Hi William

I believe that the weld is doable, but as you pointed out, the problem is most probably going to lie in the P91 HAZ wrt hardness. I am not sure why this is a big issue, but I believe that the buttering technique with two PWHT cycles will probably give you the result you are looking for.

A problem often encountered with these materials is that you end up with retained austenite in the weld and particularly the HAZ. In a single PWHT cycle, you stand a chance of transforming the retained austenite to martensite after only a single PWHT cycle. A double cycle will definately sort out this problem.

In addition, I would use a somewhat longer PWHT cycle than the minimum code requirement for the second PWHT cycle. As an example, instead of one hour, try double this time.

Obviously the double cycle on the P91 does not take away the same potential problem on the P11 material. I would think that giving both materials the double cycle may just be the safest bet. If you are doing this, you will also not have to worry about the "butter layer". Just weld it and let it cool down to the pre-heat temp. and immediately go into the first PWHT. Then perform hardness tests and radiography, and if required, perform the second PWHT followed by MPI and hardness tests.

Hope this helps

Regards
Niekie Jooste
Fabristruct Solutions

You stated
"2. Client recommendation is to use ER80S-B2 or E8018-B2 butter on the P91 side, followed by a 1350F PWHT; weld out with 1.25% Chrome and PWHT again at 1100F.
3. Construction Code is B31.1. This Code specifies a minimum PWHT of 1300F for the P11 side."

Unless my interpretation of number 2 is wrong, the last PWHT at 1100 F does not comply with the code of construction (B31.1). As you stated, the code requires that the P11 is to be PWHT at 1300F MINIMUM. Performing the PWHT at 1100F is below the code minimum requirement. In addition, code does not set an upper limit on temperature when PWHT is performed. The temperatures given in the code are minimums that must be followed. When write a procedure for P91 the PWHT is always 1375 – 1425 F and I try to target 1400 F. This is done for a minimum of 2 hours when wall thickness is 2” or less, and 4 hours when the thickness is over 2”.

My suggestion would be to Preheat between 400 –450 F, butter the P91 pipe and perform the first PWHT cycle at 1400 F for 2 hours after buttering. Weld the buttered P91 pipe (after PWHT) to the P11 using the B2 wire and a preheat of 400 F –450F. Since the wall thickness is almost 2” I would suggest a “Hydrogen Bake” after this weld. To do this I would hold the weldment at 450F for 2 hours, then let the weldment cool to below 200C (390 F). Perform the last PWHT cycle per code for P11. (1300 to 1375 F for 1 hour per inch.

Realizing that our client recommendation did not meet code PWHT requirements and opting for a single PWHT to minimize the carbon depleted zone due to the dissimilar chromium contents we have completed a test coupon as follows.
1. 12" Sch 160 (1.312") SA 335 P11 - P91 pipe; 1G, 37.5/10 degree modified single V groove, 1/16" land, 5/32 gap, flat position
2. 350F preheat (wrapped electric heating elements for thorough soak), 1/8" ER80S-B2 GTAW root 2.0 Kj/mm;2.3" of rod/in of travel
3. 350F interpass, 5/32" E8018-B2 SMAW, single pass/single layer; 0.8 Kj/mm
4. 350F-600F interpass (600F max per P91 recommended practice), 3/16" E8018-B2 SMAW, pass 3 single pass/single layer; pass 4-40 multi-pass/layer 12 layers total 1.0 Kj/mm tempering stringers
5. Cool to ambient 70F
6. PWHT 2 hours at 1375F (maximum temp for P11 per B31.1), ramp 260F/hr max, fall 330F/hr max
Hardness grid was performed at 1mm depth & 0.5mm spacing; P11 HAZ 173-183 HV10; root 212-216 HV10; cap 182-187 HV10; P91 cap HAZ 203-262; P91 root HAZ 233-274
We believe these are good numbers. The 4 readings in the P91 HAZ that exceed the 248 limit, range from 249-274; not bad for P91 PWHT at 1375F?
My question, is a 248 HV10 hardness limit on P91 HAZ's PWHT at 1375F max a reasonable expectation?
I believe we could go to the time (if we had it), and expense, to weld another coupon. We could use B2L filler, butter the P91 side using 500F preheat (maintain the 600F interpass), using 3/32" stringers followed by 1/8" stringers in a true temper bead application, followed by a 1375F PWHT, then reface and weld out the joint per the original procedure and PWHT again at 1300F and still not be sure about getting 248 HV10.
(We also prefer 1400-1420F for P91 hold times however even with the butter the client who is holding us to B2 filler does not want the lower critical temperature of the B2 filler exceeded)

Until a very few years ago, when metal hardnesses were mentioned on American standards, specifications, Codes, instruction manuals, wps etc., they were measured in Rockwell Hardness numbers (HR), either HRA, HRB, HRC etc. The most common were HRB, which used a steel ball, and HRC, which used a diamond cone.
Occasionally however, some document produced in America mentioned Brinell Hardness numbers (BHN). Vickers Hardness (HV) never. The use of Vickers hardness was limited to Great Britain and its former colonies, such as South Africa (right Niekie?) for the simple reason that Vickers way of measuring hardness, using a diamond pyramid, was developed in England by a company called Vickers Armstrong, which manufactured the first hardness meters using that unit.
Then, all of a sudden, Vickers hardness started literally flooding the United States. Nowadays, all documents produced in America refer to Vickers and Rockwell is completely forgotten.
Can anybody tell me what happened? The good old Rockwell numbers, which rendered decades of relevant services to the engineering profession, were found to be no good? Who discovered that, and WHY? What are the reasons for having put the Rockwell hardness astray?
Here at Mackenzie U. in Sao Paulo we have several hardness meters which use Rockwell and Brinell units. Should we scrap them?
Any opinion will be highly appreciated.
Giovanni S. Crisi
Sao Paulo - Brazil