[Alexa]

TECHNIQUES OF REPAIRING WELDS

The first time I open a topic ... hopefully it is in the correct section.

I would enjoy having input from everyone concerning the techniques and tools used to repair welds. It is a very encompassing topic, that moves across the various welding processes but also the various disciplines of structural, piping, pipeline, different alloys, etc.. Many fabricators have written welding procedures (WPS) that are dedicated just to the repair of the welds.

Perhaps it would be simpler if one were to explain a tricky or difficult repair that they had performed or seen performed.

Tanks.
Alexa

[noddybrian]

That's an interesting one - I do more machine shop stuff & mechanical currently so welding is secondary - but typically when I am welding it's repairs on earth moving / farm equipment- this can be routine repairs of cracks found during periodic / yearly insurance inspection or repair due to damage or reinforcement of inadequate designed parts - there is rarely a procedure for the job- though some insurance related things require one - generally it's left to my own experience to decide how to do the repair - I have a general procedure for single sided repair without backing using stick weld as most of this type of work has to be done outside - often wherever the machine is working / broke down.
General procedure is to dye penetrant test for cracks from extremeties of visible cracks or in adjacent parent metal
Grind full thickness prep at 70 degree included angle - often this will require taking the bulk metal out with oxy / fuel gouging or air / arc or micro lance first.
Weld is similar to pipeline though we almost never use 6010 here - root run 7018 in 2.5mm reverse polarity to achieve full penetration with small reinforcement on backside - dress to good standard to follow with as many runs of 7018 standard polarity in 3.25mm as needed to achieve bead - on thicker pieces 4mm or 5mm rods are accepted as long as position & access into V prep allows. there was a bunch of boring spec stuff & the test pieces had an amp recording device working out heat input connected & they were X rayed & destructive tested
My original test pieces were one vertical up and one horizontal overhead - this allowed repair to be covered by this method in all positions - the plate limits are between half and double the test piece - so by taking the test piece in 1/2" it covers from 1/4" to 1" - this is adequate for most mundane repairs on trackhoes booms etc - probably sounds weird but practices are not the same on both sides of the pond !
Strangely it's mostly left to the welder to figure why the thing cracked / broke in the first place & make corrective repairs - though sometimes it's just a case of keep the machine going to the end of the job when it can go back to the yard for more extensive therapy - especially if it's on tidal work !
Due to cost & stupid rules on maximum age of machines on jobs there are alot more weird make import machines around now and they seem to break way more than the established makes & in some very strange places - guess that's just the way of the economy.
Could probably rant on longer - but for what it's worth that's a brief view of how I earn not too much money ! I'm sure there are many others doing more glamorous work for better pay. It would make better reading if we got some people that repair nuclear stuff or submarines or the space shuttle !

[Otto Nobedder]

Man, I'll have to dig...

BUT, AWS has a recent standard for structural repairs. D.1.1.17?

Specifically about the repair of existing structure, including weld repair.

In my job, repair of a "defective" weld usually involves complete removal, and sometimes the addition of reinforcing "doublers" to distribute stresses and loads over a broader area. The repair of an "inadequate" weld in new work can be as simple as using TIG to "wash in" cold-lap or overfill.

Interesting topic!

You might have to pose specific questions, since it's so broad a subject.

Steve S

[TamJeff]

Usually have to remove all of the weld, or at least, "should have" anyway.

On aluminum, it's likely the edge of the weld and it's the parent metal that has failed. Re-welding it, typically won't even be as strong as it was originally once you figure in even more heat fatigue or how work hardened the part is by the time it has flexed enough to crack. In these structural instances, at least, you will do well to reinforce and relocate the heat zone. With pipe, an extra brace elsewhere in the weldment, or even laminating the tube either internally or externally via a pin or sleeve. Double the wall thickness, without the thickness being continuous, so that a stress crack cannot easily telegraph through it.

The hardest part I think is, repairing it without moving the problem to some other area of the structure to save from having a patchwork quilt further down the line that you will have to repair again yourself. Or one of those projects that never go away. I've gotten pretty good at spotting problem children, and have no issue with condemning them outright.

Knowing when to say "no."
Today, I had a part that they wanted a cross pipe/footrest cut out, the old weld ground and sanded smooth and left as is, in exchange for a folding footrest. . .on polished anodized pipe. Well, that's already starting out wrong. You're still going to have the scars no matter how good a job you do from there on out. Not only that, you can't F-up with the would-be grinding/sanding tools and shiz is gonna be in your hair, pockets and socks. 6 cuts with an air saw later, a little acetone and two new 30" pieces of pipe, and all is well in the world and in less time and the job looks right.

[weldin mike 27]

Usually have to remove all of the weld, or at least, "should have" anyway.

On aluminum, it's likely the edge of the weld and it's the parent metal that has failed. Re-welding it, typically won't even be as strong as it was originally once you figure in even more heat fatigue or how work hardened the part is by the time it has flexed enough to crack. In these structural instances, at least, you will do well to reinforce and relocate the heat zone. With pipe, an extra brace elsewhere in the weldment, or even laminating the tube either internally or externally via a pin or sleeve. Double the wall thickness, without the thickness being continuous, so that a stress crack cannot easily telegraph through it.

The hardest part I think is, repairing it without moving the problem to some other area of the structure to save from having a patchwork quilt further down the line that you will have to repair again yourself. Or one of those projects that never go away. I've gotten pretty good at spotting problem children, and have no issue with condemning them outright.

Knowing when to say "no."
Today, I had a part that they wanted a cross pipe/footrest cut out, the old weld ground and sanded smooth and left as is, in exchange for a folding footrest. . .on polished anodized pipe. Well, that's already starting out wrong. You're still going to have the scars no matter how good a job you do from there on out. Not only that, you can't F-up with the would-be grinding/sanding tools and shiz is gonna be in your hair, pockets and socks. 6 cuts with an air saw later, a little acetone and two new 30" pieces of pipe, and all is well in the world and in less time and the job looks right.

I do love a good remove and replace. Much easier in most cases.

Anyhow, I perform repairs on robot welds, in two types of Quenched and Tempered steels, The welds are 6mm fillets, out side corners and some angled butt welds. The defects include, lack of fusion, at either or both toes, under cut burn though and pretty much all the others as well. Our "rules" allow for TIG washing and Filling, for no more than 25 mm or 1/2 the depth of the prep in butt welds. The length limit is usually pushed to 75-100 mm. Simply for practicality . Anything over these limits is excavated and re -welded with the original procedure . The wires used are Lincoln 70s6 and Welding Alloys Tetra s 309 ss flux core. The SS wire the robot uses is called Vega 309 Metal cored. Also from Welding Alloys (it is rubbish)

Mick

[weldin mike 27]

Also 8mm and 10mm 3 run fillets on the S6.

Mick

[Otto Nobedder]

In my work, oddly, I rarely see a failed weld. Usually, I'm reparing a crack in the parent metal. It's usually under a gusset, between two fillets, which in most cases appear to be dual-shield.

It's always in 304SS, in a "formed" part. Most commonly, it's a vessel's parabolic "head". I can only assume stresses left in the metal during what appears to be a cold-forming process are amplified by the heat-stress of the welding on the gusset.

I've not investigated whether the crack travels into the gusset, as well, as the gusset is sacrificed in finding the parent-metal crack. I will attempt to do so, the next time this comes up, though, as I've developed a system of "localizing" the crack to a few inches before I cut anything.

I've discussed the repair of this particular type of failure in-depth, here:

http://forum.weldingtipsandtricks.com/v ... f=9&t=2426

and I've done several more like it, suggesting a common cause.

Steve S

[Otto Nobedder]

Alexa,

Here's one to analyze.

GEDC0866.JPG (151.21 KiB) Viewed 1303 times

Here's a shot without the dye, but the crack is still visible.

GEDC0867.JPG (154.17 KiB) Viewed 1303 times

It appears to me, this weld was done with a large gap. It was done in one pass, which shrunk leaving a sharp line in the center of the concavity. It should have been done in two passes.

Note the crack originates in a "crater" at a start or stop point.

Basic repair is not an issue here, however. This will be sectioned out. There is major support structure behind it which must be examined to be sure it's only a weld flaw, and not a structural issue.

Steve S

[jwmacawful]

when I worked high rise and bridge construction air/arc gouging was the standard method of repairing defective or cracked welds. during the construction of a high rise building there are fabrication mistakes and design changes that have to be rectified in the field, sometimes while the piece is hanging on the crane or on blocks many floors above the ground. when this happens all production comes to a halt and there are lots of highly paid guys just standing around so needless to say the walking boss is anxious to get things moving again and the a.a. comes in very handy. in my present place of employment as a maint welder/fabricator most of my work is in my shop on the workbench. in the 25 years I've been there I only had to use the air/arc once. it's way too noisy and produces too many fumes for confined spaces (and sets off the fire control alarms) so I use the 4 1/2" grinder or cutoff wheel and occasionally a die grinder with a burr.