By jwright650 
Date 12-03-2015 19:18
Edited 12-03-2015 19:22
Yes, The side bends merely give you another option other than just the root and face bends in older versions of the code.
Think about this: These bend straps only give you a glimpse in time of the actual welding that took place during this test. If you are finding issues with one of them, more than likely there are more....slicing more off the coupon isn't making the welder a better welder. The code gives you the option to slice more off the coupon when there may have been a problem with the way the coupon was sliced or bent(prepared) and may have caused the issue, but that is my opinion.
edit: Bend tests vs RT(code allows either method)....RT shows you the entire length of the joint that was welded, vs the small kerf where the saw just happened to find something. This was my reason for giving the statement above.
Agree with John
Corner cracks on a root/face bend are the only reason you could prepare another bend.. And that alternate sample would need to be a root or face strap (whichever had the corner crack)
If one component of the test fails.. The whole test fails...
For example:
If visuals fail for overlap or porosity, you can't "grind her down and see if it bends ok" Just pick an area that looks good. That would be unethical
John and Lawrence answered just fine but let's walk through this to try and specifically answer your question: "My question is if either face or root bend fails are you allowed to test another specimen using side bend out of the same test plate. If not then what does this statement mean?"
You have your welder/self run the welder performance qualification coupon at 3/8" X 7-8" length and I like to use 7-8" wide, 4" each plate, because it is easier to grip and bend in my wrap around. Plunger doesn't make as much difference in the width of the plate. You need to ask yourself some questions before even beginning as it will make a difference from before an arc is struck on which test can be run. RT takes a different backing bar. So, will you be using RT or doing a bend test?
Once that decision is made and everything is set up correctly, the test is run. Then visual is performed per Clause 4.20.1(1) taking you to 4.9.1 for your accept/reject criteria. If that passes you move on to your RT or bend test. If you opted for RT then it is easy, the whole thing gets zapped, no cutting needed.
BUT, if you opted for the bend tests you have another decision to make: side bends or face/root bends? Either way you need two specimen cut from the coupon. If you went with my choice of face and root you prepare one of each according to Figure 4.31 and Table 4.11 (1) Test on Plate. Notice as well that when you choose the Face and Root bends from Table 4.11 you need to go to Figure 4.12 to find how wide your specimen will be cut. If you chose side bends then you go to Figure 4.13 to find the size of your specimen.
All the information is there. 1) weld coupons; 2) visual exam; 3) bend or RT. If the test fails, you re-test and take the next samples from new plates and now you need 2 instead of 1 of each. If I have to perform VT and RT on 2 instead of 1 it only stands to reason the welder must weld up 2 plates not just one. That obviously goes for root/face bends as well since you could not get 2 samples of each out of 1 coupon. Thus, it stands to reason TO ME, that it also applies to side bends. You may have enough material to cut multiple samples from each coupon, but you need 2 coupons in order to meet the requirements.
Now, your last question, YES, you can use 2 side bends, 1 for each of root and face bends, the first time. But, you are seeing less of the coupon. and I want as much of the sample tested as possible. And, take into account Clause 4.9.3.1 on Specimens. On side bends you must put the outside surface of the bend being the side that shows the greater discontinuity. We test to see what's there, not to make the best chance of passing.
Having said that, I never test to fail. But the code is plain that we do test to find if the weakest link is strong enough to hold the structure together.
He Is In Control, Have a Great Day, Brent
