I developed a procedure for 4130 to A514H a little over a year ago. It was for a critical weld, failure probably would not cause loss of life (though depending on the time and place, it could) but the economic loss would be high, possibly as high as 100MM dollars. I would advise that you look to ASME IX if you want to develop a procedure - AWS D1.1 simply doesn't have the relevant structure to record parameters for temper beading. I used SMAW, because I didn't think I could get the necessary control of heat input with fluxcore. A custom consumable was used (Devasco 10018, which is specially formulated at < 1% Ni to meet NACE requirements) I know "custom" sounds expensive, but Devasco is reasonable, they cost $6.50/lb which is about what you are going to pay for anybody's E 10018. The 4130 was preheated to 400F and buttered first with 3/32 dia. rods and then with 1/8 rods and the stringer was hot passed over during the 1/8 pass. The A514 side was preheated to 200F and not reheated further during welding, 400F was maintained on the 4130 side of the weld during welding. After the face of the 4130 was covered with the first two passes, the groove was filled with 5/32 rods (the production piece was a 20" 4130 tube welded into a plate of A514H in an odd shaped assembly that we didn't have equipment to turn, so it had to be done 5G, you might be able to use 3/16 downhand). During welding the A514 side of the weld was monitored to insure it did not exceed an interpass of 400F. The toe areas of the cap (1/8 on the cap, temperbead with 1/8) were temperbeaded. The test turned out very well, the WPS is compliant with NACE and ASME, developed good impacts, tensiles slightly under 100K, no problems with the side bends even though our lab mistakenly used a 1 1/2" mandrel instead of the 2 1/2" specified by ASME. The metallography was amazing, no sign of a large grained HAZ on the 4130 side, just a line demarcating change in grain boundary orientations. Doesn't sound as if you need NACE compliance, so the initial material selection should not be as critical, we started with a 80 KSI plate at about 218 BNH for the 4130. If you really need to go D1.1, I suppose you could simply add the parameters to the PQR and WPS, but personally I would at least have a look at what ASME requires of a temper beaded procedure to get some orientation. I wasn't bold enough to play around with heat settings on fluxcore because I had to come up with a procedure under time pressure, but if you have the time, it may be possible to dial in the correct temperatures to temperbead. Be prepared to look over the welder's shoulder during production welding, because if our shop is any indication, its best not to leave them to their own devices.
It sounds like you had everything under control. I like your approach.
The only thing I would add is that close attention to the welding parameters, typically considered to be nonessential variables by ASME Section IX, be monitored and recorded. The WPS should take the heat input into consideration if for no other reason than to ensure you can replicate what was done for the test sample on the actual production parts. AWS D1.1 Table 4.5 can be used as a sanity check for the reasonable ranges of voltage, amperage, travel speed, etc. to be listed on the production WPS.
Whereas this is not your typical carbon steel alloy combination, the welding parameters are key to ensure the lessons learned when qualifying the WPS are addressed in the production WPS. Your approach to maintaining a limited interpass temperature with the A514 and higher interpass temperature with the 4130, as well as using the tempering bead technique, was well planned and executed. The trick is to addressed those techniques in the WPS. This is when a standard WPS format may not be sufficient to provide the direction needed to the welders.
Again, you have your act together and it paid off! Good show with some great recommendations!
Best regards - Al
Thanks for the kind words Al.
Something I didn't catch on first read. You give your yields as 91/82. I'm assuming the 91 is for the A514? If the plate is < 2.5" T it is not within ASTM spec for A514. ASTM specifies a min yield of 100K on T-1 material < 2.5 T, 90K over. If you are doing a 1" CJP you are probably going to be using grade B which can only be obtained with a max T of
1 1/4". Grades are not important if you don't require charpy testing, but if you do, you will need to test each grade, as normal P designation rules for welding different materials do not apply. A couple of things to consider here.
1. You say a grade Q to 4130 procedure, but if your production is 1" T/t it is a different grade, unless its a connection is involving a machined part.
2. Grade Q has a much higher percentage of Cr and Ni than other grades. In A514 different grades have different chemistry, its not like some materials where grade is about heat treatment. If you don't need charpies and you don't need NACE, the grade business is really not important, but if you do, I would steer away from Q if at all possible for any 1" CJP, because Q tends to turn out harder than other grades after welding. The chemistry of Q is formulated so that it maintains the ASTM A514 requirement of 90KSI at over 2.5 T, which in the min thickness for Q. If you are not concerned with NACE, finished hardness is not an issue and if charpy's aren't required grade is not an issue.
For the project I was talking about in the other post, we had, I think, 4 different grades involved, so I ended up doing about 10 different procedure qualifications because charpy testing was required.
Also I would consider using A517 for any PQR testing. Same material, but 517 is assigned P numbers while 514 is assigned an S number.
Forum
