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 >
Acoustic Emission Testing of existing piping systems
Career Discussions
Job Offers
Job Seeks
Classified Ads
About NDT.net
Articles & News

YXLON Copenhagen
YXLON Copenhagen is a highly specialized, award winning Danish company with 60 years of experience in portable X-ray solutions for industrial NDT.

4100 views
00:42 Aug-03-2000

Gerry May

Consultant
Kellogg Brown & Root (Halliburton),
USA,
Joined Aug 2000
6
Acoustic Emission Testing of existing piping systems

Some companies have been promoting the use of Acoustic Emission methods to monitor for cracks in existing high temperature piping systems, particularly seam welded pipe. I have performed literature searches, but have not been able to find any reported test summaries on its effectiveness. In particular, I want some idea of the answer to the question, "If there is a crack of a certain size, what is the probability it will be detected?" I have heard a lot of anecdotal answers saying the system works and many reports from users that it does not work. Does anyone have any factual information, or know of a source I can research?


 
01:06 Aug-03-2000

Rolf Diederichs

Director, Editor, Publisher, Internet, PHP MySQL
NDT.net,
Germany,
Joined Nov 1998
602
Re: Acoustic Emission Testing of existing piping systems As far as I now it is not possible to quantitative cracks with AE.
Where AE detects flaws and locates them, Ultrasonics quantifies their size and obtains their orientation.

Rolf Diederichs

-------
: Some companies have been promoting the use of Acoustic Emission methods to monitor for cracks in existing high temperature piping systems, particularly seam welded pipe. I have performed literature searches, but have not been able to find any reported test summaries on its effectiveness. In particular, I want some idea of the answer to the question, "If there is a crack of a certain size, what is the probability it will be detected?" I have heard a lot of anecdotal answers saying the system works and many reports from users that it does not work. Does anyone have any factual information, or know of a source I can research?




 
02:10 Aug-03-2000

gerry may

Consultant
Kellogg Brown & Root (Halliburton),
USA,
Joined Aug 2000
6
Re: Acoustic Emission Testing of existing piping systems : As far as I now it is not possible to quantitative cracks with AE.
: Where AE detects flaws and locates them, Ultrasonics quantifies their size and obtains their orientation.

: Rolf Diederichs

Thank you for your quick response:

I agree that AE is only a scanning method and cannot be expected to size flaws. What I am trying to find out is if there has been any documented research that would report that "If a flaw of x mm length and y mm depth exists, AE will find it some percentage of the time." We know all NDT techniques have strengths and limitations. I am just having trouble finding out what the factual capabilities are of this technique on piping.

: -------
: : Some companies have been promoting the use of Acoustic Emission methods to monitor for cracks in existing high temperature piping systems, particularly seam welded pipe. I have performed literature searches, but have not been able to find any reported test summaries on its effectiveness. In particular, I want some idea of theanswer to the question, "If there is a crack of a certain size, what is the probability it will be detected?" I have heard a lot of anecdotal answers saying the system works and many reports from users that it does not work. Does anyone have any factual information, or know of a source I can research?




 
02:57 Aug-03-2000
Dennis Sullivan
Re: Acoustic Emission Testing of existing piping systems : Some companies have been promoting the use of Acoustic Emission methods to monitor for cracks in existing high temperature piping systems, particularly seam welded pipe. I have performed literature searches, but have not been able to find any reported test summaries on its effectiveness. In particular, I want some idea of the answer to the question, "If there is a crack of a certain size, what is the probability it will be detected?" I have heard a lot of anecdotal answers saying the system works and many reports from users that it does not work. Does anyone have any factual information, or know of a source I can research?



 
00:05 Aug-04-2000

Christian Grosse

Teacher,
Technische Universität München,
Germany,
Joined Nov 2000
11
Re: Acoustic Emission Testing of existing piping systems : As far as I now it is not possible to quantitative cracks with AE.
: Where AE detects flaws and locates them, Ultrasonics quantifies their size and obtains their orientation.

: Rolf Diederichs


Your statement is not quite correct. Talking about the so-called "quantitative AE" not only the size of a flaw but also the orientation and even the fracture mode (mode I, II, III) can be determined by using a method based on moment tensor inversion. However, this method is much more complicated than the "classical AE" and requires more sensor informations and a simple geometry (so far). According to my knowledge it is not applied to piping systems so far, but to half space materials as concrete even if there might be a possibility to extent the theory. There are no commercial AE measuring systems available to determine the size of flaws in pipelines.
For more information about the moment tensor inversion technique please refer to my web-page (see the link below -> "publications") or to the article: S.Köppel; C. Grosse: Advanced acoustic emission techniques for failure analysis in concrete. 15. World Conf. on NDT, Rom 2000, (in print).

Christian Grosse


 
01:25 Aug-04-2000

John O'Brien

Consultant, -
Chevron ETC ,
USA,
Joined Jan 2000
278
Re: Acoustic Emission Testing of existing piping systems If a crack is active you can detect it with Acoustic Emission even down to microcrcaking levels i.e sub millimeter cracks. The question is then whether you can do it in your particular indistrial application taking into account interference issues. This has got better in recent years due to improved software, sensors and computing capabilities.
As has been pointed out it is also possible to determine orientation in many cases using sophisticated approaches such as Moment Tensor anlysis. If you know the localised area of a crack and you place a correct number of sensors around it you can plot activity at the ends of the crack and therefore to some degree determine size. When you get multi branching cracks such as SCC this can become more difficult.
Of course a dormant crack cannot be detected with AE and that is where UT comes into its own. However if you have a crack that remains dormant even in an overstress situation why worry about it is many peoples philosophy, There is a significant current R&D project linking the AE results directly into FFS programs such as API RP 579 so experience and relationships improve.

There are literally thousnads of paper of AE related to cracking - its a case of finding the correct ones.

Related to piping typical are:

AE Testing Of A seamed Hot Reheat Line - J E Gillen June 1987
Boiler Acoustic Emission Crack Monitor - Sherlock/Coulter/Stevens - April 1996
Lot of papers On Cracking Of Vessels & piping - Fowler T J - 1980's MONSANTO
AE Monitoring Of Cracks In Ammonia Tanks - Matti Sarkimo - May 1989
Babcock & wilcox Guidelines For AE Testing Of Steam pipes
Monitoring Crack Growth Of Hydrogen Embittlement & SCC With AE - Gerberich/Hartbower - August 1967
AE During SCC in High Strength Steels - P McIntyre - May 1978

If you can be more specific it may be possible to direct you to exact sources.
Remeber AE can be oversold but for detecting and monitoing active cracking on a global basis to identify areas for UT sizing it works extremely well.



 
05:25 Aug-04-2000

Sotirios J. Vahaviolos

R & D, -
MISTRAS Group, Inc.,
USA,
Joined Jan 2000
15
Re: Acoustic Emission Testing of existing piping systems : Some companies have been promoting the use of Acoustic Emission methods to monitor for cracks in existing high temperature piping systems, particularly seam welded pipe. I have performed literature searches, but have not been able to find any reported test summaries on its effectiveness. In particular, I want some idea of the answer to the question, "If there is a crack of a certain size, what is the probability it will be detected?" I have heard a lot of anecdotal answers saying the system works and many reports from users that it does not work. Does anyone have any factual information, or know of a source I can research?
I have read with much interest this debate.AE is ideal
for Global Monitoring of structures.Today's advancements
in AE do not allow us for "overextension of the
technology" to beyond real time "flaw" source detection
and location.In some special cases as Dr.Grosse
rightfully pointed out,advancements in Moment Tensor
Analysis using the pioneering work of Prof.Otsu go
beyond the existing AE technology and into the
detection of Crack propagation,3D location and most
importantly AE source direction.Quantative studies in
AE have also resulted in "special applications" where
more info about the crack or say composite defects can
be determined with only the use of AE.Make no mistake
about it effective use of Wave propagation studies and
Moment Tensor Analysis can go beyond the normal use of
AE.People like myself with 30-plus years in AE are much
afraid of the "overselling" of such new pioneering
methods,and that is why John O'Brien cautioned us.In
the example of the use of Moment Tensor Analysis our
company supplies this program,but cautions the buyers
to use it in thick specimens and mostly in concrete
work,where it has been found very effective it defect
detection and location and direction of these defects.
Do the same in pipeline work will be a bit difficult for
me to accept at this time unless I see sufficient proven
data verified with other NDT methods as Mr.O'Brien
pointed out.As for me at this time I use AE for Global
Monitoring and Automated UT say for defect characterization.
Complimentary NDT,or NDT sensor fusion as some might
say,is the proper approach for effective AE use.We need
and must seperate Research work from "daily use of AE"
where numerous codes exist and AE applications are
performed by qualified and most importantly by trained
and properly certified AE Technicians to such schemes
as SNT-TC-1A,EN-473,JNDI,ROSNDT,ABENDE etc.,and thus,
continue to make AE a strong Partner of the Big-5 NDT
methods.A complete Bibliography of the Pioneering work
of Ohtsu and others in Wave Propagation,but for AE work
can be supplied to interesting parties from our company's
Library and selected papers can be send to you.Sorry
for the long answer,but I am a bit concerned that we can
give a bad name to new advancements in AE if we oversell
them and make them widely available before their time.



 
08:43 Aug-04-2000

John Rodgers/Acoustic Emission Consulting, Inc.

Consultant
Acoustic Emission Consulting, Inc.,
USA,
Joined Oct 2001
8
Re: Acoustic Emission Testing of existing piping systems : Some companies have been promoting the use of Acoustic Emission methods to monitor for cracks in existing high temperature piping systems, particularly seam welded pipe. I have performed literature searches, but have not been able to find any reported test summaries on its effectiveness. In particular, I want some idea of the answer to the question, "If there is a crack of a certain size, what is the probability it will be detected?" I have heard a lot of anecdotal answers saying the system works and many reports from users that it does not work. Does anyone have any factual information, or know of a source I can research?

Answer:
We have been involved in both R&D and field testing of fossil power high energy piping, in particular seam welded piping, for the past 10 years. EPRI published AE testing guidelines in Nov 1995 for seam-welded hot reheat piping. Since that time we have monitored over 50 piping systems during online operation, startups, and cooldowns. The issue of cracking and flaw sizing is much different than you would expect going in to this application. The process of creep degradation in seam welds typically starts from development of cavitation around inclusions (sulfides, oxides, carbides) in the grain boundaries in the fusion zone (or heat affected zone if the weld is subcritically annealed after fabrication). This may occur over many feet of weld, even entire spool pieces--but typically occurs first at high stress areas such as bends/elbows. Cavitation eventually progresses to microcracking--which may be extensive before macrocracking and eventual link up occurs. What we have determined from our testing is that a very characteristic emission pattern can be detected--and located--on seam welded piping during the process of cavitation formation---well before cracking develops. This process goes on for years, and there are multiple opportunities to monitor its progression. This has been validated by advanced UT techniques and by material sampling/metallography in several cases. We have yet to find a "macrocrack" in a seam weld that would meet the traditional definition---and the absence of significant failures (of thin-walled hot reheat piping)since the mid-80's tends to validate the low incidence of such to be found. We have identified cracking in hanger support welds, at vents and drains and closure welds--and microcracking in seam weld repairs (not generated by typical creep processes). These have tended to be multiple defects, and not accurately sized for length and depth. On the more generic issue of flaw sizing with acoustic emission, we would agree that current field instrumentation and practice is not capable of either accurately sizing flaws, or yielding a minimum detectable flaw size. The reasons for this are complex and would need much more explanation than space/time available here. AE is a very good general survey tool, and can be used effectively for prioritizing follow-up inspections with conventional NDE methods. It has certainly proven so in the case of fossilseam-welded piping. A summary report of our testing experience is being prepared for EPRI. You may wish to enquire further by contacting Rich Tilley (rtilley@epri.com).


 
07:48 Aug-05-2000

G:NARDONI

Director
I&T Nardoni Institute,
Italy,
Joined Dec 1999
4
Re: Acoustic Emission Testing of existing piping systems Dear Delhaza
AE test .
AE is an excellent method to detect the dynamic behaviour of discontinuity
under stress conditions . The detection defect is based on the amount of
acoustic waves reliesed at the defect countur . We have used it for many
yars and very good result have been achieved . A global map on active
sourcing of defects may be produced and not in all cases this mup correspond
to the static map ddetected by UT , RT , MT .
In your case , to evaluate cracks on serviced components you have to
pressurized over at least 10% of the pressure
test made during the hydraulic test . This due to the Kaiser effect .


Best
regards


G:NARDONI


---------------
: Some companies have been promoting the use of Acoustic Emission methods to monitor for cracks in existing high temperature piping systems, particularly seam welded pipe. I have performed literature searches, but have not been ableto find any reported test summaries on its effectiveness. In particular, I want some idea of the answer to the question, "If there is a crack of a certain size, what is the probability it will be detected?" I have heard a lot of anecdotal answers saying the system works and many reports from users that it does not work. Does anyone have any factual information, or know of a source I can research?




 
05:49 Aug-08-2000
Claudio Allevato - Stress Engineering Services - Houston-Tx
Re: Acoustic Emission Testing of existing piping systems Gerry,

To my knowledge, the answer to your question is NO. I endorse and agree with the comments of Mr. Vahavioulos, O'Brian, Rodgers, and others above. There's been a number of efforts in the past trying to determine a relationship between crack-like defect dimensions and orientation, and their detectability by AET. Some were based on J-Integral principles, others on pure experimental work in the laboratory. I would say one could come up with a general curve correlating Stress Intensity Factors and AE behaviour, thus predicting the crack's AE behaviour under specific conditions, but it would be specific to a certain stress/load situation. I wouldn't adopt it to every case.

It would be too pessimistic to assume that cracks under a certain size/orientation are not being detected by AET, when they might actually be due to oxide cracking inside the crack, or fritting of its surface. At the same time it would too risky to assume cracks above a certain size/orientation will be detected, when they might not be, due to operator error, equipment malfunction, etc.

It is generally accepted that AET should be used very carefully, and complemented with an array of other analysis methods. As others said, it is a "screening method" and should be used as such. I can give you actual results from AET inspections on P11, P22, SS, materials where cracks as small as 1/4" long, all the way to 9' long were detected, confirmed and repaired.

Thanks and good luck.



 
02:14 Aug-08-2000
Richard Nordstrom
Re: Acoustic Emission Testing of existing piping systems Probability of Detection (POD) is great answer to have but it is not an easy answer to obtain, for any NDT method but especially difficult in Acoustic Emission. It is also very dependent on the details of the application, more on those details for seam welded pipes in HRH below. POD should not be confused with crack sizing accuracy. Being able to size a crack (in any NDT method) can occur only after it has been detected. In the case of AE, methods to size cracks often actually reduce POD because they force the investigator to focus on a limited amount of the AE which retains a high quality of information, usually within a waveform. This means filtering for specific waveforms which precludes any indication of a crack in favor of waveforms which retain the minimum amount of information for crack sizing and orientation (i.e., a significantly smaller part of the AE than the whole which predicts a crack). In any case, the highest POD that I have seen on Hot reheat line is achieved with online monitoring that requires waveguides. The dispersion of AE signals through waveguides precludes the crack sizing and orientation analysis that has so far been developed but does not prevent the measurement and evaluation now being applied to seam welded HRH lines.

A recent short description of what is meant by POD can be found in Materials Evaluation (April 2000, page 503). The useful separation of factors which reduce POD described in this articles are those which are operator oriented (human factors or HF) and those which are application oriented (application parameters or AP). Usually AE is automated, which would tend to reduce the HF, however the choice of procedure and evaluation can become large HF in AE testing.

POD in UT can be quite high in ideal cases. Certification requires 80% proficiency of the operators and some estimates (gueses) put procedural aspects of POD at 90% average.

For UT inspection of seam welded HRH pipes (which is easier to address than AE), one can point out several places where POD is limited as an example. First, UT cannot access all points on the seam welded pipe, especially under hangers. This reduces the POD of the method. Second orientation of the defect/crack effects how early it can be detected. Several initiators are known for defects in HRH and procedures were devised for the most likely crack/crack initiators but not all crack initiators. The procedures were modified after additional HRH failures, which improved overall POD from UT but it still was not comprehensive.

For establishing POD in Acoustic Emission it is much more gray. The general approach in highly successful AE applications is statistics. In pressure vessels tested under accepted procedures (ASME V, CARP, etc.) feedback indicates very high POD rates when fixed procedures were used. Operator influence is reduced because the AE system is automated (no fatigue from scanning), the most optimistic proponents will state their procedures are 100%, however the procedures have needed revision with use so it would be more honest and conservative to state that for these systems are probably in the 95-99% regime, still very high. The experience for these tests have been on the order of tens of thousands of tests.

For seam welded pipe testing in the HRH of power plants, there are limitations to be considered that are not faced in the chemical industry pressure vessel example. First, there are no common procedures that form a floor for AE testing in the industry. While there are certainly groups which perform higher quality AE procedures, there are no guarantees that the worst that perform the tests under the name "AE" don't have POD that approach zero (or at least very low levels). Some of the choices which differ among AE practitioners on HRH testing are substitution of line loading, length of testing, required load variation within testing and evaluation procedure. This list is far from complete.

Second, there is no larger data base that indicates that the best procedures can approach the high POD found in thechemical industry example. John Rodgers has pointed out that he has tested 50 HRH lines. This may be the highest number by a single group (I will leave it to the others to identify themselves), the total of all tests performed at any quality will be in the range of no more than several hundred at best. Many of these do not serve to help establish POD or improve on it. For one, HRH tests have evolved and were not performed with optimal experience on the method. Secondly, other tests were performed as inspections and have limited impact on the overall statistics for seam welded pipes. EPRI did destructively fail at least one HRH seam welded pipe (outside of an actual plant) with AE monitoring.

There is no perfect solution to HRH inspection, but AE is a viable NDT option that should be considered. I know it would be preferred to have a black and white answer to the question of POD, but I feel that this is a fair assessment of the state of affairs of HRH AE inspection, trying to stay away from hedging AE's capabilities or exaggerating them. Having said that, there are many parties which do not freely distribute information on this matter. EPRI has many internal studies with details that are not openly available (which I am not aware of) and the major AE contractors on this application keep their information to themselves for competitive reasons.

: Some companies have been promoting the use of Acoustic Emission methods to monitor for cracks in existing high temperature piping systems, particularly seam welded pipe. I have performed literature searches, but have not been able to find any reported test summaries on its effectiveness. In particular, I want some idea of the answer to the question, "If there is a crack of a certain size, what is the probability it will be detected?" I have heard a lot of anecdotal answers saying the system works and many reports from users that it does not work. Does anyone have any factual information, or know of a source I can research?




 
03:37 Aug-10-2000
Harold Dunegan
Re: Acoustic Emission Testing of existing piping systems It is impossible with AE to detect or size a static
crack. The crack must be growing with a subsequent
release of strain energy to be detected by AE(with
the exception of oxide crushing during compression).
For a given amount of new crack area, the amplitude
of the AE signal produced and therefore its detecta-
bility is dependent on the value of the stress intensity
factor at the crack tip, the path which the stress
wave must traverse to reach the transducer, and the
background noise present within the bandpass of the
instrumentation. Under ideal conditions (laboratory
specimens) we have detected crack growth in fatigue
of 1 microinch per cycle.

Regarding sizing of cracks. Techniques have been
developed to estimate the depth of a growing crack.
An example of how this works can be found on our web
page under DECI reports which describes detecting
crack depth in a CT fracture toughness specimen undergoing
hydrogen embrittlement cracking.



 


© 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|