where expertise comes together - since 1996 -

The Largest Open Access Portal of Nondestructive Testing (NDT)

Conference Proceedings, Articles, News, Exhibition, Forum, Network and more

where expertise comes together
- since 1996 -
9155 views
Technical Discussions
Paul Holloway
Consultant,
Holloway NDT & Engineering Inc , Canada, Joined Apr 2010, 189

Paul Holloway

Consultant,
Holloway NDT & Engineering Inc ,
Canada,
Joined Apr 2010
189
04:57 Dec-07-2015
Error inherent when using 6dB drop for length sizing (AWS/CSA)
zoom image


6dB drop method underestimating actual flaw length

zoom image


T-joint lack of penetration throughout entire weld

[Heads up, this is less of a question and more an opinion piece -PH]

In manual ultrasonic inspection, the commonly understood practice for length sizing discontinuities larger than the beam width is the "6dB drop method". However, peaking the signal then moving laterally to 50% signal height can often undersize the length. The result is repairing just the middle "really bad part" of a flaw, leaving two smaller discontinuities at the ends.

The 6dB drop method is best suited for ideal reflectors like side drilled holes in reference blocks and tends to fall apart when applied to real, irregularly shaped discontinuities (e.g. lack of penetration, casting voids, etc). Unfortunately, the 6dB drop method is taught without incorporating real discontinuity reflection characteristics and an appreciation for all acceptance criteria limits. As such, discontinuities are easily undersized, resulting in repeated repairs, delays for the client, and the question "how did we miss that the first time?". Not good for the optics of ultrasonic inspections.

Interestingly, the AWS code accounts for this inherent error (somewhat) by requiring the 6 dB drop not from the peak signal, but from the "rating for the applicable discontinuity classification". See 6.29.2 in the 2015 edition of D1.1.

As an example, consider a flaw in a 3/4" thick butt weld (non-tubular, cyclic service) with a "d" rating of -8 dB. This corresponds to a discontinuity of Class A, which for a 3/4" weld has a "d' limit of +5 dB and lower. To find the extremities, the probe would be moved laterally until the "d" rating was reduced not 6 dB below the peak of -8 dB, but to 6 dB below +5 dB (which would be +11dB). This is well outside the boundaries for all possible classes, and ensures that the entire flaw is captured.

However, this method fails when the span between the maximum and minimum classes exceeds 6 dB. Take for instance the same flaw in a 2 1/2" thick butt weld. At 70 degrees, there is an 8 dB span between Class A and Class D. Subtracting 6 dB from the peak level of a Class A may leave some flaw remaining at the extremities if it exceeds the allowable length for the discontinuity class. Essentially, all you did was cut the middle out and make one big discontinuity into two little ones. D'oh!

I teach technicians to recognize the limitations of the method and recommend chasing the flaw extremities all the way out to a Class D. The only exception would be if leaving some small, allowable length of a Class B or Class C discontinuity is necessary (likely due to access limitations for repair, etc.)

4    
 
 Reply 
 
Manfred R Richter
Consultant,
Richter & Tomasi Ltd., Brazil, Joined Jan 2008, 31

Manfred R Richter

Consultant,
Richter & Tomasi Ltd.,
Brazil,
Joined Jan 2008
31
11:19 Dec-07-2015
Re: Error inherent when using 6dB drop for length sizing (AWS/CSA)
In Reply to Paul Holloway at 04:57 Dec-07-2015 (Opening).

Hi Paul,
Good matter. This has been identified by the company who I worked for years, and as a result their internal and additional requirements in specs/standards states that the sizing shall be made at the extremities of the flaws, implying that the inspector shall search the whole lenght of indications.

    
 
 Reply 
 
Anmol Birring
Consultant,
Birring NDE Center, Inc., USA, Joined Aug 2011, 746

Anmol Birring

Consultant,
Birring NDE Center, Inc.,
USA,
Joined Aug 2011
746
13:27 Dec-07-2015
Re: Error inherent when using 6dB drop for length sizing (AWS/CSA)
In Reply to Paul Holloway at 04:57 Dec-07-2015 (Opening).

That is why some codes measure relative to a percentage of the the reference level and not the peak level.

    
 
 Reply 
 
Paul Holloway
Consultant,
Holloway NDT & Engineering Inc , Canada, Joined Apr 2010, 189

Paul Holloway

Consultant,
Holloway NDT & Engineering Inc ,
Canada,
Joined Apr 2010
189
13:46 Dec-07-2015
Re: Error inherent when using 6dB drop for length sizing (AWS/CSA)
In Reply to Anmol Birring at 13:27 Dec-07-2015 .

Anmol,

I am mostly familiar with AWS D1.1, CSA W59 and ASME Sec. V. I'd like to do some reading on other codes. Which ones are you referring to?

    
 
 Reply 
 
Anmol Birring
Consultant,
Birring NDE Center, Inc., USA, Joined Aug 2011, 746

Anmol Birring

Consultant,
Birring NDE Center, Inc.,
USA,
Joined Aug 2011
746
14:17 Dec-07-2015
Re: Error inherent when using 6dB drop for length sizing (AWS/CSA)
In Reply to Paul Holloway at 13:46 Dec-07-2015 .

DNV OS-F101, App D, B 337 includes the fixed level method
You have provided some very good information. Technicians should know the limitations of the 6dB drop method and understand how the shape and size (when flaws are smaller than beam width) affect the measured size.

    
 
 Reply 
 
Vadim Kniazev
R & D, Director
Wave Process Simulation System Laboratory, Russia, Joined Jun 2010, 3

Vadim Kniazev

R & D, Director
Wave Process Simulation System Laboratory,
Russia,
Joined Jun 2010
3
15:16 Dec-07-2015
Re: Error inherent when using 6dB drop for length sizing (AWS/CSA)
In Reply to Paul Holloway at 04:57 Dec-07-2015 (Opening).

Dear colleagues,
I hope that this article will be interesting for you:
Opposite Phased Transducer – OPT for Sizing of Defects
http://www.researchgate.net/publication/266618048_Opposite_Phased_Transducer__OPT_for_Sizing_of_Defects

It uses a fundamentally different technology for length sizing.

    
 
 Reply 
 
Csaba Hollo
,
Retired, Canada, Joined Feb 2010, 301

Csaba Hollo

,
Retired,
Canada,
Joined Feb 2010
301
18:25 Dec-07-2015
Re: Error inherent when using 6dB drop for length sizing (AWS/CSA)
In Reply to Paul Holloway at 04:57 Dec-07-2015 (Opening).

Early on in my inspection career, before I had much accumulated knowledge and went by the letter of the codes, I found myself making these same mistakes.

I remember one such case where after UT examination to CSA W59, I had a piling weld marked up over 50% of the circumference in 2" to 6" lengths using the 6 db sizing method required by the code. The welder brought me over after he had excavated 25% of the circumference and instructed me to put a welding helmet on. He then proceeded to show me the 100% circumferential incomplete penetration as he gouged the remaining weld out.

It really opened my eyes to sizing defects with UT.

    
 
 Reply 
 
andrew cunningham
NDT Inspector
Canada, Joined Jun 2008, 238

andrew cunningham

NDT Inspector
Canada,
Joined Jun 2008
238
19:11 Dec-07-2015
Re: Error inherent when using 6dB drop for length sizing (AWS/CSA)
In Reply to Paul Holloway at 04:57 Dec-07-2015 (Opening).

Hi Paul
I applaud you that you have learnt 6dB is not an accurate method and you will to speak out. 6dB drop is at a “that will do” lazy way of sizing discontinuities. It is a good reproducible sizing method for repeatable results. It in no-way could be classed as accurate. I defy anyone to size the 50mm SDH in the IIW V1 with 6dB drop method. When a technician gets his/her first qualification to practice he/she knows nothing but the bare minimum to pass an exam. (You would not class a person who has just passed a driving test as skilled). Knowing that amplitude is governed by orientation before size, I one wonder how much practical ultrasonic experience the writers of procedures and codes have? It is up to the technician to learn other sizing methods so he/she knows what sizing method is best suited to the discontinuity found.

    
 
 Reply 
 
Paul Holloway
Consultant,
Holloway NDT & Engineering Inc , Canada, Joined Apr 2010, 189

Paul Holloway

Consultant,
Holloway NDT & Engineering Inc ,
Canada,
Joined Apr 2010
189
19:23 Dec-07-2015
Re: Error inherent when using 6dB drop for length sizing (AWS/CSA)
In Reply to andrew cunningham at 19:11 Dec-07-2015 .

Csaba/Andrew:

I guess we can all agree that there's no substitute for field experience. Csaba, I too was "shown the way" by a welder during excavation early in my field days and it really opened my eyes.

Although it's unrealistic to think we can fully prepare fresh NDT hands for all the tricks of the trade, I think there's a lot of room of improvement.

    
 
 Reply 
 
sreejesh pillai
sreejesh pillai
05:23 Dec-08-2015
Re: Error inherent when using 6dB drop for length sizing (AWS/CSA)
In Reply to Paul Holloway at 19:23 Dec-07-2015 .

Please refer to AWS D1.1 ,section 6.13.2.1.Indications of discontinuities that remain on the display as the search unit is moved towards and away from the discontinuity (scanning movement “b”) may
be indicative of planar discontinuities, relative to the sound beam, deviates from
the perpendicular dB rating which do not permit direct, reliable evaluation of the
welded joint integrity may result. When indications that exhibit these planer
characteristics are present at scanning sensitivity, a more detailed evaluation of
the discontinuity by other means may be required (e.g., alternate ultrasonic
techniques, radiography, grinding or gouging for visual inspection etc.,)

    
 
 Reply 
 
Charlie Jackson
Charlie Jackson
12:30 Dec-08-2015
Re: Error inherent when using 6dB drop for length sizing (AWS/CSA)
In Reply to Paul Holloway at 04:57 Dec-07-2015 (Opening).

To all

Guys despite which spec you use you have to apply the basics you should have learned during training. Find the flaw which will have many reflection faces and therefore variable amplitudes along its length then find the last facet of the flaw and apply your sizing technique - do not get the biggest echo and just drop it by half. By finding the last facet and applying all the technques correctly to a flaw bigger than the beam - 6db - 20dB - Max amp you should have hardly any difference.

We can all get wrapped up in specs but dont forget just apply the basics first -

Cheers Charlie

and yes Ive applied AWS - ASME and the many varied BS specs for over 40 years -

1    
 
 Reply 
 
Paul Holloway
Consultant,
Holloway NDT & Engineering Inc , Canada, Joined Apr 2010, 189

Paul Holloway

Consultant,
Holloway NDT & Engineering Inc ,
Canada,
Joined Apr 2010
189
15:21 Dec-11-2015
Re: Error inherent when using 6dB drop for length sizing (AWS/CSA)
In Reply to Charlie Jackson at 12:30 Dec-08-2015 .


6dB drop from AWS site

"...apply the basics you should have learned during training"

Charlie, this is the main problem. The methods are often taught at a bare minimum without having to bother the trainees with pesky real-life complications. Manufactured flaw specimens often contain completely artificial defects like completely uniform lack of penetration with easily identifiable starts and stops. These specimens are used for both training and examination, they're easy to teach with and they're even easier to mark. But they're not indicative of real world defects, and students leave with a false sense of security: "I nailed that lack of pen, bro! I passed first shot, I'm the best they've ever seen!".

I'm sure it's not this way all over the world, but I think many can agree that there are some basic UT principles which are poorly understood by trainees and experienced technicians alike. Everyone *should* have learned the right way during training, but most don't.

By way of example, I've attached an image from the AWS site (https://app.aws.org/itrends/2005/01/021/). What is shown is the classic 6dB drop method, basing the method on the peak (incorrect), yet its straight off the AWS website. The text which accompanies the original image references 6.31.2 in D1.1 (2000), which *does* instruct to "drop 6dB below the rating for the applicable flaw classification", but how many NDT technicians are running around out there with a copy of D1.1, or at least have read and memorized that requirement? I'm going to hazard a guess and say less than 5%. The rest of them are going to go on what they were taught in a classroom years before, and the simple, easy to mark method(s) they used to pass their qualification exams.
1    
 
 Reply 
 
charlie
charlie
17:12 Dec-11-2015
Re: Error inherent when using 6dB drop for length sizing (AWS/CSA)
In Reply to Paul Holloway at 15:21 Dec-11-2015 .

Paul

Thanks for the reply and the link and whilst I am glad I no longer have to wrestle with the AWS D11 spec I proffer the following reasoning for the AWS approach. I can only assume their intent was to size defective areas and not the extent of the flaw. By dropping 6dB below the relevant unacceptable rating they will remove the ‘bad’ bit of the flaw which can be classified as a defect. This could leave a continuation of the flaw either side but this area would be classified as acceptable once the defective part is removed. It’s an approach which can be reasoned in a production environment.
I guess my memory isnt as good these days and I forget those years spent in construction shops tangling with the various specs and 'not so good' welding.
Apologies to all for my previous over simplification of the problem

Regards

Charlie

    
 
 Reply 
 
massimo carminati
Consultant, AUT specialist
IMG Ultrasuoni Srl, Italy, Joined Apr 2007, 691

massimo carminati

Consultant, AUT specialist
IMG Ultrasuoni Srl,
Italy,
Joined Apr 2007
691
21:11 Dec-15-2015
Re: Error inherent when using 6dB drop for length sizing (AWS/CSA)
In Reply to Paul Holloway at 04:57 Dec-07-2015 (Opening).

Very interesting topic and very well detailed. I also try to teach people to "look" into the weld despite to what sizing methodology is written in the paper. In other words, to deeply investigate when very long indications with mostly low amplitude has few points with "recordable" amplitude. Real defects are not SDH and do not behave the same along its whole length.

    
 
 Reply 
 

Product Spotlight

FAAST-PA! OEM Patented phased Array for high speed UT inspection

Multiangle, Multifocus, Multifrequency, Multibeam. Instead of stacking UT electronics and having m
...
any PA probes, FAAST-PA is able to transmit all delay laws within ONE single shot in Real time.
>

YXLON Cougar EVO

Scalable small footprint X-ray inspection systems for assembly and laboratory applications. The
...
YXLON Cougar EVO series was designed to provide the "best-in- class" inspection solutions for SMT, semiconductor, and laboratory assembly applications, while maintaining a small system footprint for maximum convenience. With optimized software and hardware, these systems produce higher quality and more consistent results than other electronics inspection systems currently on the market.
>

Ultrasonic tomograph for imaging of concrete structures А1040 MIRA

Applicable for concrete inspection allowing imaging of the internal structure of objects from conc
...
rete, reinforced concrete, different stones. The operation applies pulse-echo technique at one-side access to the object. The instrument is feasible for concrete inspection for searching conduct ducts, conduits, detection of foreign inclusions, holes, honeycombing, cracks and other concrete defects.
>

Sci Aps Z-Series Portable Handheld Analysers

The world’s only handheld analyzer that measures carbon content in stainless (yes even L-grades),s
...
teels, and cast irons. Also accepted for low Si analysis for sulfidic corrosion analysis, and is widely used in the power industry for Cr analysis, for flow accelerated corrosion applications.
>

Share...
We use technical and analytics cookies to ensure that we will give you the best experience of our website - More Info
Accept
top
this is debug window