Certifying an unusual process under AWS D1.1

admin

We are a machine shop and mfg of hydraulic cylinders. I am the "special projects" guy and have been assigned the task of getting our process certified under AWS D1.1We have a test lab available and I built a guided bend test jig to specs of D1.1 figure 4.15. Basically our process consists of welding a 1026 DOM honed tube (the hydraulic cylinder barrel) to a round base of A36.the completed cylinder is like a piece of pipe with a piece of round steel welded to the end.Because of our design, the weld joint is basically a v-groove with backing.The tube is rotated in an electrically driven lathe and the mig is hand held.We usually have multiple passes with a 1/2" wall tube being the thickest.We want to certify to the thickest tube (1/2", it has the most passes.)I have some confusion as to how to make bend test coupons. The weld needs to be in the middle of a test coupon but in our process the weld is within 3/4" from the end of the baseAny help?Thanks, Bill
Reply:Your best bet will be to obtain flat stock if possible in the material grades that you're working with, and weld out a coupon with them, rather than an in-situ type weld.  There are other possibilities, but you can get into muddy waters with that.  I'm going through a similar situation with the NRC and a rebar welding procedure right now because of it.
Reply:Thanks for the reply.Problem is we want to certify our process, and that is a semi-automated weld. It is 1G rotated position, weld deposit near the top.I actually made a test specimen out of a piece of  5" long 1026 tube, and we got a burn of A36 and machined it into a tube 5" long with the same configuration as a base to simulate our V-groove with backing. I just cut a section from it and machined it to 3/8" thick and machined off the backing.When I did a test bend over a 3-1/2" mandrel we had a crack about 1/4" long and about .020" deep on the backing sideI do suspect that the flat coupons is how we will certify our welder.If I am incorrect please let me know.Bill Attached ImagesLast edited by bbiggs; 02-01-2010 at 01:57 PM.Reason: misspelling
Reply:bbiggs- check into D1.1 for procedure qualifications. It has been awhile however i do not remember any where that you would have to perform a PT examination after the bend test. Check in chapter 4 section "B". See if a pre-qualified weld joint configuration would work like shown in Chapter 3 for the weld you are trying to accomplish. If the attached plate is perpendicullar with the tubing wouldn't this be a fillet weld?  The code does speak of qualification test for fillet welds to pipe.   Hope this helped if not please send me a private message and we can work this out.Weldor/ Certified Inspector
Reply:so you are trying to generate a PQR,  I don't work in D1.1 so take this with a grain of salt.First of all, flat plate IS generally how you certify welders. A flat test plate welded in the 1G position would certify the welder for 1G rolled pipe above a certain diameter.  Alternatively you can certify a welder with a production weld if it's an unusual joint configuration, but then you are certified strictly for that particular joint. You can also "double-dip" and certify the welder and procedure at the same time, so the guy who ran the PQR piece will be certed for that weld.now back to your questionIf I understand it your welding a cap to the end of the pipe with a v groove? and because of this you can't center  the weld bead on a guided bend test because the shoulders are unequal length.If so Iyou can qualify the procedure with two pieces of pipe and an identical joint profile so that you have the required length to cut out samples. The PQR is looking to qualify a weld on all the essential variables, the width of the shoulder of the weld is not an essential variable therefore it can be changed.   you also might be able to get away with some alternative testing methods, but I don't know D1.1 well enough to know what does and doesn't have alternative testing methods.Finally in the codes I work too if no generic shape adequately describes the weld then a production weld can be certified (we needed to do this with some strange, flange to flare bevel joint fitups) and NDT samples taken on that.Finally If I have this pictured in my mind right. Why are you welding on the such a narrow shoulder? Is there  a reason a socket weld isn't done? If your welding on a flat plate pressure isn't a concern, I thought fillet socket welds were the preferred method of capping a pipe.
Reply:This from memmory. The CC I took welding my welding classes at also did alot of outside testing for businesses and welders looking to recertify.  For D1.1 structual test they had two test 3/8" and 1".  The 3/8" test was of course v groove test but required 12" lenghth v groove weld.  Passing the 3/8" test allowed someone to be certified to weld up 3/4" material and passing the 1" test allowed the person to be certified for unlimited thickness.  Test samples were always 3/8" thick with the weld and backing ground to create uniform thickness and then the edges were radiused slightly with a file before being bent.       I think your testing procedure  is sound except for the thickness of your test material.  If your machining the test samples down to 3/8" from 1/2" you can be removing possible defects in the weld.     I think the next step is to create a welding procedure that way consistency can be maintained in the weld quaulity.
Reply:These threads interest the heck out of me. I just see so many "what ifs" in them. I think having standards is great. I also think it's fun when you do something that isn't addressed in the book. I am keeping tabs on this thread.Thank you, carry on!Lincoln Power Mig 210MP MIGLincoln Power Mig 350MP - MIG and Push-PullLincoln TIG 300-300Lincoln Hobby-Weld 110v  Thanks JLAMESCK TIG TORCH, gas diffuser, pyrex cupThermal Dynamics Cutmaster 101My brain
Reply:Machining the A36 to match the tube for a CJP groove weld with backing was the correct approach to take.  The tear you see there likely could have been avoided with coupon prep.  It looks like the bend test specimen cross sectional thickness is significantly wider at the bend area.  Uniformity will prevent those types of tears from opening up.  The edges also look relatively sharp.  I think if you repeat that test with a soft 1/8" radius on the edges of the entire coupon, you will satisfactorily pass the bend.
Reply:Thanks,I will follow your recommendations. The coupons are going to the lab today, keep you posted.Bill
Reply:Turns out the small "crack" was actually the very bottom of our joint profile. I milled the rest of the coupons to  .375"x.445" instead of .375"x.500" to remove ALL the backing and no cracks on the side bend test even with a 2" diameter bend mandrel. (see point of red arrow below) This drawing shows how we simulated the base-tube interface and had coupons long enough to do bend tests.Coupons headed to a AWS certified lab today. Attached Images
Reply:Originally Posted by SupeYour best bet will be to obtain flat stock if possible in the material grades that you're working with, and weld out a coupon with them, rather than an in-situ type weld.  There are other possibilities, but you can get into muddy waters with that.  I'm going through a similar situation with the NRC and a rebar welding procedure right now because of it.
Reply:Fat Man,Thanks for your input.We will use your suggestion to certify our individual welders but my problem is that we are trying to certify the process (PQR) and that process is 1G-rotated.For our own piece of mind we want to make sure our exact proceedure and joint is sound, although we have been doing it this way for 30 years and the joint is our industry standard.I actually did some in house side bends yesterday and the coupons cut from the specimin shown in the drawing passed with flying colors even when bent to a radius almost half smaller then D1.1 requires.I worked for one of our competitors for a few years and they used the flat plate method to certify individual welders but never got a PQR.BillThanks,BillLast edited by bbiggs; 02-04-2010 at 06:46 AM.Reason: spelling
Reply:bbiggs,  Why are you conducting side bends to qualify a procedure? If you look in table 4.2 WPS qualification CJP groove welds this should be looked at and followed. From what I see you are required to perform 2 reduced section tension test samples and either 2 Face and Root bends or 4 side bends depending on the size pipe. Check into D1.1 section 4 for understanding Procedure qualifications.Weldor/ Certified Inspector
Reply:Our pipe is 4" so falls in the 150 to 20MM range. It requires NDT (Visual & RT or UT) 2 RST and 4 side bends.We actually got our PQR today, and the welder who performed the welds on the test specimens was qualified for 1G-pipe or tube horizontally rotated.Bill
Reply:Good glad to hear it worked out for you now you can proceed.Weldor/ Certified Inspector