Determining Wt loss on rolled surfaces

[MightyJ]

Hello all,

I am performing corrosion Wt. loss test (ASTM G48) on 4 samples. I will be immersing the entire 6-sided rectangular sample into the solution, however am wondering how I could obtain the corrosion Wt. loss of just the rolled surfaces (2 sides). The other 4 surfaces have been ground down. I realize I could weigh the sample, mask off the ground surfaces, test the sample, remove the masking and weigh the sample again. But was asked to not mask anything off and just immerse the entire sample. And I want to know the Wt. loss of the rolled surfaces for myself. Can any one help guide me through this?

I think the key is that I have 4 samples, two are almost identical in size with the rolled surfaces being equal to the 2 other sides and the 2 ends are about half the size. The other 2 samples both are different in thickness (the ground sides are smaller). The rolled surfaces of all 4 samples are equal. I can post exact dimensions if needed. Any help greatly appreciated.

I'm thinking of coming up with an equation and plotting maybe a ratio of the sides with wt. loss but just not sure how to go about doing this, if it would be right at all.

 

[EdStainless]

What alloy?
It is critical that you rig the test so that you get NO corrosion on the edges. We usually do this by polishing the edges.
The trouble comes in if you get end grain or midwall attack. There is nothing that you can do to prevent this, and you can't correct for it in the weight numbers.

G48 isn't a weight loss test at all. There is no general corrosion, only pitting. You are supposed to be looking for indications of pit initiation. You just use the weight to tell you which samples you can ignore. One very small and very deep pit will give a very small weight loss and still be a test failure.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[MightyJ]

Ah, I understand! And it is super duplex S32760. And we will be grinding the edges. So grinding will most likely prevent pitting on these surfaces? I now understand even if it doesn’t we would still examine it to determine a failure.

And I know burns from cutting would affect the test but would they corrode in the form of pits?

 

[EdStainless]

No, grinding won't help. You would need to polish like preparing a metallographic sample, through 0.5 micron diamond.
If there are some mid-wall inclusions or intermetallics then you will still get corrosion.

Are you going an A923 also? You really should. It is how you verify that the heat treatment was correct.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[MightyJ]

I see, makes sense. And we are detecting the percentage of ferrite near the rolled surface and mid thickness but not doing the ASTM A923. I guess while doing the % ferrite we could see if there was sigma present and then would know if the heat treatment was wrong? I think sigma occurs when heating 1500-1600 C and forms during cooling?

And yes we usually do these tests together (ASTM A923 Methods A & C, G48 Method A, and % ferrite). In ASTM A923A we look for affected/unaffected/possibly affected structure, which I also noticed the more sigma usually the more corrosion in the other tests. I guess sigma corrodes more readily?

 

[EdStainless]

Yes the sigma will corrode fast.
You won't see it in a micro unless it is really bad.
For the A923 you want all surface freshly ground with no passivation or treatment.


= = = = = = = = = = = = = = = = = = = =
Plymouth Tube