GMAW of A387 Gr 22

Date 07-09-2002 14:30

We usually use 90T-B3 for FCAW, and do not retard the cooling rate from welding, but always PWHT on 2 1/4 Cr, 1 Mo material. Was there something specific that you are asking about?

By Tim Buyle (**)

Date 07-09-2002 14:57

We have to weld a PQR, 90 mm thick coupon, followed by PWHT (step cooling cycle). Normally for this combination of material and PWHT only SAW and MMAW are used. I can not find any literature about mechanical properties obtained in 90 mm thick plates and after such PWHT. Normally thinner sections are welded for constructions.

These materials suffer from temper embrittlement during exposure long time at high temperature. This is simulated in a step cooling cycle PWHT. Resulting in low Charpy V impact testing values, due to brittle phase forming.

Question: who has experience with 90 mm plate and GMAW resulting in good properties after step cooling cycle PWHT ?

By GRoberts (***)

Date 07-09-2002 19:03

I don't have exactly what you want since we don't use GMAW much here. Why do you want to use GMAW over FCAW? You can usually get a better deposition rate with FCAW for thicker weldments like you are talking about. We only have one test with CVN testing. It is not very common to require impact testing on these type of alloys since they are used at high temperatures. The one test we have is on 1 1/2" (38mm) plate with a 1250 PWHT for 20 hours. These are our results

SPECIMEN NOTCH TEST TEMP ENERGY LAT. EXP. PERCENT
ID LOCATION (Degrees F) (ft lbf) (mils) SHEAR
AH HAZ +35 66 53 40
BH " " 54 49 40
CH " " 65 48 40
DW WELD " 52 36 50
EW " " 77 51 60
FW " " 55 38 50
GW " " 111 71 85
HW " " 89 44 60
A2 BASE " 49 43 35
B2 " " 50 45 35
C2 " " 47 40 35

Our base metal was A217 Grade WC9, and we used ESAB 9018-B3. Tensiles were 88ksi though. We do have FCAW test result, and 4" (102mm) thick plate results too, but no CVN tests on those. I don't think your CNV results will vary though, if you have tesing on 1 1/2" or thicker plate. The cooling rate of the weld metal varies by thickness, but by 1 1/2" or so, the cooling rate curve has flattened out. Usually they are taken at the surface anyway, unless your specification requires something else. Have you contacted any manufacturers to see if they have any data like what you want? I don't know who would be best for GMAW wire, but if you're considering FCAW, I'd contact ESAB, Tri-mark, or Select-arc. (if in the USA)

Also, could you describe your heat treat in terms of time and temperatures? I am interested in this requirement. Is the impact testing due to cold start up temperatures?

I can also see that the table didn't turn out after posting the way it was before posting. Hopefully you can figure it out.

By Tim Buyle (**)

Date 07-10-2002 07:01

As I describes this material and the welds are susceptible for embrittlement. To simulate this phenomenon there's a special PWHT, called step cooling. This PWHT is as follows :(example) warming up to 600°C at 50°C/hr, holding time 1 hr, followed by cooling steps : cooling down at 5.5°C/hr to 540°C, holding during 15 hr // cooling down at 5.5°C/hr to 525°C, holding during 24 hr // cooling down at 2.8°C/hr to 500°C, holding during 60 hr // cooling down at 2.8°C/hr to 470°C, holding during 100 hr // cooling down at 10°C/hr to 315°C, followed by air cooling.

After this cycle CVN values are measured.There's a shift in the 54 Joule transition temperature, compared to normal PWHT : example for normal PWHT there's 54 J at -40°C, but for PWHT with step cooling there's only 54 J at -20°C !!! The material is more brittle after step cooling. The acceptance criterium is formulated : transition temperature after PWHT + (1.5 x transition temperature after PWHT with step cooling) <= 38°C

Not every welding filler can give these results. If FCAW is used then step cooling results are very bad and the criterium fails. This is due to impurities (talking about ppm's P, As, Tn, Sb) in the flux. There are so called "embrittlement factors", describing the maximum amount of allowable impurities : Bruscato and Watanabe factors. There's no fabricator capable to produce flux cored wire with this low impurity level. Normal flux cored wires for this material of course exist but can not meet the requirements of step cooling !

For this reason there's only MMAW and SAW, step cooling tested, for most fabricators. Also GMAW should not be a problem ! So, I am looking for some field experience on thick plate !

If you want to read about this, look for "Science and Technology of Welding and Joining 2000 Vol.5 No.5" : good article.

Further : http:\\www.msm.cam.ac.uk/map/mapmain.html
Further : http:\\www.btwcan.com/html/ppf8.html

By GRoberts (***)

Date 07-10-2002 14:51

This is very interesting. I have only seen the X and J factors referenced on the Thyssen site one other time, but it was on a 5 Cr, 1/2 Mo project we were working on. The customer backed off from that requirement though. I know it can be very difficult to meet. You might try getting ahold of Bill Newell or Roger Swain from Euroweld. (704) 662-3993 they do know quite a bit about Cr-Mo steel welding. Unfortunately, I don't think I have any info that may help you.