Tuesday 24 July 2018

Tom Hall's Comment & Associated Ideas ~ A Response

Here is the response to Tom's interesting comment about the perrenial subject of the early Zero colour.

TH: Nick, won’t you comment on the article in Gakken’s Rekishi Gunzō for February 2018? 

I have not read it yet, although I have been made aware of those new ideas from various other sources. (Ed. I've since read it)

TH: The article ostensibly states that the paint of a Mitsubishi-built Zero was tested at a fairly prestigious lab for cultural properties in Japan. That lab found only two pigments in the top layer of paint: white and black. (This test was just for a Mitsubishi-built machine.)

Samples from just one machine tested? And if black and white pigments were identified did the article state what they were? There are several types of black and white pigments which can result in significant differences in the resultant grey colour. Did the lab reveal the nature of the paint binder? I understand (and I might be wrong) that the paint sample(s) examined are from A6M2  Houkoku-515 which had lain exposed in the New Guinea environment for 66 years. If so the residual components of paint on the airframe are unlikely to represent the original and are possibly contaminated by other intrusive elements. The absence of evidence of certain pigments (or tinters) might not be evidence of their absence. Even with Raman microspectroscopy the results are not always absolutely conclusive. 

TH: White and black in the right proportions can make J3. 

So you say but how do you know? The swatches of J3 in the Kariki 117 document and the Kugiho 0266 report do not appear as pure neutral grey. It is quite possible that the swatches have yellowed with age producing their current appearance and that originally J3 was indeed a pure neutral grey. However, GSI (Gunze) Mr Color 35 (IJN Gray Mitsubishi) as measured is not a pure neutral grey. Without knowledge of the official formula for J3 it would be bold to assert that black and white in the right proportions can make J3!

TH: However, to get a shade of khaki, you need additional pigments; or there must be damage to the paint that changes it from a neutral grey to a shade of khaki; or you need different lighting; or there must be some combination of the three. 

Khaki is your word not mine. I would prefer amber grey - a grey paint with a slight amber or yellowish tint as well as a fugitive olive undertone. That seems more consistent with the Kugiho 0266 report which describes a tinted grey. Your brief summary of how to get khaki and what can happen to paint and the pigments in it is overly simplistic. May I refer you to 'The Chemistry and Physics of Coatings' by A R Marrion, published by the Royal Society of Chemistry in several editions. That will provide a better introduction to and understanding of the multiple factor complexities of the subject.  There you will also learn about sacrificial anode pigments and tinting dye-like pigments, amongst many other fascinating scientific facts about paint technology, rather than ideas about it where enthusiasm to find something new might romp ahead of established science. . 

TH: Mr. Taizō Nakamura has demonstrated that lighting with a flash for a camera can produce the sort of gold tones that would qualify as “ameiro”. 

How the paint surface might look under different forms of illumination has no bearing on the constituent polymer of the paint or its colour values when measured. Rather the reverse is true - the polymer will determine how metamerismic the surface appears. And I do not believe that the Kugiho 0266 authors were making flash photographs of the paint inside their hangars in order to describe its appearance. But please read on.

TH: So, when the Japanese naval officer said in a footnote that the Zerosen then in use was J3 and had some ameiro, he was exactly right. He was referring to an unusual finish that exhibits a lot of metamerism.

Yes, well, I have no disagreement with the paint being prone to metamerism - it very much is - but there is more to the naval officer's description than simply a 'footnote'. When you commented on the translation of that 'key passage' in 2005 you agreed that the meaning was that the paint 'has somewhat of an ameiro tint. In other words, a warm gray as opposed to a cold one'. You also asserted that the 'genyo' (sic) in the statement referred to the Zero rather than the paint. 

In fact there are other references throughout the report which make it clear that 'genyo' refers to the paint and that it was not simply J3. The 'genyo' paint is referred to as 'ameiro' throughout the rest of the report. In the report's recommendations the Head of Kugiho states (Exhibit A):-

'At present, my opinion is that we may continue using the currently used ameiro'

And it appears distinct from J3. One of the test aircraft - Yo-105 - was specially painted with J3 for the trials, described as hai iro (ash colour - grey) and another - Yo-107 - painted with J3 and D1 (Deep Green Black colour) in patches. The J3 chip in the report shows a slightly greenish-grey appearance similar to the J3 chip in Kariki 117. The methodology used in painting all five trials aircraft is described in the report as follows (Exhibit B):-

'The paint is camouflage paint: Standard Chi-57A, Type # l.  The paint was applied directly over the currently used ameiro.  The film of the paint was rendered matte ['frosted'] and was produced by adding talc or clay powder without making the paint body coarse.'

Exhibit C:-

"If we expect to be in a superior position or battles to be advantageous for our side, we may continue to apply the currently used ameiro in order make it easy to distinguish our side."

Please note 'continue to apply' suggesting the paint was ameiro in appearance to begin with and not some colour shift later.

Exhibit D:-

"Blue Ash-Gray (J2)/3/ It’s little more difficult to notice than the currently used amiero."

Exhibit E:-

"Stationing the five camouflaged aircraft, as mentioned above, two aircraft covered by the camouflage cloth, and a currently used ameiro aircraft on the concrete of the airfield or on the dry grass, we compared the degree of difficulty of spotting each by inspecting from a height of 1000 meters as well as 2000 meters."

Exhibit F:-

"Having the five experimentally camouflaged aircraft listed above as well as the currently used amiero airplane fly side by side, we compared the visual effects of the camouflage from both smaller and larger aircraft."

Exhibit G:-

"Having the experimentally [painted] aircraft, that were rendered matte thereby resulting in their having a rough surface, and the currently used ameiro aircraft that had a smooth surface, fly under equal conditions, we studied the effect of the rough surface paint and examined the difference of their speed."

The other aircraft specially painted for the trials were Yo-109 (D2), Yo-113 (J2) and Yo-151 (M1). There were four other aircraft used in the trials Yo-102, Yo-108, Yo-110 and Yo-112 referred to as being in currently used ameiro and not specially re-painted. In other words the test paints, including J3 and J2 (described in the report as blueish grey), were not the existing factory-finish colour of the A6M2 aircraft being used in the camouflage trials. They were paints applied over the existing amber grey finish. The report is dated February 1942 and I do not believe that the authors were referring simply to the effects of metamerism (which would vary anyway) when they consistently describe the paint as ameiro.  Occam's Razor tiptoes into the room at this point, hopefully not to become an elephant.

I've recently had correspondence from a Japanese interested party who went to extraordinary lengths to try to prove that ameiro as referenced in the report meant a blue grey colour tinted to become light blue green as a result of adding clear yellow. He also asserted that the amber grey explanation in Japan is not popular - which doesn't make it wrong.  It seems that the starting point for some is to want to believe in an absolute grey paint as a preference with the evidence being marshalled to try to prove that, rather than an objective analysis of all the evidence to draw a reasonable conclusion. The former endeavour involves either disregarding the Kugiho 0266 report entirely or attempting to distort its rather obvious meaning to mean something else.

The report's description is also consistent with the appearance of extant paint samples from several different aircraft as examined by different researchers in the USA, Japan, Australia, Canada and the UK. Mr Nakamura himself has shown extant paint samples of his own juxtaposed in comparison to the JPMA colour deck chips C27-50D and C27-60D. Those two chips are also matched to Munsell colour values 7.5 Y 5/2 and 7.5 Y 6/2 - Munsell Y = Yellow - and neither neutral grey nor grey-green. 

Most of the samples physically examined by those various researchers were originally collected and stored more or less contemporaneously with the particular aircraft's demise. As such they have not been exposed to UV light or the resultant photo-chemical activity and retain much of the paint's original integrity. The binder in six A6M2 and A6M3 examples, analysed in two other fairly prestigious laboratories, is consistently identified as a plant-type resin modified with styrene. The resin is made from the sap of trees now called Toxicodendron vernicifluum, also known as Chinese laquer tree, Japanese lacquer tree or Varnish tree. The old name for it is Rhus verniciflua. The 'urushi' type varnish made from the sap is naturally amber to dark brown in untinted form and will darken and brown further with age without UV exposure. It would have an immediate tinting effect on a paint containing a predominant quantity of white pigment.  The combination with black and yellow-brown anti-corrosive pigments creates the fugitive olive undertone but there are multiple influences going on there as the paint ages in storage. The styrene modified resin binder forms a thin film which air cures and cross links to create a very hard and durable surface as new, almost like a ceramic. But the majority anatase form of Titanium Dioxide (white) pigment in the paint compromises that durability because it is so prone to chalking. The paint also incorporates magnesium silicate (talc), an extender with reinforcing properties which improves the impermeability of the paint film (but also exacerbates chalking). With exposure to UV light the amber tone would gradually bleach out as the top surface of the paint absorbed radiation, the amber tinting preserved only in the underlying and chalking protected paint strata where it would gradually become slightly darker and browner. Was that process deliberately intended? Dunno! But the paint is by no means crude in its composition.

The dope on the fabric components of the Zero is cellulose acetate butyrate which is clear and does not readily yellow with age. It also contains triphenyl phosphate which is a plasticiser and fire retardant. There would be no need for anti-corrosive protection of light metal in the fabric dope. That is one explanation for the different tones of rudder, ailerons and elevators apparent in some monochrome photographs, also suggesting that the amber appearance of the metal paint is not simply age related degradation.

Also, Jim Lansdale has metal artifacts painted in J3 which show no sign of becoming amber or olive grey in appearance.  

It is quite possible that some aircraft were painted or re-painted plain ol' J3 from stores on repair or overhaul by Japanese naval units, especially perhaps in training units, whereas Mitsubishi applied a more complex anti-corrosive factory finish. In 2008 an artifact from s/n 3647 said to be from V-103 in Saburo Sakai's own collection was compared to Munsell N 5.5 (N = Neutral).  That's quite a dark grey - not far off FS 36251! Did Sakai do something to it? And Owaki-san once presented a piece from A6M2 W1-187 which he compared as close to FS 16314. When taken together with all the other artifacts they are the exceptions rather than the rule.  

By the way, have you raised your interesting neutral grey suggestion with David Aiken or were you being 'target specific'? ;-) David promotes a distinctive and rather bright light grey-green for the Zero, not far off RAF Sky, which is rather harder to reconcile with both your neutral grey idea and the tangible scientific facts accumulated from many extant samples.  Maybe it's that metamerism again, an idea which appears as a convenient fig leaf in order to promote any colour desired.

Post-Blog Note. Since writing this response I have received a translated summary of the Rekishi Gunzō article referred to by Tom. Contrary to his suggestion of paint samples from a Mitsubishi-built Zero being analysed, it reports analysis of the 2-6 paint chip from the 8609 standard (the successor colour to J3) by Dr Hashimoto of the Tokyo Cultural Properties Research institute, with a presumption that J3 was the paint colour applied to the Zero. In fact no black pigments were identified in the analysis and the other chemical elements identified included more than just white pigment. The relevance of the 1945 2-6 chip to the early A6M2 paint finish is moot, although it has been reported as identical to J3, for the reasons described above.  

Disclaimer: Nothing in this blog article is intended to prevent or discourage anyone painting Zero models or illustrations in any colour they might wish to. 

References: Yokosuka Kaigun Kokutai (KuGiHo) Report No.0266 Revised Translation November 2010 via James F Lansdale
CAL# 5370 Conservation Analytical Laboratory Report on A6M3 paint, Smithsonian Institution, May 1992  
Molecular Microspectroscopy Laboratory Report, Department of Chemistry and Biochemistry, Miami University, January 2006. Paint samples tested from A6M2 Lt Iida PH, A6M2 Lt Iwama Midway, A6M2 PO1/c Shigenori PH, A6M3 c/n 3023 Buna, A6M3 c/n 3268 Lae, A6M5 USS Lexington (Kamikaze) & B5N2 F1/c Kitahara PH  
Painting the Early Zero-Sen by Nicholas Millman, self-published 2011.
A6M Zero Painting Guide by Ryan Toews, unpublished thesis, 2011.
Paint samples loaned to author for analysis and measurement by James F Lansdale, Bob Alford and Greg Springer. 
Degradation models © 2018 Aviation of Japan




WilliamR said...


Thank you for the excellent analysis of this important and contentious subject. I found the last part regarding the pigments found to be especially interesting.

William Reece

Ken Glass said...

Thanks for the detailed review, Nick.