09:18 Jun-12-2008 Philippe Rubbers Engineering SCM, South Africa, Joined Nov 1998 21
Ultrasonic A-Scan interpretation
Dear colleagues
There are 2 approaches for setting the range on an A-Scan: 1) at the position of the peak amplitude. 2) at 10% of the peak amplitude on the leading edge (flank).
Historically (for very narrow bandwidth probes) with long ring down times, this 10% story makes a bit of sence, as the peak is not always evident. However in this day and age, this is no longer an issue, and the peak is the correct mathematical location of the reflector (can be proven with broadband probes and CSSP, as all frequencies are in phase at the peak). Since beam skew will give longer measured ranges, an error in the shorter direction is also beneficial (statistically) if the operator does not have the correct wedge angle or does not maximise as he should. If the gain is then changed, the range for the '10%' setting will change as the leading edge will come closer.
Now comes the question: In phased array images, the overlay is nearly always placed at the 10% location. To me this is incorrect... Why isthis done?
00:09 Jun-12-2008 Ed Ginzel R & D, - Materials Research Institute, Canada, Joined Nov 1998 1208
Re: Ultrasonic A-Scan interpretation Philippe: Good question. I suspect that the assessment of phased array pulses is a carry-over from the tradition of manual single element A-scan assessments. It might be imagined that the "leading edge" of the pulse is the first point of contact and this rationalised the use of "flank". But some of the blame must be put on our trend (over the past 40-50 years) for a preference for rectified A-scans. I suspect that if we were to revert to the real data (unrectified waveforms) we would be more inclined to look for the first well defined peak. On a broadband pulse this will be easier to locate than on a "ringy" narrowband pulse. In fact we seem to have rediscovered the advantage of unrectified waveforms and most of us use a peak for TOFD assessments (phased array generated or single element TOFD is treated the same). Cheers, Ed
----------- Start Original Message ----------- : Dear colleagues : There are 2 approaches for setting the range on an A-Scan: : 1) at the position of the peakamplitude. : 2) at 10% of the peak amplitude on the leading edge (flank). : Historically (for very narrow bandwidth probes) with long ring down times, this 10% story makes a bit of sence, as the peak is not always evident. However in this day and age, this is no longer an issue, and the peak is the correct mathematical location of the reflector (can be proven with broadband probes and CSSP, as all frequencies are in phase at the peak). Since beam skew will give longer measured ranges, an error in the shorter direction is also beneficial (statistically) if the operator does not have the correct wedge angle or does not maximise as he should. : If the gain is then changed, the range for the '10%' setting will change as the leading edge will come closer. : Now comes the question: : In phased array images, the overlay is nearly always placed at the 10% location. To me this is incorrect... : Why is this done? : Best regards ------------ End Original Message ------------
01:10 Jun-12-2008 mustafa GONULAL Engineering, CIMTAS / ENKA, Turkey, Joined Jun 2008 8
Re: Ultrasonic A-Scan interpretation ----------- Start Original Message ----------- : Philippe: : Good question. I suspect that the assessment of phased array pulses is a carry-over from the tradition of manual single element A-scan assessments. It might be imagined that the "leading edge" of the pulse is the first point of contact and this rationalised the use of "flank". But some of the blame must be put on our trend (over the past 40-50 years) for a preference for rectified A-scans. I suspect that if we were to revert to the real data (unrectified waveforms) we would be more inclined to look for the first well defined peak. On a broadband pulse this will be easier to locate than on a "ringy" narrowband pulse. In fact we seem to have rediscovered the advantage of unrectified waveforms and most of us use a peak for TOFD assessments (phased array generated or single element TOFD is treated the same). : Cheers, Ed : : : Dear colleagues : : There are 2 approaches for setting the range on an A-Scan: : : 1) at the position of the peak amplitude. : : 2) at 10% of the peak amplitude on the leading edge (flank). : : Historically (for very narrow bandwidth probes) with long ring down times, this 10% story makes a bit of sence, as the peak is not always evident. However in this day and age, this is no longer an issue, and the peak is the correct mathematical location of the reflector (can be proven with broadband probes and CSSP, as all frequencies are in phase at the peak). Since beam skew will give longer measured ranges, an error in the shorter direction is also beneficial (statistically) if the operator does not have the correct wedge angle or does not maximise as he should. : : If the gain is then changed, the range for the '10%' setting will change as the leading edge will come closer. : : Now comes the question: : : In phased array images, the overlay is nearly always placed at the 10% location. To me this is incorrect... : : Why is this done? : : Best regards ------------ End Original Message ------------
First of all I would like to thank you for such a discussion, My concern is mostly for sizing of defects in Phased Array. For planar defects like crak, interpreting the Phased Array tomographic view resembles to the TOFD image and you can use tip diffraction signals. However for other type of defects standards saying 6dB drop method for flaw sizing. For Phased Array my opinion is different because, you are using a fan of beams coming in different angles with different metal paths. So using 6dB method is too old fashion.
872 views
21
