= Tensiles But Different Elongs; Why? 5

[tc7]

We have been struggling with qualifying a weld procedure on a PH material (13-8). We welded two coupons and heat treated to condition H1050 (solution treat + 4 hour aging). In the same oven at the same time we also heat treated an unwelded plate of the same exact width and thickness and from the same mill stock as the welded plates. Welds (GTAW, Groove welds) were made with matching 13-8 filler (chemistry of the filler is virtually identical to the plates). We tension tested the three coupons and obtained the following:
Welded Plate #1: 183ksi at failure; 5% elongation; failure in weld.
Welded Plate #2: 185KSI at failure; 6% elongation; failure in weld.
Unwelded Plate: 187ksi at failure; 17% elongation.
(All tests were done by a very reputable independent test lab.)

The target tensile strength for 13-8 in condition H1050 is 175-180ksi. Manufacturers data (Allegheny) for unwelded mill stock elongation in H1050 condition is 12%. I don't have a reference elongation for welded 13-8 (should I expect an equal elongation?). By my reckoning, I have passed my weld procedure tension requirement, but cannot understand why the welded specimens elongations can be so different when HT was identical. It seems like the weld metal may have suffered some overaging, but the solution treatment should have addressed this????? Can someone offer an explanation on what is happening here please?

Also note that we were VERY careful to control welding heat input to 40-50 kJ/inch. I was also very, very careful about consistent grain orientation and tension direction of all plates.

By the way, we have not done the bend test yet, but will base the mandrel diameter on the 5% elongation.

Thankyou for your interest, all ideas are welcomed.

 

[EdStainless]

This look typical to me, based on my experience with other PH grades.
The bend test will be the critical one.
If you don't need the strength you could raise the age temp 25F and see if you get more elong, but I doubt that it will help.

= = = = = = = = = = = = = = = = = = = =
Still trying to help you stop corrosion.
formerly Trent Tube, now Plymouth Tube
eblessman@plymouth.com
or edstainless@earthlink.net

 

[redpicker]

I think I'd assume the differences in % Elongation are just due to the fact that you are testing weld metal as opposed to wrought plate. The weld metal is essentally a casting where the plate has undergone considerable hot-work. One of the major reasons for performing hot work is to improve the ductility (% Elongation) of the material.

rp

 

[GRoberts]

If the test is breaking in the weld as you indicate, you will never get as much elongation as the base metal. The simple reason is that if for instance, on a 2" gage length, the weld is only 1/2" wide, the elongation is only taking place in 1/4 of the gage length when the weld is softer than the base metal. But on a base metal sample, it can elongate over the whole 2" gage section. That is why there is never a requiremet in any code for elongation on transverse weld tension tests.

If you are qualifying in accordance with ASME Section IX, you have to do the bend tests at the minimum specified base metal elongation. For instance, the XM-13 requirement in ASTM A693 for H1000 is 10% for over 3/16" thick. However, just because you didn't get that elongation in a transverse tensile doen't mean you can't get it in a bend for the reasons mentioned above. If you want to know the true elongation of the weld metal, you would have to do an all-weld metal tensile tests.

 

[tc7]

Well GRoberts,
Based on what you are saying, then I should calculate my bend test radius on a specification elongation of 12%. Since I know that I have not achieved 12% in the transverse tensile, I don't see how I can possibly pass the bend tests unless I increase the bend mandrel diameter. Does the mechanics of bend testing somehow equalize the elongation between the base metal and the weld metal?

.

 

[metengr]

tc7;

Does the mechanics of bend testing somehow equalize the elongation between the base metal and the weld metal?

No. The bend testing in ASME Section IX was designed as a go/no-go threshold criteria for minimum ductility (22% fiber elongation). If our base material cannot achieve this fiber elongation, the radius has to be increased to accommodate the minimum base metal fiber elongation. If it is a minimum of 12%, this is what you would bend test for threshold fiber elongation.

 

[GRoberts]

TC7,
As I mentioned, you do not really know the ductility of your weld metal because transverse tensile tests are not homogeneous, and the results are not indicitive of real elongation properties. So unless you do an all weld metal or bend test, your weld metal ductility will not be known.

 

[tc7]

GR-
I understand what you are saying – the elongation calculation using a 2-inch gage length results in a superficially low value because only the weld metal is stretching with very little contribution from the base metal. And that was the point of my original question: Why is my weld coupon not behaving homogenously?
EdStainless and Redpicker have indicated that my results so far seem typical for PH materials, yet with matching filler and a full heat treat as described, I certainly would have expected similar behavior in base and weld. Hardness measurements in the weld area were consistent with readings (HRc40-42) in the base. I can only guess that the explanation must have something to do with differing grain structures between the weld metal and base metal, which the solution treat and the aging heat treatment did not equalize. Is this a plausible explanation?


If I adjust the elongation calculation and instead use a gauge length of ~½-inch (which is equal to the widest part of the V-groove (45 degree included angle in 3/8 thk plate, 3/16-inch root)) then I get an elongation estimate of about 20%. This would indicate a bend test using the 12% specification minimum mentioned by Metengr should have a good chance of success. What do you think?

 

[NickE]

"I can only guess that the explanation must have something to do with differing grain structures between the weld metal and base metal, which the solution treat and the aging heat treatment did not equalize."

Exactly, in redpicker's first post he mentions:

"The weld metal is essentally a casting where the plate has undergone considerable hot-work."

That is why you wont get the same properties from weld metal as the base metal. The base metal was extensively hot worked after casting and has a particular grain structure, and properties as a result. I would expect a much larger grain size and odd segregation of the alloy in the welded area, both of wich would reduce elongation.

Processing=Structure=Properties are permanantly interconnected, as you have discovered.

Nick
I love materials science!

 

[GRoberts]

One thing the SA+age will never be able to make up for even though it does typically help the weld come closer base metal properties, is the improved propterties in the rolling direction of the plate. The weld grains will be approximately equiaxed, or slightly elongated in the direction of the weld.

 

[tc7]

I'd like to provide a final update to anybody interested on my 13-8 PQR: All bend tests passed at the required 12% elongation mandrel diameter. However, three of the four bends did exhibit minute surface defects due to porosities opening up at the surface. I believe these porosities may have simply been due to high humidity (highest of the summer) when we welded the coupons last July because we took many precautions to keep everything clean (even wiped down the filler rods and each weld pass with alcohol). Anyway, I plan to RT a couple of preproduction welds to be sure we have resolved the porosity problem.
Thanks to all who responded to my questions, all of your advice helped ease many concerns that I had.