Hi Spedly78

The (relatively) easy second part first - and my reply starts with a question!

When you say correction factors, do you mean constructing a Distance-Amplitude Correction (DAC) curve, usually achieved electronically on modern digital flaw detectors? Incidentally such flaw detectors usually have a Time Corrected Gain (TCG) function which compensates the display for material attenuation and sound beam characterisitics so that discontinuities of equal relative size at different depths in the tes piece will give an equal amplitude response on the display. Generally DAC curves are constructed using DAC or test sensitivity blocks ideally made of the same material in the same heat-treat condition and of the same surface roughness as the material under test. Several artificial reflectors (side-drilled holes (SDH), flat-bottomed holes (FBH) or notches are stragetically placed in positions corresponding to several different beam path locations on the production test. So already several drawbacks here - do the acoustic properties and coupling conditions of the sensitivity block adequately match those of the item(s) under test. You will need to run attenuation checks and transfer correction test to minimalise any discrepancies. How consistent are such properties across the range of test items? There could be wide variation especially if tests are on in-service items.

Then the sensitivity setting targets are artifical reflectors with a perfect geometry (spherical, circular, notch-shaped), hardly representative of general defect morphology and orientation to the ultrasonic beam. How does one relate the response from a perfect geometrical reflector to irregularly shaped inclusions? A programme of comparison of UT responses from various different real indications with real-time radiographic images or salami-slicing sectioning will build up a data bank of expected echo amplitudes for various morphologies, but it will be an expensive and time-consuming business.

And the DAC curve is only suitable for a particular test equipment/lead/probe combination and must be checked regularly to detect any deviations in echo amplitude caused by probe shoe wear or equipment deterioration. And no two UT operators using the same equipment will ever get the same primary gain setting for the curves.

Just some of the factors to be taken into account when assigning a confidence level to sizing accuracy.

The first part of your post made me nervous with that word "downgrade", as if you had an unacceptable indication on the original test and you were trying to classify it as acceptable. But perhaps I am just misreading your post and what you are actually referring to is the principal reason why DAC curves are used, i.e. in the far field of the ultrasonic beam the defect response from a discontinuity falls off with increasing beam path distance. Don't forget that discontinuity sizing should not take place in the beam's near field due to sound pressure fluctuations.

For a technical discussion on testing of railroad wheels. follow this link:

http://www.ndt.net/wshop/forum/messages-1/218.htm

Also I am certain a search on the ndt.net site will produce much other useful information.

Cheers!

Nigel Armstrong
Karachaganak Petroleum
Kazakhstan