  It did work, I never cut myself.

The edges and tabs came out well.

  I then assembled the end plates on the skin and used the end plate as a stop for the skin notches.  It was much more awkward to do this with the whole tank in hand but it worked.  It doesn't cut as deep because the cut edge of the end plate hits the housing of the nibbler.  It's a small difference, about 0.020", and my experiments say it should not cause a problem welding.

Again I cleaned up the tabs on the skin just like the end plates.

 The top seam was done the same way and everything is Clecoed back together ready to tack weld.  So far the plan is working.

 There are 2 or 3 tack weld between every tab.  I found that a tack every 1/2" - 3/4" was needed or the skin will pucker out creating a slight gap when the torch heats it.  Oh yeah, I'm Oxy-Acetylene welding the tank.

 Some of the things I've learned Oxy-Acetylene welding 0.040" Aluminum are:
- You only need to flux the edges of the assembled parts to weld,
- It's easier to weld aluminum that is fit very tight, the aluminum is thin and gaps lead to holes, even more so when you are trying to fusion weld a corner,
- You can almost gas weld steel in the dark because it makes so much light, You need bright light at the weld for Aluminum,
- Use the biggest tip you can control, you want to minimize the size of the heat affected zone to prevent holes forming, if the metal doesn't quickly puddle you are using to small a tip,
- The rod can be used as a heat sink and to stir metal to get it to flow together,
- With these 2 pieces of sheet aluminum at 90 degrees to each other start the torch at the outer edge, flame parallel to the outer skin, it will take more heat to get the outer skin to flow than the piece which is perpendicular to the flame, which will easily heat enough to make a hole,
- If a gap opens either stop and hammer it shut, or quickly get the rod into the puddle to cool it while you use the rod to run the bead in the area of the gap, then you can go back to doing the fusion weld.
- If a hole opens get out of there and let it cool, work your fill puddle in steps around the edges,
- If the weld is not smooth and well bonded it might be strong enough but it will leak.

 The tabs were easily cut off with the snips leaving a nice tight spot for a tack weld.

 A nice looking section of welding.

While washing and scrubbing the inside of the tank to get rid of the flux I found a few leaks where the water sprayed through.  Those were a quick fix.

 I then used Gorilla tape to seal the access panel.  I also had some tape on the inside which the air pressure blew down tight.  I found before that the blower end of my vacuum will only put out 2 psi so it's a quick way to pressurize the tank.  Put a pressure gauge, or water manometer, on the outlet of the tank and use a dimmer control on the vacuum to slowly bring it up to pressure.

Bubble solution quickly finds small leaks.  I still plan to pressurize the tank and let it set to see if there are any leaks so small I missed them with the bubbles.  You can't do that with the vacuum blowing.

 With the leaks fixed the tank was treated with Alodine to give a little better corrosion protection and the access panel was screwed on.

I applied sealant around the screws and with 2 beads between them.
I started the assembly with 4 screws in position through the clamp ring and carefully started them before starting all the others. I didn't want the sealant smeared, just compressed to make a good seal.  Once all the screws were started I put some more sealant under the edge of the tank opening.  With the screws tightened I used the round end of a craft stick to clean up the sealant.

 The front and back of the finished tank.  The weld on the forward seam is the best and it's the one I did first.  I had a lot of gaps open on the aft seam.  I eventually figured out that if I kept the rod close and in motion, even though I didn't use it in the fusion weld, I could more quickly get it into the puddle at the start of a gap, if it formed, and keep on welding.  The gaps this worked for were only about 0.010".  When the gap opened you suddenly had a small puddle going on the edge of each piece but not enough metal to bridge the gap without getting the rod quickly in the puddle.  Any hesitation and the skin, being perpendicular to the flame, would form a hole.

 The square bottom and tapered top of the tank can be seen here.

The outlet fitting turned out to be easy to weld.  The small welding lip on it fused easily with the flange on the tank.

 In the fuselage setting on the supports.  I need to make the straps which will hold it in place.

 There are some bolts in the front which limit how far forward the tank can go as well as the forward struts of the Cabanes.

The straight edge shows the cowling line.  It looks like the filler will be out of the cowling.  I may need a slight bump in the cowl.  Once the tank is strapped in place I can put the filler neck hole in the cowling and work out any bump needed.

Fabricating The New Gas Tank

 Plywood form blocks were made for the end plates.  I bought a piece of 1/2" hardwood plywood for the actual form blocks and used some scrap 1/2" plywood for the backing blocks.  The form blocks have a 1/16" radius on the bend corner.  I used the belt sander to put a 5 degree spring back relief angle on the edge.

The blocks are made undersized to allow for the end plate thickness, in this case 0.040" thick.

 I did all of this twice.  I started with a plan to use 0.040" 5052-H32 aluminum but could never get good at welding it.  I still think I can learn to weld it.  I finally gave up and remade everything with 0.040" 3003-H14 aluminum.  The guys making cool aluminum bodies use 3003 because it's easy to work with.

I put a 1/2" lip on the plates even though I don't want that wide of a lip for welding, you'll see later.  To lay it out I just used a pencil and a buffing pad from my Dremel Tool.  It was quick and it worked.  I found I only wanted a 1/4" lip at the 2 tips so I used a smaller pad there when I made the parts the second time.

 The aluminum gets clamped between the 2 blocks so it's centered well.  The aluminum is soft enough it doesn't take a lot of clamps.  The backing block allows you to form a nice tight square bend.

 As you start forming the aluminum with a rubber mallet it gets quite wrinkled.  Just form the wrinkles into fairly distinct wrinkles and then hammer them down with a plastic or wood mallet.  You're shrinking the aluminum to form a smooth flange.

The rubber mallet works great on the inside bends at the 2 bottom corners.

 I probably could have used a slightly flatter spring back angle since the aluminum had a fairly tight grip on the block.  It did come off without bending it.  At the same time, I would rather have it over bent slightly then under bent so the skin fits tightly when trimmed for welding.

 The skin pattern was used to layout the skin.  I made the seam offset to one side to allow an access panel at the filler neck.  I reused the panel I made for the old tank.  It worked well to allow access if you needed to repair a leak, etc.

Because the top is sloped to fit the cowling there is some curve to the edges of the skin.

 I try to avoid using tools for their intended purpose.  I'm bending the end curves around the tubes on the ShopSmith.  It worked perfect.

 The skin is clamped to the end plates with straps so that holes can be punched to Cleco clamp the parts together.

 I don't plan to rivet the thank together.  The Clecos are just to hold it together while I tack weld it.  I plan to nibble away the aluminum between the Clecos.  Then I'll tack weld between the tabs with the Clecos.  Once it's all tacked I'll snip off the tabs and put a tack weld where each tab was.  I don't know how else to hold it all aligned while tacking it without a jig.  The straps held because there was a wide flange.  If the flange ended at the corner radius I don't think the straps would hold it as you tack weld.  There are a lot of odd forces as you heat each spot to tack it.  It really wants to bend and twist.

It's looking like a tank.

 Even though the bottom is curved I want a sump to help assure water find its way to the drain.

I made a form block and hammered a shallow sump.

The next problem was how to make a fanged hole for the drain fitting.  The flange will get welded to the flange on the fitting.

 I discovered one of my ball peen hammers was the exact same diameter as the fitting.  I sawed a hole in a piece of particle board.  Amazingly the hole saw was the exact size to allow clearance for the 0.040" aluminum flange.

I sawed a hole in the aluminum to allow about a 1/8" flange and just pounded the hammer into the form block with a mallet, awesome.

 The fitting fits perfect with just enough flange to easily weld the fitting to the tank.

 The next problem was how to locate the hole in the tank bottom so the flange would be the same height all the way around.

I drew a circle on the aluminum to use for aligning the hammer.

For the form block I cut out a piece of foam rubber to fit the hole and drew a circle on it concentric with the form hole.

 With the foam in the block you just position the aluminum so the circles are aligned.

 Align the hammer with the circle on the aluminum and give it one quick wack with a sledge hammer.  I probably could have used the mallet but I wanted to be sure I did this in one blow.

It worked perfect.

 The fitting is in tightly and ready to weld.

The last thing was to shape the old access panel so the filler was close to vertical and so the curve matches this tank.  There is a ring on the bottom of the panel with nut plates riveted on and a ring that sets on top of the tank to clamp everything for a tight seal.

I think it's time to weld the tank.