Newbie 3

[Weldtek]

Why do welds fail?

[weldrwomn]

IMHO, welds fail because of defects. Defects are unaccaptable discontinuities. Examples of discontinuities are porosity, slag inclusions, and cracks. Porosity is caused by too much heat, lack of shielding, and/or contamination. Slag inclusions are caused by improper cleaning/preperation of the weld joint and improper welding technique. Cracks are caused by any number of things not the least of which are, contamination, wrong filler metal, too much heat, not enough heat (as in preheat, postheat), etc.

Also, welds can fail because of poor engineering. If the stress placed on the welded joint is more that the joint is designed to handle, it will fail. Example: World Trade Center. Welds were not designed to handle the stress of being hit by an aircraft.

[zapster]

i cant really say i've had a weld fail..
but what does happen on occasion (rare) is that the
material around the weld failed but the weld itself was just fine..

...zap!

[backuproller]

good point zap. how many times have we welded and 3/8" next to the weld, or so, a crack appears. is this a heat (post/pre) problem. or is it bad engineering and we need to relook at the design. or just grind/gouge out the crack and weld it and maybe put in a scabbing plate.

[MAC702]

Originally Posted by Weldtek

Why do welds fail?

Poor design or bad application.

[weldrwomn]

Originally Posted by backuproller

good point zap. how many times have we welded and 3/8" next to the weld, or so, a crack appears. is this a heat (post/pre) problem. or is it bad engineering and we need to relook at the design. or just grind/gouge out the crack and weld it and maybe put in a scabbing plate.

I might be inclined to think that such a crack is due to heat-related stress.

[backuproller]

example: the cat 345 that carries the hydraulic hammer we use. the jib boom developed a crack on the end of the boom. well CAT blamed it on operator error. yeah ok. well it was warranty work. they gouged out crack, chain hoist with padeyes and welded it up. well, 4 weeks later the edge of the weld area started cracking. well, back to the shop it goes. 3 weeks later same thing. but the cracks always appeared next to the weld area, never in the weld.
just one exmple of the area next to the weld cracking. this was ongoing for 6 months. the last time it broke i was working in ATL and yancy cat had a full load, so we arranged for us to do the warranty work. i repaired the weld like they did, but i scabbed 1/2" plate arround the boom. i used preheat and post cool. tractor is still running and haven't had no problems.

[Weldtek]

Now this is more like it!

Weldrwomn - Thanks for that very insightful reply. I agree with all that. But, those can all be grouped into "Quality". Which is the number 1 answer.

Zap, never had a weld fail? I want to be just like you when I grow up.

What I was going for here (for the newbies)... Lets say youre weld is done by the book. Why do welds fail? Weldrwomn says also, Stress.

That's what I was looking for. Would someone please explain the five types of stress and the 3 ways they can be applied. Please.

[zapster]

Zap, never had a weld fail? I want to be just like you when I grow up.(quote)


no..i cant say i have in all reality..


if it did i'd be the third one to know...

the customer would be first..
my boss second and then me..

nope (knocking on my desk now) knock on wood..

...zap!

[Weldtek]

Oh, yeah Backuproller; Hi. I'm with you on the hydraulic stuff. I assume that's what you do? I've come to the conclusion that hydraulic equipment in general is capable of destroying itself no matter what. I've done that kind of welding repair for years and CAT, IMHO, really has there stuff together. I've used the CAT welding standards for warranty repair work for years and they're well written and intuitive. Seems to me that the equipment operators run 'em like they're indestructable, which they're not.

Anyhow, please, go on.

[prop-doctor]

most problems, cracks, stress, or failure do appear in the base metal on either side of the weld..... unless its a poor weld
could it be from the heat an changing in the base metal

[Weldtek]

Most people think of the HAZ as the base metal. But it would not exist without the welding process. The HAZ or heat affected zone is the weakest part of a perfect weld. The size and shape of the HAZ depends on material, thickness, joint geometry, welding process used and technique of the welder. Stringer beads will have a smaller HAZ than weave beads, because of travel speed. What actually kills the material is not how hot it got. It's the trifecta -
How hot, how long at temp., and cooling rate. Those guys that do heat treating know about this stuff from a technical perspective; "Joe Welder" doesn't. So when you "Run it hot" or "Burn it in" or "Weave" cause it's faster, your monkeying around with the same laws of physics and chemistry (metallurgy) that made it what it is in the first place.

There are few welding processes that won't have an HAZ - friction stir welding is one that comes to mind.

Oh, yeah the stress thing...

1. Tension
2. Compression
3. Bend
4. Shear
5. Torsion

Ways of application:

a. Impact
b. Static
c. Dynamic

The Navy spends a gazillion dollars annually on research about harmonic resonance (vibration). These are the things those Geeky engineers that don't know nuthin are designing to fight.

[backuproller]

so, could the cracking be caused by the arc gouging, or the heat of the weld in the HAZ. combo both maybe. yes, those machines take a whooping, 24/7/ 365.

i did some work for a company on a hydraulic motor, but they wanted the work done on the shaft w/o taking the motor off. so their plan was to make several 1 inch weld and rapid cool with compressed air to keep the seals from melting. well, my boss told me to do it the way they wanted. seals melted anyway. but could that process of rapid cooling put extra stress in the weld zone.

[weldrwomn]

In rapidly cooled steel, there is a higher occurence of martensitic structures making the steel harder and more brittle. That is why high heat then rapid cooling of steel makes it more prone to cracking.

[Weldtek]

Backuproller, that's a good possibility.


Weldrwomn - Ding! Ding! Ding! Alex, tell our contestant what they've won!

I don't know where you learnt ta weld, but I'm impressed. So many welders only see the artistic side of welding; the science is more important than how pretty the bead looks. A good looking weld doesn't guarantee a sound weld.
And a qualification test doesn't mean every weld you make for the rest of your life is a good one.

I hope the newbies followed this thread, 'cause you guys are really sharing it now!

[weldrwomn]

Weldtek, I went to Hobart Institute of Welding Technology, but it is sad to say, if you asked this stuff to most of the other students there, they would be clueless. Basically, it boils down to being interested and paying attention. What you will hear from most ppl at school is "I don't care about the bookwork, I just want to weld". Bummer eh?

[Weldtek]

HIWT - Best school in the country! I have friends that went there. Yeah, real bummer about the books, but what ya gonna do? I guess us welder geeks can stick together though, eh?

[Steelwill]

Well weltek keep it up because I am learning alot. What would be a good book that you would recomend to read to learn more about what you are talking about.

[Weldtek]

Steelwill, are you a reader? I'm not. Start with the Lincolnelectric website. They are relatively free with their printed word. They have an excellent library of .pdf articles from GMAW how to's to "Why do welds crack?" Hobart Institute of Welding Technology (HIWT) they have the best welder's handbook. Paperback, about 75 pages. It covers welding basic theory, Hobart rods and even has the old metal identification spark test guide in it. Other than that, go to the AWS to get a list of their library of books. i.e.
Welding Handbook vol. 1, 8th ed. Now that one is a tough read unless you're a welding geek like me. Otherwise it's sominex....

[Steelwill]

Hey thanks and please keep the knowledge rolling.

[Weldtek]

Your welcome. I'll try, long as I don't get runnoft.