where expertise comes together - since 1996

Web's Largest Portal of Nondestructive Testing (NDT)
Open Access Database (Conference Proceedings, Articles, News), Exhibition, Forum, Network

All Forum Boards
Technical Discussions >
Interpretation of 2235-9 maximum permissible flaw size
Career Discussions
Job Offers
Job Seeks
Classified Ads
About NDT.net
Articles & News

1275 views
06:30 Sep-03-2008

Nigel Armstrong

Engineering, - Specialist services
United Kingdom,
Joined Oct 2000
1094
Interpretation of 2235-9 maximum permissible flaw size

Respected colleagues,

out of interest, what is the basis for the values used in the Code Case for Table 2, a/t, (columns 2 to 4) for surface and sub-surface defects required as a prelude to determining the maximum permissible flaw length vs height?

What value should be utilised in the case of a subsurface defect of say 4,8mm height. Fig 1(c) annotates this dimension as 2a. Does that mean that for subsurface flaws only half the flaw height i.e. 2,4mm, should be utilised in calculating a/t and a/l?

Lastly can anybody provide a clear and simple explanation of Figs 2 thru 5 (flaw interaction).

Please if you have the answer, or an opinion, on any of these questions I woulod be grateful to hear from you.


 
06:44 Sep-07-2008
Massimo
Re: Interpretation of 2235-9 maximum permissible flaw size I daily read all the posts and answers on this forum since few years. I have observed that most of the posts regarding this ASME code case looks "ignored" by most of the very active and helpful participants, despite to the fact that this code case introduction brought a significant "revolution" in the PV testing processes.
I would give my opinion about the "a", "2a" definition.
I don't think that the intention was just to give a different name to the defect height, therefore I would input half of the measured defect height for subsurface defects acceptance calculation. ASME XI acceptance is like this. It would be very beneficial if someone who knows the answer would post it.




 
00:18 Sep-07-2008

Ed Ginzel

R & D, -
Materials Research Institute,
Canada,
Joined Nov 1998
1185
Re: Interpretation of 2235-9 maximum permissible flaw size Massimo:
You are correct. The concepts of "a" and "2a" are fundamental in fracture mechanics. The "a" value might be considered the "stress size" allowed. Stress concentrations for surface flaws are more critical and for embedded flaws a larger flaw is tolerated because the stress concentrations to the surface are reduced.

The same concepts (and the same use of "a" and "2a" terminology) are seen in other codes using fracture mechanics-based flaw-size acceptance criteria. BS 7910 is perhaps the most detailed description but API 1104 and CSA Z662 also use it for pipeline applications.

What is odd about ASME CC2235 is the limit on percent wall calculated. Typically half the wall thicknss is the limit calculated for; i.e. the length allowed when loss of 50% of the wall occurs. NORMALLY, provision must be added for potential for NDT sizing errors.
ASME CC2235 has significant conservatism built into it. You see in Table 1 that the maximum wall loss considered for surface flaws is 8.7% and14.3% for subsurface flaws. Table 2, up to 64mm wall thickness uses the same upper limits of 8.7% and 14.3%, then reduces this to 5.2% and 7.6% (surface and subsurface respectively) for the range from 100-300mm. Of course for embedded flaws this implies that the actual flaw size may be double that (i.e. "2a") so this increases the allowed flaw height for subsurface flaws to 28.6% (or 15.2% for 100-300mm) but that is still a long way from the 50% considered by other similar Codes.
There is no explanation for this in the rationalisation paper by Rana et al. Also there is no mention of NDT sizing error tolerances in the Code Case document (nor in the paper by Rana that was the foundation for the updated Code Case).

I have been unable to get explanations from "experts" in fracture mechanics as to why ASME would elect to opt for this treatment of the fracture mechanics principles. But I find it "odd" that the resultant maximum length ends up at nearly the same maximum length (or even LESS) as the workmanship values given in Section VIII Appendix 12 for the maximum vertical extent in the tables! When calculating allowed lengths for the range of thicknesses from 25-150mm, all the values end up back to about 12-17mm maximum length for a 0.5 aspect ratio of "a/l". This seems to be too coincidental to the one-third "t" or 3/4 inch imposed by the workmanship to be mere coincidence.

Yet for the same steel strengths and applied stresses, significantly greater flaw sizes are "tolerated" in pipeline applications.

In spite of the conservatism in 2235 there still seems to be some advantage using the Code Case...especially in wall thicknesses over 50-60mm.
Ed

----------- Start Original Message -----------
: I daily read all the posts and answers on this forum since few years. I have observed that most of the posts regarding this ASME code case looks "ignored" by most of the very active and helpful participants, despite to the fact that this code case introduction brought a significant "revolution" in the PV testing processes.
: I would give my opinion about the "a", "2a" definition.
: I don't think that the intention was just to give a different name to the defect height, therefore I would input half of the measured defect height for subsurface defects acceptance calculation. ASME XI acceptance is like this. It would be very beneficial if someone who knows the answer would post it.
------------ End Original Message ------------




 
02:02 Sep-08-2008
Ed T.
Re: Interpretation of 2235-9 maximum permissible flaw size ----------- Start Original Message -----------
: Massimo:
: You are correct. The concepts of "a" and "2a" are fundamental in fracture mechanics. The "a" value might be considered the "stress size" allowed. Stress concentrations for surface flaws are more critical and for embedded flaws a larger flaw is tolerated because the stress concentrations to the surface are reduced.
:
: The same concepts (and the same use of "a" and "2a" terminology) are seen in other codes using fracture mechanics-based flaw-size acceptance criteria. BS 7910 is perhaps the most detailed description but API 1104 and CSA Z662 also use it for pipeline applications.
: What is odd about ASME CC2235 is the limit on percent wall calculated. Typically half the wall thicknss is the limit calculated for; i.e. the length allowed when loss of 50% of the wall occurs. NORMALLY, provision must be added for potential for NDT sizing errors.
: ASME CC2235 has significant conservatism built into it. You see in Table 1 thatthe maximum wall loss considered for surface flaws is 8.7% and 14.3% for subsurface flaws. Table 2, up to 64mm wall thickness uses the same upper limits of 8.7% and 14.3%, then reduces this to 5.2% and 7.6% (surface and subsurface respectively) for the range from 100-300mm. Of course for embedded flaws this implies that the actual flaw size may be double that (i.e. "2a") so this increases the allowed flaw height for subsurface flaws to 28.6% (or 15.2% for 100-300mm) but that is still a long way from the 50% considered by other similar Codes.
: There is no explanation for this in the rationalisation paper by Rana et al. Also there is no mention of NDT sizing error tolerances in the Code Case document (nor in the paper by Rana that was the foundation for the updated Code Case).
: I have been unable to get explanations from "experts" in fracture mechanics as to why ASME would elect to opt for this treatment of the fracture mechanics principles. But I find it "odd" that the resultant maximum length ends up at nearly the same maximum length (or even LESS) as the workmanship values given in Section VIII Appendix 12 for the maximum vertical extent in the tables! When calculating allowed lengths for the range of thicknesses from 25-150mm, all the values end up back to about 12-17mm maximum length for a 0.5 aspect ratio of "a/l". This seems to be too coincidental to the one-third "t" or 3/4 inch imposed by the workmanship to be mere coincidence.
: Yet for the same steel strengths and applied stresses, significantly greater flaw sizes are "tolerated" in pipeline applications.
: In spite of the conservatism in 2235 there still seems to be some advantage using the Code Case...especially in wall thicknesses over 50-60mm.
: Ed
: : I daily read all the posts and answers on this forum since few years. I have observed that most of the posts regarding this ASME code case looks "ignored" by most of the very active and helpful participants, despite to the fact that this code case introduction broughta significant "revolution" in the PV testing processes.
: : I would give my opinion about the "a", "2a" definition.
: : I don't think that the intention was just to give a different name to the defect height, therefore I would input half of the measured defect height for subsurface defects acceptance calculation. ASME XI acceptance is like this. It would be very beneficial if someone who knows the answer would post it.
------------ End Original Message ------------

"a" is the vertical height of a flaw. "2a" is 2x the vertical height. It is more or less a bonus for being subsurface and away from the surface. As Ed G. said, less of a stress riser.
This has been ASME's philosophy for vessels for many years in the nuclear industry. (ASME Section XI)


 
02:15 Sep-08-2008
Bill Blanshan
Re: Interpretation of 2235-9 maximum permissible flaw size My company recently made a program which calculates a flaws "accept/reject" per the 2235 code case; in fact Ed.G helped me with the program. If you our anybody else wants this program, please email me at bill@autsolutions.net and I will send it to you.

Regards,
Bill Blanshan


 


© NDT.net - The Web's Largest Portal of Nondestructive Testing (NDT) ISSN 1435-4934

Open Access Database, |Conference Proceedings| |Articles| |News| |Exhibition| |Forum| |Professional Network|