Need welding tips

So I'm welding in some small sections of new floor sheet metal into my fiat and Im having a lot of trouble getting a good weld. I've cleaned to the bare metal on both sides and the old metal and the new metal are touching already, but I keep getting stuttery sporadic welds. When I practice off car with the same new sheet metal I can get perfect (perfect for my begginner standards) beads with little effort, but every time I try with the same settings on the car I get about 1/2 inch of bead, then *pop pop pop pop* then the whole stick out of the wire burns off and I have to stop. What am I doing wrong?

ps. 18 gauge steel, 135 amp Eastwood MIG, .30 wire 

Is the path between your welder and the floor much more convoluted than to where you did the testing? If you have a buddy who can watch the spool while you weld, see if it is jittering when the weld goes wonky.

I am impressed that you can lay a 1/2" long continuous bead on tin. I have to use the trigger pulse method when doing 18 gauge.

Pete

NOHOME said:

The problem I have with flux core is that sheet-metal requires a stop-start method of welding. In essence, connecting a series of dots to form a long scar.

The reason that this wont work with the flux core is because the bead is covered in slag (flux) when you stop. This slag would contaminate the nest dot of weld.

But...I see where you are willing to go for gas. Get MigMix: that being argon CO2 mix.Below is a quick tutorial on a process that works for me. It revolves getting your rig dialed in so that a short burst with the gun held in one position will give you a weld dot that almost falls out the backside. You then start welding off the front sloping face of that first dot and let the molten metal slide/flow into the seam.

Hope this helps.

The other thing I will mention is clean metal. When you grind or flap wheel the seam, you need at least 1" either side to be clean shiny metal. And the backside is just as important as the front side.If the molten metal comes in contact with anyting that it can vaporize, it pollutes your gas shield and the boiling/evaporating smegma turns your puddle into fizzy pop that just drops out the backside of the seam. Wiping the seam with acetone to remove any grease makes a huge difference.

I was already quoting this post from NoHome when he posted here because it's one of the most educational posts I've seen on the site and I refer to it pretty often when I'm welding. 

How's your light inside the car? How well can you see what you're doing vs on the test surface?

There should probably be an amendment to that post because Crackers has proven that you CAN in fact use the same technique with a flux core welder.

The OP does not mention if this is gas or flux welding.

A thought: Is the tip on the torch the right size for the 0.030" wire? 0.035" is a more common tip.

With light sheet metal, I find it far easier to increase the heat a bit, and forget about laying a bead.  A series of overlapping spot welds is much easier.

2 You could have some odd steel in the body of the car.  Maybe a light galvanising.

3. Make sure your torch cable isn't kinking or getting stressed.  I have a long one, and too many loops in it will restrict wire feed.

Where is the ground? I found when doing seam welding and rust repair that the ground made more difference than the weld prep. I took to welding a chunk of angle iron within 6-8 inches of what i wanted to weld. Attached my freshly cleaned ground clamp there. Made a world of difference. 

Dusterbd13-michael said:

Where is the ground? 

This.  Sometimes it's hard to get a good ground when working sheet metal.  An alternative to welding on a piece of metal is a magnetic ground post.

Magnetic ground post

My first thought was also a kinked torch lead but I'll second ground issues.

When doing bodywork you want your ground attached pretty much to the panel you're welding. Especially if it's a pre 90's car. 

Like Duster, I've also welded small bits to a panel for a good ground. Can be any size/shape just no thinner than the base metal you're welding. Hold the new piece with the ground clamp itself, buzz it on and move along. 

In reply to NOHOME :

I've been thinking about my successes with flux core and have made some mental notes about my process. 

I think the biggest thing is I weld a lot hotter than most guys do while butt welding and never give the flux time solidify. Even when pulse welding while filling holes I don't let my bead cool to black. 

The problem with welding this hot is you put a lot of heat into the panel, and inevitably warp it. I'm just not afraid of distortion because I know I can go back with a hammer and dolly and fix it.

If I'm welding in an area with poor access to the backside for dolly work I just skip around 1-2" and make dots then hit it with a wire brush in between passes. Wire brushing between passes sounds tedious (and sometimes it is) but ultimately you need to wait for the previous pass of welds to cool anyway. 

I think the other problem people have with restarts on flux core is they don't focus enough heat into the previous bead/dot. If you're going to restart with flux core without cleaning, you *have to* make sure the previous dot and the flux on the surface is penetrated well enough to melt that old flux down and bring it back up to the surface with the new flux.

Something I've noticed I do that I wasn't really ever aware of until sometime last year, is I literally wipe contamination out of the welds with my glove.

With flux core, contamination will float to the surface and bubble. Simply wipe that bubble off before it solidifies and hit it with welder again while the bead is still glowing. This removes the garbage, and the extra weld on top will replace the flux you just removed so the weld doesn't get brittle. 

I guess they're not really mental notes anymore. LOL

In reply to Crackers :

Alright here is some pictures to give you guys a better idea of what I'm doing wrong hopefully. I am using flux core wire but keep in mind when looking at these pictures that outside of the car with the same metal and setting I am able to get decent long beads with no stuttering. But on car there is this:

I am ground clamped right out of frame on both pieces of the metal like one of you suggested and still no dice. Obviously a couple of these holes are total blow through, woops, but a majority is not and that is my problem

What are your settings? I’ve found D-3 is close to where I need to be on my Eastwood for thin sheet metal (with gas and .030 solid wire.) Maybe not enough filler? I’d also recommend against trying to run continuous beads.

If you really have the same thickness material and fitup in the car and out I'd be taking a close look at the torch lead kinking.    That can be a real issue with some machines.  My ancient Snap-On YA212A industrial gets very fussy about twists and turns in the lead when the liner gets old.

Those panels aren't clean enough. 99% sure that's all it is. 

You can see paint or galvanizing adjacent to the seam. That looks like soot off a galvanized panel. Galvanizing will polish up and can look like bare steel (especially for the color blind like myself) and will definitely cause spatter. 

On second look, that could be an old leaded seam too. Getting all the lead off to prep for welding is next to impossible.

In reply to Crackers :

Muriatic acid will remove zinc, galvanized, cadmium coating and rust from steel. Use caution and be well ventilated. Better than grinding the coating off IMO.

Can't tell from the pics whether it was zinc or galvanized coated. If there was white deposits or 'white feathers' while welding then that could be the problem. 

Edit: Muriatic acid will remove lead as well.

In reply to Crackers :

Then what am I to do? Also just for clarification the bottom is the car, the top metal is the new metal, and the white stuff directly below the weld is just weld dust.

This looks like the burnt edge of some sort of coating. 

There could be contamination on the backside adding to the issue, but I think if you get all of that coating off the issue will be reduced greatly. 

It can be ground off with a fiber/flap disc, but it's difficult to get it all without gouging the base metal, so take care doing so. 

I know muriatic acid is an option as fasted58 mentioned, but I've never used it. 

Edit: i know it's likely too late to do anything about it now, but that second layer underneath should also have been cleaned out. If you get a puddle drop down into that it will immediately bring all that gunk to the surface.