Stress Relieving 9% NI, Tensile Properties

[AvusBlueM3]

Here's the situation. We cold form our heads for pressure vessels using 9% Ni plate (A553 and A353, mostly A553 though), and stress relieve the heads after forming. We closed a plant in Plaistow, NJ and inherited (paided for) a bunch of heads from that facility. Problem was, they didn't have any heat numbers or markings on the heads, thus they were not traceable as per ASME code. (We only found this out after it was too late to get a refund). But ultimately we narrowed it down to two possible heat numbers, both of which the original material passed all code requirements.

Thus to pass ASME code, we had to retest the heads. Chemical Analysis, Charpy Impact Test, and Tensile Tests had to be conducted.

Upon retesting, the heads failed its tensile test, but did pass the impact tests and the chemical analysis. Before the forming process the UTS was at 106.8ksi or 101.9ksi. After these values fell to 97.28-98.27ksi. (2 tests, one on each head tested) ASME code requires +100ksi.

Is it possible that the forming process or the stress relieving process affected the tensile strength of the heads? I know that the forming process does manipulate the mechanical properties of the steel a bit, but we haven't had this problem with the tensile strength after forming as ASME code doesnt require us to retest properties after forming if the heat numbers of the original material follow each individual head.

Any suggestions or comments? Any help would be great...

Ryan Felsenthal
Quality Engineer

 

[metengr]

AvusBlueM3;
The web site below contains excellent technical information for 9% Ni steel - composition, formability, heat treatment, welding, etc. The information was complied by the International Steel Group. I reviewed the file and there could be several reasons for the lower tensile strength that you reported. When you click on the web site, scroll to the link labeled 9% Ni steels. A pdf file will be available for download.

http://www.intlsteel.com/content/products/prod_plate_lit.aspx

 

[metengr]

After reviewing the ASME SA 20 Specification for General Requirements for Steel Plates for Pressure Vessels and the technical information from the ISG web site, I can think of two possible scenarios;
-you could be observing scatter with the ultimate tensile strength data because of your location of sampling (where did you remove the samples and how were the tests conducted?)
-the heads were tempered outside of the narrow temperature band mentioned in the ISG publication for 9% Ni steel.

You did not report in your post other tensile test results like yield strength (0.2% Yield Strength or extension under load method) and % Elongation. How did these results compare with the original tensile results in the MTR'? If the recent test data shows the same trend - lower yield strength and higher %E , it would seem that the heads could have been tempered outside of the narrow band recommended by ISG of 1050 to 1125 deg F.

If the only anomaly is UTS and the other tensile data is at or above minimum values, I would sit down and discuss this with the material supplier of the heads to determine what can be done using ASME SA 20 as a guide. Review this Specification carefully.

 

[EdStainless]

Depending on how severe the forming was you may have some directional effects. What was the orientation of your tensile samples? Have you checked multiple orientations?

The other thing is to check some microstructure. As I recall this grade is supposed to have some retained austenite to help the toughness (like ~10%). If there is little to none they screwed up the heat treatment.

= = = = = = = = = = = = = = = = = = = =
Corrosion never sleeps, but it can be managed.

http://www.trenttube.com/Trent/tech_form.htm

 

[AvusBlueM3]

metengr & edstainless,

We sampled from the top of the head as to reduce the influence of the forming on the properties of the metal. According to our suppliers, the A353 was double normalized and tempered all within spec of the ISG guidelines. Here is some other data:

Original MTRs:
Yield Strength - 84.0/83.2 ksi
Ultimate Tensile Strength - 101.9/100.9 ksi
% Elongation - 31/32%

Post-Forming:
Yield Strength - 91.5/86.4 ksi
Ultimate Tensile Strength - 98.3/97.3 ksi
% Elongation - 30/32%

Normalizing: 1st @ 1690F, 2nd @ 1450F, Tempering @ 1095F. Hold times 1H/inch, Still Air Cooled.

I'm not sure of the orientation of the sample with respects to the directional effects. And I will look into the austenite so see if the heat treatment was done correctly. We have sent in another sample to the former to get stress relieved again. This will provide us with some more information to see if these heads are recoverable.

Comments/Suggestions?

~Ryan

 

[metengr]

Based on your latest information, it would appear that the heat treatment is probably correct in comparing the YS and %E results between post forming and original heat treatment - the austenite percentage (5 to 15% range) is probably acceptable as well because of the impact values you stated.

It looks as though with the UTS on the low end from the original heat treatment, it does not surprise me that you fell slightly below this value after forming and tempering. Statistically speaking, 97-98 Ksi is darn close to 100 Ksi, especially when the original heat treatment resulted in a low end UTS of 101-102 Ksi, and you factor in post forming directionality effects, as well.

Let us know what the second set of samples reveals.

 

[unclesyd]

As stated above your values are quite close and I think well within the scatter and variation due to coupon orientation. I have a table in a United States Steel booklet) showing quite a difference in Charpy Impact values for full and reduced size specimens and longitudinal vs traverse with the higher values being in full size longitudinal specimens on NNT (normalized, normalized, and tempered) material.

More years than I like to remember the lab where I was working at the time performed some physical testing for CBI on material from their failed cryogenic test vessels. Based on some very old notebooks the numbers show almost a 15%, ±7.5%, scatter on UTS, YS, and El. These vessels were 3/8" so all the test coupons were reduced section by our standards. I did not mention any testing the head material and don't think any was tested as the heads were hemispherical.

There appears no difference between the QT (quenched and tempered) and NNT material except on EL. The NNT material has a slightly high elongation than the QT material, but the scatter overlaps. I know some of these vessels were stressed relieved after fabrication.

We were primarily looking at values at LN2 temperature and RT.

All the test vessels had failed in a ductile manner at four to 5 times the design hoop stress.

meteng

I know you probably caught this on the news, ISG just got bought out by Dutch company. Probably going to be a loss of lot of good information, again.

 

[AvusBlueM3]

Test results are in! These results are after a 2nd heat treatment on the sample.

Post-Forming, Post-2nd Heat Treatment:
Yield Strength - 81.4 ksi
Ultimate Tensile Strength - 99.3 ksi
% Elongation - 32%

So pretty much didnt change after the second heat treatment. One thing noteable happened though, the Transverse Charpy Impact failed miserably this time...

Longitudinal Average - 39.3 (lateral expansion, mils)
Transverse Average - 8.3 (lateral expansion, mils)

Compared to the previous Impact results...

Longitudinal Average - 50.3(lateral expansion, mils)
Transverse Average - 51.3 (lateral expansion, mils)

Any ideas why the charpy impact was affected so much by the 2nd heat treatment? Again we tried to keep the sampling area as constant as we could.

Thanks,

~Ryan

 

[metengr]

AvusBlueM3;
What was the second heat treatment? I thought that you were going to conduct a second set of tensile specimens to evaluate the degree of scatter, and to assure proper sample orientation.

The only explanation I have for the sudden drop in impact properties, assuming all test samples were obtained from proper orientation, is that the level of retained austenite had decreased.

To confirm if this is the case, you need to perform a microstructural evaluation. At this point, these formed heads sound like a nightmare. Are you sure they weren't tagged for non-conformance originally??

 

[unclesyd]

If possible you might send a coupon or contact Dr. Manahan at the following Website.

http://www.mpmtechnologies.com/CharpyFit.htm