Charpy Impact Testing on reduced size super-duplex tubing

[AnsAvi]

Hi

I have a super duplex stainless steel tubing that I would like to test for toughness using Charpy Impact Testing. Because of the dimensions of the tubing, the sample cross section will have to be lower than what is recommended by ASTM. My understanding is the "absorbed energy" requirement for smaller samples is lower compared to standard samples. We have a minimum requirement for the absorbed energy requirement for super duplex bar stock material. Using it, How should a correction factor be used to determine the requirement that the tubing should meet? Can it be energy absorbed reduced in the same proportion as the cross section areas of the tubing vs the bar stock or is it more complicated than that?

Thanks!

 

[EdStainless]

Are you doing this for A923 compliance? Why not use the corrosion testing?
The only way to do this for odd sizes is to make some from known material.
Start with a piece of plate and make a series of bars in various sizes and test them all.
Small bars have a number of issues. They are more sensitive to the notch (broached vs ground) and location and alignment are more delicate (LT vs ST, near surface vs mid thickness, and so on).

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

Thickness of a subsize Charpy coupon needs to be considered in establishing an equivalent toughness to that of a full-size coupon of equivalent material.
The only Code guidance I know of is found in Section VIII-D1, Table UG-84, which imposes a penalty in terms of test temperature for ferritic materials.

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[metengr]

Since your CVN impact testing will not be in accordance with subsize testing requirements per ASTM E23, your information will not be comparable to any full size samples, period. Yes, you can develop your own correlation but this will require plenty of test samples to ensure statistical meaning and even in this case there could be reluctance in trying to correlate full size specimens.

To give you an idea of the magnitude of developing a subsize CVN correlation see the attachment

https://www.sfsa.org/publications/misc/ASTM A1-13 presentation.pdf

 

[AnsAvi]

Thanks for your responses. The testing was not for A923 compliance, but dictated by the application. Regardless, its very useful to know that a simple proportionality based on cross section area maybe good as a first pass, but we will need much more data to get a good correlation. I also like the idea of getting at least some of this data from "known" bar. Thanks for that EdStainless.