Japanese researcher Katsushi Owaki made the following illuminating comment at the Fuku BBS forum (translated):-
This is something that I posted on j-aircraft.com several years ago.
I polished the exterior of a piece of a Zero model 22 (Mitsubishi serial #3353) that I happened to have with a cloth. As a result, an inner paint that is shown on the bottom left appeared. From my past experience, I have no doubt that the color that appeared was 'I3' (the 3 indicated as a subset) of IJN Kariki 117. Of course, the top coat gray is still there.
It hasn't been so long since 'Ameiro' disappeared in comments abroad, and I'm afraid that there will be another uproar saying, "The Zero's color is I3!!!"
Researcher James F Lansdale maintains that the grey top coat is a matt "chalking" caused by exposure/weathering and may be rubbed through to reveal the original exterior colour of I3. Some of the artifacts in his collection which are said not to be weathered show an original glossy colour similar to the appearance of the I3 chip in the Kariki 117 document.
A difficulty arises because the nearest standard Munsell value for these artifacts, Munsell 5 Y 5/2, is actually very close to the nearest standard Munsell value for J3, Munsell 7.5 Y 5/2. The DE2000 difference between them is only 1.71 (2.0 or less = a close match). However when one compares the very precise fractional Munsell values of the chips from the original colour standards Kariki 117 and the 1945 8609 document the differences are much more pronounced with measurements of 10.7 (early J3 to I3) and 7.76 (late J3/2-6 to I3) respectively.
Given the known effects of age darkening and ambering of the original paint it is perhaps misleading to insist that the current appearance of the relics evidences a topcoat of I3 'Tsuchi iro' (earth or clay colour) rather than J3 'Hai iro' (ash colour), especially as the Zero technical manual refers to 'Hai nezumi iro' (grey rat colour) and the YoKu Report No.0266 of February 1942 describes the paint colour of current Zeros as "J3 (grey) (but) leaning slightly towards ameiro (amber or candy colour)". In other words the colour was a slightly brownish or warm grey. J3 changed to a warmer colour over time and the effects of age on the paint have exacerbated the similarity between the colours.
J3 has a little more green and may have originally been more grey in appearance. The use of plant resins as binder may have made the colour more amber over time and this appearance has been increased with age.
More images of the Type 97 Nakajima Ki-27 in Royal Thai Air Force Service. This camouflaged example has the red, white and blue rudder stripes combined with the white elephant wing insignia. See also 25 March 2008 posting for another image.
The model is by Gary Wenko, leader of the Japanese Aviation SIG of IPMS (UK) and is built from the Hasegawa (formerly Mania) kit in 1/48th scale.
In the instruction manual for the A6M2 Type 0 fighter, the Kaigun Kokuo Honbu (Naval Air Headquarters) specified the following painting procedure:
"Toryou wa kouzou naibu ni tomei toryou (tan ao iro) wo mochii gaimen wa keikinzokuyou tokushu toryou (hai nezumi iro) wo hodokoshi hyoumen wa migaki shiage nari."
Which translates as:
"The paints to be used are transparent paint (light blue colour) for the interior and special paint for light metals (grey rat colour) for the exterior, and the surface is to have a polished finish."
2,000 copies of this manual were published and distributed to the commands of all air related units for use by pilots and ground crew.
Noted Japanese researcher Katsushi Owaki advises that the early Zero was generally painted with a midcoat of colour I3 'Tsuchi iro' (earth or clay colour) and a topcoat of J3 'Hai iro' (ash or grey colour). He cites a Munsell colour value of 7.5 Y 5.5/1.5 for the J3 chip in the IJN colour standards document Kariki 117 and notes that this is identical to the later 2-6 chip in the 1945 8609 document which rationalised Army and Navy colours.
I have displayed a rendered chip of this colour. The closest standard Munsell colour value is 7.5 Y 5/2 @ 2.7 (too brown). The closest FS 595b comparison is 16350 @ 4.79 (too dark) and the closest RAL is 7034 Gelbgrau (Yellow grey) @ 3.33 (too light). In comparing the DE2000 differences a calculation of 2.0 less = a close match.
FS 16350 is a value often cited for the early Zero, based on the current appearance of extant artifacts, but it is a little too dark. A model version should be lightened.
Munsell 5 Y 5/2/FS 34201 is also cited for some extant Zero & Rufe artifacts and is very close to RAL 7002 'Olivgrau' (Olive grey) @ 1.37. Revell paint colour 32145 Helloliv (Light Olive) is supposed to be the equivalent to RAL 7003 but is closer in appearance to 7002. Again the colour is a little too dark for a model and should be lightened.
Mitsubishi and Nakajima built Zeros demonstrate variations in the colour applied which will be explored further on specific airframes. For modellers please bear in mind that all extant samples are probably darker and more yellowed (browner) than the originals, including the paint chips.
All these greys might be legitimately described as "grey rat colour".
With gratitude to IM for translation and appreciation to Ken Glass for assistance in calculating Munsell fractional notations.
Prolific Japanese researcher Summer recently posted this image of a Ki-43 Hayabusa in Thai markings that gives us, perhaps for the first time, a clear impression of the camouflage pattern. Back in 1970, Richard Ward's profile in the Aircam Aviation Series No.13 'Nakajima Ki.43 Hayabusa I-III in Japanese Army Air Force - RTAF - CAF - IPSF Service' (Osprey Publications Ltd) depicted the aircraft in a stylised "herringbone" pattern that has been copied in many representations, including models, ever since.
Revell's 1967 box art for the Thai issue of their Hayabusa kit also reflected this idea, although quite anachronistic for the Ki-43 I represented by the kit (of which more anon). The photograph shows a much more random pattern of at least two colours, sprayed at close range over the natural metal. Some depictions represent the glimpses of the natural metal surface as deliberately applied swirls of light grey but I don't believe that to be the case.
At the end of the war the camouflage was stripped from some Oscars and the pre-war Thai roundels and rudder stripes were applied. On these aircraft the fabric covered control surfaces remained in their original camouflage finish. When L A Manwaring of the Royal Corps of Signals visited Don Muang airfield in 1945 he described the Ki-43s as "Natural metal overall with Thai Air Force markings" although noting that "some of these fighters still wore their Japanese green finish". A photograph taken by Manwaring (and reproduced here) shows a Ki-43 II late production Hayabusa, apparently crudely stripped of paint to leave a dull, stained finish. The ailerons and elevators appear to retain their former camouflaged mottle. The entire rudder is painted in Thai colours and the aircraft has large Thai roundels on the upper surfaces of the wings. A small black (?) number 17 (or perhaps a Thai character) is painted on the fin. The Hayabusa retains its aerial mast which appears to be painted black and white.
Carl Molesworth, author of the latest book in the Osprey 'Duel' series 'P-40 Warhawk vs Ki-43 Oscar' and several other authoritative CBI related titles, has very kindly sent to me and given me permission to post here this revealing photograph of Ki-84 Hayate of the 9th Sentai, probably taken at Nanking, China at the end of the war. In the background may be seen Ki-44 Shoki aircraft possibly from the same unit.
For years the design of the 9th Sentai tail marking has been questioned, despite being depicted in profiles and artwork, because of the apparent absence of photographic evidence. According to Minoru Akimoto the marking was representative of the historical spear of warrior Kiyomasa Kato during Japan's invasion of Korea around 1596 and the character 'na' from the name of the Sentai's first commander Lt Col Seisaku Namba. A simpler explanation is that the marking represents the Kanji character for '9' (with thanks to Katsushi Owaki).
The distinctive dark stain on the undercarriage cover was caused by the lowest sets of the ejector exhausts each side of the cowling but has been incorrectly interpreted as a painted marking in some profiles.
This contemporary painting depicts the severely wounded Lieutenant Yokesuke Fukuyama of the Army's 2nd Hiko Daitai (Air Battalion) struggling to keep his Kawasaki Ki-10 biplane fighter in the air following combat with Chinese aeroplanes on 10th April 1938. Lieutenant Fukuyama managed to reach his base and crash land but unfortunately died of his wounds four days later.
The artist has depicted the aircraft in a very cold pale grey scheme with little hint of blue and/or green. Does it represent a silver-grey dope? In addition he has provided a tantalising glimpse of the interior colour, not the expected sandy brown associated with Kawasaki Hien cockpits but instead the Malachite green (Methuen 26B6) of earlier and popular Japanese references.
As artwork it may be tempting to dismiss these colours in preference to the tangible and seductive evidence of extant artifacts from later Kawasaki aircraft. However, this large painting was displayed alongside the actual remains of Lieutenant Fukuyama's Ki-10 (refer to the photograph on page 6 of Arawasi Magazine Issue #5). As such it seems unlikely that the artist would have depicted colours so far removed from the original. The colours he has used for the pilot's flight clothing and equipment appear absolutely correct.
Russian pilots encountering the Ki-10 over Nomonhan described the pale grey-green colour scheme as "almost white" in the bright sunlight.
A pair of Yokosuka 'Ginga' (銀河 - Milky Way or Galaxy) and a Zero in Republic of China Air Force (RoCAF) markings bask in the sun on a Formosan airfield circa 1948. Of note are the different shades of blue visible in the Kuomintang 'white sun on blue sky' insignia and the yellow and the white tail code on the Ginga, difficult to discern from b/w photographs - the Zero appears to have a yellow tail code. The aircraft out of view with a striped rudder may be a Ki-48 or possibly a Ki-45.
Gary Wenko, leader of the Japanese Aviation SIG (Special Interest Group) of IPMS UK has kindly sent me a photograph of his model of a 'Frances' in Chinese markings, built from the classic Revell/Takara kit.
Image credits: Bill Ward via Warren E Thompson; Gary Wenko
Dan Salamone very kindly sent me photographs of some of his models to show here. Here is a superb Zero night fighter, described by Dan as follows:-
"This is the 1/48 Hasegawa A6M5 night fighter, built pretty much from the box, with small improvements and additions. The kit instructions will have you build these aircraft, regardless of unit, as A6M5a, but close inspection of available photographs in publications such as Koku Fan #96 show that they were more than likely standard A6M5 variants.
This particular aircraft, from the 332nd Kokutai based at Iwakuni, had non-standard antenna rigging, and what I chose to depict was based on a best guess from the few available photos. The aircraft used by the 302nd Kokutai at Atsugi in this role also sported different equipment, more than likely due to the loss of the loop antenna behind the pilot with the installation of the oblique weapon.
The model was finished using Mr Color #15 for the upper surfaces (FS 14077), and a mix of Tamiya acrylics thinned with Mr Color lacquer thinner - XF-20 and XF-49 - to arrive at a variation of FS 24201 for the lower surfaces. Tamiya XF-25, lightened with white, was used for the fabric surfaces, to replicate FS 16314.
Nearly every A6M5 with an oblique cannon shows different metal work on the rear canopy, so close attention to detail on the specific aircraft you are building reaps benefits. The kit as molded is a good generalisation of what was seen on the various airframes."
Owaki-san very kindly sent me a splendid painting of a Ki-27 in RTAF camouflage and markings provided by Mr Tomoaki Shimamoto who had visited the RTAF Museum. In Thai service the aircraft was known as the Ki-27 Ota (after the location of the Nakajima factory) and Type 12 or Type 15.
The colour scheme appears to be a reddish brown and olive green over blue-grey.
Image credit: RTAF Museum via Tomoaki Shimamoto & Katsushi Owaki
In his report on Ki-27 colours at j-aircraft.com, James F Lansdale also referred to "a small piece of aircraft metal which was a close match to FS 16160 (Munsell 5 Y 5/4). This color is similar to the color called I3 or 'grey-poupon'". This artifact was associated with Frank Lawlor and may have been recovered from one of the 77th Sentai 'Nates' brought down over Burma, but the provenance is not definite.
FS 16160 and Munsell 5 Y 5/4 are not close; in fact the DE2000 difference calculation is a significant 7.49 (where 2.0 or less = a close match). The Munsell value 2.5 Y 4/4 is closer @ 4.40, but in reality neither may be considered good matches. FS 16160 is a colour that has also been cited for artifacts from Nakajima-built Zeros. It does indeed appear like the I3 from the IJN colour standards book known as 'Kariki 117' but it is also similar to the Army paint colour #30 'Karekusa Iro' (Parched or Dried Grass Colour).
It may be going too far to presume that an ochre colour like #30 was factory applied as an overall scheme on the Ki-27 but it certainly figured as one of the predominant colours available for the camouflaging of Army aircraft. After being on the receiving end of unwelcome attention from the AVG (the Flying Tigers) at its airfields in Thailand, the 77th began to hastily camouflage the upper surfaces of the wings and tailplanes of their Ki-27 aircraft, using at least two colours but possibly more. At least one aircraft was completely camouflaged. These ad hoc camouflage schemes have been depicted in various colours but it is not beyond the realms of possibility that #30, or an ochre/earth colour very much like it, was used.
Whatever the explanation, it is unlikely that a colour such as FS 16160 represents the factory applied 'Hairyokushoku' (Ash Green colour) of the Ki-27. Descriptions such as "olive gray" or "gray-poupon" are misleading terms for the straightforward ochre/earth colours specified in the Kariki 117 and JAAF colour standards as I3 and #30 respectively.
An indigo blue is a colour associated with several Japanese Army aeroplane types and indeed JAAF paint colour #28 'Sei Ran Shoku' (Indigo Blue Colour) is a dull, dark indigo blue. The manual specified that the camouflage paints to be used on the Ash Green base coat of the Type 97 were yellow brown, indigo blue and green. On other types the blue was sometimes applied as a thin dividing line between the green and brown segments but a photograph reveals that the 24th Sentai may have applied it in the manner shown above.
The scheme is usually depicted as two greens and a brown but we like this interpretation better!
In his report on Ki-27 colours at j-aircraft.com noted Japanese aircraft colour expert and relic collector James F Lansdale compared the appearance of the rudder from the 77th Sentai Ki-27 #842 brought down in Burma (see Part 5 below) as "a close match to FS 16314 in bright sunlight. However, the Munsell visual match by MIKESH inside the Museum rendered a better match to Munsell 5 B 5/1 (close to FS 16270)".
It is a little difficult to reconcile a visual measurement inside a museum with being a better match than one taken in daylight but in any case Munsell 5 B 5/1 is not close to FS 16270 - in fact the calculated difference between them is 4.93 where 2.0 or less = a close match. A closer match to 16270 is Munsell 2.5 B 5/0.5 @ 4.49, but not by much! Munsell 5 B 5/1 is actually much closer to FS 24237 @ 2.78.
A piece of rudder fabric presumed to be from #3218, another 77th Sentai Ki-27, "appeared to be a close match to FS 16307 (Munsell 7.5 Y 5/1)". The difference calculation between these quoted FS and Munsell values is a whopping 8.46 where 2.0 or less = a close match. FS 16307 is actually closer to Munsell 2.5 GY 6/1 @ 1.91, which is almost an identical match. These comparisons raise the question as to whether the quoted FS or Munsell values represent more closely the appearance of the original artifact, a common problem when multiple colour measurement systems are referenced.
Neither of the rudders appears to show the pale blue-green of JAAF #1 Hairyokushoku (Ash Green Colour) but are more towards warm or blue-greys. In 1972 South Africa's well known aviation artist Ron Belling conducted a series of experiments with colour, including the colour Sky (a mixture of white, yellow oxide and Prussian Blue pigments), using BSC 1964 cellulose paints. He found that after two months of exposure matt Sky was the first colour to change "and six months later it was very slightly more grey than the gloss sample which had remained unaltered. After sixteen months matt Sky had become a more blue-grey than the gloss because the Yellow had evidently faded out of the surface layer, having been broken down by ultra-violet light."
So it might be reasonably concluded that ultra-violet exposure may tend to fade the yellow pigment in grey greens, making them appear more blue-grey over time. Once a painted surface is removed from exposure long term storage may result in the "ambering" typical for cellulose paints, making it appear darker and browner.
Fellow Japanese aircraft enthusiast William Reece has very kindly sent me photographs of two of his beautiful aircraft models.
The Zero was made using the 1/48th scale Hasegawa kit and represents an A6M2 Model 11 of the 12th Naval Air Group in China during 1940. William painted the model using Polly Scale #414317 Concrete for the 'Olive Gray-Green' (following Jim Lansdale's artifact matches) and Tamiya XF-12 IJN Gray for the light gray. The kit decals were used.
The second model, a stunning Army Shoki of the 47th Hiko Sentai (Flying Regiment), was also built from a 1/48th scale Hasegawa kit. William finished the model in Floquil Bright Silver for the natural metal and Tamiya XF-14 IJA Gray for the fabric covered control surfaces.
Click on the images to see the full size versions. Comments are welcome.
Thanks very much, William, and I look forward to being able to post more photographs of your excellent models here. I hope that others will contribute theirs too - just send them via my contact details on the sidebar.
During the first months of 1942 a number of 50th and 77th Sentai 'Nates' were brought down over Allied airfields in Burma. Parts from these aircraft were souvenired by members of the AVG (the Flying Tigers) and taken back to the USA, where two rudders, one from aircraft #842 and one from aircraft #3218 still survive.
Colour matches for these extant artifacts have been published and will be examined in due course, but what is less well known is that some parts from these aircraft also found their way back to the UK. Many such artifacts, in private collections or small, provincial museums, are almost unknown and have never been published before, mainly because the focus of these collections is not about the colours of JAAF aircraft! The artifacts are incidental, in some cases wrongly attributed and often held in storage rather than on public display.
I compared the identified value range of JAAF #1 Hairyokushoku (Ash Green Colour) to Thorpe's identified Army colours and at first glance there seemed to be little correlation. The upper picture shows the #1 value range as a gradient in the top chip and below it the two closest Thorpe colours also rendered as a gradient. Both A4 Light Green and A5 Light Grey Green are described as being used for the upper and lower surfaces of Army aeroplanes.
When I calculated gradients of the #1 low end value to Thorpe's A5 and the #1 high end to Thorpe's A4 a different picture emerged. The gradiated chips are shown in the lower image. This suggested to me that the Thorpe colours are variations of the same colour and in turn variations of the standard Army #1 Hairyokushoku.
The photograph shows the very well preserved Hairyokushoku (ash green colour) paint on an Army aerial camera. This is a valuable example of Hairyokushoku because it has not been subject to extreme ultra-violet exposure, dirt or weathering. Adjacent to the photograph is the rendered chip for the median of JAAF paint colour #1 Hairyokushoku as calculated by Owaki-san from the original Hess Ives RGB values in the JAAF paint colour standard documents. Visually the colour of the camera lies midway between Munsell 10 GY 8/1 and 10 GY 8/2, the former being slightly too grey, the latter slightly too green. The closest FS 595b values are to the right of each chip respectively, marked with the DE2000 difference calculation in parantheses. In each case the FS value is within the score of 2.0 needed for a close match.
As published the photograph appears slightly more grey and not quite as green as the original. It is almost identical to Owaki-san's chip. The appearance of the colour is similar to the RAF Sky Type 'S' but paler and brighter. Many references describe the paint as "blueish" and certainly at the low end of the range the colour has a definite blue-green quality, almost like a very pale turquoise. As a result of the pigment mixes, as with Sky, variations would be towards blue, towards green and towards grey. The warm brownish caste seen on some extant examples is an aberration from standard not yet fully explained but possibly due to age discoloration or the use of plant resins as binder.
A consensus is emerging that the Army Hairyokushoku was more blueish than the Navy version of this colour, the latter appearing more yellowish-green.
Taking the Munsell values as the low and high ends of the tolerable range I compared them to the Army colours identified by Thorpe. At first there appeared to be no exactly similar colours in the Thorpe chart. (To be continued).
In a posting at the Fuku BBS A6M232 demonstrates the change in appearance caused by flash photography.
The photograph shows a piece of fabric from a Toryu (Ki-45) wing laid on a piece from a Type 97. A flash has been used for the picture on the left which makes the fabric look more blueish and the 97 surface more greenish. Without the flash the pieces are closer in appearance.
All illuminants, even various forms of daylight, will affect the appearance of colour. Most colorimeters require the illuminant to be set so the readings can be made against various forms of light and will therefore vary accordingly.
Another example of Ki-27 colour comes from an extant rudder in the collection of the Hiko Jinja (Flight Shrine) in Yahata City, Japan. Owaki-san has very kindly provided a photograph and the measured Munsell value of 7.5 Y 7.5/1.5
The closest FS595b value is 26559 @ 1.62 which is almost identical (2.0 or less = a close match) whilst the nearest RAL value is 7032 Kieselgrau (Pebble Grey) @ 3.20
Owaki-san notes that this example is doubtful because the kanji characters painted on the rudder refer to a Navy unit at Otsu air base on Lake Biwa north of Kyoto. This base was associated with Navy seaplane training - maybe the rudder was from an Army Type 97 fighter- trainer that crashed or force-landed there?
In 1970 Dr Izawa commented "Some Type 97 fighter-trainers that had their spats removed were painted in a cream-gray".
Japanese researcher Owaki-san has very kindly provided a measured Munsell value from another extant Ki-27 flap currently in the collection of Mr Kenzo Taniguchi in Japan.
The colour is a distinctly blueish grey. The closest FS595b match to Munsell 5 BG 6.5/1 is 36473 with a DE2000 difference calculation of 2.17 (2.0 or less = a close match) so it is a reasonable comparison.
Owaki-san also provided the Munsell measurement for the interior colour of this particular example which I shall post in due course.
Japanese researcher A6M232 has posted images of the Ki-27 flap in his possession at Fuku's Forum.
The blueish-grey and slightly greenish appearance of the paint (close up) is apparent. In the second photograph pages from the Japanese Paint Manufacturers Association (JPMA) colour standards are being held against the artifact. The JPMA deck cross-references each colour to the Munsell standard.
Dr Yasuho Izawa described the finish as light grey overall "this was the factory finish and, in my opinion, it is light grey blue plus a drop of yellow".
In recent years there has been pioneering colour research in both the USA and Japan focussed on extant painted artifacts from Japanese aircraft. In the West these came to prominence in the late and much lamented 'Asahi Journal' with fascinating revelations about the true colours of the famous Zero fighter by James F Lansdale.
At first the evidence from these relics was somewhat disembodied. In some cases the provenance of the pieces was not precise enough to draw definitive conclusions about the appearance of the colours. Were they really from the aircraft in question? Had they been re-painted, repaired, affected by extremes of temperature or was the chain of evidence from the time of their recovery continuous? More recently research in Japan has revealed primary documents, aircraft manuals, paint standards and reports, that allow some of the artifacts to be placed in context and mapped to specific paint and colour designations. The picture is by no means complete but the situation is now far advanced from that existing only 8 years ago.
Underpinning all this is the pioneering work of Donald W Thorpe, whose two seminal books 'Japanese Army Air Force Camouflage and Markings World War II' and 'Japanese Naval Air Force Camouflage and Markings World War II', published in 1968 and 1977, set out to record and identify the actual colours in use by the Japanese Air Forces. Recent studies have reinforced the quality of Don's work and his books remain essential aquisitions for anyone interested in the subject. Although long out of print they may still be obtained through the used book market.
Writing in 1980 Don recorded he had "personally examined several hundred samples taken from wrecks throughout the Pacific, the majority of them catalogued as to type, unit and serial number."
The main problem with the books was that the colour charts included in them were printed and the Munsell values which had been matched to the colours were not quoted. In 2005 the late Bill Leyh and myself prepared a table showing all the Thorpe colours, their assigned Munsell values, sRGB values and the closest FS 595b equivalents. This table is still available in pdf format. When we examined some of the colours we were struck by how closely they matched both recently identified artifacts and the documents then emerging from Japan.