[Coolidge]

I'm working on a project that does not require tight tolerances so in terms of metal moving/sucking towards the heat I figured I'd use this as a learning opportunity. Material is 2.5 by .375 CRS.

Here's what I'm fabricating



These T joints were no problem. I didn't clamp this together, I just used alignment pins in my table to position it square. The two fillet welds really did not suck the metal towards the heat, they came out surprisingly flat and square.



Now here's basically the same T joint. First I tacked welded the outside fillets, then I welded the inside fillets, then the outside fillets. As you can see it sucked the metal towards the heat pretty significantly.

Question: Are there any techniques to keep this from happening or at least reduce it? Death grip clamp it to a much thicker piece to hold it flat?



These welds won't win any TIG competition. It was surprisingly more difficult to weld this than running test beads. It didn't help that its freaking 98 degrees outside I can't wait for winter.

[Otto Nobedder]

If you want those joints square, post-weld, you have to actually pre-load them the wrong way, to about the degree of flex you see in the finished weld. This is where a Master fitter shines over a first-year journeyman. Developing the feel for how much a weld will "draw", and compensating in the fit-up.

I have to do this with pipe welds all the time. Since the inside of a 90 will draw just as much as the outside, but the circumference on which it draws is different, I must fit the 90 at 91-93 degrees, depending on thickness/passes, so it shrinks into the configuration I want.

Steve S

[Louie1961]

I am no expert but I think you also put more heat into those welds than necessary. It looks like you were maybe holding too long an arc, and/or not giving it enough amps. If you hold a tighter arc, and give it more amps so you can move along faster, you will put less heat into the metal, causing less draw. That's my theory anyway and I am sticking with it.

[big gear head]

Looks like the top is under cut some too, so you might have needed more wire. A tighter arc, more wire and good torch angle. It takes a lot of practice to make this look good. I use aluminum blocks when I can and clamp everything together to hold it square. I try to tack so that it can't move much after I unclamp.

[Metal_pig2001]

You can also have a go at straightening it by running a bead on the back. Don't add any filler just apply heat. Similar process to what Jody used to demonstrate in a video on straightening a shaft.
Used to be done by heating with an oxy acetylene torch
Regards
Ralph

[Boomer63]

For me, I would do as Steve (Otto) suggested. You can clamp the base plate to something stronger, and move your heat around more; do a bit on one side, go to the other side, let it cool, repeat. Or as Metal suggested, post heat on the reverse side of the base plate; if it were me, I would heat an area of about the size of a quarter (size based on the discoloration fo the metal), beneath each of the vertical elements. Move from heat area to heat area to spread the heat. Sometimes! Sometimes that takes out the distortion.

[Rick_H]

This is a good thread, welding is usually the easy part. I back bow large frames to help keep the straight like Steve mentioned above. I also skip around a lot and will sit stuff off camber and allow it to pull. Just takes experience and time to learn what works.

I do a lot of stainless and it will move I you look at it wrong...lol pulsing can help as well on thin wall tubing or sheet.

On something like you pictured above I'd leave it off camber a few degrees, and depending on how much bevel you put on the plate you can control where it goes. I also tack in a specific order, use backing plates etc...

[Coolidge]

Next noob question...selecting amps when welding different thicknesses together, in this case 1/2 thick plate to 3/16 wall square tubing, 304 stainless. A 4x4 1/2 inch plate will be drilled and tapped for a caster, then welded to a machine base made of 3x3 3/16 wall 304 stainless. Or I may weld it up then drill and tap it.

I'm guessing if I go much hotter than the tubing requires I'll make a mess of that, but will I get stick to the 1/2 inch plate at those amps?

[Coolidge]

For me, I would do as Steve (Otto) suggested. You can clamp the base plate to something stronger, and move your heat around more; do a bit on one side, go to the other side, let it cool, repeat. Or as Metal suggested, post heat on the reverse side of the base plate; if it were me, I would heat an area of about the size of a quarter (size based on the discoloration fo the metal), beneath each of the vertical elements. Move from heat area to heat area to spread the heat. Sometimes! Sometimes that takes out the distortion.

I suspect this was largely the problem. It was so hot in the shop, add the welding jacket and helmet I just blasted through the job as quick as I could without allowing time for it to cool down.

[Rick_H]

You could preheat the 1/2" plate a little if your worried, but you basically will set it for the thinner material easiest way for a beginner. Or go more amps but try to keep the heat in the 1/2" and just wash and dab the material to your tube.(my preferred method to ensure penetration.

Off top of.my head I'd use a 3/32" tung, 180-200 amps and see how it welds. For comparrison sake 1/2" plate id be in the 300amp range

[Otto Nobedder]

Next noob question...selecting amps when welding different thicknesses together, in this case 1/2 thick plate to 3/16 wall square tubing, 304 stainless...

This is another clear case of "everyone does it different." I've read some of the other replies.

I'd dial it high for penetration in the 1/2" material, keep the arc on the half-inch so the puddle wants to "climb" the 3/16", and use a quick "swoop" (if that's a word) as I add rod, to tie in the thin side. The addition of rod AS you move to the thin side will help chill the puddle and prevent undercut.


Steve S