Curtiss AT-9, Wlliams Field, Arizona, Early Summer, 1942

[Old Man]

 

The Curtiss AT-9 was designed to be difficult to fly and land, or perhaps more precisely, it was designed to mimic the handling characteristics of much larger, heavier, and speedier airplanes. The 'AT' in AT-9 stood for 'advanced trainer', and the aircraft was intended to teach cadets fresh off of primary trainers how to fly high performance twin-engine aircraft, which by 1940 were set to become a large portion of Army Air Force combat equipment.

 

 

The AT-9 was a product of Curtiss-Wright's Airplane Division at St. Louis. Originally the Curtiss-Robertson company, after Curtiss and Wright merged in 1929 it became the new firm's civil aviation shop.  In 1934, while the original Curtiss facility at Buffalo was designing the Hawk 75 pursuit ship, the St. Louis designers were building the CW-19 Coupe. Though this was a two seat aircraft meant for civil customers, with an engine of just 90 hp, the CW-19 was a low-wing cantilever monoplane of all metal stressed-skin construction, and was quite capable of further development with more powerful motors. The lines of its fuselage, and its wing-plan, can clearly be seen in subsequent single-engine designs out of St. Louis, such as the CW-21 interceptor, with its 900hp Cyclone engine, and the CW-22 Falcon trainer.

 

 

 

 

When the design team at St. Louis, led by George Page, set out in 1940 to create a twin-engine transition trainer for the Army (under the company designation CW-25), what they produced was in essence a CW-19 scaled up and altered to accommodate two engines instead of one. The prototype, which first flew early in 1941, had a squared-off fuselage of steel tube and fabric, owing to concerns over whether the supply of aluminum would be sufficient for both combat and training planes, but when the Army accepted the design, this was discarded in favor of the modern all metal construction employed by the Airplane Division ever since the ancestral CW-19. The Army initially ordered 150 AT-9 aircraft, which were among the first items produced at a mammoth new facility being built for the St. Louis branch of Curtiss-Wright with government assistance. These aircraft reached training units early in 1942, and were followed by another 341, plus an additional 300 examples of a slightly modified version, designated AT-9A, ordered in 1942.
 

 

The AT-9 had a wing loading similar to that of the CW-21 interceptor, but its engines provided only two thirds the horsepower. This relative lack of engine power was at the root of most of the AT-9's 'tricky' flight characteristics. Poor stability in pitch and yaw owed something to this, as the aircraft just was not moving fast enough to give the stabilizing surfaces all the bite they needed, and so did the AT-9's proneness to sudden stalls, for even at its fastest, an AT-9 was not going much faster than its stall speed. A steep descent was required to keep up speed when landing, and pulling up the nose to do a 'three point' landing could easily result in a stall and a wreck. On the other hand, the AT-9 was fully aerobatic, and there were some pilots, cadets as well as instructors, who liked the type immensely, describing it as a 'hot-rod', and a real treat to fly. The AT-9 was certainly a fast and responsive machine by compare to basic trainers, and to other twin engined advanced trainers serving alongside it, such as Beechcraft's AT-10 and Cessna's AT-17. There is room to wonder how the AT-9 might have performed and handled had it been fitted with a pair of 450hp Wasp Juniors, rather than a pair of 295hp Lycomings.

 

 

This model represents AT-9 41-5891 early in its service at Williams Field in Arizona. This aircraft was one of the very last of the first production batch of 150, and still bears the factory applied national markings, in a style which lingered in training units for some months after removal of red centers and stripes was ordered. The letter 'Y' on the fuselage indicated Williams Field in the sometimes opaque coding system employed by Army training commands, introduced around May of 1942. An in-flight photograph shows this machine with the fuselage codes, but with its cowlings in bare metal, and with no anti-glare panel.

 

 

 

Williams Field during WWII became a center for training pilots to fly the P-38 Lightning. Construction of the airfield began in July, 1941, and proceeded under the initial name of Higley Field, after a tiny town nearby (which may account for the 'Y' field code, as the southwest training command often assigned a field a letter taken from the name of an adjacent  town). The first group of instructors and ground crew arrived at the new facility on December 4, mere days before the Japanese attack on Pearl Harbor. Training operations commenced in February, 1942, by which time the Williams name was officially in place, memorializing an Army pilot born in Arizona who had died in a crash in Hawaii in 1927.

 

 

The initial twin-engine complement at Williams Field was a mix of Beechcraft AT-10 and Curtiss AT-9 aircraft. Within a few weeks it became clear that, whatever difficulties might be encountered with the AT-9, those besetting the Beechcraft twin here were far worse --- the dry desert heat of south central Arizona wreaked havoc on the plywood structures and glues of the AT-10, resulting in a rash of crashes which killed ten men.  The Beechcraft twins were packed off to more humid stations, and the Curtiss AT-9 became the predominant twin-engined type at Williams, since the only alternative, the Cessna AT-17, was deemed unsuitable for training fighter pilots.

 

 

By the start of 1943, there were nearly 200 AT-9s assigned to Williams Field, fully a quarter of the whole number built. Cadets put in a seventy hour course on the AT-9 to become qualified for the Lightning. Accidents, it may be safely said, were frequent. One training command in Mississippi, operating 40 AT-9s, recorded 47 accidents involving them in an eighteen month period, with 10 of these wrecking the plane completely, and four causing fatalities. One fairly common cause of fatal crashes was low-level flying, a necessary part of training, but something instructors dreaded because even slightly mishandling the controls of an AT-9 could produce a sudden stall, with no room at all for recovery. Engine failure could lead to fatal crashes as well, for the AT-9 could not maintain itself in the air on one engine, and often it was difficult to open the doors against a stiff slipstream.

 

 

Early in 1944, a trainer version of the Lightning became available, and the role of the AT-9 at Williams was accordingly reduced, with cadets doing their solo work on the Lightning trainers. The subject of this model, 41-5891, was out of the picture by then. It was wrecked in a crash due to engine failure on May 27, 1943, near Casa Grande, where an auxiliary runway for Williams was located. It was piloted by one John M. Proos, who survived by taking to his parachute. He appears to have been born in 1921 in Michigan, married a young woman while in Arizona, fathered a son born there in 1944, and then gone on to graduate from Notre Dame in 1950. It would be a decent bet that he met the young woman at the USO in Chandler, the nearest town of any size to Williams Field at the time, and the chief resort for men on liberty from the base to socialize.

 

 

 

The model is constructed from the old Pavla limited run 1/72 kit. There has been much improvement in the Czech kits, and people have grown less forgiving of the earlier offerings. In its day, this would have been considered pretty decent for a limited run kit. It certainly requires a bit of work. Two things are worth noting particularly. First, unless some work is done on the mating surfaces, the wings will have almost no taper in thickness from the nacelles out to the wingtips. Second, the cowlings should not be straight cylinders, but taper back to the nacelles. Since the cowlings are very thick, and thus the nicely formed resin motors cannot fit inside them, I shimmed the cowlings to increase their diameter a bit, which let the motors fit, and let me file and sand in a good taper to blend the cowlings back to the nacelles. I scratch-built an interior, though I used the kit's photo-etched instrument panel front. On-line reviews of this kit call the panel too wide, but actually, the instruction sheet illustrations, and the kit's placement of the crew's seats, put the panel too far forward. Put where it ought to go, it fits well. To get it there, the seats have to be put where they actually were, right back against the rear bulkhead, not a couple of millimeters before it, as the kit pieces and instructions would have you do. The kit has nice resin blades, but the hub pieces provided are useless, and I made my own.
 

 

 

I did not set out to make a vignette with this model. I will say that I am not partial to opened doors and hatches as a general thing. I certainly respect the work many people do to show normally hidden mechanisms, but when the plane is simply there with its innards exposed it just doesn't look natural to me. The sight calls out for some guy in overalls reaching in with a wrench, or at the very least nearby taking a smoke, to my mind. On this kit, the window frames of the cockpit's 'car doors' are moulded on the fuselage halves, while the canopy is a vacu-form piece. Instructions suggest trimming off the door frame and simply placing the canopy, but this would put a seam in a prominent place that I doubt I could clean up properly for a bare metal finish. Nor did cutting the canopy to fit the outer curves of the frames, and cutting out window pieces to place and fit into the door window openings on the fuselage, seem a promising course. So I cut the doors out whole, filled their window frames with clear sheet, then cut the door area out of the canopy, and posed the doors open. By my own lights, I then had no choice but incorporating some people. The figures are from a Preisser set of civilian air and ground crew, and passengers, of the 1930s. All have been greatly altered, and parachutes and headphones added to the instructor (standing on the wing) and to the student (seated in the door). I have seen a couple of photographs showing men going to a flightline of AT-9s, and they did not have on helmets, just caps, and prominent headphones. What the senior officer on the tarmac is doing there I have no idea: anything from offering congratulations on some crackerjack flying just completed, to enquiring whether the cadet really thinks he's going into the air with those unpolished shoes, is possible....

 

 

The model is finished in kitchen foil, dulled a bit by boiling in water that eggshells have been boiled in previously, and affixed with MicroScale foil adhesive. National markings are from the kit's decals. The rudder striping decals were much too small, necessitating more touching up than is decent. The kit's 'Army' decal disintegrated after placement when handled, something I have never seen before. I found a suitable replacement in the spares. Wife made the decals for the field code alphanumeric, and the serial run.

 

 

 

 

Edited August 7, 2020 by Old Man

[Cookenbacher]

Wow, that's a great metal finish and a very interesting subject, thanks for the historical background too.

[SAT69]

An incredible model and the background information is fascinating. 

[RMCS]

Superb 

[RidgeRunner]

Superb! I was searching for the scale, expecting 1/48 at least! 

 

Martin

[bigbadbadge]

This is absolutely stunning, great work on the kit and what a lovely metal foil finish which certainly gives it a massive Wow factor.  Great background information and a cracking base and figures .  

Fantastic work.

All the best

Chris

[Wulfman]

Excellent build and accompanying information, very interesting !

 

Wulfman

[Old Man]

Thank you, gentlemen, for the kind words.

 

I am of the 'nothing looks so like metal as metal' school when it comes to doing a bare metal finish.

 

Here is the picture of the actual machine....

 

 

It was a bit of luck the name of the pilot who eventually crashed it was an unusual one. If it had been Pruit or Peters or something like that, such clear traces could not have been readily found.

 

Here is a link to an excellent site with a lot of information on U.S. Army trainers and bases at the time.

 

http://www.fuselagecodes.com/index.html

[russ c]

Superb. Fantastic build and finish

[Old Man]

Thank you, Russ.

 

Foiling is easier than people seem to think. The very front was tricky, and took a while to figure. I did a wrap, finally, that left a the tip bare, and covered that with a small circle.

[stevehnz]

A brilliant piece of work in every sense of the word, there is a lot to like about it, not least that great back ground. Thanks.

Steve.

[Old Man]

Thank you, Sir.

 

Under the lights especially it is shiny.

 

I encountered this plane when quite young, reading war-time Life Magazine's in the school library. Curtiss featured them in an advertisement, three stacked up in close echelon. I thought it looked awful neat. Still do.