Just some questions about MT and PT

October 2010

On June 11th, 2010, one of our readers sent us the following message:

‘‘A Quality Control coordinator, an NDT specialist in UT, but also an MT and PT Level 2, I am always happy to read your papers when my NDT Web searches make me go to your Website.

The reason why I come to you now is a simple practical question as a result from several findings.

When seeing an MT linear indication (length: 10 mm, ca ½”), we are in a classic customer/supplier “disagree situation” about its danger and its nature. I decided to carry out a local PT inspection starting from a simple and pragmatic idea: if it is wide, the final PT indication will be a further information that this indication is due to a volumetric flaw; the contrary is not necessarily true as the type of the flaw as well as its direction/shape added to the 2 or 3 MT (with black magnetic ink) and the UT couplant used previous to PT may impair the test! As you may guess, the supplier claims that this indication is due to a plane flaw while I suspect it to be deep in the 5 mm (0.2”) range.

My idea was to try and fill up the flaw with penetrant: I did not follow any written procedure (at least none I had knowledge of), but proceeded in a very pragmatic way after my ten-year practical experience on different kinds of materials.

I therefore did my best to fill up this defect with a colour contrast penetrant. Some minutes later, I applied again penetrant as I thought that the aspect of the penetrant around the crack may make it useful to put more penetrant. Though I thought in a pragmatic way that there is no way to put more in a glass that it can contain (a truth often said in bars) I left time enough for the penetrant to fill the defect (penetration time: 30/45 minutes, depending on whether we count from the first or the second spraying).

Excess of penetrant removing was performed using a dry and lint-free rag, moistened with tap water supplied by the water network of the French city where the parts were stored.

I let parts dry alone, due to the 25°C (77°F) of the air. I then applied a homogeneous, thin and even (as much as possible) layer of developer. The questioned linear indication very quickly became a spot of about fifteen mm².

Let us go to my question (or my questions):

I do not expect to know exactly the volume of the flaw, as I know that it is today (and will probably be in the future) impossible: too many parameters are to be considered. However, being pragmatic once again, I think that a small glass of water turned upside-down will give less work to the housewife than a mop bucket.

• Has anyone already tried to make a relationship between the end surface of an indication and the volume of the flaw?
and

• Are there some parameters that one could use to guess the volume of the flaw (apart the inspector’s insight)?

I hope that you will have time to answer me."

Our answer:

"Thank you very much for your e-mail; we are grateful for the interest you show in our Website.  

This topic has already been, at least partially, written about in our DPCNEWSLETTER N° 022(*): we wrote about the indications’ length but we have been very careful not to write about the indications’ surface or diameter.

Indeed, our experience showed us that:

• An indication which appears and quickly spreads is the proof that the discontinuity is widely opened.

• An indication which slowly appears and finally which gives a thin line, is the proof that the discontinuity’s opening is tight.

However, in this second example, there is no room to say that beneath the opening the flaw is not a large one with almost closed edges. In the same way, a quickly “bleeding” indication may be due to a shallow but wide opened flaw. Making a parallel between the depth or the volume of a flaw and the quantity of bleeding penetrant or with the indication width, to be honest...is utopian!!!  The only way to get an idea is either "to spotface the metal" by removing a thin layer of material (if allowed), or to use another non-destructive testing (NDT) method. If ultrasonic testing (UT) gave an unambiguous answer about the presence of a deep discontinuity, it IS deep.

UT and MT performed before PT obviously raise a major trouble. PT confirmed that the discontinuity is opened ... however it may be polluted by residues (and it is certainly!) which modify any interpretation as for the size, the depth and the volume of the defect!

Reapplying penetrant is generally useless. If the penetrant began to "ripple”, it is because of pollutants; we think of hydrocarbons used in MT materials products, or of some constituents of the UT couplant.

Depending on the penetrant/developer used for inspection, the same discontinuity will not give an indication of the same width. Renowned products will give the same length, but they do not spread out the same way from both linear or rounded indications. Then, if as an example, the relevant specification states that any indication less than 3 mm diameter is acceptable, and that at more than 3 mm, it shall be considered as unacceptable, it may occur that a 2.5 mm indication is got from a couple of colour contrast products and 3.5 mm with a couple of a competitor’s materials, both couples being approved as per the American specifications or approved as per the French RCC-M Code (Design and construction rules for mechanical components of PWR** Nuclear Islands) requirements. A real experience.

There are many parameters to carry out reproducible PT inspections. The same very seasoned operator, on the same parts, with exactly the same materials (the same spray cans from the same batch numbers), may very well detect "borderline" discontinuities on reference test blocks in the morning ... and detect only 60 or 70% of the discontinuities in the afternoon.

We hope our answer meets your needs.

Our reader sent us a further message:

‘‘Coming back from a professional assignment, I would like to thank you for your complete and well-argued answer.

I expected your answer, but I wish to enhance this topic, as I have another set of questions for you. I’ll likely write about something I read in one of your papers(*): ‘‘guess that as a maximum 50% of the penetrant gets out of the crack.’’

I think it useful you know that the problem I am facing is about fifteen moulded parts which almost all have linear indications in the same machined areas; in my opinion, these indications correspond to defects that occur during the cooling phases.

My first idea is that’s for sure, and I fully agree with you on this point, a weak indication gives no clue to the volume of a defect, especially in my case where 2 MT inspections and some UT couplant have probably partially clogged the discontinuities. However, on the opposite, a wide spot may or may not be due to a deep defect or to a volumetric one (long, yes, but deep at the same time), but surely from a widely opened one; it may be deep ... or not.’’

Notes

(*) Patrick DUBOSC and Pierre CHEMIN, Penetrant testing: is more too much? (revisited), DPCNewsletter N°022 – March 2010. On our Website:
http://www.ressuage-magnetoscopie-penetranttesting-magnetictesting-dpc.info/

(**) PWR means Pressurized Water Reactor.