Bend Test Problem Face vs side bend

Date 04-25-2008 15:06

I'm guessing you have combination of two things going. An axis orientation issue (thats why you get one direction failing) and the fact that you are bending materials of considreably different yield strengths. One issue alone may not be problematic, but the two together may be a different story.

By eekpod

Date 04-25-2008 15:24

Thanks for your reply, Let me clarify, the tests are either A36 to A36 or A572-50 to A572-50. I didn't mean to inply it was A36 to A572-50, sorry bout that. Chris

By js55

Date 04-25-2008 15:46

Where does it look like the fracture is initiating?
At what point in the bend is it fracturing (estimated bend angle at fracture)?
What filler are you using?

By eekpod

Date 04-25-2008 16:41

It's a GMAW process with ER70S-4 filler wire. Because it's only 3/8" thk it's always fracturing/ breaking at the intersection of the cap pass to the intermediate pass. So it's about 3/32"-1/8" down from the top surface and it's either at the heat affected zone, or where the weld meets the base metal on one side of the bevel. I'd say it's gets bent to about about 1/3 the distance into the jig then POW, it breaks the coupon right in half, two pieces. Ive had coupons crack before but never break right in half like that.
Could it be lack of fusion of the cap to the side wall of the bevel? and I guess because it's towards the top that's why the face bend is failing versus the root bend? The roots go fine, they bend right around no problem.

By hogan

Date 04-25-2008 16:54

do you have the ability to perform wet fluorescent mag on the cut piece

By jwright650 (*****)

Date 04-25-2008 17:05

Chris are the passes blending together or just laying beside/on top of one another?....Make for certain that they are chipping the silicon islands off of each pass. That little bit of silicon will make it fail..........

By bmaas1

Date 04-25-2008 17:15

Is the base metal direction of roll going in the correct direction?

Brian

By js55

Date 04-25-2008 17:43

I don't believe its a fusion issue for two reasons. Fusion, if the material was still ductile would tear not break(or at least open up progressively as the elongation in the test increased), and especially not at about 90deg or less in the bend which is an elongation of maybe 10%, or less (and thats on the outer fibers-the face itself-not 3/32" underneath). There is in my opinion an embrittling process going on, which I could possibly explain with Cb/V alloyed 572 (though this is welded everyday without embrittlement) but it really does not seem likely at all with A36, or S-4 filler which has to be spec't at 22% elong. You sure its happening with A36?
Breaking the way it is seems that it has to be embrittled. But to my knowledge, there's nothing in A36 or S-4 to embrittle in that manner.

By jmdugan10 (*)

Date 04-25-2008 18:46

I don't know if this would lead to embrittlement, but could the weld restraint cause enough stress to be an issue? If there is significant heat input with the thinner material, it may be cooling to fast causing embrittlement.

Could someone wiser comment?

By js55

Date 04-25-2008 20:58

Seems to me with an A36 material you really gotta get it hot to do any damage. And I don't know how hot that would have to be. I've run SAW PQR's on A36 at well over 100 kj/in and not embrittled it. I can't think off hand what phenomena would do it. In fact, I've seen high heat inputs improve the toughness of this material. Its already a large grain material, so depending upon where your high heat input falls you can very well broaden the fine grained heat effected zone and depending on where your Charpy notch falls get better impacts(the very phenomena Section III impact regimes had in mind). You can enlarge the grains further but I don't know how bad it would have to be to reduce ductility to below 10% or lower. Not sure it can even do that just by welding.
Martensite embrittlement is caused by rapid cooling. But in A36?
Drop it in water maybe while its still above transformation. I don't how how fast you gotta cool that stuff to get martensite but I bet its getting in on. Besides,the description seems to imply its more related to the weld deposit. But the martensite problem here is the consistency with which it fails in the same location all the time. How good do you have to be to get your martensite to form at the same spot in every weld?
With 572 I can see the martensite argument. With Cb and V. And grain enlargment becomes a real problem because Cb is a grain refiner (or at least inhibits grain growth)in processing and in very high heat input welds would tend to precipitate large hard carbides in the HAZ which I suppose could be fracture initiation points. But still, in the same location every time?

By pipewelder_1999 (****)

Date 04-25-2008 20:22

What position are you welding in and how many passes/layers ?

By eekpod

Date 04-26-2008 04:27

Thank you for all the input so far.
Flat position, 3 passes, 1 root, 2 bead cover split side to side.
I spoke to the foreman of the shop who made the coupons for me, he's not sure if the plates are A572-50 or A36. Yeah I know I need to keep track of this type of thing, it's just a busy week and I thought he'd do that automatically, I'll keep a closer eye on it in the future. I'm sure the backing bar is A36, it's just regular 1"x 1/4" bar stock. the bevels were cut on the band saw nice and clean and sharp, no contamination.

I wish I had brought them home to take a picture of them. the breaks look like they took alot of strain before they let go. I use a manual porta power to bend them in the jig, and I swear just by the feel that these coupons got harder to "bend" just before they split, like it was resisting the deflect any more and just ripped loose.

I still want to get into the size of our plunger and die, I think that might be it, I measured it real quick and it was like 1" 1/2" plunger and I know 2"3/8 die, I have to go through the figure in the code to compare it. Has anyone else used the same die setup for side bends and face bends with any problems?
Chris

By pipewelder_1999 (****)

Date 04-26-2008 10:51 Edited 04-26-2008 11:04

Those are some pretty big passes. What diameter wire type/shielding gas are you using and voltage/amperage are you running? It is possible in my opinion that the welds beads could be so large that the heat from the arc is being used to superheat the puddle as opposed to heating/melting the base metal and underlying weld metal. A slow travel speed and large puddle can cause fusion issues. If you are confident you have complete fusion then maybe I am up the wrong tree. As indicated above, the fusion is an unlikely culprit. Maybe try one with more beads and see.

If your travel speed is such that you are staying towards the leading edge of the puddle you should be OK.

Per D1.1 - for material greater than 50KSI yield the plunger should be 2" and the die 2-7/8" so if you have 572-50 or A36 you should be fine however the yield strengths for these materials are specified as "Miinimums" (50KSI for 572/50) and therefore could exceed 50Ksi. Your bend radius MAY be too tight but you would need MTR's for the material or get some tensiles done.

Have a nice day

Gerald

By thirdeye (***)

Date 04-26-2008 13:24

This is interesting......make sure to keep us posted.

Is there any chance you can (or did you) polish a couple of edges and etch them? A look under magnification can sometimes give you some clues.

~thirdeye~

By Shane Feder (****)

Date 04-26-2008 13:40

Gerald,
I was thinking along the same lines as you as soon as I saw GMAW.
Witnessed some weld tests on 12mm (1/2") plate that were GMAW in the 1G position.
The "instructor" had them doing 3 runs only and although I disagreed I was there purely to supervise the tests and not train the welders. They were putting a root in and then filling about 8 mm (5/16") in one go.
When we did the bend tests they all broke down the bevel face,2 of them you could still see the grind marks on the bevel.
Felt sorry for the young guys because they were only doing what the "instructor" told them to do.
Regards,
Shane

By 803056

Date 04-26-2008 18:34

There are several items that come to mind, most of them have been touched upon already.

1) Are the plates cut properly, i.e., is the direction of roll perpendicular to the longitudinal axis of the weld? I purchase 3/8 and 1 inch bar stock 8 inches wide so there is no question they are prepared properly.
2) Are they using short circuiting transfer or spray transfer? Short circuiting transfer, using large passes that don't employ the proper weave technique often result in incomplete fusion along the groove face to to previous weld layers as Shane noted.
3) If the cover pass is wider than the groove opening, but the cover is thin at the toe, the weld often pulls out of the base metal at the toe. Fusion is evident, but it appears the higher tensile strength of the electrode and lower ductility when compared to the base metal, thus the weld pulls out of the base metal.
4) Are the grinding marks parallel to the weld's longitudinal axis or are they parallel to the direction of plate's roll, i.e., transverse to the longitudinal axis of the weld? The grind marks (scratches) should be transverse to the weld face and transverse to the longitudinal axis of the weld.
5) Was there undercut present along the toe of the weld? Both undercut and grinding marks act as stress risers and can initiate cracks along the toe of the cover bead.
6) Were the samples cooled to room temperature when they were bent? Attempting to bend the samples at temperatures that are too hot to hold comfortably seems to result in the samples fracturing prematurely. I ran a series of informal bend tests many years ago. It seemed that all the samples bent at temperatures above 150 degrees had a high failure rate even though there were no obvious discontinuities present. Samples allowed to cool (from the same test piece) bent without mishap.

Good luck - Al

By swsweld (****)

Date 04-27-2008 01:38

I took a dual shield FCAW groove test a few years ago and busted twice as did the other welder taking the test. It ran and looked great.
Then an old timer walked by and suggested that we pull the puddle and don't push it. I thanked him for the advise, ran it as he suggested and passed the bend test. I know this is apples and oranges, just thought it worth mentioning.