A Division of InspecTech Analygas Group Inc.
450 Midwest Road, Toronto, Ontario, Canada, M1P 3A9
Tel: 416-757-1179      Fax: 416-757-8096
Email:    Web:



Ultrasonic weld testing in ERW tube mills has been common practice for over 40 years. Product is tested for defects and flash condition both on the mill (on-line) and on the finishing floor (off-line). Over the years ultrasonic systems have improved in performance, and to-day’s P.C. based systems provide unambiguous results with simple set-ups and include data storage capability.

The ultrasonic method is ideal for testing ERW tube, as signal to noise ratios are excellent (very few false indications) and sensitivity is unaffected by wall thickness (ID and OD defects are equally discernable). Unfortunately, although the method is ideal, the application is imperfect.

The imperfection is because the transducers must operate in the exact relative position to the weld, as they were when the system was calibrated. Lack of precision in the transfer from the calibration standard and the natural “wander” of the weld line on the mill or the conveyor both conspire to make this condition almost impossible to meet.

The signal from a test notch can fall by 30dB (50:1) and disappear totally with just 3mm of weld line shift. This problem is well known in the welded tube industry and considerable effort has been made to find a solution over the years.

The “Involute Search Unit” was an early attempt to compensate for weld line wander. It was a multi-crystal compound transducer, which was unfortunately costly and inefficient.

Other methods have included the use of multiple transducers, which is very demanding on set-up; and even by the use of phased arrays. These latter systems can produce sequential scans at varying angles in the weld zone, but at a repetition rate limited by the material acoustics. The possibility is open for short defects to pass undetected when the sound beam is focussed elsewhere in the weld zone.

InspecTech has long used special high-aspect-ratio transducers, which focus the ultrasonic energy to maximize the beam width through the weld zone, while at the same time minimize the beam width along the weld length. These transducers, when used at the maximum possible repetition rate, have provided good weld wander compensation coupled with excellent discrimination for small defects. Penetrators 1mm and smaller in length are routinely found using this transducer technology

Nevertheless high aspect ratio transducers are not totally immune to weld line shift, especially in heavy wall products.

InspecTech is pleased to announce a totally effective solution to this vexing problem in the form of monoblock array transducers which require no independent adjustment and therefore easier to use than conventional systems.

A series of high-aspect-ratio transducers are mounted as a discrete array, with a pre-calculated offset which gives just a few dB variation in response over a very broad range of positions. Unlike in phased arrays, the transducers are “fired” in unison and at a prf of several KHz, which ensures total coverage of the weld zone and the heat affected zone over a broad range of weld line positions. Figures 1, 2, 3 & 4 illustrate the system’s capability to capture and display flaw indications within a few dB over a distance range of ± 1” (25mm). The system will maintain the ± 1” tested by ultrasonic methods (usually 3” to 24” diameter).

Figure 5 shows the new Array Transducer Assembly for 6-5/8” pipe weld.

Computer-aided processing of the ultrasonic data gives the user a conventional A-scan presentation, a familiar format for most operators.

In order to improve noise rejection, it is common practice with ultrasonic test systems to “countdown” a series of defect indications prior to showing the alarm status for a flaw. Since the array system consists of discrete transducers all working in unison, the countdown approach remains perfectly valid as a means of enhancing performance on line.

The attached Table 1 shows how the countdown settings used can impact the length of defect, which is theoretically detectable. In the area of the table which contains no numbers, the countdown setting has no influence on defect detectability whatsoever, it becomes an issue of defect size and orientation only. Therefore, the equipment operator can use any countdown setting within the blank area with confidence. Table 1 is based upon system clock rates close to the maximum useable for this type of testing without the possibility of wraparound signals.

The discrete array therefore offers the user 3 brand-new features, which represent important advances in welded, tube ultrasonic testing:

Confidence in the test even with an imprecise weld location


Detectability of very small flaws


Very easy set-up and calibration procedures.

Fig. 1, 2, 3 & 4.

Response To Standard Defects As Measured Over

A Range Of Transducer Positions



Figure 5 Array Transducer Assembly for 6-5/8” Pipe Weld