[ Follow Ups ] [ Post Followup ] [ Help ]
Re: Can TOFD replace conventional UT on welding?
Posted by: Ed Ginzel , E-mail: eginzel@mri.on.ca, on May 21, 1997 at 17:54:42:In Reply to: Can TOFD replace conventional UT on welding? posted by : Rolf Diederichs on May 20, 1997 at 16:55:43:
: Many advantages of the TOFD (Time of Flight Diffraction) method
: are reported in literature.
: Since the method works fast it can be an economically alternative to pulse echo UT or X-ray.
: What are the draw backs especially for welding testing?
: Is the method just a useful addition?
: Is there already a standard established?
: Rolf Diederichs
: References:
: APPLICATION OF MECHANIZED ULTRASONIC INSPECTION TO MANUALLY WELDED PIPELINE GIRTH WELDS
: http://www.ultrasonic.de/article/wsho0597/ginzel3/ginzel3.htm
: Ultrasonic testing and image processing for in-progress weld inspection
: http://www.ultrasonic.de/article/shaun/shaun.htm
TOFD has been proposed as a viable option to Pulse-echo methods by some practitioners. However, it suffers from several shortcomings that can limit its effectiveness in some situations.
The lateral wave is usually used as a reference indicating the entry surface. It has the typical ringdown associated with pulse length at any interface. This ring time is a function of several factors including probe damping, element nominal frequency and angle of incidence. Near surface defects are either not detected or not possible to size due to the dead zone formed by the lateral wave. Advocates of the method point out that a disruption of the lateral wave can be seen where near surface nonfusion and undercut are present, however, the advantage of precise depth assessment associated with TOFD advantages is not seen here.
In a situation where a high-low (poor fitup or mismatch) exists between two welded pieces, the transmitted pulse forms two signals off the opposite wall, one from the higher side arriving prior to the one from the lower. The amount of high-low can be accurately assessed but if a defect occurred in the lower of the two sides the backwall signal from the higher side could obscure the defect.
In both cases a pulse-echo technique can be adapted to inspect these areas to provide improved coverage where TOFD is not ideal.
In the specific application described for pipeline girthweld inspections the problems are compounded by the relatively thin wall used. Typically 9mm to 16mm wall thicknesses are used in North American. The lateral wave duration for the 4 MHz probe used by one of the service providers is about 5mm. On a 9mm wall over 50% of the metal is not ideally inspected using TOFD. On a 16mm wall the situation improves but it is not totally corrected. I have managed to reduce this dead zone to about 2mm using a 15 MHz highly damped probe but grain structure starts to be detected. It is possible to apply digital signal processing to effectively cancel the lateral wave by adding a phase-reversed equivalent signal to the lateral wave signal. This is an attempt to look under the entry "noise" and see near surface signals. This is done off line and would therefore require more than doubling the evaluation time.
TOFD is more than just a useful addition. Because TOFD is not as sensitive to defect orientation it is a necessity to assess irregular volumetric flaws more common to manual welding. Specifications normally require assessment of indications based on amplitude which, although reasonable for planar flaws of a fixed orientations, suffers for non ideal reflection surfaces. Until now, pipeline inspections were satisfied with the pulse-echo techniques as the welding process was a mechanized Gas Metal Arc Welding process where primarily planar nonfusion was the concern.
Standards and proposed standards for TOFD exist but these are not referenced in the pipeline industry. For example, British Standard BS 7706 is a Guide for using TOFD and prEN583-6. There is a draft for Acceptance Criteria for TOFD based on workmanship and is mentioned in Insight (April 1997 vol. 39 #4) by F.Dijkstra et al.
Image of situation where TOFD is not ideal