Charpy Help

Date 06-05-2013 00:12 Edited 06-05-2013 00:18

I looked at your report, and there are at least 3 pieces of info/questions that stick out.

1. What is the average value that you have to hit? What did the one sample that passed hit?

2. "Absorbed energy levels below 6J are outside the verified range." Out of 6 values, you have 1@7J and 1@6J, and the rest are below 6. 66% of your sample values are suspect.

3. There is a fair amount of info on the interweb (I know, I know) that shows the NDT of A36 at about -40 F, and you are testing at -45 F. Gonna be tough. (or not)

BTW, you have to chop off the characters after the ".pdf" to see the report

There seems to be 3 main impact measures, related to the %ductile vs.brittle fracture. The report shows 5% ductile failure, so you are virtually at the end of the curve.

http://www.uvm.edu/~dhitt/me124/CharpyImpactNotes.pdf

I would be shopping for some different base metal

By ltfab Date 06-05-2013 09:47

Thank you for confirming my base metal suspicions.
The one sample that passed was the Weld at centerline notch location sample. It hit 98, 93, 80, for an average of 90.
Lateral Expansion was 81,76,64, for an average of 74.
Percent Shear was 95,95,95 for an average of 95.

We have to hit the minimum of 15,15,15 and average 20.

Do you have a different base metal recommendation? Our customer seems to be open to qualifying a grade 50; I was thinking maybe A572 grade 50, with PWHT (1150 for an hour) to stabilize the grain structure in the HAZ.

Link is awesome info - thank you.

By fschweighardt (***)

Date 06-05-2013 13:03

Was the one the passed done at -45? Dont see how the centerline is 15 x tougher.

By MMyers

Date 06-05-2013 13:24

If the weld metal is appreciably tougher than the base metal or HAZ, getting those sorts of results is possible. I worked on a project where adjacent impact tests pulled from the same weld would vary by as much as 280 ft-lb. We were literally getting hammer stoppers right next to samples that broke below 20 ft-lb. That was a frustrating problem.

Have you done base metal impacts at that test temperature? If not, I'd get some done. When doing impacts, it's always wise to do a set of base metal impacts along with your weld and HAZ impacts. This way if you end up with non-conforming results, this allows you to identify where the problem is: the welding process and resultant HAZ, or the base metal itself. If you have a base metal that can't hit your minimum required impact energy, then getting conforming HAZ impacts is less likely. I'd offer a suggestion for another base metal, but unfortunately I don't play in the structural steel world.

By fschweighardt (***)

Date 06-05-2013 22:35 Edited 06-05-2013 22:47

It looks like the weld metal is ER70 S-6, which should not have any stunning impact properties. I think one of the fails was 1mm from centerline, and that sample should have been all WM in a 45 deg included angle on backing with a 1/4" root opening.

By MMyers

Date 06-05-2013 23:11

The report states the notch locations are +1 mm and +5 mm from the fusion line, not from the weld center line. That would put them in the HAZ, or at best a mix of weld metal and HAZ because of the groove angle.

By fschweighardt (***)

Date 06-06-2013 03:05

I just checked the Lincoln site for L56, and it does show very nice impacts @-40. MMyers is right on about the failure locations, your base metal just isn't up to snuff.

By ltfab Date 06-05-2013 14:12

Yes it was. This has been a source of frustration for me these past couple of days.

By ltfab Date 06-05-2013 14:14

Mike, thanks for the suggestions about base metal impact tests. We have not done any base metal impacts. What you said makes great sense. Having base material impacts most definitely would help with narrowing the possibilities. Thanks!

By MMyers

Date 06-05-2013 14:42

Sounds good. Keep us posted.

Along with doing those base metal impacts, see if you can chase down the certs for the heat of base metal you tested and compare it to the A36 standard to see if it actually conforms to A36. This falls under the heading of "trust, but verify".

By fschweighardt (***)

Date 06-06-2013 00:24

http://www.keytometals.com/Articles/Art61.htm

some ideas on low temp steel

By ltfab Date 06-06-2013 09:21

Thanks for the info. You guys are a great help! I will keep this thread updated in the hope this headache can be alleviated for someone else in the future lol.

By js55

Date 06-07-2013 13:07

The thing about HAZ impact testing is that the CVN notch will always overlap into the weld metal or the UBM. If you have a great concern for plate CVN' S/SA-516 is the best choice. It has to do with the starting grain structure. S/SA-516 is a fine grained steel, as processed. Once you weld it the fine grain IS GONE, unless the 'normalizing' of the heat regime re-establishes some semblance of fine grain. Meaning, in general, low heat input. For large grain materials such as A/SA-36 it is a little different. Your HAZ will actualy demonstrate finer grain than the UBM, unless you really run hot and so your HAZ will actually improve the CVN's.
But, and this is the secret, even though ASME Section IX and all the others utilize heat input maximums as their standard of qualification, sometimes increasing heat input will improve CVN's. As you weld hotter and hotter the extent of the HAZ will expand. The trick is to maximize the 'normalized' width while minimizing the grain growth width.
Of course, this is all thinking in terms of single pass welding. When multiple pass welding it gets complicated beyond belief.

By ltfab Date 06-10-2013 16:14

Thank You for the information! Unfortunately, this falls in the multipass category, I believe. We have decided to use a different base material - I will certainly look into the S/SA-516. We will resubmit a test plate in the next couple of weeks, and with the info yourself and others have provided I believe we should qualify this time. You guys have been a great resource! Thanks!