Procedure Qualification for Ultrasonic Testing of Austenitic Welds
A procedure for ultrasonic testing of austenitic stainless steel welds is being qualified. The weld is 25 mm thick, and the procedure specifies a 5 MHz, 70-degree shear wave probe. The qualification test block is made of the same material and contains electro-discharge machined (EDM) slots and side-drilled holes. The procedure is being reviewed for compliance with ASME Section V, Article 4. The reviewer notes that the procedure does not address the anisotropic and coarse grain structure of austenitic welds, which can cause beam skewing and attenuation. The procedure also does not specify a reference reflector for sensitivity calibration. The qualification test results show that the signal-to-noise ratio for the EDM slots is 6 dB, which is marginal. The reviewer must decide if the procedure is acceptable or requires revision.
1. Is the procedure likely to be acceptable as written? Why or why not?
Read the worked answer
No. The procedure does not address the specific challenges of austenitic welds, such as grain scattering and beam skewing, which can lead to missed defects or false calls. ASME Section V requires that procedures for austenitic welds consider these factors, often by using lower frequency, dual-element, or focused probes, and by qualification on test blocks with realistic defects. The lack of a specified reference reflector for sensitivity calibration is a deficiency. The marginal signal-to-noise ratio of 6 dB is generally considered too low for reliable detection; a minimum of 12 dB is often recommended. Therefore, the procedure requires revision to include these elements and improve SNR.
2. What revisions would you recommend to improve the procedure?
Read the worked answer
Recommendations: 1) Use a lower frequency probe (e.g., 2.25 MHz) to reduce scattering; 2) Consider a dual-element or focused probe to improve signal-to-noise; 3) Specify a reference reflector, such as a side-drilled hole or notch, for sensitivity calibration; 4) Include a qualification block with realistic defects (e.g., lack of fusion, cracks) rather than only EDM slots; 5) Address beam skewing by scanning from multiple directions; 6) Increase the signal-to-noise ratio requirement to at least 12 dB. These changes would make the procedure more robust for austenitic welds.