Not logged inAmerican Welding Society Forum
Forum AWS Website Help Search Login
Up Topic Welding Industry / Inspection & Qualification / Crack sizing
- - By triplethreat Date 10-18-2007 20:03
An engineer in my company was recently in England and was told by the people there that they could size cracks as small as 1mm (not centimeters but millimeters).  He asked me if I had ever heard of such capabilities and what would be the best technique to use.  I said they were pulling his chain or the technique was something new.  I said multi transducer UT scaning was probably the best method but the big issue with all the UT methods was the operator.  Has anyone heard of any other ways?
Parent - - By Joseph P. Kane (****) Date 10-18-2007 20:24
Technically they should be able to detect a crack as small as 1/2 wavelength. 

I think an operator of a single transducer would be Very Very Weerrrrry  good to be able to do it!
Parent - - By triplethreat Date 10-18-2007 20:33
What I was really wondering was "realistically".  I figure anything less than an 1/8" will be hard to do consistently.
Parent - - By hogan (****) Date 10-18-2007 20:46
+ or - .005" with phased array
Parent - - By triplethreat Date 10-18-2007 20:53
please elaborate on the phased array system I don't know too much about it
Parent - By hogan (****) Date 10-18-2007 21:00
try a search on this forum. what is your leve of understanding of ut?
Parent - By hogan (****) Date 10-18-2007 21:00
try a search on this forum. what is your level of understanding of ut?
Parent - - By CWI555 (*****) Date 10-19-2007 15:27
An array of transducers that typically can be individually tuned to an overall system and are typically mounted in one head.
finding a 1mm crack is not unheard off, IGSCC is often around that size. It does take a skilled technician to pick it up, and even
then, is dependant on several other factors.
Parent - - By ndeguy (*) Date 10-20-2007 05:36
Those lying limeys, take everything they say with a grain of salt! But seriously what was meant by that remark? Was it that defects as small as 1mm can be detected or that defect dimensions as small as 1 mm can be accurately measured or that the accuracy of measurement is equal to +/- 1mm. In the days before Code Cases such as 2235 and 181 there was always a problem of relating RT acceptance criteria, especially for porosity, to UT results. Since the advent of Engineering Critical Assessment (ECA) based on the fatigue strength of a material as determined by Crack Tip-Opening Displacement (CTOD) tests Fitness For Purpose/Service (FFP/S) acceptance criteria have been derived which rely on accurate measurement of both lenght and height of defects.

I have been taught and also teach others that unless the 20dB drop method is used then only dimensions bigger than the ultrasonic beam cross-section can be measured by UT, otherwise the technician is just measuring (and reporting the defect as) the beam dimension. The British BINDT produced a series of small handbooks in the 1970's enttled "Capabilities and Limitations of.." The UT handbook has a table showing the expected accuracy in defect sizing by manual ultrasonics for differing thicknesses and in general it is +/- 3mm.

For defects with dimensions smaller than the beam dimensions, signal amplitude from the defect is  COMPARED to that of an artificial reflector best representing the nature of the defect (side-drilled hole, flat-bottomed hole, surface notch) and an estimate of the size made from such comparison. All indications are modelled as relative to a perfect circular reflector. The German DGS system, which is an example of the amplitude comparison method, takes no account of the impact of defect orientation or surface morphology on echo amplitude.

With manual defect sizing the only MEASUREMENT techniques I know of are 6dB drop and MaxAmp for length and 20dB drop for defect height. The latter relies on having plotted the 20dB limits of the beam. AUT of girth welds uses zonal discrimination and is not much more accurate than manual in the hands of a good inspector. I dont know what Phased Array uSome yars ago trials known as PANI demonstrated the variability in manual UT results in both detection and measurement of defects. A new series of PANI trials (PANI 3) has taken plece recently to try to understand the reasons for this variability and ways to overcome it but I havent read anything about the results yet.
Parent - - By CWI555 (*****) Date 10-21-2007 02:12
"I have been taught and also teach others that unless the 20dB drop method is used then only dimensions bigger than the ultrasonic beam cross-section can be measured by UT, otherwise the technician is just measuring (and reporting the defect as) the beam dimension."

That part is correct insofar as it concerns standard manual UT. As for the days before code cases, in those days, UT was not allowed in lieu of RT.

"signal amplitude from the defect is  COMPARED to that of an artificial reflector best representing the nature of the defect (side-drilled hole, flat-bottomed hole, surface notch) and an estimate of the size made from such comparison. All indications are modelled as relative to a perfect circular reflector."

Again correct in regards to standard manual UT.

What I see missing out of your post is consideration for phased array and TOFD. I suggest some further research on the matter. You seem to have a good understanding of the standard methods.

Regards,
Gerald
Parent - - By NDTIII (***) Date 10-22-2007 04:36
TOFD is currently  used to size flaws down to 1mm with reasonable accuracy.
Parent - - By CWI555 (*****) Date 10-23-2007 00:43
Absolutely
Parent - - By ndeguy (*) Date 11-01-2007 11:04
Point taken Gerald - I am ordering Olympus Advanced PAUT book to bring myself up to date. I hope to attend formal TOFD training in the near future also.

Hogan 0,005" = 0,125 mm. - twice the thickness of the average human hair. Do you mean in a laboratory?  I can imagine that defects that size MAY be detected in the field but not that such defects WILL be measured to an accuracy of +/- 0,125 mm. And the problem is that FFS is making demands on the techniques for such precision.  Reminds me of the claims by "super" UT techs that they could size a defect accurately to 0,5 mm with manual UT. I am certain that some TOFD and PAUT techs will be making corresponding claims. Or am I even further out of touch and now that we are in the digital age all weld files are gathered by a tech and returned to an engineer in his office who tweaks the parameters to give maximum resolution?

The PANI test series rattled everybody's cages when manual sizing capabilites of experienced manual UT techs were accurately assessed. The new PANI rounds is to to try to find ways of improving capabilities.

I hope others who work regularly with PAUT and TOFD will post their realistic opinions here to help us understand the true limits and capabilities of the tools we work with (or not) in the future.

Regards

Nigel
Parent - By CWI555 (*****) Date 11-01-2007 16:14
Having worked with phased array, and TOFD, I believe it's realistic to go down to a mm reliably with a properly trained tech. In a lab, it can go lower. Standard manual methods are as you say.
Parent - By CWI555 (*****) Date 11-02-2007 06:34
Another note on "capabilities of the tools we work with (or not) in the future."

I believe your going to find little information on that one. Many of the larger firms have/are/or plan to develop these technologies in the future. As a point of history, Look to the SEL research and others. It has become a very competitive market, with names such as GE, Westinghouse, Olympus, Applus/RTD, Areva, Sigma, Shell global, and many others getting involved.
There are also a host of individuals looking to scoop these organizations on new techniques, (including myself). The only real information I expect that you will get out of them is in generalized form.

Which btw, since you've brought it up, some people on this side of the pond are not familiar with PANI. For those who do not know what it is:
Programme for the Assessment of NDT in Industry. It was initiated by the British Health and Safety Executive. which is a governmental body similar to a cross between the U.S. OSHA, and the EPA. http://www.hse.gov.uk/aboutus/index.htm.

Of note is the use of PISC III data for the PANI trials. (Programme for the Inspection of Steel Components) as a multi national collaboration. 
(I believe it was PISC III report number 41 and a couple of others that were used.)

Further information is readily available in many books and papers, so the intent is to give the people unfamiliar with it a starting point for research.

Regards,
Gerald
Parent - - By 357max (***) Date 11-06-2007 14:52
I submit a crack for measuring
Attachment: WeldingCrack.jpg (214k)
Parent - By CWI555 (*****) Date 11-06-2007 22:06
I believe that qualifies for reject. Crack, lack of a cover pass/underfill. I would recommend an OSHA compliant PPC buckling device to prevent these occurrences in the future. I would also recommend the photographer seek immediate psychological evaluation.
Up Topic Welding Industry / Inspection & Qualification / Crack sizing

Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill