Re: Ultrasonic Defect Sizing Techniques Dorin:
The three methods you describe are fairly basic but perhaps some "fancy titles" on traditional stuff is a bit confusing. Many of the items have appeared in previous writings on the NDT.net so a search of the website might help. John Drury's text, Dr. Harumi book on Tip Diffraction, Mark Davis' book on Sizing and the ASNT Handbook are just some of the texts more information can be found in.
Flaw Tip Echo (Tip Diffraction) refers to a technique whereby the backscattered echo from the end of a flaw is measured in arrival time. If you know the angle and signal is indeed the TIP echo then you need only read the travel time off the scope and do the trigonometry to get the position of the tip of the flaw. For an imbedded flaw (i.e. not surface breaking) this gives only a single point so you may need to try to locate the other tip from another direction. Alternatively, the flaw may be known to be surface breaking (using a "creeping wave" to confirm this or visual or MPI or other such surface technique). Then a corner reflector can be seen and you can look for the tip diffracted wave to occur as an early arrival signal. The difference in time between the corner signal and the tip signal together with the refracted angle can be used to get the approximate depth of the flaw from the surface it was observed breaking. These principles are best described in a book by Dr. Harumi "Japanese Tip Diffraction Handbook".
The amplitude sizing techniques are of 2 sorts, one is a simple linearisztion of size to amplitude where a known target sets 100% and the smaller amplitudes assume a smaller flaw (e.g. 50% amplitude is 50% of the calibration target size. This method is filled with problems and it should be considered a sorting technique rather than a sizing technique. The other is more scientific and is based on the work of Drs Krautkramer in 1959. THe AVG system (called the DGS system in English) rates a flaw size based on the echo amplitude of an ideal circular reflector on the beam axis and well into the far zone. Curves are made up for specific ranges and specific probes and specific instruments. These indicate the amplitude of a small flaw (must be much smaller than the beam size) with respect to an infinite reflector. Since it is an area ratio the effects of irregular shapes are not accounted for but it is a reasonable approximation.
The Echo Dynamic method uses the drop of a maximised signal to a specified amplitude difference rleated to a probe displacement. This is often done using Distance Amplitude Correction Curves generated from side drilled holes at different distances. Alternatively a "Maximum Amplitude" option might be used whereby you just drop the maximum signal from each flaw facet by 10% of the maximised signal. Probe displacement is marked on the test surface for each location (maximum amplitude and forward and backward displacement points) with sound paths also recorded. These coordinates are used to position the flaw "edges". This method is again filled with unaccountable variables such as flaw shape, flaw lateral extents, flaw orientation effects, reflectivity variations, etc., etc.
In many cases the "sizing" we do in UT is not so much sizing as an evaluation tool whereby we estimate if the flaw is critical or not and record it because the Code or Specification has put a limit on amplitude for a flaw. Generally (but not always) a larger amplitude indicates a larger flaw.
Any sizing method is subject to limitations. Even DETECTION of flaws has limits (we typically consider half the wavelength as the flaw size limit of detection....but DETECTION is NOT sizing).
When a UT technician is given the task of Sizing a flaw, the people using this information must be cognisant of the REAL tolerance of the method. In some cases the accuracy that is "perceived" is in fact many times greater than the tolerance of the method. e.g. A target used as a sensitivity reference target is a known size (say 3mm diameter). Company XYZ states that they will assign a size based on Amplitude and decide to use 5% increments of the reference amplitude to assign a size to any signals detected. Because multiplying 3mm x 5% gives 0.15mm it wrongly gives the impression that an accuracy of 0.05mm is being achieved.
Both the Amplitude Sizing and the Echo dynamic sizing methods are subject to considerable variations and results from these must be treated with caution and suitable tolerance must be stated up front.
PS I will be presenting more on the pitfalls of sizing in a later article
: Hello to whom ever this may concern,
: > My name is Dorin Brodersen. I'm currently attending the
: > Northern Alberta Institute of Technology, this is in Edmonton, Alberta,
: > Canada. I am researching flaw detection using ultrasonic inspection. I have
: > ran into some problems when researching 3 methods used. I have spent time
: > looking through magazines and books at the library and have found nothing.
: > The three methods I am researching for at this point are the Flaw Tip Echo
: > Method, Echo Amplitude Method, and the Echodynamic Method. I am wondering if
: > you have an idea of what these methods are. If so, would you mind sending me
: > some information on these 3 methods. I would be very appreciative of any
: > help that you may assist me with. After graduation in April 2000 I am
: > planning to work in the NDT field (ultrasonics).