Sizing of flaws using both penetrant and ultrasonic methods

I have recently been involved in casual discussions relating to the sizing of flaws using both the penetrant and ultrasonic methods. In both cases, it must be stressed that the conversations were not referring to published standards or company procedures but were between ISO 9712 Level 2s and Level 3s, so please do not take the following as a definite statement.

For penetrant testing, the debate was as to whether the measurement should be of the total dimensions of the bleed-out at the end of the bleed-out period specified in the company’s procedures and instructions, or whether it is necessary to wipe off the indication and then measure the visible flaw. There were concerns that if the penetrant were to be removed, a very tight indication may not be visible and would be undersized. Where a flaw had a significant amount of volume, particularly subsurface, the bleed-out would be more significant than the surface profile would indicate, which is both good, warning that there is probably a significant subsurface volume, and bad, as the surface dimensions after wiping off the penetrant and developer would not be that significant. It may be appropriate to record both the bleed-out and the wipe-off dimensions, but that then relies on the recipient of the data understanding the significance of what is being presented; a couple of photographs may also help. Different industry sectors may also adopt different approaches, with the aerospace sector using a wipe-off approach more readily, but large casting manufacturers being less critical on the sizing and more likely to have greater bleed-out from cavities. One approach would be to discuss the situations with the company Level 3 if the work instructions are ambiguous.

The second related to the sizing of an indication found using ultrasonic testing. Extensive work has been carried out on the cross-sizing and associated measurements of both the top and bottom depths of flaws; however, the length has not been as carefully considered. One of the parameters that needs to be considered is the potential tapering away of the flaw at either end. If a reference level is being used, does the defect length go to the point where the signal height is below said level or does it travel to the point where there is no evidence of a response? Many defects, such as slag, will taper off, giving a reducing signal response. Beam spread can be used for cross-sectional sizing and, with phased array, the different beam angles help with time-of-flight diffraction, again providing more accuracy. How many testers have generated the transducer’s beam spread at 90° to the usual? Not many, if any, as it is not typically requested. The ultrasonic phased array is also not set up for longitudinal measurements. If there is a crack, it is more important to understand the full defect length so an appropriate repair procedure can be used. Putting heat into metal can encourage it to propagate and not removing all of the crack will also cause problems. Again, you need to refer to the company’s procedures and discuss what is required with the company Level 3.