|Rebecca Wuellner |
Lateral resolution of TOFD method
I'm a graduate student at the University of Wisconsin-Madison studying the ultrasonic time of flight diffraction (TOFD) method. I would like to determine the lateral resolution of the setup I am using. Essentially what I would like to be able to calculate is the required difference in length of two ultrasonic paths of travel so that both signals will be detected in a single scan. As an example, assume inspection of a flat plate. How do I calculate that smallest thickness of plate that will produce a waveform containing both the near surface wave and backwall reflection? Any referces to literature about determining the lateral resolution of TOFD would be helpful.
|Steven Johnson |
USA, Joined Mar 2004, 16
Re: Lateral resolution of TOFD method I can give you the practicle hands on approach.
Testing of carbon steel piping under 12 mm in thick using TOFD doesn't very well . The receiver transducer can't recover any useful information due to the intensity of the transmitting transducer. And even at 12 mm thick and using 15 MHZ transducers there a good chance something could go undetected. Calculations are nice but can't take the place of hands on.
|Ed Ginzel |
R & D, -
Materials Research Institute, Canada, Joined Nov 1998, 1236
Re: Lateral resolution of TOFD method Rebecca:
The resolution you describe sounds like you are looking at the temporal resolution to resolve the "depth" difference between 2 points.
This would be a compromise of parameters (like much of the TOFD technique is).
Some of your concerns may be addressed by the information found in "Engineering Applications of Time of Flight Diffraction", Second Edition, by Charlesworth and Temple, Research Studies Press, 2001.
This will be determined by the probe separation, nominal angle, the frequency of the pulse used, and the number of cycles in the pulse (to mention a few).
The occlusion of flaw upper tips by the lateral wave can be reduced using DSP to eliminate the lateral wave too (I have seen an effective rudimentary version of this by RTD). Dr. Honarvar at the University of Toronto did recent work using Weiner filtering to improve resolutuion too.
In a 1989 paper presented at the 17th Symposium on NDE in San Antonio, Texas (April 1989) Karl Lindenschmidt, Michael Moles and Tony Sinclair presented a paper that described TOFD with a 20MHz probe and used the shear mode for analysis on 4.2mm thick Zirconium. They detected and sized flaws on the far side having depths of 2.07 to 2.6mm, i.e. the deepst crack was only 1.6mm from the entry surface.
But a similar dead zone will need to be addressed due to pulse-width off the backwall so you may have to estimate the same sort of thing for resolution of far surface breaking flaws as separate from the backwall signal. This is generally smaller than the deadzone presented by the lateral wave but not zero.
----------- Start Original Message -----------
: I'm a graduate student at the University of Wisconsin-Madison studying the ultrasonic time of flight diffraction (TOFD) method. I would like to determine the lateral resolution of the setup I am using. Essentially what I would like to be able to calculate is the required difference in length of two ultrasonic paths of travel so that both signals will be detected in a single scan. As an example, assume inspection of a flat plate. How do I calculate that smallest thickness of plate that will produce a waveform containing both the near surface wave and backwall reflection? Any referces to literature about determining the lateral resolution of TOFD would be helpful.
------------ End Original Message ------------
|Jan Verkooijen |
Sonovation, Netherlands, Joined Nov 1998, 29
Re: Lateral resolution of TOFD method Before starting a head on clash between two methods again, I think it would be good to get the starting point right: Neither RT, nor UT, nor TOFD and not even Phased Arrays will solve every NDT problem in every configuration at all times. Each of those has advantages and disadvantages which may differ under different circumstances. Also the operator is not totally insignificant to the success of the exercise. After all, even if Ullrich would have had Armstrongs' bike, it is still not certain he would have won the Tour de France, or is it?
We will never claim that we can find everything in every weld, be it thinner or thicker than 12 mm, but we have excellent results of TOFD inspections of welds down to 6 mm thick carbon steel, which were verified by destructive testing.
It is also important to know what the object of the exercise is: for a fabrication inspection the object of the exercise is NOT to find all the defects. The object of the exercise is to maintain the quality level for the fabricator, specified by the end user. If the object of the exercise is to establish the status of a plant during an in service inspection, than damage mechanisms, utilisation of the plant, operating temperatures, expected extent of damage and many other factors come into play, and in many cases it will not suffice to use just one inspection method. A combination of methods will in most cases be necessary to come up with a sufficiently reliable result.
In both cases, the defect size and type that will likely pose a problem is important to enable the proper choice of NDT technique. Why use a magnifying glass to find sandgrains with a diameter of less than 0.1mm possibly obstructing our path if we know we would only hurt our toes if the diameter would be in the order of 100 mm? Or, the other way round, if already a 0.1mm sandgrain would damage our bearing, why spent money on a technique that will only find small rocks from 100mm diameter onwards?
The other thing is that probably the real question posed here is different to what it says litteraly: lateral is perpendicular to the weld direction and going on the rest of the question lateral resolution is probably not what Mr. Trinidad really wanted to know. I suspect the true question is the depth resolution of TOFD. Obviously, near the lateral wave this is not the favourable place to look for the best resolution of TOFD. Again, one should know what the object of the exercise is: we have excellent results in detecting and sizing defects in the root region and weld body of duplex steel welds, using validated procedures. This is not to say that we will have the same quality of results in all types of duplex welds or even in the top layer of the same weld. We simply cannot say this, as we would have only validated our procedures for this type of weld and this type of welding process in this particular region for this particular exercise. In carbon steel this would probably suffice: there is no reason to assume anything else, strengthened by the exhaustive tests and practical work that has been done in the past 15 to 20 years. For duplex and other materials this means it is extremely important to validate procedures by inspecting testblocks and verifying the outcome by comparison with destructive testing, as we do in most cases. The same goes for any other technique, be it manual or automated UT (including phased arrays). Contrary to what Mr. Johson says, the theory behind TOFD works quite well and can be used to set up the initial procedure, predict practical results, and in many cases extrapolate practical results. One should however use the right models and make sure the instrumentation, the rest of the equipment and the procedures are optimised for the exercise. Remember that although TOFD is an ultrasonic method, it is still sufficiently different from conventional UT to justify a considerable amount of theoretical and practical training and let us not forget practical experience.
So what is the answer to the original question? Without knowing the full details, it is hard to say, and I can understand why different people would give different answers, as it depends upon the circumstances and the experiences they have had. Fact is, that we should all realise that Duplex and full austenitic stainless steels are difficult to inspect with any ultrasonic technique and one should know one's theory and apply the equipment and technology there where he knows it works or where he can validate the procedure. Simply put: openess about the possibilities and limitations of the technology, and working with the client to arrive at a validated inspection procedure which, under the given circumstances would give acceptable, or may be better put: realistic results is probably the best we as NDT people can do.
----------- Start Original Message -----------
: I can give you the practicle hands on approach.
: Testing of carbon steel piping under 12 mm in thick using TOFD doesn't very well . The receiver transducer can't recover any useful information due to the intensity of the transmitting transducer. And even at 12 mm thick and using 15 MHZ transducers there a good chance something could go undetected. Calculations are nice but can't take the place of hands on.
------------ End Original Message ------------
NDT, RESEARCH AND DEVELOPMENT OF NEW METHODOLOGY, TRAINING CENTER
I&T NARDON INSTITUTE was idiated
in 1988 by Giuseppe Nardoni, now
president and owner of the company, based on his multi-annual and
international experience in the field of NDT.
Cygnus 6+ PRO Multi-Mode Ultrasonic Thickness Gauge
The Cygnus 6+ PRO thickness gauge is the most advance gauge within
the Cygnus range with key featu
res including: comprehensive data
logging; A-scan and B-scan display; manual gain control; Bluetooth
connectivity; and much more. With its unique dual display and three
measuring modes (Multiple-Echo, Echo-Echo and Single-Echo), this
surface thickness gauge offers maximum versatility for inspections.