TIG penetration and other questions

[Kubs]

I'm currently in the middle of a project and I'm working on some fillet welds with the TIG process. The material I'm working with is 2"x1.25"x.125" rectangle tubing and I'm mating a 2"x.120" round tubing perpendicular to 2" side of the rectangle. It's a form fit with a hole saw cut on the end. The rectangle tubing is ~19" long so theres a big base there. The round tubing is 1-15/16" long centered on both ends of the rectangle. So basically I have a fillet weld thats 19"x7/32" and obviously circles around the 2" side of the rectangle tubing.

Equipment setup
DCEN
Amps 120'ish
3/32" 2% thoriated tungsten
Argon gas w/ gas lens and a 3/8" cup
3/32" copper coated ER70s Filler / 1/16" non coated steel

My problem
The fillet weld isn't going smoothly. I'm undercutting the crap out of round tubing. The arc doesn't seem to wanna focus for penetration and goes wide. The arc also wanders a bit. I tried to feed it with filler rod fast but it seems to wanna climb round tubing and give me problems.

I haven't dabbled with TIG on a regular basis since I first started welding 5 years ago. I took a class just to get a base to learn. My first machine was my TIG and immediately I was producing near perfect welds, including fillet. I dunno where I lost it in the recent years and it's actually kinda agitating. I tried switching between the two different filler rods I listed and notice no difference really. I sharpen my tungsten properly and give it a long point and stick it out about 1/4"-3/8" from the cup. I am new to the gas lens and have been told it allows the tungsten to stick out further. I'm beginning to wonder if sticking it out exposing the tungsten more is causing the wide arc but then again I have had the tungsten out further when I didn't have a gas lens and had no problems before. So I'm not sure.

My flat weld portion of the project seems to be good although I did note my weld pool was wider than normal but I was also juicing it to get good penetration which I know I have there but I'm not sure about covering the rounded portion.

Also note that the weld area is cleaned well.

Looking for any insight as this is a structural weld. If theres anything I haven't covered feel free to ask and I will give my input so I can get sound advice. Also if theres any tips to ensure penetration for a fillet weld using the TIG process I'm all ears.

[DesertRider33]

It sounds like you've done a good job of covering all the bases to a tee. I'm not a TIG expert, still on the learning curve myself, so the only thing that comes to my mind is maybe you're holding too long an arc length (tungsten to work distance). The closer the better, about 1/16" is good. The other idea I'm thinking is try adding the filler to the top part of the puddle, up on the tube, to lessen the undercut on the tube.

Like I say, I'm no expert, but I doubt your gas lens is causing the problem.

Hope you get it worked out. Sounds like you've thought it out well already.

[Rojodiablo]

Originally Posted by DesertRider33

It sounds like you've done a good job of covering all the bases to a tee. I'm not a TIG expert, still on the learning curve myself, so the only thing that comes to my mind is maybe you're holding too long an arc length (tungsten to work distance). The closer the better, about 1/16" is good. The other idea I'm thinking is try adding the filler to the top part of the puddle, up on the tube, to lessen the undercut on the tube.

Like I say, I'm no expert, but I doubt your gas lens is causing the problem.

Hope you get it worked out. Sounds like you've thought it out well already.

Some pretty good advice there. I would add, try thicker filler rod. Try 3/32 filler. Also, concentrate the arc on the main tube, and like DR said, try to add the filler on the high end of the puddle, to bring it down into the base pool.

[Kubs]

Desert,
I'll try feeding from the top next time to reduce undercutting

Rojo,
I was using the 3/32 filler rod primarily, I switched to the 1/16 briefly to see if a control difference would help and other than just feeding faster there wasn't a change.


Here's some pics of the finished product. They're control arms for a camaro so I'm looking for some good sound feedback if the weld job is adequate (looking for overkill) for the application they'll be used for. Like I said, it's been awhile since I've TIG'ed and maybe I'm just paranoid that my welds don't look perfect. Sorry I didn't get any shots before I painted. Excuse the quality, my digicam didn't wanna focus and the shots turned out "better" using my cell phone instead. So bare with that. You can see in the 2nd pic where I had some issues.



Should prolly also add some scrap pieces I welded together and did a lil destruction test of my own. I welded both sides of the round to the rectangle and grinded off the side I thought had better penetration going for the what I believed to be the weak side to be tested. This is more of a demonstration of the flat weld of the 1.5" weld of the project piece, except on the scrap I was on the 2" section. This result of this was tearing of the rectangle.

[papasloan]

Like you said the arc goes wide. Try regrinding the tungsten. A little contamination can do that. Also try a little different angle. Dirt steel seems to contaminate the tungsten very quickly.

[Supe]

On material only .120 thick, especially near an edge, you can typically reduce your amperage quite a bit. 120 amps is on the high side for that. Stick with the 1/16th filler, by the looks of it, you want to keep the puddle a bit smaller. You shouldn't have to dab filler from the topside, that's just a way of compensation for a puddle that is too wide. A 1/16 tungsten wouldn't hurt from a control standpoint, either.

[David R]

MY Opinion, too hot! BACK OFF a little, run a smaller bead. or a few smaller beads. 3 little stringers instead of one wide bead. Arc should be not more than one electrode diameter.

You are close, try on some scrap.

David

[Kubs]

Thanks for the input guys.

OK here is some weld in the raw. I decided to start building another set and as I go post my progress and see what you guys think. I dropped the heat down considerably and have several different welds with various heat ranges. I resharpened my 3/32 tungsten and went at it. Here's the results:

Here is 80 Amps using a 3/32 ER70s copper coated filler rod.


Shot of the inside of the round tubing




Here is 75 amps using a 1/16 non coated steel filler


Inside the round tubing again




Here is one I did at 70 amps using the 1/16 filler. I didn't finish the pass because I wanted to demonstrate a look down the valley so I could get some opinions to benchmark my performance


Inside again




And lastly here is my fillet weld that I'm most concerned about. 70 amps here with 1/16 filler rod


Here's a couple shots from either side to show you so you can examine the corner of the fillet




I'll appreciate the feedback.

[William McCormick Jr]

Looks strong.


It looks like you could use a bit more gas. I actually like a golden brown color if I can get it.

You might also want to try 1/16" filler wire. It will make the weld much hotter, and narrower. Allowing you to penetrate the same, but move along at three times the speed. As you add the thinner wire it does not cool the puddle as much. When you are done adding the wire, and letting it burn in and level out, you can cool the puddle by coming off the pedal.

It looks like you are moving along really slowly and creating a really wide hot area. Which is not terrible. But most prefer a more narrow weld.

Sincerely,


William McCormick

[William McCormick Jr]

I just realized you are using 1/16" filler wire. It looks like it is going on cold to me. Like you are using a larger wire.


Sincerely,


William McCormick

[William McCormick Jr]

Ha-ha, now I understand you are using both types. Where you used the 3/32" it looks like it went on just a little cold.



Sincerely,


William McCormick

[OldSparks]

Kubs
In most cases the weld you made below would be perfectly accepable in the sense that it is is tied in nicely and flat to convex shaped. However in this case the item is structurally significant. I think of tig as a 'finess' form rather than just rapid production. If I was to do that joint I would put a 45 degree bevel on the rectangular tube, run a quick pass to fill groove, then run a fillet like you've done over top and never think twice about it.

[Kubs]

Originally Posted by William McCormick Jr

Ha-ha, now I understand you are using both types. Where you used the 3/32" it looks like it went on just a little cold.



Sincerely,


William McCormick

But you think the heat for the several sections I used 1/16 is adequate?

Originally Posted by OldSparks

Kubs
In most cases the weld you made below would be perfectly accepable in the sense that it is is tied in nicely and flat to convex shaped. However in this case the item is structurally significant. I think of tig as a 'finess' form rather than just rapid production. If I was to do that joint I would put a 45 degree bevel on the rectangular tube, run a quick pass to fill groove, then run a fillet like you've done over top and never think twice about it.

I did put somewhat of a bevel on the radius of the rectangle tubing. I know the photo's aren't the best so I drew a line to maybe help paint a better picture



I dunno if that would be considered an adequate bevel or not. But you think a small root pass then a cover pass would be the optimal way to go with this? I too don't look of TIG as a means of production but more along the lines of precision. Do you think it would be possible to do this in a single pass if the bevel was a bit wider, perhaps something like 60*?


Also, speaking of bevels. Do you think this particular setup to weld the flat part is adequate? I have it beveled, I would think thats fairly visible despite the pic quality, and the penetration on the round tubing seems pretty decent and the weld is about 50/50 so on each piece. But there is that hair line gap at the very base of the root.

[David R]

I think your welds look good. MUCH better than the first attempt. In MY opinion, they are fine. I would dip more often. With the heat down, it looks much better and I see no undercutting.

Is that a control arm bushing mount?

David

[William McCormick Jr]

The 1/16" filler wire definitely looked better. But still a bit cold for something very important. You might want to as the other fellow mentioned dip more often. Or Heat add wire cool, and move many more times in tiny increments. This creates better mixing of the filler and base metal. Because you will not be adding globes of filler wire, and overpowering the base material.

They are two different things. The filler wire is supposed to be added to rejuvenate the base material, add cleansing agents and to bring the weld up to the level of the surface. It is not really made to be globed on.

Years ago a weld was a weld, one pass. Many passes was considered something you needed a black smith license for.

One thing I was always taught was that on structural it is better to be a bit hot then a bit cold. A little undercut is not your enemy. Take a look at many factory made parts they know that hot is better then not undercut.

I also noticed that you put most of the heat into the rectangular tube. When most of the heat should have been into the round tube.

Which ever piece meets with a cut edge that is the edge you want to keep the heat away from. You want to add the heat to the flat uncut surface. That is why you were getting the undercut in the tube. It was cold and feeding the puddle.

You might be surprised how long it takes to get a puddle like that though without filler wire. Most people get nervous and just start to add filler wire or put the heat on the cut edge and add wire.

Sincerely,


William McCormick

[Kubs]

Originally Posted by David R

I think your welds look good. MUCH better than the first attempt. In MY opinion, they are fine. I would dip more often. With the heat down, it looks much better and I see no undercutting.

Is that a control arm bushing mount?

David

Thanks and yes, its a control arm bushing mount.

Originally Posted by William McCormick Jr

The 1/16" filler wire definitely looked better. But still a bit cold for something very important. You might want to as the other fellow mentioned dip more often. Or Heat add wire cool, and move many more times in tiny increments. This creates better mixing of the filler and base metal. Because you will not be adding globes of filler wire, and overpowering the base material.

They are two different things. The filler wire is supposed to be added to rejuvenate the base material, add cleansing agents and to bring the weld up to the level of the surface. It is not really made to be globed on.

Years ago a weld was a weld, one pass. Many passes was considered something you needed a black smith license for.

One thing I was always taught was that on structural it is better to be a bit hot then a bit cold. A little undercut is not your enemy. Take a look at many factory made parts they know that hot is better then not undercut.

I also noticed that you put most of the heat into the rectangular tube. When most of the heat should have been into the round tube.

Which ever piece meets with a cut edge that is the edge you want to keep the heat away from. You want to add the heat to the flat uncut surface. That is why you were getting the undercut in the tube. It was cold and feeding the puddle.

You might be surprised how long it takes to get a puddle like that though without filler wire. Most people get nervous and just start to add filler wire or put the heat on the cut edge and add wire.

Sincerely,


William McCormick

I did favor the rectangle a hair bit more than the round, but with the round being thinner and I was getting penetration real good on my flat welds I figure I could lay off just a smidge. When I first start the weld I work the base metals into a pool then I add filler and move in about 3/32 maybe 1/16 increments and push a lil filler into the pool then move. I think the other reason I was getting severe undercutting, aside from the heat, was the way I had my torch slanted going around the radius. In my recent attempt I kept it focus in the valley pretty much spot on while holding a 45* angle favoring the thicker material slightly.

I'll head out in a minute, bump up the heat a few amps, and do another pass and see what you all think. Like I said they're control arms and I need them to basically take alot of compression force with a bit of torsional force as well. Manufacturers use thin 1/8" or so, concaved welds on their products and don't seem to have issues with strength. Here's an example:

[Supe]

If all the welds are done like that 1/16th @ 70 amps, on a set of rear control arms, you will have no problems with reliability at all.

[Kubs]

Originally Posted by Supe

If all the welds are done like that 1/16th @ 70 amps, on a set of rear control arms, you will have no problems with reliability at all.

Just to confirm, you're talking about these correct?

[David R]

I think Supe is saying go around the whole bushing mount just like that both sides. I don't see Any undercutting.

David

[William McCormick Jr]

Originally Posted by Kubs

We always leave that edge straight and do not bevel it. The edge that is left will wash in and you will need less filler wire. That is how we do it. Sometimes if we machine or bore the part, it causes sharp edges. If it does we grind them back straight. But we do not chamfer or grind them past 90 degrees.

I did not realize the square tubing was thicker then the round tubing. That is not always preferable when welding things together. It is best to keep the material the same thickness when designing if possible.

It would be like welding a thick wall pipe to a thin flange. In the thin flange scenario you realize, that is not such a good idea with all that leverage on a thin plate.

One and one quarter inch pipe, schedule forty, welded with 7018 rod, to an approved flange 11/16" thick, five and one quarter inches round, is rated for 600 pounds using true engineering standards. I just posted this to give you an idea of the strength of the lever. Often when destructive testing and designing they lighten up the arms so they give or flex as well. And do not just put an unyielding tearing force on the weld to the main branch pipe or tube.

Sincerely,


William McCormick

[Supe]

Originally Posted by David R

I think Supe is saying go around the whole bushing mount just like that both sides. I don't see Any undercutting.

David

Yep. Do it like those two pics and weld around the entire perimeter and you'll be fine.

[Kubs]

Originally Posted by William McCormick Jr

We always leave that edge straight and do not bevel it. The edge that is left will wash in and you will need less filler wire. That is how we do it. Sometimes if we machine or bore the part, it causes sharp edges. If it does we grind them back straight. But we do not chamfer or grind them past 90 degrees.

I did not realize the square tubing was thicker then the round tubing. That is not always preferable when welding things together. It is best to keep the material the same thickness when designing if possible.

It would be like welding a thick wall pipe to a thin flange. In the thin flange scenario you realize, that is not such a good idea with all that leverage on a thin plate.

One and one quarter inch pipe, schedule forty, welded with 7018 rod, to an approved flange 11/16" thick, five and one quarter inches round, is rated for 600 pounds using true engineering standards. I just posted this to give you an idea of the strength of the lever. Often when destructive testing and designing they lighten up the arms so they give or flex as well. And do not just put an unyielding tearing force on the weld to the main branch pipe or tube.

Sincerely,


William McCormick

I apologize, they are both the same thickness. Relatively, .125" rectangle, and .120" round (see my OP). I kept thinking it was .095" as of late which is a chromoly tubing I have for something else entirely. I'm sorry for the confusion.

But back to the bevel... Judging by the pics I posted with the ends of the fillet weld visible it doesn't look like I'm going any deeper than what I beveled, I though the purpose was to go all the way or most of the way through for solid penetration? Maybe I'm missing a few things, but thats what I'm hear for

[William McCormick Jr]

This is a test piece I did. It is hot rolled and a little pitted. I cleaned it up, but it was still pitted. It came out pretty good though. I would be happy if a run of parts looked like this.






I went a little heavy on the square piece myself, Ha-ha.



Sincerely,


William McCormick

[William McCormick Jr]

Yea, at that angle it creates an optical illusion or something. If I had not said something to you this morning about putting a lot of heat into the tube. I would have gone out and undercut the tube by a mile. That was with me purposely keeping the heat on the pipe. Ha-ha.

Good practice test!

Sincerely,


William McCormick

[William McCormick Jr]

I used pretty low amperage to that. I would say about 90 amps at most. Probably less. I start off slowly letting the part get some heat in it. And then I go a little faster once it is hot. I used a number 7 gas lens and 20 cubic feet an hour.


Sincerely,


William McCormick