bekks
Materials
- Mar 9, 2013
- 31
I am helping a fab shop qualify a WPS and I realize there are other easier ways of doing this, but I have the weld coupons in hand and am going to test them. I'm looking for advise on root bend tests. Here are the details:
7/8" diameter pipe with 0.125" wall thickness.
Partial penetration weld of about 2/3 of the wall thickness, let's call it 0.9" of penetration.
Square butt with no gap.
Laser weld.
Dissimilar material, 316 to a martensitic stainless with a min. %el of 14%. This creates some kinking potential, but let's assume I can figure that out.
Figure QW-462.3 (a) - General note (b) of ASME Section IX (there is also a clause that states this as well) allows the bend specimens to be one quarter of the pipe or 3/8" wide.
I've done a face bend successfully the following way: Imagine the cross-section of a quarter section of the pipe and place a caliper across it to measure from the cap surface to the two cut locations of the quarter circumference - this is what I am calling the "thickness", which was 0.180", rather than the 0.125" wall thickness. I then calculate my 3 point bend mandrel diameter to be 1.1" for an outer fiber elongation of 14%. I realize that this strains only a finite point of the bend at 14% and the rest is less than 14%. That is my problem to justify and I'll probably try to decrease the mandrel diameter to test more of the circumference at 14%.
Here's where I'm looking for advice:
The root bend needs to have the ID surface machined off to the point of penetration, which I'm trying to find a way around as that is not a simple machining task. Should I try to flatten the sample first? Anyone have any clever suggestions?
7/8" diameter pipe with 0.125" wall thickness.
Partial penetration weld of about 2/3 of the wall thickness, let's call it 0.9" of penetration.
Square butt with no gap.
Laser weld.
Dissimilar material, 316 to a martensitic stainless with a min. %el of 14%. This creates some kinking potential, but let's assume I can figure that out.
Figure QW-462.3 (a) - General note (b) of ASME Section IX (there is also a clause that states this as well) allows the bend specimens to be one quarter of the pipe or 3/8" wide.
I've done a face bend successfully the following way: Imagine the cross-section of a quarter section of the pipe and place a caliper across it to measure from the cap surface to the two cut locations of the quarter circumference - this is what I am calling the "thickness", which was 0.180", rather than the 0.125" wall thickness. I then calculate my 3 point bend mandrel diameter to be 1.1" for an outer fiber elongation of 14%. I realize that this strains only a finite point of the bend at 14% and the rest is less than 14%. That is my problem to justify and I'll probably try to decrease the mandrel diameter to test more of the circumference at 14%.
Here's where I'm looking for advice:
The root bend needs to have the ID surface machined off to the point of penetration, which I'm trying to find a way around as that is not a simple machining task. Should I try to flatten the sample first? Anyone have any clever suggestions?
