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?

DPCNews 022 - Penetrant testing: is more too much?

Written by Administrator
Monday, 01 March 2010 08:13

March 2010


1- Introduction

Penetrant testing is an outstanding Non Destructive Testing method: simple in its principle, easy to carry out and which does not require any electronic means to excite parts, to analyze signals or still to record data.

For decades, PT has had several purposes: detection and localization of the open to surface discontinuities.

Since several years, PT has been asked for performances for which it is not designed. This leads to specifications, requirements in particular concerning the size of the indications which seem to us incompatible with the purpose of PT.

As an introduction, let us remind that when we carry out a non-destructive testing, we do not directly observe the discontinuity, but the picture given by a medium: screen of a UT or ET flaw detector, film or screen for RT, detection medium in MT, penetrant in PT. It is this picture that we have to qualify before any "accept" or "reject" conclusion.

In PT, pictures or indications can be misleading. Thanks to experience traps are generally avoided.

Some examples:
• A very tiny indication, which bleeds out very slowly, may in fact be a large discontinuity whose opening is tight;
• A ‘‘blurred’’ indication which quickly spreads out, which even may give an impressive spot, may be just a porosity without any consequence;
• An area a bit more "porous" on a foundry part which generally displays an even background may be seen as an area of porosities and not as a background;
• An experienced inspector within a quarter of second will detect aligned indications- no artificial vision system is able to duplicate such an expertise.

2- Beware


The outstanding machine comprising the stereoscopic detector (the eyes) and the data processing system (the brain) is able to achieve surprising achievements.

Sight is therefore crucial for PT. Corrected or not, it has to correctly respond to colours. Eyes need to work under good conditions.

The right viewing conditions are summarised underneath:

2.1- Colour contrast PT

• White light whose spectrum is very close to the solar's one;
• Non-glaring lighting;
• Avoid reflective surfaces;
• Illuminance of 500 lx (ISO 3059:2001) or 1,000 lx (American specifications) minimum on the surface under inspection;
• From experience some filters may increase contrast.

2.2- Fluorescent PT

• UV-A irradiance greater than 10 W/m² (1,000 µW/cm²) on the surface under inspection (ISO 3059:2001 and American specifications).
Furthermore American specifications require to measure UV-A sources' irradiance at 15" (38 cm). If the 10 W/m² (1,000 µW/cm²) requirement is met, why is it mandatory to increase Quality Assurance burden by this measurement at a specific distance?

- Illuminance (visible light) less than 20 lx on the surface under inspection in the real working conditions (ISO 3059:2001).

In this standard it is mandatory that all the UV-A sources be in service while measuring illuminance; no visible light or UV-A radiation source shall be in the operator's field of vision. Further the illuminance of the ambient visible light shall be less than 20 lux: this requirement prevents an operator to inspect through a window parts which would be in a small box with UV-A sources while his eyes would be in the plain visible light of the workshop. And yet!!!

- Some specifications have additional requirements: as an example, measure the illuminance (visible light) of EVERY UV-A source set at a distance from the surface under inspection or the sensor such that the UV-A irradiance is 1,200 µW/cm².

Don't you think the main point is: carry out illuminance and irradiance measurements on the surface under inspection, and illuminance at the operator's eyes level, in the real working conditions?

Why ask for "artificial" conditions which, once more, come atop Quality Assurance requirements, and give NO ADDITIONAL GUARANTEE about the quality of the process? Better to check that the operator is in an inspection booth and not in the workshop performing inspection through a window!

Two other points of interest:

• Abide with the eyes adaptation time when entering a UV-A radiation inspection booth. Depending on applicable documents, on the inspector's age, on the illuminance outside of the booth 1 to 5 minutes are needed before beginning any inspection-- especially if tiny indications are looked for;

• Going into or out of the booth not too often: every eyes' adaptation time (from plain visible light to UV-A radiation and from UV-A radiation to plain visible light) is very tiring for eyes.

3- The drifts

After reading the above paragraph about these basics and the very important precautions to be taken let's have a look at what we call "the drifts".

Three of them are the topic of the underneath paragraphs:
• Tools used to measure indications;
• Penetrants' performance and sensitivity;
• Human eyes' performances.

3.1- The means of indications measurement

Indications dimensions are given as length or diameter. Rules, gauges, callipers, magnifiers comprising a reticule, probably others gimmicks are used! Shall we call them "measuring equipment" or "indicators"? (see the glossary at the bottom of the paper).

If an indication measurement figure is close to the acceptable criteria, maybe it is better to use another means.

3.2- The performance and the sensitivity of penetrants

The ISO 3452-2:2006 standard requires that the "couple" colour contrast penetrant/developer, the penetrant being a Level 2, shall show at least 75% of the discontinuities on a Type I, 30 µm test block as detailed in the ISO 3452-3:1998 standard (30 µm is the depth of the discontinuities.)

The 2006 version states that "indications the unbroken indications covering at least 80% of the panel width, clearly visible to the unaided eye (with glasses if usually worn)’’are to be taken into account. This sentence is to be understood as: the number of all the indications comprising one or several segments that display at least 80% of the panel's width as far as
it is not stated that the 80% should be as a single segment.

When using colour contrast penetrants one may assume that a 0.3 mm long indication is visible (hence detected) at a distance of 30 cm (15") (we will explain the calculus in the paragraph dealing with eyesight).

What's the volume? What is the quantity of the dye in the crack?
• Depth: 30 µm (0.03 mm);
• Width: 1.5 µm (maximum);
• Length: 0.3 mm.

Only a part of the penetrant which entered the flaw bleeds out. Guess that as a maximum 50% of the penetrant gets out of the crack.
This comes to 6,750 µm³ i.e. 6,75 10 -6 mm³ or 6.75 10 -9 mL.

This volume is the volume of the penetrant! The dye (the molecule) is dissolved at a ratio of ca 1%. If we assume a density of 1,000kg/m³ we go to a mass of 6.75 10 -11  gram rounded up to 0.07 ng (nanogram)!

Let's make a similar calculus with fluorescent penetrants.

We use a 10 µm Type I test block as described in ISO 3452-3:1998 standard.

• Depth: 10 µm (0.01 mm);
• Width: 0.5 µm (maximum);
• Length: 0.2 mm.

This length is the smallest one on the fluorescent comparator and is the smallest indication generally considered to be non-acceptable.

Further it almost matches the visual acuity at a distance of one meter. At 30 cm, a more common distance for inspection, especially for critical parts, the visual acuity is close to 87 µm, i.e. 0.087 mm (an angle of one minute).

For very high sensitivity penetrants the concentration of the fluorescent dye + optical brightener complex is ca 1.5% weight/volume.

The volume of the discontinuity comes as:

0.010 x 0.0005 x 0.2 mm = 0.000001 mm³ or
10 -6 mm³ or 10 -9 cm³.

If only half of the penetrant bleeds out, i.e. 0.5 10 -9 cm³, the mass of dye + brightener is:  0.5 10 -9 x 15  10 -3 g = 7.5 10 -12 gram, the enormous mass of 7.5 pg (picogram) or 0.0075 ng!!

3.3- Performances of the human eyesight

Our topic is not to teach you about the eyes; it's only to remind you of some data.

- 3.3.1 Response to colours
The top of the human eye's response curve to colours is at 555 nm (in photopic conditions), which is yellow-green.
In scotopic conditions (when light is dimmed) the top is at 505 nm (green/green-yellow).
Our eyes are far less sensitive to the red and to the blue.

- 3.3.2 Sensitivity to contrast ratio
The human eyes are sensitive to contrast ratio: a black indication against a white background is far easier to detect than a deep yellow one against a faint yellow background.
If the background is very dark, almost invisible, if the indication is green, green-yellow or yellow, contrast will be high and even a very tight and small indication (as those from fluorescent penetrants) will be seen when several conditions are met:

•  Very low ambient visible light: any visible photon lowers the signal/noise ratio. It is not acceptable that the eyes be in a high illuminace area when inspecting parts with fluorescents indications.

•  The eyes shall be adapted to the low level of visible light (illuminance). This may need several minutes. The human eyes are among the most sensitive of living beings: 4 or 5 photons are enough to set off a signal to the brain.

Interesting to know: in some countries the comparison of the fluorescence of an in-use penetrant with the brand new same penetrant is still performed with the naked eyes; the two penetrants must be diluted at 1/1,000, even at 1/5,000 so that they emit very few visible photons. Doing so the eyes will be able to detect a very small difference of fluorescence. If penetrants are undiluted we challenge anybody to detect even a 50% difference! Eyes are flooded by photons!!

This is an explanation why using a filter paper moistened with undiluted penetrants for a direct comparison never led to any detection of a loss of fluorescence!

Nowadays comparison of the fluorescence intensity of fluorescent penetrants shall be carried out using a fluorimeter as per:
• Either ISO 3452-2:2006 standard Annex A.
• Or the paragraph of the SAE-AMS 2644E specification, test method described in ASTM E 1135 ‘‘Standard Test Method for Comparing the Brightness of Fluorescent Penetrants’’.

- 3.3.3 Eyesight attraction
When using fluorescent penetrants, after eyes' adaptation to the low illuminance, the eyes will be attracted by the glowing indications displayed against a dark background. Inspection's reliability and speed are dramatically increased.

- 3.3.4 Linear arrangement fixing
A seasoned professional is able, within a quarter of a second, to tell that indications are in a line and that the total length is to be considered, instead of the length of every small indication. No artificial vision system is able to duplicate such an expertise. In the same way a human inspector is irreplaceable when coming to "feel" that an area which seems very similar to the current background as found on castings is in fact an area of porosities.

- 3.3.5 Resolving power
The average human eye is able to see 2 points 1 minute of angle apart.
A minute of angle comes as 0.3 mm at 1 metre (perimeter of a 1m radius circle: 6.28 m; 360° x 60 = 21,600 minutes. 6.28 m= 6,280 mm divided by 21,600 = 0.29 mm!)

4- Other borders

One may understand that penetrant testing, though it seems very simple, though many people consider it as obsolete, is able to show outstanding performances-- given it is used in the right conditions.

Every batch of penetrant is validated "on the border". The ultimate performance is checked on every batch.

Further it is known that the same inspector, on the same day, on the same parts, using the same products (the same spraycans from the same batch number, for instance) may get different results. That is especially true when using colour contrast PT on Type I, 30 µm test blocks: this method is then "at the limit" and the smallest difference in precleaning parts or while spraying the NAWD leads to a noticeable difference of detection capability.

Sometimes after a very gentle grinding the indication is no longer visible. On the other hand the same gentle grinding will open a reservoir under a very tight indication.

More and more useless requirements are to be met: check EVERY batch of penetrant's performances at 50°C (122°F), while qualification tests demonstrated that the penetrant withstands 114°C (237°F) for an hour. Or to demonstrate the drying oven's "overshoot" on start up or in any circumstance does not exceeds the nominal maximum 71°C (160°F) plus the allowed 8.3°C (15°F), even when the oven is empty (quite often this is the situation when a PT line is put in service); keep in mind that penetrants have been qualified at 114°C (237°F) for an hour! Watch out, you, the subcontractor who is not able to display the records! This is a major non-conformity! On a parallel way no auditor will have any doubt about the ability of companies which carry out radiometers/luxmeters calibration/verification. Obviously the certificate is the proof that everything is OK!

So why not accept that the certificates of conformity to the ISO 3452-2:2006 standard issued by renowned PT materials suppliers demonstrate that the penetrants may be used at 50°C (122°F) without any further specific test?


‘‘Accuracy’’: some documents issued by the aerospace industry require that any item used to measure indications shall have an "accuracy" of 50 µm.
Some of the measurement items to check fluorescent indications are plastic comparators comprising lines, dots, circles of different dimensions drawn with a fluorescent ink. No way to have an "accuracy" of 50 µm!

Measuring devices: Quality Assurance documents require these devices be calibrated or verified on a regular basis and that a test certificate/report states after which standards, documents, procedures this test is performed.
More and more some people want to format with accurate figures a method whose only purpose is to state: "here there is an open-to-surface discontinuity, which may be due to this or that reason".
Depending on the equipment the "regular basis" may be from daily up to once every 12 months, or even more.

Indicator: an indicator does not need any calibration/verification.

Calibration: The English wording applies to two different requirements. Basically a calibration is only a record of the gaps between a reference and the device under test without stating whether the unit is able to meet the user's needs. Many auditors use also this "calibration" word instead of the "verification" as detailed underneath. This is a confusing situation.

Verification: recording of the gap between the figures displayed by the reference unit and by the unit under verification, adjustment if needed, judgment on the capability of the equipment to meet the user’s requirements/needs.

5- Acknowledgement

The authors of this paper want to give their acknowledgments to:

• SOGI COMMUNICATION which agreed we use a paper by PATRICK DUBOSC titled: " PENETRANT TESTING: IS MORE TOO MUCH?" published in the technical journal CONTRÔLES ESSAIS MESURES January 2007 issue, pages 83 to 85.

• The Research and Development Department of BABBCO which also agreed we use this paper as a basis for an update.

We, Pierre CHEMIN and Patrick DUBOSC, welcome any comment, any idea. If you have some examples you would like to see discussed here, please give us all the useful indications. If you require confidentially, we would modify locations, names and some parameters to prevent any traceability.
Nevertheless, we are convinced that our site may be a kind of surge-valve: the topic is NOT to target this company, or that auditor; but it is always to make users think, to make them ask themselves, or others, the right questions.
We may also give advice, once again on a confidential basis if needed: please, feel free to ask questions, to document our data basis: about Material Safety Data Sheets (MSDS), about environment, a chemical name you don't understand, a Penetrant process you have heard about, etc.
We have plenty of examples, some being out of all the specifications/standards, which led to the discontinuities detection, when the "current, normal, processes" prevented discontinuity finding.

Last Updated ( Saturday, 21 May 2011 09:52 )