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 >
Basic rules for ultrasonic weld inspection of butt welds
Career Discussions
Job Offers
Job Seeks
Classified Ads
About NDT.net
Articles & News

Tecnatom has more than 50 years of experience providing services, equipment and systems devoted to NDT testing in several sectors and markets.

01:55 Feb-18-2010
David Bunch
Basic rules for ultrasonic weld inspection of butt welds

Please feel free to add, criticize, or correct my method of ultrasonic examination. I've been working in the NDT field for two years, performing angle beam inspection on my own for approximately one. I'm trying to get a list necessary fundamental rules to follow for ultrasonic inspection of butt welds, not only for myself, but to help others at my small lab who want to learn.

In no particular order:
1- Identify the type of weld joint to be inspected (single-vee, double-vee, compound)

2-Take thickness readings of the base metal on each side of the weld.

3-If doing a weld profile, take thickness readings on weld cap and adjacent to weld, and using a contour gauge, draw the weld on paper.

4- Using the formula 2T X Tan B (b=angle used), position the probe far enough back for second leg inspection

5- Scan with a 45, 60, and 70 degree wedge. If flaws are found, mark the area on the weld. Investigate the flaw with the wedge that produces the highest signal.

6- When an indcation is found, examine it from both sides of the weld. I know that lack of sidewall fusion shows up well in the first leg (when room permits) from the opposite side of the weld, and shows up well in the second leg from the same side of weld. Lack of root penetration will have different surface distances on each side of the weld.

7-Examine a nearby area of weld to see if the signal disappears. I know this goes on the assumption that the majority of the weld is good, but then again, it usuallly is.

8-Going by the specifications, determine whether the signal is natural weld geometry (weld root, weld cap), spurious signals (mode conversion by damping the weld cap, surface waves by placing a finger in front of the transducer), an acceptable discontinuity, or a rejectable flaw.

9-When necessary, plot the flaw on the weld profile by using the 3-angle system.

10- Be aware that flaws smaller than beam-spread will plot the beam spread instead of the actual flaw size. Use satellite pips or signals that travel together to more accurately size through-wall thickness.

03:12 Feb-18-2010

Michel Couture

NDT Inspector,
Joined Sep 2006
Re: Basic rules for ultrasonic weld inspection of butt welds In Reply to David Bunch at 01:55 Feb-18-2010 (Opening).


One thing you seem to have forgoten is to do your lamination check in the HAZ and base metal on both side of the weld (if you have access).

The SW angle is dependent on material thickness. I'm speaking here specifically regarding CSA W59 or AWS D1.1.

Both side of the weld again if you have access should be examine before drawing you conclusion.

Now a word of caution regarding # 7: The lack of indications and defects in a weld is directly relevant to the dexterity of the person laying the weld bead. I've met both welders and rod burners. Never assume that a weld will usually be good.


09:30 Feb-18-2010


NDT Inspector
Joined Aug 2009
Re: Basic rules for ultrasonic weld inspection of butt welds In Reply to Michel Couture at 03:12 Feb-18-2010 .

A seperate root scan with the most appropriate probe should also be done. I think you have simplified it by just saying scan with 45, 60 & 70 as there is exact scanning methods that should be followed depending on configuration.

Also pre/post-calibration using correct block, range and other variables should be done.

09:41 Feb-19-2010
Jon Wallis
Re: Basic rules for ultrasonic weld inspection of butt welds In Reply to anjafo79 at 09:30 Feb-18-2010 .

As the last contributor mentioned, a separate root scan should always be carried out. I am assuming that the welds in question are not accessible to the root side.
A root scan should be performed by choosing an appropriate angle of probe (search unit) bearing in mind the wall thickness and not forgetting corner effect especially with the 60 degree probe. You place the probe at the appropriate distance from the weld and find the root echo, then you measure carefully and work out where the signal is coming from (expected root, lack of root fusion edge etc.). Assuming the root signal IS the root signal, you scan NOT backwards and forwards, but carefully and slowly staying parallel to the weld cap. In the way you can follow the weld and you will learn to recognise in time the difference between a goor root and a bad one.

Having said this, there are many pitfalls - beware! I mention here 2 of them

Hi-lo (misalignment) should be measured when possible (visible from outside, wall thickness difference etc.)
Shallow roots as are often present with the popular GTAW (tungsten arc weld) can present problems.

All in all, weld root examination is not easy and is one of the aspect of ultrasonic examination that has more to do with 'Fingerspitzengefühl' than simple probe operation. If at all possible, look at the root visually. In the case of a pipe, climb inside if possible with a torch and compare the root profile with what you saw on the screen. Build up as much background knowledge as you can - it helps.

19:30 Feb-19-2010

Carlos Correia

R & D, - -
Joined Oct 2008
Re: Basic rules for ultrasonic weld inspection of butt welds In Reply to Jon Wallis at 09:41 Feb-19-2010 .

Hello Brothers. Very interesting discussion.
As Jon Wallis do, we perform a scan at a fixed distance from the weld center line, but using a high angle, in my case 80 or 85 degrees. This approach has shown be good to detect root problems, as wagon tracks (slag) or inadequate penetration.
This geometry approximation could be helpful when high risk of cracking is suspected after post weld heat treatment in Cr/Mo welds.
Using a high angle has shown an interesting practice since the root reflection could not interfere (disappear or is strongly minimized) and only ID penetrations or discontinuities could be detected.
Phased Arrays are a good way to achieve a high angle without the need of built or to buy a high angle wedge. Grating lobes could interfere !!!!


© 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|