Weak tig welds on aluminum

[Evan Fabrication]

I have a Diversion 165 and I'm new to tig welding. After going through the 2% ceriated tungsten and collets the package came with, and making horrific welds. I am getting the hang of it. I have only used it on 6061 aluminum using 4043 filler. My local miller dealer gave me 1/8'' pure tungsten w/ collets and 1/8'' filler. I am able to lay a decent bead but think I'm having penetration problems.The manual shows that size tungsten way out of amperage range. I am able to beat the two 1/8'' angled aluminum apart, and seems to break on the outsides of the weld. After looking at the manual I see I should be using 3/32''. I bought 3/32'' 2% ceriated tungsten, collets and filler. I am welding alot of 1/8'' angle to angle and 2''sch 40 pipe to 1/8'' angle. Also other sch 40 pipe from 1'' to 1 1/2'' to each other at times. I have my gas 100% argon @ 17-20cfh. And have the amperage setting at the line of 1/8'' and 3/16'' on the machine aprox. 140 amps. I have also had a few time where I had a hard time starting an arc. Seems after the machine has been used for a bit. Evaen with a new tungsten and collets. Could the 1/8'' pure tungsten size be the problem, being out of range for the machines capability?

[GTA/SPEC]

I believe that sch40 pipe in any diameter will be more than the diversion could handle.

Also, dont use pure tungsten for anything. 1/8" is too big - you wont need anything bigger than .093

[con_fuse9]

2" schedule 40 should be just a hair thicker than 1/8" Might be on the edge.

BTW, this is why structural stuff with aluminum scares me... I wanted to build a car trailer from aluiminum but I think I will pass for now.

One thing to remember about 6061 aluminum is that it very often comes T6 condition. Whenever you get the temperature above melting, you will, at least for some time, end up with T0 which is about 1/2 as strong. Give it a couple days and I believe it gets to T4 by age hardening (I'm a little fuzzy on that. Besides T4 isn't much stronger than T1). Bottom line you have a weak spot.

The other thing is you need to ensure the weld has penetrated through the material. Aluminum doesn't flow much and it requires quite a bit of heat (Amps) to get the puddle big enough (and thus deep enough) to penetrate all the way through. If you don't ensure that, you will be laying a bead on top of what you want to weld.

You did mention the failure is right next to the weld bead (not in it). I would think that this is in the HAZ (Heat affected zone) and you have basically softened the area. If it didn't break in the the bead area you are doing well.

(Assuming your beads are solid, through and through, and no porosity...)
You have two choices. 1) design with this in mind or 2) Have the piece solution heat treated to build back full strength.

[Donald Branscom]

I believe that sch40 pipe in any diameter will be more than the diversion could handle.

Also, dont use pure tungsten for anything. 1/8" is too big - you wont need anything bigger than .093

You CAN use the pure tungsten for aluminum with a transformer welding machine with a ball formed on the end of the tungsten.
like the Miller Syncrowave 250.

Trade with someone.

.093 is a 3/32 tungsten.

BTW almost any weldment that is only welded on one side can be easily broken.
If it is the end of pipe welded to a plate then it will not break if welded properly.

[Donald Branscom]

I have a Diversion 165 and I'm new to tig welding. After going through the 2% ceriated tungsten and collets the package came with, and making horrific welds. I am getting the hang of it. I have only used it on 6061 aluminum using 4043 filler. My local miller dealer gave me 1/8'' pure tungsten w/ collets and 1/8'' filler. I am able to lay a decent bead but think I'm having penetration problems.The manual shows that size tungsten way out of amperage range. I am able to beat the two 1/8'' angled aluminum apart, and seems to break on the outsides of the weld. After looking at the manual I see I should be using 3/32''. I bought 3/32'' 2% ceriated tungsten, collets and filler. I am welding alot of 1/8'' angle to angle and 2''sch 40 pipe to 1/8'' angle. Also other sch 40 pipe from 1'' to 1 1/2'' to each other at times. I have my gas 100% argon @ 17-20cfh. And have the amperage setting at the line of 1/8'' and 3/16'' on the machine aprox. 140 amps. I have also had a few time where I had a hard time starting an arc. Seems after the machine has been used for a bit. Evaen with a new tungsten and collets. Could the 1/8'' pure tungsten size be the problem, being out of range for the machines capability?

Yes the 1/8 tungsten is considered large. Out of range for that machine.
The 3/32 middle size tungsten is good from 0-150 amps.

Penetration of a joint in a weldment has a lot to do with how the JOINT is prepared.
If you are welding on a flat plate is not the way to determine penetration.
Practice on test coupons.Two plates side by side with a tack weld on each end.
Or practice a filet weld on two plates perpendicular to each other.

[karlcrew11]

well assuming your beads are solid there are a few things you can do as of now. have the piece heat treated... if you think that's too much of a hassle then just go with what you have and work from there.

[mooseye]

Some of you guys are out of the ball park. The guy is beating these practice coupons apart with a hammer. I don't think it is something he is going to have heat treated.
It is possible to achieve a quality weld with the equipment and materials he has on hand, but may take more expertise with the border line output.

edit Oh, and I forgot, practice practice practice!

[Evan Fabrication]

Thanks Guys

All good advise guys. I know I have a machine that is limited. It was all I could afford to get into a tig.I now have the right tungsten and filler. I'll have at it again tonite. After looking at some of the welds, it does seem like the bead is on top of the base material and not penetrating. And my original post was wrong. Machine is set to just under 135 amps. So I'm gonna crank it up and slow down a bit. Wish me luck! Do you guys think the tungsten @1/8'' being too big and out of range was causing my intermitting problem with getting an arc started?

[shovelon]

You can certainly still use the 1/8" tungsten, not what I would use, but you may as well use it up. Just grind a taper on it to help the gas flow through the opening of the nozzle. Also I can't imagine that the large tunsgten would cause poor arc starting, but ceriated does start easier than pure from my experience.

[shovelon]

Oh, and be careful of cheap chinese/import collets. They are soft copper instead of alloyed copper and tend to crush and block off the gas flow.

You have to ask for high performance/high temp collets. They are usually made in USA, and last almost a lifetime. They cost twice as much, but are worth it.

[SundownIII]

Evan Fab,

You've gotten "some" good advice along with a bunch of BS. Unfortunately, sometimes it's hard to sort it out.

First off, your machine is somewhat limited in the work you're trying to do. Add in an "inexperience" factor, and you've got a lot working against you. Remember that aluminum likes to be welded hot and fast. What you're doing with a limited output machine is overheating the base metal and causing a larger than necessary HAZ. This is due to the fact that you have to "wait" for the puddle to form. While this is happening, the heat (due to the thermal conductivity of alum) is spreading to the area adjacent to the weld, weakening the material. The goal in aluminum welding is to keep the HAZ as small as possible. A properly executed tig weld on aluminum "will nearly always" fail in the HAZ, rather than the weld bead.

I weld the type of material you're dealing with on a regular basis. I use amps you can't get out of your machine. With aluminum tubing, you've got to consider the overall mass of the material as well as the "thickness". When starting your bead, keep the majority of the heat (arc) directed at the "non coped" tubing. This will take the heat better than the edge of the tubing you're joining. This will also reduce the HAZ on the coped tubing (which is where most of your failures will occur).

Loose the pure tungsten on an inverter. Inverters do not like "ba11ed tungsten". Ceriated or 2% Lanthanated are a much better option. With your machine, a 3/32" tungsten should be plenty. Arc starts will be easier with the 3/32" than they are with the 1/8". I generally use a 1/8" 2% Lanthated tungsten, but I'm also using a lot more amps than you have available. Sharpen the tungsten to a "crayon point" not a "pencil point". With the Lanthanated, a small ba11 will form naturally, but not like the ball on pure.

I suspect what you're doing (most common) is overheating the coped tubing while not getting sufficient heat in the solid tube. This will cause the bead to "lay on top" of the solid tube. In other words, minimum penetration.

Also, keeping your arc length short will also allow you to better focus the heat where you need it.

One other thing you may want to try is using 3/32" 5356 filler. For your machine, you'd be cooling the puddle too much to use 1/8" filler.

For learning/practice, you may be better off starting with tubing cut square and welded to flat plate. This will give you the experience of controlling the heat without having to be concerned with the varying thickness of the coped tube as you progress around the joint. It's also easier to learn how to maintain your "angle of attack" as you progress around the tube.

Kinda goes without saying that the whole process is easier if you've already become "really proficient" with filet welds on flat material. If you haven't done that, you're trying to skip a major step in the learning process.

No substitute for arc time. Remember, what you're trying to do isn't "easy", and it's not going to happen overnight.

[GTA/SPEC]

You CAN use the pure tungsten for aluminum with a transformer welding machine with a ball formed on the end of the tungsten.
like the Miller Syncrowave 250.

Donald, the only place that pure tungsten still has in tig welding is if you are using a balanced sine wave AC machine. If you are using syncrowave, you have a square wave powersource and can benefit from an alloyed electrode. Discard the pure tungsten and dont Pre-Ball your electrodes.

[Donald Branscom]

Some of you guys are out of the ball park. The guy is beating these practice coupons apart with a hammer. I don't think it is something he is going to have heat treated.
It is possible to achieve a quality weld with the equipment and materials he has on hand, but may take more expertise with the border line output.

edit Oh, and I forgot, practice practice practice!

I agree MOOSEYE LOL heat treated...lol...lol

I believe he was breaking a test weld that was only welded on one side.
Since there is no photo, who knows what is going on and he does seem inexperienced.

[Donald Branscom]

Donald, the only place that pure tungsten still has in tig welding is if you are using a balanced sine wave AC machine. If you are using syncrowave, you have a square wave powersource and can benefit from an alloyed electrode. Discard the pure tungsten and dont Pre-Ball your electrodes.

I have used pure tungsten with my syncrowave 250 with better results on alumimum that 2% thoriated tungsten.
That is what I was taught in welding school.
Think about this:
Why would Sylvania even manufacture pure tungsten (green code) if it was of no use???

I think there may be a difference between the new SQUAREWAVE machines and the SYNCROWAVE machines.

I have heard that the new inverter SQUAREWAVE machines can weld aluminum with 2% thoriated tungsten instead of the pure tungsten and I accept that.

MAybe an expert on the subject can tell us all something about that.

[MMW]

Listen to Sundown III. He gives good advice.

The diversion is an inverter based machine not transformer based. The OP has a Diversion. Miller does not recommend using pure on any inverter based machine.

As far as using pure tungsten on syncrowaves you can keep it. I have been using a sync 250 since the early 90's & used pure & 2% thoriated. Pure doesn't work as well, a/c or d/c. I also have a Lincoln transformer squarewave machine & 2% works better than pure on that too.