Alloy 825 internal surface cracks

[Hansac]

Dear all.

We have an Incoloy 825 pipe, just received from the mill with internal surface cracks first detected by visual (plain eyes!) then confirmed with dye penetrant (PT).

Questions:
1. From ASTM B423 and then ASTM B829, it seems that these surface cracks would not be able to be detected by UT. Other than PT, how else can we screen/assess these surface cracks?

We discussed internally on PAUT, pulse eddy current, TOFD, what else?

Please note that these are full length pipes and we discovered by chance these surface cracks after cutting the pipes to spools.

2. We surmise that being seamless construction, these surface cracks occured during fabrication. Is this a common thing? Considering 825 metallurgy?

3. The cracks are roughly 5-13 mm in length and about 1-1.5 mm in depth, and they are numerous!!!! Seems like those ductile surface cracks you see on copper when subjected to cyclic stress.

Appreciate everyone's input, thank you in advance.

 

[EdStainless]

So I presume that you bought this to the requirements of B423.
What condition did you specify (cold worked annealed, HF?)?
What testing did you specify (ET, UT, or hydro?)?
UT would easily detect these, the UT reference standard uses notches that are 12.5% of the wall though.
So I can heat the argument that these are below that level. But you should be able to detect much smaller ones, especially if this product has been cold finished.
Have you tried flaring a sample or doing a reverse flatten test? Even though these are not part of B423 they should open the cracks and help show how bad this is.
Longitudinal crack in seamless product are not uncommon. Depending on the processing they may be at a slight angle to the axis of the pipe.
Have you been to the mill that made the pipe? Ever witness any production? Ever audit their QA system?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy

 

[Hansac]

Hi there Ed. Thank you for your reply.

1. Yes it is under B423.

2. As per the mill certificates:
a. Cold finished
b. Delivery Condition: Annealed Pickled Passivated

3. Testing (all passed):
a. UT as per ASTM E213
b. Hydro as per ASTM B829
c. Etching test as per ASTM E381
d. PT for beveled ends and 100 mm from ends as per ASTM E165
e. Huey Test as per ASTM A262 Practice C
f. Flattening Test as per ASTM A530

4. Our UT technician could not detect the cracks that we found and verified using PT. Thickness is 14.27 mm so 12.5% notch is 1.78 mm. Both the crack width and crack depth are below this dimension of 1.78 mm. I understand your argument but somehow our UT technician could not detect/separate the UT signal from the back-wall echo. UT probe diameter is 10 mm, and not twin-crystal (I think).

5. Flattening test was done (see 3(f)) above, and passed. Flaring test was not done. Will explore "reverse flattening test"?

6. That's my exact question: is internal surface/longitudinal cracks on a seamless internal pipe common? I think it that is quite shocking to have visible surface cracks for newly procured pressure containing service parts (in this case piping).

7. Personally I have not been to their facility, and I don't think the production was witnessed by any of our representative, but I believe their QA/QC system has been audited by our counterpart, but that also I am not sure.

Thank you again. I am very concerned with these cracks. Looking forward to be enlightened and further educated on this.

Regards.

 

[EdStainless]

When they do the UT at the mill the pipe is submerged in water and the transducers are on the outside. They send a signal in both circumferential directions. With cold finished pipe I would expect the background noise to be less than 40% or the rejection level. So these indications should be very clear. I have done UT on seamless using 5% notches and it worked.
What has the mill said?
At the least this is very poor workmanship.
Does the remaining wall under these indications still meet the required minimum? If not then there is no question.
I would take a short piece (1 to 1.5x the OD) and cut it lengthwise into a half pipe. then flatten it in a press so that you stretch the ID surface. This should clearly how how extensive and deep the indications are. If any cracks grow (tear) when you flatten it then I would not use any of it.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy

 

[JR97]

Hansac,

Best UT technique would be an ID creeping wave from an L70 wedge on the OD. The technique was developed to find IGSCC in Nuke plants back in the 1990s. If you have the right fixturing ET from the ID should also be a good choice to find very shallow ID flaws.

JR97