Maybe undercut failure

admin

Ahhhh, fall in the midwest, machinery in the fields, contractors turn it up a notch to finish before bad weather, and things begin to break.The shop next door had two real good failure examples yesterday, so I brought my camera today. They had already fixed the one good example of a "good" weld failing last on a 1040-45 plate with a tube welded through it on an angle.They were fixing to go at this one when I showed up and said I needed it for a minute. Pictures tell the story. People who build things (welders included) often don't get to see failures to study so I thought I'd post up.Matt Attached Images
Reply:Originally Posted by Matt_MaguireAhhhh, fall in the midwest, machinery in the fields, contractors turn it up a notch to finish before bad weather, and things begin to break.The shop next door had two real good failure examples yesterday, so I brought my camera today. They had already fixed the one good example of a "good" weld failing last on a 1040-45 plate with a tube welded through it on an angle.They were fixing to go at this one when I showed up and said I needed it for a minute. Pictures tell the story. People who build things (welders included) often don't get to see failures to study so I thought I'd post up.Matt
Reply:thank you for posting these.  they are interesting.
Reply:Thanks for taking the time to post these. Makes you wonder if these were ever inspected. I'm used to the oil patch where every (structural) weld is visually inspected (at minimum - sometimes mpi, xray, etc) before leaving the shop, and at specified intervals.BTW, what application was this. A pic of the equipment showing the broken piece would be neat. I know, I'm an ungrateful *******dXMT 350 MPA Dynasty 280DXSuitcase 8RC / OptimaSpoolmatic 30AThermco Ar / C02 mixerAssorted O/A gear
Reply:Noob here so bare with me please.  I knew what undercutting looks like but I wasnt exactly sure what caused it, so I Googled it and came up with...Undercutting is the burning away of the base metal at the toe of the weld. Undercutting may be caused by the following conditions: a. Current adjustment that is too high.b. Arc gap that is too long.c. Failure to fill up the crater completely with weld metal. I am guessing that travel speed could also add to the problem?What is it about the undercutting that causes the weld to fail?Thanks in advance;SteveI'm not a welder, but I do sit next to one at the bar.
Reply:Nice pictures. I see a lot of that myself. I guess when they put them together on a line it does not matter. Oh well more work for us field guys. I get a lot of calls where they break the hydraulic lines(steel) on the boom. We fix them in a pinch, but the welds they break on them are nasty and usually undercut.
Reply:Ever see what the guys at CAT want for repairs? Guys around here don't refer to the local CAT dealer as "Gangsters and Robbers" for nothing.... Thats a factory preengineered failure point for future service..No government ever voluntarily reduces itself in size. Government programs, once launched, never disappear. Actually, a government bureau is the nearest thing to eternal life we'll ever see on this earth! Ronald Reagan
Reply:Thanks everyone, I can spend a long time looking at failures, weld, heat treat, machining, material selection etc. always hoping I may learn something from another persons troubles, as sad as that may seem. I suspect that the weld is a hard wire pulse, the operator is young or new, has very poor/lazy supervision in the weld shop and there is no sample inspection in the shop. Everything done in design, material puchase, WPS etc. has now been wasted, very sad, but here is a weak link in the chain. Originally Posted by 1969 VernonHey Matt,       That is some bad undercut Someone needs to learn how to weld Undercut is the #1 reason a break happens next to a weld. I weld for years on roller coasters no undercut was allowed for go reason as you can see. Thanks for the photo   Vernon
Reply:Just looked like CAT yellow.  All in all we've never had any issues with our CAT's or with the rentals we've had from them. Quality wise I'd say they were the best. Everything we've broken came from heavy use, not poor standards..No government ever voluntarily reduces itself in size. Government programs, once launched, never disappear. Actually, a government bureau is the nearest thing to eternal life we'll ever see on this earth! Ronald Reagan
Reply:Matt,I enjoy seeing good photos taken of failures and the fracture surfaces.  You can diagnose quite a bit about the nature of the failure just from looking at the surfaces of the break.  I suspect you know this stuff already, but for others that don't have much experience diagnosing these things...The contrast in the photo you took is good for a macro(normal type) camera.  Sometimes a flashlight head almost parallel to the surface of the break can reveal clearer details.  The patterns in the fracture surface show up in the areas of light and dark created by the peaks and valleys on the surface.If you look closely, you can see the texture of the fracture has a pattern.  There are lines that point back toward the origin of the failure.  I agree with you that the undercut at 12 o'clock is a stress riser, and that this is the origin of the break.  The surface texture of the fracture proves this.There's no rust or smeared areas on the fracture surface.  This tells me that the failure was due to overloading, and not fatigue.  Fatigue cracks grow slowly over time, and are often filled with rust and contamination.  A clean break happens fast and if checked right away shows no rust.  Fatigue cracks will frequently have areas that are smeared, where the crack opens and closes under tension and compression loads.  The only clear spot on a fatigue failure is where the final catastrophic break occurs.Someone was also careful(or lucky) about handling the piece.  Putting the pieces back together and often does destroy the delicate surface features that tell the story of what happened.Oh, last thought...that area you indicate at the bottom of the fracture as maybe a 'fine-grain' structure.  In my experience that type of detail can't be seen with the naked eye.  It's tough to say for certain why it appears different.  It could be a shear lip, which is techno-babble for the last place where fracture happens.  It's usually raised up and stretched out, and if the material is ductile enough then it's bent over.  Because the way the failure spreads changes in the shear lip, the appearance of the fracture surface can change.  But this is just guesswork without a light microscope or an SEM. Originally Posted by Matt_MaguireAhhhh, fall in the midwest, machinery in the fields, contractors turn it up a notch to finish before bad weather, and things begin to break.The shop next door had two real good failure examples yesterday, so I brought my camera today. They had already fixed the one good example of a "good" weld failing last on a 1040-45 plate with a tube welded through it on an angle.They were fixing to go at this one when I showed up and said I needed it for a minute. Pictures tell the story. People who build things (welders included) often don't get to see failures to study so I thought I'd post up.Matt
Reply:If I may chime in here...Basically undercut is considered the same as a crack; especially when it is as deep and long as shown above. I agree with A_DAB-will_do. Initially UC is just ugly, but eventually, Stress, especially in dynamically loaded structures will cause a crack in the UC.Generally any hydraulic machinery (backhoes, cranes, etc.) the hydraulics is stronger than the structure itself and anyone of those machines can tear itself apart when abused.Weldtek
Reply:When in doubt, follow the Chevron marks!  That sucker failed top dead center!
Reply:Notice that there is no rust evident in the photo, i.e. 100% new break. There are 2 curious dark islands roughly on the horizontal center line, 1 near top, 1 near bottom.Without more info and knowing Cat welding procedures and QC like I do, I'm kinda leaning toward that weld is a repair job not original work. Can you verify?Weldtek
Reply:Assuming that is CAT equipment as referenced by the yellow paint...Weldtek
Reply:I know i may get some heat from this comment but - the undercut showed on this pic diddnt have any effect on this break. if the break is looked at closly i, it didnt start or travel from the UC. It may not be a great weld , but it did maintane its integrity and not break.   If we had more or better view of the equipment then more could be understood.  The break seems more abuse than anything to do with weld quality.
Reply:i was thinking the same thing swamp rat
Reply:to help explain it,  undercut acts like a score line in glass.  the crack will almost always start there.
Reply:No heat from me.  But if I'm wrong,  Where do you think the break started?  And Why that location instead?To better illustrate what we've been talking about, I edited a copy of the OP's first photo.  I added a whole bunch of thin black lines that parallel the features I keyed off of.  Notice that these lines point towards a common area on the surface of the break.  That undercut is the area indicated by the texture on the surface.  Supe mentioned the chevron marks, which is the technical jargon used to describe these patterns.  [I think]They're really hard to recognize on this particular photo of this fracture.Compare the markup photo with the one in the original post.  See if you can't see what I'm trying to describe.A agree that this part failed due to an overload condition.  But the question is, did the overload that caused this failure exceed the design load for the part?  Or did the undercut cause the part to fail prematurely?  (Which most of us suspect)  None of this means that even a perfectly welded part would have survived.  Depending on how bad the abuse was, it might have broken even a perfect part.  What I can say is that the undercut doesn't help.If you look at the original photo, you can also see a second crack  parallel to the other weld toe line on the other piece.  Cracks open up under tension, so I'm going to guess and say the load was applied vertical down, 12 to 6 o'clock, as oriented in the first photo.CAT quality is probably pretty good; they have a good reputation in my mind.  But bad parts will slip through.  The proof would be to talk to a service tech and see how many similar parts he's seen fail in the same manor.  A couple or none?  Then this is a fluke.  Several or a whole bunch?  Then something in the Factory got out of whack.I'm betting this was a factory weld.  I didn't see any obvious signs of a repair in the vicinity of the break.  I'm going out on a limb and say that any welder who'd leave undercut like that would have also left heavy grinding marks, burned paint, or some other obvious evidence of the repair work. Originally Posted by Swamp ratI know i may get some heat from this comment but - the undercut showed on this pic diddnt have any effect on this break. if the break is looked at closly i, it didnt start or travel from the UC. It may not be a great weld , but it did maintane its integrity and not break.   If we had more or better view of the equipment then more could be understood.  The break seems more abuse than anything to do with weld quality.
Reply:Thanks DABer, I gotta get caught up a bit here. I had more pics, so I put them in the GIMP and made them B&W and brought the contrast up.First I don't think this is a CAT part, the only thing I can rule out is CASE yellow, most everybody else tries to closely copy cat.Second, I no longer think it is T-1. I looked at the second pic in my first post and the side is not flame cut, so this is bar stock. In B&W this pics don't show the texture near the roller quenched surface of T-1 that looked more like a medium carbon break. I said grain in the first post, thats a poor choice of words, should have said finer texture ie: less ductile. (After that I went to the PC with the pics in it and studied better)The part had a new bar welded on when I went for coffee this AM, so tomorrow I'll see if I can pull the part from the scrap bucket and bust a brinell on it. (shear pins are getting rusty).Matt Attached Images
Reply:Oh I forgot, the fracture sure started in the weld area and was pulled apart. But something also banged against the toe of the weld and the toe of the weld on the other bar. There is a 1 1/4" hole in the other bar and the rest of the broken bar is lost.It looked like something to accept a draw bar. There was no up or down upset near the break and the other bar was still straight with no deformation of the hole. Maybe a drawbar pin worked its way out?I just took the pic to post up for study and discussion of failures.Oh, forgot, forgot. Dab this little digital is like 1/2 to 1 inch away on these super macros? I don't know how it works like that (not like a telephoto on a 35mm or 120 type film camera with macro (stopped down for max depth of field). I tried some small things with side light and when I saw the result it was like "WAZZAT"?MattLast edited by Matt_Maguire; 10-21-2010 at 10:32 PM.
Reply:Originally Posted by Scott Youngto help explain it,  undercut acts like a score line in glass.  the crack will almost always start there.
Reply:Matt,Thanks for posting the B&W photos.  Those do an even better job of showing the fracture surface.  The second B&W photo in particular really shows the texture on the surface.If you don't mind, would you post the brand name and model number of that digital camera you used to take the photos.  Hand-held cameras that take good closeup photos are hard to find.  I have an older Kodak camera that takes decent closeups, but none of the low- cost cameras on the market now seem to do a good job.-DaveBenson's Mobile Welding - Dayton, OH metro area - AWS Certified Welding Inspector
Reply:Originally Posted by A_DAB_will_doMatt,Thanks for posting the B&W photos.  Those do an even better job of showing the fracture surface.  The second B&W photo in particular really shows the texture on the surface.If you don't mind, would you post the brand name and model number of that digital camera you used to take the photos.  Hand-held cameras that take good closeup photos are hard to find.  I have an older Kodak camera that takes decent closeups, but none of the low- cost cameras on the market now seem to do a good job.-Dave
Reply:looks like the crack radiated from the weld.  great close up pictures.
Reply:I would have to beleive from the complete lack of observable (in the pix at least) plastic deformation, that this must not be mild steel.Even with a severe notch, I would think you would see some elongation around the bottom portion of the break.XMT 350 MPa, w/D52-DTA 185 TSWHarris of