I've begun a thread to discuss welding techniques and tips, to try and help out all the non-welders out there. First post was split off another thread - please keep all thread-specific problems in the original thread, and keep this thread clean and full of useful universal information.

Basic MIG theory:

1. MIG (or GMAW) is based on a constant voltage (that's the "heat" dial) and a consumable wire electrode (wire feed dial - if fancy, it's in inches per minute - this controls both wire feed and current (amperage)).

2. The shielding gas only performs a shielding and cleaning function - this is fundamentally different from say TIG welding. It keeps the molten weld pool in a different atmosphere while the weld is forming and during initial solidification. Set it and forget it.

3. Material thickness controls amperage for all weld processes (that involve electricity) and for GMAW, 1 amp per thou of thickness is a general rule for steel.

4. Wire is consumed at a rate that is proportional to the wire size, amps and the voltage applied. For example, if your welding sheet metal, I'll assume you're using either 0.030" or 0.035" wire. The guidelines indicate that for 0.030" wire, the wire feed rate should be 2" per minute per amp of current. So the "heat" dial sets the voltage (think depth of penetration - voltage is a force) and the wire feed controls the rate of wire consumption (current or rate). That's how they compute the charts on the inside of your machine that tell you the settings for a given gauge by a wire diameter.

5. The length of the arc that you maintain controls the current...this gets a little tricky, so let me try to explain it this way. All of what I've said so far makes sense (voltage, wire feed, amps) if and only if the length of the arc is a constant. If you need to "cool off" your weld, lengthen your arc or to heat it up, tighten the distance. One of the mistakes I see with beginners is holding the MIG gun so that they can see the weld and therefore holding the arc a long way off the weld. Push that sucker down into the puddle. This is what's called "Stick Out" which refers to the distance from the electrode being consumed to the weld puddle. A normal arc length is around 1/4".

For practice, I suggest you do the following:

1. Grab some scrap material (1/8" flat bar) and set the voltage and wire feed based on the guidelines with your machine (I'm assuming you're gas is set properly). Then hold the gun with both hands and set a distance 1/4" off the target (get a feel for that distance by setting your tip on a few pieces of scrap that make up that height). Now move off your support, and with your helmet down, pull the trigger and move (push/pull...I don't care) and see what happens when you hold the gun straight down. Now you'll need a bit of angle, so try 10-15° and go again. Adjust only the wire feed knob until you get the wire melting but not bumping into the puddle and not melting back into the tip. It should sound like frying bacon...now you're cookin (sorry - couldn't resist).

2. Go weld - you're now a pro.

Craig

http://www.millerwelds.com/pdf/mig_handbook.pdf

Great idea for a thread, Craig! Sooper Genius!



Can you elaborate on what the shielding gas does for TIG please? I assumed it did exactly the same thing and nothing more. I know the gas also cools things down but that would be a peripheral effect, and unless the gas flows are significantly different I would expect the effect to be the same between MIG and TIG.

Yes, I come up with these great ideas from time to time.

While the gas for MIG performs a shielding and cleaning function (so an atmosphere for it to cool in that is more appropriate for proper weld formation) and to provide a degree of cleaning of the metal (not like removing rust or paint...you can't weld rust or paint), but to the base metal. If you weld with insufficient gas the nitrogen, oxygen and hydrogen (in that order) will cause weld porosity and excessive splatter. Try it - turn your gas way down, then inch it up and see what happens.

The kind of gas used for MIG also controls the nature of the penetration of the weld (that's why we'd switch between different mixes for different situations as well as different materials). The arc occurs within the shielding gas - so it plays a role in arc stability and focus. Pure CO2 would produce a very deep penetrating weld (from my understanding) and the addition of other inert gasses are used to reduce this effect. Helium can be added to really heat things up.

OK - back to the question. For MIG, (and I'll be really specific here as we are talking about short-circuit transfer not spray arc (I'm the only one in the group that has a MIG capable of spray arc - which is an awesome process but not used in welding the thicknesses of steel/aluminum we're talking about). This is what I recall from several schooling sessions from about a million-years ago (and should be googled, but I don't have time right now). Short-circuit means that the wire will contact the puddle (as a molten blob), causing a momentary short, which kills the arc, which is then restarted when the blob transfers to the puddle, which starts the arc again. I recall seeing this in super-slow motion on YouTube. If you've spent a lot of time behind a MIG gun, you get a sense of this action - you shouldn't feel the wire hit the puddle, that's too much wire, but should be just a little slower than that. OK - so this is the sort of weld that we're performing - a short-circuited arc where there is physical contact between the two electrical poles.

http://youtu.be/Rsr_rzAP35A

The shielding gas provides an appropriate atmosphere for molten metal, focus and penetration and for MIG is not particularly sensitive to flow. Again, try it - don't ever take my word for it. Noting much bad will happen if you crank up the gas flow...the arc is still maintained in the stream of gas, it's just that you're wasting gas. Too little is bad and will cause crappy looking welds, but too much doesn't do much. You don't see a list of different critical gas flow rates on any charts.

Depending on the size of the nozzle I use, I usually set it around 15 CFM and forget it. If you let your nozzle get all snotty - crank it up a little or clean your nozzle silly .

TIG is not ENTIRELY different as I've boldly stated, it's similar (MIG after all was hatched from TIG), but the gas is used as the medium of electron transport (the plasma) for the weld as well as the inert atmosphere. So in that regard, they are somewhat similar. Think of it this way, in MIG not enough gas is bad, but it's pretty relaxed once you hit "good". TIG has two slopes to this relationship, not enough produces unsatisfactory results every bit as much as too much and, not unlike the Three Bears, one setting is "just right", where you have enough gas to do the shielding and arc transfer but NO MORE. More gas for TIG makes things unstable and basically cools the arc - it doesn't cool the work piece as you're implying in your question - I breathe harder than 20 CFM while at rest (and don't google that for heaven's sake ).

A good example of this was last week when I was building a camera mount for an airplane. I was TIG welding 6061 seamless 0.030 wall tube (1/2 diameter) and I didn't change my flow from 20 CFM (which I was using a moment before to weld some 1/8" 6061 angle). 20 CFM wouldn't allow a stable arc (plasma) at the low amperage and I ended up blowing a hole through the tube to just light the electrode (I stomped on the pedal like a moron). So with that expensive piece of scrap tossed out, I dialed it down to 10 CFM and was happily welding (or as happy as you can be welding tin foil). No change in amps.

That's a golden nugget right there. I'll do my best to keep that in mind, as I tend to "set it and forget it" with the gas. I suppose I should invest in a flow gauge and a chart of some sort (and maybe dust off my TIG one of these days).

Your TIG should have come with a flow meter...I prefer the floating ball type over the dial type and have a "T" set up and two flow meters for back-purging stainless. For TIG, the most important things to consider are:

1. Gas flow (yes, it's first).
2. Arc length - you have to maintain a "tight" arc, which means you're very close to the material so that the arc is focused. If you see a cone, you're too far back. This one's a fun test, Set your TIG so that the max pedal is what you'd use to properly weld a test part...lets say 1/8" steel, butt joint, flat position. So set it for 125 A (DC of course). Then start welding by using full pedal with your normal arc length. Weld 1". Stop, let the part cool...then do it again with a longer arc - see how it takes longer to get going? Note that the penetration isn't very good? No tighten up the arc until you're at the point where you're not seeing much of a cone and you're scared you're going to hit the tungsten. That's about right...see the difference? Most people think they need a bigger TIG, what they need is better technique. This was a key lesson for me.
3. Don't let the filler rod out of your shielding gas. Use a gas lens set up - especially for stainless and aluminum.
4. The PUDDLE melts the rod, not the arc (well, not directly). I feed rod to the puddle offset to the arc - this is critical when welding material of different thicknesses as the arc will be hotter towards the thin material. You can offset feed the puddle to control the heat into the thinner material.
5. Use just enough torch angle to see the puddle and no more.
6. Fill the crater, especially on aluminum.
7. TIG requires very clean work pieces and rod...but acetone is only if you're building a pressure vessel...I don't bother.
8. MIG can "tolerate" welding junk - but why would you weld rust, paint, mill scale? I prefer to weld steel/stainless/aluminum. TIG won't tolerate contamination. I see this all the time, poor weld prep, welding through mill scale is probably the most common or paint/rust on bodywork.

Such is your influence, I literally ran right out and bought a bubble-type flow gauge. I forgot to ask if they had any laminated charts or such for amperage & flow rate, but that'll come next I guess.

I may dust it off and practice some this weekend. I've got a pile of things I want to build that'll depend on the TIG, so at some point I'll have to tear that scab off and get welding.

I've been playing with MIG settings and techniques for sheet metal. I will have to start a metal shaping/bodywork thread at some point in the near future, but in the mean time, I thought I'd share a few observations on MIG welding thin metal.

HEAT in the material is your enemy. It builds as you weld and if you're welding full penetration sheet metal - it's hard as you'll start to blow holes pretty easily and that get's old quickly. I've been developing my metal shaping skills and metal working welds so that they are flush with panels. This is sort of the holy grail of body work - panels needing little to no filler over repairs. If you can get your panels to fit really well, then you don't have to use much body filler and that's my goal. So I want a small bead. This is not what's usually produced.

The current technique for welding sheet metal successfully is to string together a large number of tack welds. This technique controls the heat input, reduces the risk of blowing holes in panels, produces a large weld deposit and is relatively easy to accomplish (relatively - you need to develop a feel for heat). This is by far the most popular technique, in fact my big Miller 252 has a specific setting that will allow me to lay these tacks without taking my hand off the trigger. It pulses the wire and is very fast and works great.

Now using this technique you still can't weld more than a couple of inches at a time as the front edge of the weld puddle is still gaining heat as you move. It's important to remember to get out of an area even if it's not glowing any longer. You'll be tempted, then start blowing holes, then get ticked off and frustrated. Not to mention a big ugly blob of nastiness to clean up with your grinder as you fill the holes. Then you'll pound the heat into the panel with the grinder, shrinking it more and probably be left with a thin, weak spot.

Some guys try to control this by cooling the weld with a blast of compressed air. Don't do this...it makes a hard weld even harder. Just move and stay relaxed - if you're starting to make a mess, stop...tomorrow is a better day.

Here's the back side of a panel that I was adding a 3/4" strip to...this was welded using the multiple tack method. Note that there is no way to metal work this seam as the bumps won't form to the panel. They have to be carefully ground and planished to make them perfect. When you grind, only grind the weld. There's hardly any metal there to start with, it doesn't take much grinding to ruin a panel.



If you grind and carefully file the joint, you can finish with a weld line like this.



Now, to get to that point with this technique isn't easy - and the weld technique is the issue. The commonly held belief is that for sheet metal, the weld bead will be 4 to 5 times the material thickness. This same adage does not hold true for welds in thicker material where your weld bead (on a butt joint) would be no LARGER than twice the material thickness and more like 1.5 if your careful. So why do the rules change with thin stuff...just think about laying a 1" wide bead on 1/4" material...that's insane. Well, I'm going to tell you my theory.

It's easier to lay tacks and get consistent, full penetration welds on sheet metal for technicians without much skill. You then cover your tracks with body filler and it's good, strong and will last. That and it's damn difficult to control things the way I'm going to propose as an alternative.

So what's my gripe with this method. Well, it lays on a relatively large weld deposit that you then have to grind. This can lead to thin spot, brittle areas and everyone knows that body filler is the answer. We're working in our garages, body filler is not evil, but it isn't metal either! If the panels are carefully fitted, there's no reason to not have a nicely finished weld.

With the large weld deposit, lots of heat is applied to the weld zone. This causes the metal to shrink as it cools, this causes a regional low on either side of the weld bead. As you grind the weld deposit off, how do you know that you're not girding too much and causing the sheet metal to get thin? How thick is the weld? The grinder also adds to the heat loading and subsequent distortion of the panel. I've done this plenty of times and found that I hit the panel trying to flatten the weld, left it thin and still had lots of weld. It'll drive you nuts or you just level it out and fill it with body filler. I've never been satisfied with that as an answer to what should be a small bit of filler can turn into a ton of the goop.

So here's the technique I've been playing with. I'm going to start by tacking a panel together every couple of inches and make sure that the panels are butted together relatively tightly (as close as you can get). Guys in body shops flange panels and I've done this for years and it works, but the overlap means that working the joint is almost impossible and sealing the back side of the panel is difficult. So let's not flange unless you're just spot welding (plug welding) a panel. A tight fit is critical to this working. Spend the time to make your panels fit with no gap (or as small as possible). Larger gaps mean more heat.

This is what I mean by tight gap.




You're going to pull a bead on this panel. Yes you are...you're going to move fast, short and carefully. Think and plan each hand move before pulling the trigger. Now start with the tip of the wire on a fully cooled tack on the edge towards the direction of weld. Pull the trigger and MOVE. If the puddle doesn't get established, it will skip and sputter, but not make that big a mess. Once you try it a few times, you'll see that you can pull a small bead about an inch at at time. These welds will be about twice as wide as the material is thick. Perfect.

Penetration will be just through the panel - easy clean up if you can access it, if not, there are no bumps to work around when working the panel. There is also a lot less heat input and less distortion.

Tips to make this work.

MOVE relative fast...you have to start on the tack to stabilize the puddle.

Don't get greedy, you can still only do an inch at a time due to heat input.

You'll still blow holes every once in a while, but fewer than with tacks.

Use two hands on the MIG gun...one on the trigger, the other directing the nozzle.

Keep the stick out really tight...really tight. This is also really hard, but critical. Try different stick out lengths. You'll find it works best if you're jamming the contact tip almost right into the puddle...which makes things hard to see. You'll figure it out. Try it with longer stick out, this cools the weld and it's not going to penetrate. Get it tight and you get full penetration. Get the gun angle so you can just see what you're doing. The welds shown below were done on a fender well - so some of it was welded fully overhead - don't grumble about flat panels being difficult until you've tried overhead.

Here are a few examples. Look at the size of the heat affected zone (it's small). It's easy to dress these welds with a grinder and they aren't showing as much distortion. Give it a try and see if you like it. It demands more of the fit-up and that takes patience, but I've metal worked a few feet of these sorts of seams and it's pretty rewarding to have them come out needing only the smallest little bit of body filler.

Here's one that I've run the grinder over just to knock the top off of it...note how consistent the height of the bead was. This is going to clean up really well.




Here the HAZ shows where I started and stopped and also picked up speed or slowed down. It shows the start and stop points really well.

That's cool, Craig. It looks like it takes a lot more skill and control, but the results should speak for themselves. I'm sure most of the time our problems are down to fitup and not working on sound and clean metal, and you've shown how preparation benefits the job. At Coburn Performance, no sitchation is too problematic to prevent us from gettin er done!

I finished that seam this morning...took 1.5 hours from grinding (inside and out) to completely happy. There's barely a need for any filler, I think just some primer surfacer would be fine. So much easier to deal with than the spot welded one that I did the day before. I'll have to continue to improve the technique, but so far, I'm sold.

Here's the same seam ground from the inside - note the difference in weld bead width between this seam and adjacent seam that was welded with the multi-tack method.



I can't even feel the seam once it's worked. I'm using my slapping file to level out the sheet metal. You really need to use a slapper or heavy spoon not a hammer to planish the panel. The file teeth on the slapper leaves a mark on the highs...and you work around the panel until it's evenly marked. At this point it's smooth.



The back of the panel. Body filler would just take a minute.

Craig's master class on warp-free sheet metal repair

1: Get the fit perfect:


2: What you're shooting for:


3: Start tacking, about 1 to 1.5" apart. The parts will start to pull together downstream of the direction your tacks are travelling, so after a couple tacks... stop welding and start grinding the tacks flat.


4: Backing the weld with a dolly, tap the last tack with the edge of your special hammer to stretch it a little. This is the classic hammer-on-dolly method, what for to stretch material. Stretching it relaxes the gap that previously pulled too tight. The tack MUST be ground down to make this work - hitting a big blob of weld does nothing but make a mess.


5: You may need to get your fuzzy beard in the way of the camera man in order to adjust the bottom of the panel, assuming you have access (or a fuzzy beard, or a camera man)


6: Keep tacking, grinding, tapping, checking the fit. This takes forever.

7: Now comes the sporty bit: you start to weld, FAST. Tiny thin beads strung between alternating tacks. The speed controls the penetration: you start relatively slowly and as you build heat in the panel you need to move faster. Keeping the speed and direction is tough enough when your panel fit is perfect, but if you hit a thin spot or a gap you'll blow through in a hurry. Craig makes this look easy.


8: Use lots of air to cool the tacks and keep the work piece temperature down.


9: Go back and string weld between the beads you made earlier. Overlap the start and finish to eliminate pinholes.


10: Using the EDGE of the grinding wheel (never the face, which delivers too much heat to the panel and will warp it) slowly grind back the weld bead to the surface of the panel. Go slowly, using air to keep the panel cool, and work the grinder down the weld string.


11: Don't go a lot further than this. With a couple tappa tappas from the hammer you can get your surface pretty darn flat. When you're this good, a little high build primer will finish the job. If the fitup wasn't perfect you will rough up the surface and start the kitty hair & surface filler process.

Awesome tutorial guys...

Can you tell what kind of welder and what settings ot had just to get idea of the amperage and the wire feed settings? .023" wire I assume?

I have some fairly decent patch panels to put into my wagon so sheet metal welding is of great interest.

One question: As the gap gets tighter is it better to grind and stretch the tacks as shown, or trim / cut metal so that as it closes gap itself?

Rabin

It's the ubiquitous Lincoln 100 amp 110v welder. We both switched between current setting 1 and 2 depending on the application: I find the tacks for setting 1 a little cold for the thicker stuff, but for 20 gauge 1 is fine. On my machine I have to turn up the speed a little for out of position welds, I'm sure it's the same with every MIG welder but this is the only one I have real experience with. I'm using .030 wire. I'm sure .023 would be alright in my machine but it has trouble pushing through the torch and can create welding inconsistency when the torch is wound around your legs compared to laying it out straight. Craig's is longer than mine (it's genetics, guys) and .023 was impossible to use on his machine. His is super fancy and has much finer adjustment in current and welding speed which makes mine easier in a way - you just have to get used to it.

Not an expert by any stretch, but I think I can answer this question:

In a couple places, most notably on the lower rear fender, we had a lot of shape in the original panel that wasn't desired (impact damage). We needed about 10 more hands to hold these two long edges together for accurate fitup, and even then the fit wouldn't have been really good. In this case we tacked a couple and found a tight spot, ground that back, tacked a couple more, found another tight spot, ground it back, etc. I was pushing on one panel from the backside while Craig was stabilizing the other panel from the outside until he could tack it. Helpers required, for sure, and it would be impractical to expect to get the fitup perfect on a part that doesn't have the right shape in the first place.

In this case, Craig was using a higher amperage on the welder (position 2) to get solid and quick tacks, and dropped down to 1 for some of the welding. Some of it was done in position 2, which made it easier when the road got a little rocky.

With the part in the example I showed in this thread, the fitup was perfect to begin with so why change it with the grinder? Neither part needs a change of shape, so if you reverse the localized distortion at the tack with a hammer & dolly, the final distortion of the part will be reduced while fitup returns to normal. The weld was easily controlled so position 1 was used and he breezed through it once the tacks were in place.

Think about this: If you have a straight strip of metal 3/4" wide and you lay a series of tacks right along the edge, what happens? The metal shrinks right at the tack and tensile stress is added to the strip, unevenly, at the edge. This pulls the metal together in those spots, and tries to curve the strip. This puts compressive stress into areas around the tacks which try to resist that shrinkage, and the part turns into a lasagna noodle. By stretching the tacks you relieve the tensile stress so the compressive stress automatically goes away and the part goes flat again.