Little shop project

I've had this portable A/C unit that every so often decides to leak water all over the place. I've had it apart about a half a dozen times to fix this and that, but it will always find a new way to leak. My final solution was to put a baking sheet under it. It can go weeks with nothing, then one day the pan will be full. I have a drain hose on the A/C, but it still leaks in the pan every so often. Sometimes when it's humid the pan will fill just from condensation on the outside. The pan works good, but I have to check it and vacuum out the water from time to time. So, this morning, I decided to finally do something about it.

Since the sides are angled, I welded up a boss to give me the correct angle, and some meat to drill and tap.
Attachment 7575 Attachment 7576
Drilled, tapped 1/8 NPT and filed the top flat.
Attachment 7577
Screwed in a fitting to attach a drain hose to.
Attachment 7578 Attachment 7579
1/16" 4043 filler, 1/16 pure tungsten, 20-80 amps with about 15 cfh argon. No idea what the alloy of the baking pan was, but even being pretty thin, it welded nicely. I think this just might win the battle between me and this A/C.

That's a pretty neat idea!

You just inspired me, I'm about to start on a full custom gas tank for my Ironhead chopper. They sell bungs for the petcock for $13-$14 but now I'm thinking of building up an area and drilling and tapping it like you did. One less seam to fart around with and I guarantee that it won't take $13 worth of filler rod :)

Not to mention copying your exact idea for a coolant catch pan on my lathe :)

If using aluminum for tanks ect, Some conduit fittings are made from aluminum and they are threaded NPT, I have used a lot of them in the past to build fuel tanks and live wells for boats. Makes it handy that something from another industry will a lot of times work on a project.

The thing I like about this idea is that almost all weld-on bungs I've seen will still leave a small bit of seam on the inside once welded due to the joint design. The way he did it eliminates that entirely from the inside of a fuel tank. I wonder if done out of steel if annealing it would be a good idea afterwards to eliminate stresses and prevent cracks...

I doubt I would ever do something like this with steel. Building up aluminum is no big deal, but on steel sheet metal I would always weld on a fitting or maybe even braze it on, depending. Then the fitting could even be brass. By the time you got any steel built up the sheet metal would no doubt be warped pretty good, possibly overheated, and the weld might be pretty tough to drill and tap. Annealing would fix that part but it just seems like way too much on the con side for so little on the pro side. Now if you want to build an aluminum tank, then you might go that way.

Most conduit fittings I've seen are die cast and not the best material for TIG welding. Lots of impurities. They are also usually threaded NPS, same thread form and pitch but no taper. You can screw tapered fittings into them, and it will usually work, but not the best match. I always keep my eyes open for things that can be repurposed from other industries. Why reinvent the wheel.

The only reason I did this the way I did was that the sides were tapered and I needed the fitting to be straight. At first I actually made a little tapered fitting to weld on, but I wanted all the welding on the outside, and with the lip on the pan it was just too tight to get in there and weld all around. So I just built up the boss from weld metal, since that was an easy option.

I like how you built that material up and turned it into the just fitting you wanted! Nice work. And, you didn't waste time locating a specialty weld-on bung at the store or online, and have to wait for it to get shipped to you.

I've cut apart aluminum "AN" fittings before and welded them onto other aluminum parts, using them as makeshift weld bungs before. The alloy welded and held up nicely. I've done similar with steel "AN" fittings, and even with brass fittings. Brass doesn't TIG weld as nicely because some of the tin/zinc in it vaporizes, but you can still leave nice, solid weld deposit underneath the surface. Same problem exists with welding pot metal / aluminum-zinc alloys (but to an greater extent, IME). (I wish there were a way of TIG welding tin/zinc laden alloys without the vaporization problem...)

As a side note, threads I've machined from built-up aluminum I've noticed have been much stronger (and able to take much more torque/force) when the filler rod used for the built-up material is 5356 instead of 4043. For this application it didn't matter, but something to keep in mind for those fasteners without a lot of thread depth that you want to bring up to full torque and really hold.

I almost always use 5356 for everything, but I had a bunch of 4043 to use up and this was an unknown alloy. I actually was surprised that it seemed very pure, I guess the stamping process requires good material. I was talking with a rep from Alcoa once who told me that some of the best made aluminum is the alloy used for cans. The high speed stamping and drawing process requires a perfect combination to work at the speeds they run and with the absolute minimum of material. His comment was that the next time I was flying, remember the plane was built with the material that wouldn't pass QC for making cans. ;)

Quote:

---

Originally Posted by Rambozo

I almost always use 5356 for everything, but I had a bunch of 4043 to use up and this was an unknown alloy. I actually was surprised that it seemed very pure, I guess the stamping process requires good material. I was talking with a rep from Alcoa once who told me that some of the best made aluminum is the alloy used for cans. The high speed stamping and drawing process requires a perfect combination to work at the speeds they run and with the absolute minimum of material. His comment was that the next time I was flying, remember the plane was built with the material that wouldn't pass QC for making cans. ;)

---

And thousands of moving parts assembled by the lowest bidder...