French (Fr)English (United Kingdom)

DPC NEWS: a website dedicated to Penetrant Testing and Magnetic Testing



visits on site since April 2008

Log in


Receive HTML?

Mail Inbox - July 2012

Written by Administrator
Sunday, 10 June 2012 14:33

The smallest detectable crack width

July 2012

Recently, we have got a very specific question, asked by one of our female readers:
"What is the minimum crack width detectable by MT?"
She was surprised not to have found the answer on our Website.

This question is not without reminding us of a similar one asked about PT:
"How small a crack can be found?"
In the January 2012 Penetrant Professor issue(1), you may read an answer.
As a conclusion, on a humorous note, the author mischievously asks:
"How long is a piece of string??"

Bear in mind that PT, as MT, is an NDT method designed for the detection and location of surface discontinuities. At best, the indication length may be measured, which may well be not the actual crack length.

However, some answers may be given.

First, about PT.
We may use the Type 1 reference test blocks of the ISO 3452-3 standard. The cracks on these test panels are essentially straight and open to surface.
The four test panels mentioned in this standard display cracks, the depth of which is respectively 10 µm, 20 µm, 30 µm and 50 µm. The crack depth/width ratio is ca 20.

However, in the past, similarly-designed test panels with 5 µm crack depth and 0.25 µm crack width have been marketed. In one of our papers (2), we reported that, on these test panels, the most sensitive penetrant system detected 80% of the discontinuities.

We are, then, very close to the PT detection limits, if we can assume there is a limit, knowing that viewing conditions are essential for such a result. Nevertheless, keep in mind that these laboratory conditions are much more favourable than that those met in workshops. Further, the test panel surface is perfectly smooth.
Therefore, it would be unrealistic to state such a detection requirement.

Let us consider now MT: for MT, unlike for PT, there are no such test panels which, incidentally, would not be very useful.
Indeed, in MT, too many parameters are involved, among others: the type of magnetic field applied to the part (AC, HWDC, FWDC, etc.), the magnetic characteristics of the material, the surface condition of the parts, the geometric characteristics of the discontinuities, the applied tangential magnetic-field strength, the discontinuity direction with respect to the direction of the applied magnetic field, the magnetic particle size, the characteristics of the carrier liquid, the viewing conditions, etc. It is, then, easy to understand why no figure for the smallest discontinuity width detectable by MT is stated in documents, whatever they are. For instance, no figure is given for the Type 1 reference block of the ISO 9934-2 standard.

We know well of a very thorough study that was performed in 2002 by simulating interactions between magnetic particles and defects. It appeared that the defect width was not an overriding factor. For example, the defect orientation, the flaw depth and many other parameters have a greater influence than its width.
Under these conditions, it does not seem then reasonable to give a minimum detectable width figure. Only tests on real defects could help giving an informed opinion, and yet, as noted above, by stating many parameters. Note that MT is particularly well suited for detecting fine and deep discontinuities (cracks).
Our experience is that some micrometer-wide defects could be reliably detected.
For example, the artificial defect width, on the Type 2 reference test block of the ISO 9934-2 standard, is 15 μm. It is easy to detect, especially at its both ends. We do not believe that tests have been performed to give a definite figure, though we are in the opinion that a width of 1 to 2 µm could be detected under optimal conditions, subject to laboratory testing and not in workshop conditions.

As you can see, be it PT or MT, these NDT methods shall not be asked for things for which they are unable to give an answer. However, in a world where one wants to quantify and measure everything, inevitably, we are convinced that this kind of question will come up again!

One would almost forget the basics...the qualification, expertise, skills and experience of operators and inspectors, knowing that the automatic inspection processes of indications have not yet been proven, at least in real working conditions.


(1) The January 2012 Penetrant Professor issue: On this Website.

(2) Pierre CHEMIN and Patrick DUBOSC, Ultra high sensitivity penetrants, December 2009: On our Website.

Normative references

• ISO 3452-3:1998 Non-destructive testing - Penetrant testing - Part 3: Reference test blocks, International Organization for Standardization, Geneva, Switzerland, 1998.

• ISO 9934-2:2002 Non-destructive testing - Magnetic particle testing - Part 2: Detection media, International Organization for Standardization, Geneva, Switzerland, 2002.

Last Updated ( Sunday, 10 June 2012 15:19 )