Welding: Blowing holes in sheet metal

[Vegaman_Dan]

Okay, I've been reading up all I can on here on GJ and other sites to figure out what can be done about reducing blowing holes in sheet metal when trying to weld pieces. Yes, I've searched the forums. All the suggestions deal with power, speed, stickout, etc. I've applied these with very limited success.

I believe the metal of the little British car is just too thin to be welded with the equipment I have.

Hobart 120 MIG
Century 80 Flux core

Neither one of these can dial down low enough. I've tried adjusting stickout, power, speed, etc. I can do all the tests I want with the replacement sheet metal I'm using and can join that without too much work, but the original sheet metal is just blowing away and I'm left chasing holes across the surface.

It's not a rust issue that has affected the material. I've cut all of that away, formed new patch panels and can't keep them joined.

I'm wondering if this is an area where TIG really is the practical solution since you can control the heat better? I hate to spend all that money on a TIG welder as well (okay, I don't mind THAT much to have an excuse for a new tool), but since the replacement body panel is $800, I'm very much interested in saving this one.

 

[Professur]

The hobart with gas should have no trouble with thin stuff. What gas and diameter wire are you using?

 

[Vegaman_Dan]

The hobart with gas should have no trouble with thin stuff. What gas and diameter wire are you using?

Argon and .024 wire.

 

[HMCFab9]

Flux core is Useless for welding on a car body.
I'm not familiar with the Hobart, but is there a dial for wire speed & voltage control, or does it have the 1,2,3,4 type selections for voltage?

If there is a dial, you should be able to set it to make perfect welds.
If it is a selector switch, it's hard to get a real good weld with that.
You should be using er70s6 wire.
I've never used argon for mig welding. CO2 or C25 is most common.
Why are you using argon?
Are you lap welding, or but welding? Lap should work pretty well. **** welding takes patience & a tig is much better for that.
The only other things I can say is: only weld an inch or so at a time (or less) & the other thing you can do is put a piece of brass right behind your weld (the brass will pull some of the heat away & if you do blow through, it will still weld up better with the brass behind it.

 

[stinkity stoink]

try backing up the joint with thin strips of copper. It might help with it blowing out.

 

[Kevin54]

Flux core is Useless for welding on a car body.
I'm not familiar with the Hobart, but is there a dial for wire speed & voltage control, or does it have the 1,2,3,4 type selections for voltage?

If there is a dial, you should be able to set it to make perfect welds.
If it is a selector switch, it's hard to get a real good weld with that.
You should be using er70s6 wire.
I've never used argon for mig welding. CO2 or C25 is most common.
Why are you using argon?
Are you lap welding, or but welding? Lap should work pretty well. **** welding takes patience & a tig is much better for that.
The only other things I can say is: only weld an inch or so at a time (or less) & the other thing you can do is put a piece of brass right behind your weld (the brass will pull some of the heat away & if you do blow through, it will still weld up better with the brass behind it.

My guess he's using the 75%-25% Argon mix

 

[chevroletfreak]

I am not a professional welder, but I believe you should be using 75/25 (Argon/CO2) and not 100% Argon.

If you TIG, you would need the 100% Argon.

 

[arrowhead]

+1 on using the MIG instead of the flux core (although TIG would be best)

Don't try to just run a continous bead, it builds too much heat and blows through. Try stiching multiple hot tack welds together in a short 1 inch row. I'll hit it for a second and stop, move a little and hit it again. I try to hit it just as the last one starts to cool off, but not too much.

 

[Vegaman_Dan]

No option to back the weld joint. No access. It is behind structural members of the trunk lid. I've made copper spoons for this purpose in other areas of the body, but this trunk lid is being a persnickety thing indeed.

Voltage is a stepped multiple position rotary switch. Wire speed is a dial potentiometer. Please disregard the settings in this photo- I was last working with some angle iron when this pic was taken, but it shows the settings available.



The trunk lid is 19-20 gauge. Yes, it really is that thin. And that's not a rusted area.

 

[Vegaman_Dan]

Correct, it is a mix. I've always had a mix and didn't even think about it being considered a pure solution.

 

[gorilla]

The wire is your problem, try .023" and Ill bet you a beer you'll have better results.

 

[Vegaman_Dan]

+1 on using the MIG instead of the flux core (although TIG would be best)

Don't try to just run a continous bead, it builds too much heat and blows through. Try stiching multiple hot tack welds together in a short 1 inch row. I'll hit it for a second and stop, move a little and hit it again. I try to hit it just as the last one starts to cool off, but not too much.

I'm familiar with and actively stitch weld on other sheet metal repairs on other vehicles. That's not the issue.

This will blow through on the very *first* tack weld. There's too much heat going in too fast. I can weld 18 ga no problem with this Hobart, but this stuff is thin- 19-20 guage. And this welder just doesn't seem to be able to dial down that far. I've tried increasing stickout, but that doesn't seem to affect enough to be helpful. Shielding gas is good, no issues with porosity. Definitely dealing with too much heat being generated even in that spot weld.

 

[rsanter]

The problem you are seeing is that you can likely weld a short bead and then you get the blow through. The thin metal can only hold so much heat before you will burn through and at some point if you turn the welder down too far you will not get penetration.

You need to stitch weld the panel in place. Make a short bead of a 1/2 inch to an inch depending on what the material will handle, skip an inch and do it again.
Go back and fill it in.
You will also get less warpage this way

Bob

 

[Professur]

With that machine, wire and gas you shouldn't be blowing out unless your technique really *****, but you seem to know your stuff well enough. I'd make some coupons with different thicknesses of metal from 16ga to 22ga .. your welding supply may well have stock. Set the machine up for the 16 and see if you can't get that one going ... then work your way down. It could be the machine isn't dropping power going from 2 to 1 because of a defective control. Alternately, take the machine to the shop and have them run a few tests. But that thickness is well within the machine's range.

 

[Vegaman_Dan]

The problem you are seeing is that you can likely weld a short bead and then you get the blow through. The thin metal can only hold so much heat before you will burn through and at some point if you turn the welder down too far you will not get penetration.

You need to stitch weld the panel in place. Make a short bead of a 1/2 inch to an inch depending on what the material will handle, skip an inch and do it again.
Go back and fill it in.
You will also get less warpage this way

Bob

Can't even start it with a tack weld without a blowthrough, so a bead is out of the question. I've stitch welded panels before without issue and know the method.

This is a case of trying to start even lower than the welder can do, it seems. That sheet metal is seriously challenging! I'd be much happier with 18 ga or larger like normal.

 

[HMCFab9]

It sounds like you have everything covered.
Only thing I can think of is: do you think your machine isn't going to the lower voltage setting? It should work otherwise, even on 19-20 gauge.

 

[lametec]

How 'bout a longer stickout?

Edit: Oops. I guess you tried that already. But how much stickout have you tried?

 

[gorilla]

You're trying to weld .035" steel with .035" wire by the time you get a puddle the base metal melts away. You need a smaller wire, .023" or .030" are most commonly use for auto body sheet metal.

 

[rsanter]

Sounds like you should see if you could try a different welder. Do you have a friend or fellow GJ member that you can work with?
Either that or you need to go to a TIG

Bob

 

[rsanter]

If this area is completely hidden such that you cannot get a view of it then you can take a strip of metal that you tack to the seam so you have a sort of step joint. Once welded this piece will become part of the car but if you can't see it then no matter

Bob

 

[longlivepunk]

Do you have an oxy/acetalene setup? You could always braze it. Just a thought.

 

[Vegaman_Dan]

With that machine, wire and gas you shouldn't be blowing out unless your technique really *****, but you seem to know your stuff well enough. I'd make some coupons with different thicknesses of metal from 16ga to 22ga .. your welding supply may well have stock. Set the machine up for the 16 and see if you can't get that one going ... then work your way down. It could be the machine isn't dropping power going from 2 to 1 because of a defective control. Alternately, take the machine to the shop and have them run a few tests. But that thickness is well within the machine's range.

That's not a bad idea. And I'm totally willing to believe my technique *****. That's why I'm here asking for ideas.

It's *possible* the welder could be the issue, but I'm far more likely to believe it's me. I'll see if I can snag some 20 and 22 ga sheet stock to do more practice on, see what it is that I'm doing wrong.

 

[Vegaman_Dan]

Do you have an oxy/acetalene setup? You could always braze it. Just a thought.

I do, but I don't like the idea of brazing bodypanels. I've seen the way the paint can react later with metals that can't expand at the same rate and would want to avoid that.

 

[Vegaman_Dan]

How 'bout a longer stickout?

Edit: Oops. I guess you tried that already. But how much stickout have you tried?

3/4" stickout, and that's from the tip, not the gas shield, which I know is a common mistake. I've tried increasing it without any real effect. Shortening it certainly does increase the heat quickly.

 

[Vegaman_Dan]

You're trying to weld .035" steel with .035" wire by the time you get a puddle the base metal melts away. You need a smaller wire, .023" or .030" are most commonly use for auto body sheet metal.

I thought I mentioned I was using .024" wire earlier, but I may have neglected that. Yes, picked up a spool specifically for this purpose. The Hobart makes changing out spools and drive wheels easily.

 

[Vegaman_Dan]

If this area is completely hidden such that you cannot get a view of it then you can take a strip of metal that you tack to the seam so you have a sort of step joint. Once welded this piece will become part of the car but if you can't see it then no matter

Bob

Funny you should suggest that. I did that as a possible solution early on. It would still blow through the original sheet metal leaving the backing piece intact.

I have a good strong joint, but a whole lot of little holes at the edges of all the tack joints that I keep chasing around like termites.

 

[wumpus138]

Is your polarity set for DCEP for running solid wire?

 

[gorilla]

This is from way out in left field but is it possible that the panel your trying to weld is aluminum?I know that you said you cut away rust or was it corroded aluminum? It's not unusual to find a odd aluminum panel on a British car, the hoods on MGB are aluminum, as are the doors and trunk lids on some old Jaguars. Easy way to check with is a magnet. One other possibility is that your input voltage from the power company may be high, my utility company supplies power at 254 volts which is 5% high. Not sure if that would really make a difference.

 

[BD1]

Check the manual for correct ground / gun connection. Try the lowest setting and work up from there. Maybe the machine has a output problem and stuck on a higher setting.

 

[ADSR]

Is your polarity set for DCEP for running solid wire?

The same question I was going to ask. If OP went from flux to solid and didn't change polarity, that's the problem.

 

[pepi]

I am not a professional welder, but I believe you should be using 75/25 (Argon/CO2) and not 100% Argon.

If you TIG, you would need the 100% Argon.

You are correct 75/25 mig.

 

[ADSR]

You are correct 75/25 mig.

100% argon on mig makes a much nicer weld. But you need more power to do the digging.

 

[ejm1961Tbird]

Read all the entries by MP&C in this thread:
http://www.garagejournal.com/forum/showthread.php?t=53534

 

[HMCFab9]

As stated before.... have you checked polarity?
It's different between flux core & gas shielded.

 

[crewchief888]

If this area is completely hidden such that you cannot get a view of it then you can take a strip of metal that you tack to the seam so you have a sort of step joint. Once welded this piece will become part of the car but if you can't see it then no matter

Bob

this is what i was gonna suggest.

i found steel banding/strapping works well.


but then again, i've welded sheetmetal with .035 flux core.


what if you clamped a piece of copper or brass near the weld to act as a heat sink of sorts? maybe draw away some heat.
just a random thought.

 

[03protege]

I also think you should try the copper heatsink idea.

Also is there not a way to check the voltage with a DMM? That way he can confirm the power settings without welding and comparing.

 

[MarkG]

+1 on using the MIG instead of the flux core (although TIG would be best)

Don't try to just run a continous bead, it builds too much heat and blows through. Try stiching multiple hot tack welds together in a short 1 inch row. I'll hit it for a second and stop, move a little and hit it again. I try to hit it just as the last one starts to cool off, but not too much.

Do this!! This is good advice and has worked for me in the past. Leave your nozzle down there so the gas shield is effective, scoot back and do an overlapping tack just as the orange from the previous one starts fading.

The pic is of some very thin wall tubing I did this way. Why I have that little 'divot' on each tack, I have no idea, but it worked. (if someone can let me know about that divot, please let me know! These were just short, overlapping tacks with ER70S2 and 75/25 gas in a Handler 120.

 

[ADSR]

Do this!! This is good advice and has worked for me in the past. Leave your nozzle down there so the gas shield is effective, scoot back and do an overlapping tack just as the orange from the previous one starts fading.

The pic is of some very thin wall tubing I did this way. Why I have that little 'divot' on each tack, I have no idea, but it worked. (if someone can let me know about that divot, please let me know! These were just short, overlapping tacks with ER70S2 and 75/25 gas in a Handler 120.

Those bug eyes wouldn't pass inspection and look like hell. Those holes are from trapped gas leaving the hot weld while the wire is still in there. Adjust your angle. Turn your rig to kill and adjust wire speed fast enough not to let bead build up. It takes time but a continuous bead on thin sheet metal is the only way to go.

 

[Vegaman_Dan]

I will check to see the polarity. It's a used system, so the previous owner may have set it incorrectly. It's never run flux core that I know of.

There is no option to back the weld with a welding spoon. It's behind sheet metal brackets for the structure. I'd have to do a lot of very expensive damage to get to that area- and then still have to weld back in the bracket made of the same material and still have no access. Catch 22.

I'll keep at it.

 

[R.Anderson]

Do this!! This is good advice and has worked for me in the past. Leave your nozzle down there so the gas shield is effective, scoot back and do an overlapping tack just as the orange from the previous one starts fading.

The pic is of some very thin wall tubing I did this way. Why I have that little 'divot' on each tack, I have no idea, but it worked. (if someone can let me know about that divot, please let me know! These were just short, overlapping tacks with ER70S2 and 75/25 gas in a Handler 120.

Those bug eyes wouldn't pass inspection and look like hell. Those holes are from trapped gas leaving the hot weld while the wire is still in there. Adjust your angle. Turn your rig to kill and adjust wire speed fast enough not to let bead build up. It takes time but a continuous bead on thin sheet metal is the only way to go.

The "divots" pictured are craters not created by trapped gas, just lack of fill/metal shrinkage. Bubbles are usually created by lack of shielding gas and or impurities/dirt.

I agree tack welding a weld like that is not the way to go. There should be no reason not to weld a straight bead on the diamond plate to that square tubing pictured. Increase your travel speed.

Argon and .024 wire.

The wire is your problem, try .023" and Ill bet you a beer you'll have better results.

.001" difference is the problem I think not.

I see people posting on this without reading the previous posts, how many people posted about wire size being wrong after Vegaman_Dan posted .024.
Read people, read

With thin metals I travel speed and wire speed are the common problems. You have to move pretty quick not to blow through thin metals.