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 -
3053 views
Technical Discussions
Jeff Gibson
Jeff Gibson
12:12 Jan-14-2013
Uncertainty calculation method for UT (and related)

Hi there,

I am looking for a document, a comprehensive example or standard for calculating the uncertainty in UT measurement owing to factors such as (but not limited to):
1) Temperature effects changes in speed of sound
2) (non-)parallelism, curvature and surface profile (internal and external)
3) roughness (or scaling)
.....
I have BS EN 14127:2001 wherein Table C.1 in Annex C gies the influencing parameters, but does not go as far as to give some illustrative examples.
I am also interested in any useful references / expericiences on EMAT.

Are there any other useful standards/ documents that folks would be willing to share with me?

My background is in oil and gas and NDT is fairly new to me, but I have experience with uncertainty calculations. My interest is in a sensor that can track changes in wall thickness due to erosion or scaling. Absolute wall thickness is of less importance as getting the change in wall thickness right.

Thanks in advance,
Jeff.

 
 Reply 
 
creed
NDT Inspector,
n.n., Joined Jul 2009, 33

creed

NDT Inspector,
n.n.,
Joined Jul 2009
33
01:15 Jan-15-2013
Re: Uncertainty calculation method for UT (and related)
In Reply to Jeff Gibson at 12:12 Jan-14-2013 (Opening).

Hi Jeff,
Please find ASME section V-2007, Article 23, SE797 for UT measurement.

 
 Reply 
 
John
John
01:46 Jan-15-2013
Re: Uncertainty calculation method for UT (and related)
In Reply to Jeff Gibson at 12:12 Jan-14-2013 (Opening).

Calibrations are used to eliminate the uncertainties of measurements. These are in most codes and procedures
1: cal block temp should be the same as the test piece
2: curvature corrections can be used and are usually std on most equipment
3: direct comparison can be used for to compensate difference in cal to test piece surface (transfer loss)

& see docs listed by "CREED"

 
 Reply 
 
Ed Ginzel
R & D, -
Materials Research Institute, Canada, Joined Nov 1998, 1286

Ed Ginzel

R & D, -
Materials Research Institute,
Canada,
Joined Nov 1998
1286
00:11 Jan-16-2013
Re: Uncertainty calculation method for UT (and related)
In Reply to Jeff Gibson at 12:12 Jan-14-2013 (Opening).

Jeff, generally you would attempt to minimise uncertainties, but they are inevitable. Most UT instruments are digital so digitisation rate of the timebase is one factor (typically +/- 1 digitisation sample could be used). If you are making measurements over a range of temperatures you would need a dV/dT and this would suggest you are interested in an absolute velocity. Then you would need to have a smooth parallel sample of the component tested to determine its velocity (and monitor its change over the temperature range. Couplant gap is another variable. If scale present at the point of measurement there could be a problem if its bond condition changes. Of some interest will be the potential for corrosion at either surface. Small irregularities could affect the assessed arrival time. The point on the signal (e.g. positive or negative peak) that you measure can cause variation and if pitting has formed between measurements the shape of the pulse could alter where this point forms.
If the sensor is not placed permanently in one fixed position you can introduce another uncertainty.
Each item identified will provide a thickness error tolerance. It would be unreasonable to simply add the individual error tolerances to derive uncertainty so perhaps the RMS error or Standard Deviation of all the values would be used.

 
 Reply 
 
Jeff Gibson
Jeff Gibson
17:31 Feb-18-2013
Re: Uncertainty calculation method for UT (and related)
In Reply to Ed Ginzel at 00:11 Jan-16-2013 .

Dear Ed,
Many thanks for your reply to me in January; I am only getting back to you now, but better late than never!

The material thickness may be, say 15 mm thick or so (that is to be decided), but it is certainly this ballpark. Its likely we have 4degC on the outside of the pipe (its subsea!) and the higher temperatures of up to 300 degC or so on the inside (in extreme cases).

I guess dV/dT is to allow a correction with variable temperatures, but its also dawned on me that the there will be a varaition in temperature through the metal itself due to conduction and convection on both inside and out (i.e. T = f(thickness), which will not be linear) and, therefore, there will be a variaiton in acoustic velocity with thickness also.

I would assume that its not a simple matter to correct the acoustic velocity simply on a=f(T), but also that the gradient dV/dT might also affect things. If its ballaprk figures then a stab at calculating the effect would be to calculate teh acoustic velocity based on the average temperature through the thickness of the material.

Any feedback much appreciated. Finally, what do you mean by "absolute" velocity?

Regards,
Jeff.

 
 Reply 
 
Ed Ginzel
R & D, -
Materials Research Institute, Canada, Joined Nov 1998, 1286

Ed Ginzel

R & D, -
Materials Research Institute,
Canada,
Joined Nov 1998
1286
20:22 Feb-21-2013
Re: Uncertainty calculation method for UT (and related)
In Reply to Jeff Gibson at 17:31 Feb-18-2013 .

Jeff, I would direct you to another forum thread http://www.ndt.net/forum/thread.php?&rootID=44132
Gerhard Splitt notes that for all practical purposes the difference in steel is negligible. e.g. a 20 mm thick pipeline wall, we will get a correction of 0.2 mm over a 100°C range. Typically the difference is 1% /100°C.
By absolute, I merely meant that you wanted a final velocity at a fixed and known temperature, not a measured difference (e.g. you would probably not take readings at a starting and end temperature).

2
 
 Reply 
 

Product Spotlight

Research and Applications Development For NDT

The Research and Applications Development (RAD) group is a newly formed team within Acuren dedicat
...
ed to tackling challenging inspection problems. Our focus is the development of novel, field deployable, advanced inspection techniques for use in cases where standard NDT methods are ineffective. We don't wait for new innovations, we engineer them. From concept to commissioning.
>

GEKKO - Portable Phased Array Testing with TFM in Real-Time

The portable phased array testing system GEKKO provides 64 parallel test channels. On creating testi
...
ng parameters the operator is assisted by the CIVA software. Due to its modular set-up the GEKKO instrument is suitable for operators of all skill levels.
>

ISAFE3 Intrinsically Safe Sensor System

ISAFE3 intrinsically safe sensor system of Vallen Systeme is especially targeted at the petrochemica
...
l - as well as oil and gas transportation industry. The sensor system is designed for permanent monitoring or periodic inspection tasks. Sensors are available for different AE-frequency ranges optimized for corrosion and fatigue crack detection and other applications. The ISAFE 3 sensor system consists of an AE-sensor (model ISAS3) certified according to ATEX/IEC for installation in zone 0, gas group IIC, IP68, 20 to +60 °C, and a signal isolator (model SISO3) certified for installation in zone 2. An ISAS3 sensor can be mounted in atmosphere or submerged, e.g. in water or crude oil. It is supported by mounting tools for temporary (magnets) or permanent (welded) installation. ISAFE3 supports automatic sensor coupling test and can be used with any AE signal processor supporting 28V supply at 90 mA peak, e.g. Vallen Systeme ASIP-2/A.
>

Navic - Steerable Modular Automated Scanner

The Navic is a modular, motorized, steerable scanner designed to carry multiple attachments used
...
in various scanning and inspection applications. The Navic is capable of weld scanning (girth welds and long seam welds), automated corrosion mapping, and tank scanning.
>

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