Mig welding thin metal and distortion

[optima21]

thanks for the reply guys, I think on my original welding the distortion could have been increased because I was welding very close to a section that would have been stretched to get the shape on the original boot lid and the heat from the welding may have released some of the built in stresses.

here's a picture of the me welding up the boot lid before I cut the number plate recess out, and its looking much better at the bottom before Ive welded it

and here is a standard boot lid

its looking like my welder is welding too hot on its lowest setting, so Im thinking about modifying the welder to have a pulse function to lower the heat input. when I say pulse it will only be on and off but thinking along the lines of having a longer first pulse say 1 second to build up enough heat for the first tack and then to pulse on and off at 1 pulse a second but with variable on/off times to control the heat input. pulsing circuitry can be done fairy easily with 555 timer chips and I'd just need to use the connection from the trigger to operate it and then have the output from pulse circuit to operate the relay on main board (which is what the trigger in the torch handle does).

does this sound reasonable to have a play with, trying to automate the pulsing that I did with the trigger by hand?

 

[mike 109444]

Don't know that the relay will withstand being operated at that rate.
You might have to go solid-state to get that sort of switching rate.
You may have to sort some means to stop the wire sticking when it is in the weld pool but at same time the power is removed (by the switching).

 

[Chris Stevens]

thanks for the reply guys, I think on my original welding the distortion could have been increased because I was welding very close to a section that would have been stretched to get the shape on the original boot lid and the heat from the welding may have released some of the built in stresses.

here's a picture of the me welding up the boot lid before I cut the number plate recess out, and its looking much better at the bottom before Ive welded it

View attachment 71650

and here is a standard boot lid

its looking like my welder is welding too hot on its lowest setting, so Im thinking about modifying the welder to have a pulse function to lower the heat input. when I say pulse it will only be on and off but thinking along the lines of having a longer first pulse say 1 second to build up enough heat for the first tack and then to pulse on and off at 1 pulse a second but with variable on/off times to control the heat input. pulsing circuitry can be done fairy easily with 555 timer chips and I'd just need to use the connection from the trigger to operate it and then have the output from pulse circuit to operate the relay on main board (which is what the trigger in the torch handle does).

does this sound reasonable to have a play with, trying to automate the pulsing that I did with the trigger by hand?

Sort out a pulse on wire feed motor.

 

[optima21]

solid state relays, sounds like it might be a plan, and the relay powers both the transformer and the wire feed on my welder (I think). think its worth a try as I can try it out without modifying the welder and it will cost lest than £5 to see if it will work.

or maybe I'd need to keep the power on all the time and just switch the wire feed on and off. that way Im thinking that the wire wouldn't stick to the weld when the wire feed is turned off.

 

[Wendelspanswick]

I'm wondering if you trigger the wire feed on and off but not the welding voltage whether you will end up with excessive burn back on the filler wire?

 

[mike 109444]

You may end up having to have two position switching. Position one = weld power on then position two feed motor on (pulsing) then when stopping back to position one, wire stops feeding then back to zero for weld power off.......hummmmmm
Getting complicated ant it

 

[Chris Stevens]

Why not look at circuit diagram from transformer set with wire pulse feature, and simply copy how its done? Might though be an awful lot easier to sell the set you have, and buy something like a used Tradesmig or Kempomatt from Ebay?

 

[Coo]

Remember the Clarke will only go down to 30amps...... maybe a more expensive machine would give better thin metal results ??

 

[Chris Stevens]

MIG sets with cheap aluminium windings dont seem to work that great for thin metal. Maybe worth trying copper wound set, with wire pulse, and seeing if it works better than the Clark?

 

[optima21]

modifying the mig welder is looking more like a winter project now. I guess I could buy a different welder, but some of us just prefer to modify and adapt things. In the mean time trying to thin metal with a tig would be good practice as it can be done on the bench. mig is much more convenient for doing out of position welding on cars

 

[Chris Stevens]

modifying the mig welder is looking more like a winter project now. I guess I could buy a different welder, but some of us just prefer to modify and adapt things. In the mean time trying to thin metal with a tig would be good practice as it can be done on the bench. mig is much more convenient for doing out of position welding on cars

I would suggest hiring a good quality copper wound set with wire feed pulse feature, and comparing it directly to what you are using now. If you like it then used sets along those lines often come up on Ebay for £250-300ish.

 

[optima21]

I've been playing this weekend, first of all trying to work out how the circuit board inside my welder works (a clarke 151te) and gave up, so then looked at being able to control the circut board indpendantly of the trigger switch in the torch handle. I ended up doing this by playing with the wiring, adding a 12v dc mains led driver and a 12v relay to make sure I could do this and it seemed to work ok. the trigger in the torch was operating the relay (which was powered by the led driver), and the relay operated the circuit board.

last week also got delivery of a usb digital oscilloscope off ebay, for use with an old netbook,as I thought it would give me a help with working electronics of the pulse generator.

I also found this circuit on the net and with a few tweaks looks to do what I'm wanting

the 555 timer chip on the left if for setting the frequency, I'll be looking at having a pulse between 1 and 6 seconds. the 555 chip on the right is for setting the length of the pulse, so I'll be looking at times between 0.3 and 2 seconds. my thought are that a long pulse and long on time can also be used as a timer for plug welding if necessary. so I'll be looking of on pulses of 0.3, 0.5, 0.75, 1.0, 1.5 and 2 seconds, and the time between pulses of 1.0, 1.5, 2.0, 3.0 and 6 seconds, both set using rotary switches instead of variable resistors.

and playing on a breadboard

and the oscilloscope software in use

hopefully it'll work when I get round to trying to weld with pulse, as its all a bit of guesswork at the minute.

 

[optima21]

suprisingly I got it to work and was definately easier to use than manually using the trigger on the torch to do the pulses for overlapping tacks.

I used the circuit above as a base and the circuit board ended up like this

and for the period of the pluses (1/frequency) and on time of the pulses I used rotary switches instead of potentiometers (space ends up being quite tight)

and the modified front of the welder,the button next to the min/max button turns the pulse on and the dials below are the rotary switches to control the on time of the tack welds and the time between the tack welds

there is a few more modifications to the wiring inside, but I wouldnt want to be responsible for people copying me as electrics can be dangerous.

when I first used the pulse I found that the off time after the first tack was about one and a half times the rest of the off times (after the tacks) and later realised it was due to the design of the 555 timer chips, that the first pulse on an astable oscillator is always longer then the rest.

the problem with the longer gap between the first and second tack welds is that the first tack will cool more than subsequent tacks. also burn back of the wire didnt appear to be an issue, and I didnt manage to make any holes on my short test.

so its back to the drawing board, with the 555 (on the left in the above circuit diagram) that controls the frequency being replaced with a op-amp oscillator. I've got to this stage and so far seem to be working better as the first off time is the same as the rest.

and this is the new prototype

 

[Wendelspanswick]

Home brew at its best!

 

[sapperb]

I concur with Marlinspike.. He uses copper behind the two piece, i use aluminum. It helps alot and cools down the area of weld..Providing you can get the piece of aluminum or copper in there..

 

[optima21]

I finally got the pulsing to work properly, but found that under testing that better results acheived my manually using the trigger to overlap the tacks. I also tried to TIG weld the steel with a small gap of about 1mm between the two edges of the panels and found that didn't work too well.

so its time to start again with another section of metal to be let into the boot lid. the section that was going to be welded into the boot lid is almost flat,but it is slightly curved in both directions.

after seeing this video of Lazze using cleco clips

I ended up using them. the plates were made of 2mm zintec. I was pretty impressed with how they work as the springs in the clecos let them adjust to the curve of the panel rather than trying to flatten it (as if it was clamped)

the panels were then tacked between the cleco's, some of the tacks needed hammering flat as the gap between the panels closed in some places

the shape of the panel was then check and the tacks ground flush with the panels. more tacks were then added between the panels leaving 1/2" between the tacks. the holes for the clecos were welded up using some flattened 15mm tube for backing, and the welds were then ground flat and the panel profile checked again.

so then its time to fully weld the two panels together using overlapping tacks starting at one end and continuing around the join of the two panels

and the reverse of the panel

annoyingly its distorted again,the gap between the panel and the piece of wood, was originally the same distance but with the edge being proud rather than recessed.

it is going back to the correct profile with some panel beating. Im trying to get the panel within a couple of mm of the final shape so keeping the amount of filler used to a minimum. now I need to work out how to do the numberplate recess and hold it for welding. and this is the new frame that the boot lid will go on, with internal hinges rather than the internal hinges as standard

 

[Chris Stevens]

Maybe an idea to have joddled the repair piece put it behind and plug or spot welded the 2 parts together, then leaded to prevent corrosion? A lot less distortion, and not so much filler needed.

 

[Wendelspanswick]

Did you cool your welds with compressed air or water?
I do on critical stuff as I find it keeps the HAZ really small so you get less distortion.

 

[Woody.v8]

Did you cool your welds with compressed air or water?
I do on critical stuff as I find it keeps the HAZ really small so you get less distortion.

Air to quench every time

 

[optima21]

neither, it was fine until I did the overlapping tacks. I'll try air using air to cool the welds next time I have a got at something similar and possibly only weld an inch or so, cool that and then grind it back, check the profile and then weld another inch