Contact - Links - © Pierre CHEMIN et Patrick DUBOSC - 2008

I- INTRODUCTION

For many years the military specification MIL-I-25135, the first issue of which is dated 1959, to which was associated a list of approved materials, was the reference document regarding penetrant testing in the Aerospace industries.

As a matter of fact, the American aeronautical manufacturers have always exercised a part, more or less important, of their activities in the Military sector with the DOD (Department of Defence)

Besides, the tied-up links, on both sides of the Atlantic Ocean, between the manufacturers, the subcontractors, the part maufacturers, the maintenance and repair shops brought the totality of them to take into account this specification as long to apply it as write their own specifications.

Considering the specificity of their manufacturings, the aircraft manufacturers indeed wrote their own instructions as soon there were liable vis-à-vis their customers for control procedures,maintenance procedured and recommended or qualified products

So was universally adopted the MIL-I-25135.

II- EVOLUTION OF THE MIL-I-25135

The MIL-I-25135 as well as the relevant list of approved products (Qualified Products Lists) were updated through amendments along the years, For several decades Amendment C of the MIL-I-25135 was in force before being replaced by Amendment D, on June 24th, 1984, then by Amendment E, on June 26th, 1989.

While the aeronautical technology quickly progressed, processing new alloys, in particular those from titanium and from nickel, it is rather surprising to notice that the MIL-I-25135C was in force for so a long time while it did not address new requirements due to technological progress. The gaps of the MIL-I-25135C were felt already in 1972 and 1973, when the Quality Direction of a French aeroengine manufacturer began to state require maximum allowable contents of chlorine, sulphur impurities in penetrant materials.

III- GAPS OF THE MIL-I-25135C

The following considerations had not been foreseen by the MIL-I-25135C.

III-1- CASE OF TITANIUM ALLOYS

The increasing use of titanium-based alloys, which led to lower weight and to interesting metallurgical characteristics, made aircraft manufacturers forbid use of the chlorinated solvents.

Indeed, under the effect of some physico-chemical phenomena, free radicals of chlorine, in the form of atomic chlorine, appear, for example due to photolysis. The chlorine so released turned out to induce, under hot conditions, phenomena of stress cracking of the titanium alloys. General Electric (USA) suspected even methanol to exercise an adverse effect on these same alloys. This suspicion was in fact because was it the radical OH or, less plausible, the radical H (atomic) which was responsible for it? The problem was solved by itself as far as methanol, being classified toxic, is

practically no longer used.

III-2- CASE OF NICKEL-BASED HEAT RESISTING ALLOYS

The constant search for improvement of performances required to increase the thermodynamics yield of gas turbines. According to the second of Carnot’s principles, an increase of the temperature of the hot section of gas turbines required to resort on one hand to the technology of hollow air cooled gas turbine blades and on the other hand to heat resisting alloys, able to resist to higher and higher temperatures, such as those based, among others, on cobalt, chromium and nickel. At temperatures above 900 °C approximately, nickel may undergo an "oxidation" by sulphur. This chemical reaction is known as the sulphidation which leads to the formation of black corrosion products which are nickel sulphide-based. This phenomena is called “Black Plague”.

The catalystic action of alkaline metals, such as sodium and potassium, was put in evidence in the sulfidation reaction :

Ni + S → NiS

or K

Sulphur is present in organic compounds and as impurities in oil products. Alkaline metals are present in surfactants, such as those of the anionic type, among which we can quote one of the most known: the sodium dodecylbenzenesulphonate (CAS N° 27323-41-7)

It is necessary to point out that the organic synthesis of some nonionic surfactants may require the use of potassium sulphate as catalyst. So, such surfactants contain unwanted impurities:which are sulphur and potassium. As penetrants and hydrophilic penetrant removers contain surfactants, it is necessary to carry out a rigorous selection of them. To get rid of the alkaline impurities, some manufacturers, at the beginning of the 80s, even made them undergo a long treatment on ion exchange resins.

III- 3- THE “HEAT FADE” PHENOMENON

The drop of the fluorescence brightness of crack indications detected by penetrant testing is a phenomenon known under the name of‚‘‘HEAT FADE‘‘. It mainly appears at the drying stage after water rinsing: because either the drying temperature is too high

or the drying time in the air circulated oven is too long. It is interesting to note that Pierre CHEMIN presented a paper on this subject titled "Thermal effects in Penetrant Testing processes" at the second European Non Destructive Testing Conference held in Vienna, from 14 till 16 September 1981.

III-4- A HIGHER FLASH POINT

Most of red dye and fluorescent penetrants approved according to MIL-I-25135C had a PMCC (Pensky-Martens closed cup) flash point of 70 °C. This relatively low flash point was due to the presence, in the penetrant formulation, of relatively low boiling range aliphatic hydrocarbons, easily available on the market at a somewhat low cost.

Indeed, higher the flash point higher the hydrocarbons price.

These fluorescent penetrants which had a flash point of 70 °C, had a major drawback when used by immersion. Indeed, in tank, there was a slow but real evaporation of hydrocarbons all the more important as the ambient temperature was high. This also caused the evaporation of the other nevertheless less volatile constituents of the penetrant. The same phenomenon could also occur by dipping parts in the penetrant tank if, the parts were not allowed a sufficient time for cooling down to room temperature after trichloroethylene vapour phase degreasing. This lead to a general imbalance in the penetrant causing on one hand: a viscosity increase and on the other hand an increase of the drag out losses together with a fluorescent background on the part surfaces.

So, to reduce this excessive fluorescent background, operators instinctively went to overwash the parts, while these penetrants exhibited a poor resistance to water overwash. Overwashing consequences were the loss of crack indications, the drop in sensitivity and finally the loss of the reliability of detection.

It is necessary to note that the replenishment of the penetrant baths was usually done by adding, after analysis (distillation of flammable products!), a standard mixture of hydrocarbons and the other relatively volatile materials of the penetrant such as: methylisobutylcarbinol, odourless kerosene, etc. … Nowadays, this replenishment, which was a complex and risky operation, is no longer performed or even necessary.

IV- THE PRATT &WHITNEY SPECIFICATION

The first serious works on quantification of the penetrant process sensitivity were undertaken in the seventies. Among these, let us name those of Norman H.HYAM (1972), then later those of Lormerson Junior (Two fold congruency test) of Pratt & Whitney USA and a little later those of Jean VAERMAN (SNECMA).

All these important and remarkable works allowed to determine the influence of the variation of operating parameters of the penetrant processes on the sensitivity of crack

detection.

These works brought Pratt and Whitney, in the absence of the revision of the specification MIL-I-25135C, to write a new Penetrant Testing specification which came into effect at the end of the 70s.

Pratt and Whitney above all took care of flash point figure. Would not one of their factories have been then on fire? However Pratt and Whitney withdrew the qualification and refused all the chemicals, whatever they were, of which the PMCC flash point was lower than 200 °F (93 °C).

As an example, let us quote the compressor washing fluids (designed to restore the thrust of aeroengines and industrial gas turbines) although these fluids are used at the strength of 20% by volume in water.

The Pratt and Whitney specification also required:

– Maximum allowable contents in impurities: chlorine, sulphur, sodium and potassium.

– Resistance to the thermal effects (heat fade).

These new requirements made the penetrants manufacturers design new penetrants because none of those who existed on the market, at that time, meet this specification. So Pratt and Whitney distanced himself by causing certain problems. The penetrant manufacturers worked hastily wanting to be quickly approved.

If new penetrants were approved according to this specification, they gave rise to surprises; in particular, problems of water washability and other problems associated to penetrant removal even by using the hydrophilic penetrant remover at the excessive strength of 20% by volume in the water.

So, the suppliers and the subcontractors working for this aeroengine manufacturer had then to double, at that time, the number of their penetrant process lines, ones to meet Pratt and Whitney’s requirements, the others to meet requirements by the other aeroengines manufacturers which had not concurred with Pratt and Whitney. Same situation in the Repair and Maintenance shops worldwide.

The future praised Pratt and Whitney ideas because its specification became the core of the MIL-I-25135D. Nobody can dispute all the relevance of the requirements which, Pratt and Whitney pushed for some years before, and which were introduced in the so long for waited specification MIL-I-25135D on June 29th, 1984. It thus brought a big relief as it twas largely recognised in the Aerospace industries.

V- MAIN DIFFERENCES OF PRODUCTS BETWEEN THE MIL-I-25135C AND THE MIL-I-25135D/E

Let us point out here that the differences between the MIL-I-25135D and E are small.

With regard to penetrant materials listed in the QPL of the MIL-I-25135C, those listed in the QPL of MIL-I-25135D/E have the following advantages:

– Guarantee that products were checked about hygiene and safety (in 3.3.1) for operators. Complete absence of asbestos required in developers.

– Higher flash point 200 °F (93 °C) minimum (in 3.3.3), higher safety. Furthermore, there is an empirical rule of thumb establishing a relationship between the flash point and the maximal temperature of penetrant application. The rule is: PMCC flash point less 20 °C is roughly equal to the maximum temperature of penetrant application. For instance, if the flash point is 93 °C, the maximum temperature for penetrant application on the part surface is 73 °C. As consequence it is possible to reduce the cooling time after parts hot degreasing before penetrant application.

– Thermal stability (in 3.4.5.4) which allows to evaluate the penetrant resistance to thermal effects (heat fade).

– Penetrant tank stability (in 3.4.7).

– High temperature titanium stress corrosion (in 4.5.2.2).

– High temperature (1,850 °F ± 50 °F) or (1,010 ± 28 °C) corrosion of cast nickel alloys (in 4.5.2.3).

– Fluorescent brightness test in accordance with ASTM E 1135.

– Ultraviolet stability.- Etc…

VI- ASSESSMENT

The American specification MIL-I-25135 D/E was since replaced by the AMS 2644 and it is the AMS 2644E, revised in October, 2006 that is the curent revision. The last

revision of the list of the qualified products is the QPL-SAE-AMS-2644-4 dated October 1st, 2004. This one should be revised in the course of the first half of 2008.

This new QPL will list only the penetrant materials which are still manufactured and marketed.

The penetrant materials which will be no longer manufactured and marketed will be phased out from this QPL, the manufacturers shall list them and the list will be sent to the Department of the Air Force, AIR Force Research Laboratory, Wright Patterson Air Force Base.

The penetrant materials approved to AMS 2644E but not listed in QPL-SAE-AMS- 2644-4 will be listed in the new QPL, provided that they are still manufactured and marketed.

The French aircraft (airframes and aeroengines) manufacturers more or less adapted their own specifications to AMS 2644 and the penetrant materials which they approve are listed in the QPL-SAE-AMS-2644-4.

The AFNOR: NF EN 571-1, NF EN ISO 3452-2, NF EN ISO 3452-3 and NF EN ISO 3452-4 standards took into account of the AMS 2644E specification.

It is interesting to note however that the AFNOR NF EN IS0 3452-2 of May, 2000 standard did not include some requirements of AMS 2644E such as: thermal stability, high temperature titanium stress corrosion, high temperature corrosion of cast nickel alloys, etc.

Nevertheless considering the fact that the Americans did not take into account this standard ISO, it was decided that AFNOR NF EN ISO 3452-2 standard would almost duplicate AMS 2644E. That is the main reason the September, 2007 edition of the AFNOR NF EN ISO 3452-2.

VII- THE FUTURE

Nowadays, there are practically no Research & Development works on penetrant and

there were extremely few communications on the subject during the last National, European or International conferences: the situation again seems to have come to a standby.

Nevertheless, works of optimization should be pursued to improve: sensitivity of detection, penetrant washability along with overwash resistance and with detection reliability.

When the AMS 2644E is revised, it is desirable that it requires:

– Banning use of halogenated solvents in penetrant materials.

– A standard test method for resistance of penetrant to overwashing.

– Maximum allowable contents of: chlorine, fluorine, sodium, potassium and sodium impurities.

Finally, since the beginning of the 70s, very few tests and theorization were made to have an automatic reading equipment able to accept, reject parts or send them to repair. There are some prototypes, but operator’s expertise in deciding in a very short span of time, what is acceptable or not, that no penetrant process line comprises such a system! That’s why one should think Penetrant Testing lags behind in ability and reliability when compared to other NDT methods.

Keep in mind that Penetrant Testing was probably the first ‘‘nanotechnics’’ as only 25 to 75 nanograms of dye come out of the tinest cracks accepted on engine blades. This very small quantity is easily detected and processed by this very old equipment: human eye (sensors) + human brain (signal processor equipment) without any electronics attached, no computer bug risk.

The Specifications Which Changed