Toronto Aviation History - Authors

deHavilland Canada’s STOL Aircraft – “The Beginnings”

A paper by George Georgas

About the author:
George was born in Lindsay, Ontario attending public and high school there. He later graduated from the Aeronautical Engineering School, University of Toronto. While attending U. of T. George joined the RCAF University Reserve Training Plan and trained as a pilot on the Texan, Harvard, T-33 and Vampire.

After receiving his wings and the Queen’s commission as a Pilot Officer he joined 400 City of Toronto Aux. Sqdn. where he flew Harvards, T-33’s, Expeditors and DHC Otters over the next 22 years. George was C.O. of 400 Sqdn. from 1957 to 1970 and was posted to No. 2 Air Reserve Wing, Downsview, which he commanded from 1973 to 1976, when he retired from the Air Reserve.

During 1955 George was hired by deHavilland Canada where he eventually worked for 35 years. He was employed in the Engineering Design Department on the Caribou and in the Engineering Experimental Flight Test Dept. on the Caribou, Turbo-Beaver, Buffalo and a variety of experimental projects including the FHE Hydrofoil. George later joined the Marketing and Sales Department.

“The Beginnings”

STOL has had a flexible definition over the years. The DHC-2 Beaver was conceived as a rugged, easily maintained bush plane that could land and take-off in confined areas, particularly lakes, both in summer and winter (Fig. 1). As with all of the subsequent DHC aircraft types, it was designed to meet and operate under one or another set of established national civil air worthiness regulations. When the prototype DHC-2 Beaver, CF-FHB-X first flew, on the 16th of August 1947, the term STOL hadn’t yet been invented! In the case of the Beaver, the regulations were the British Civil Airworthiness Regulations, Section K, Normal Category. Initially, its gross weight was 4,500 lbs. but this was eventually increased by 13 percent to 5,100 lbs.

At its original weight the Beaver landplane has a take-off distance to a 50 foot screen height (theoretical tree or obstruction, etc.) of 925 feet. It’s landing distance is 1087 feet over a 50 foot screen height. All of this is at sea level, in no wind, on a standard day (that is 15 Degrees C.). To reduce the landing distance from 1087 feet to the magical 1000 feet, the weight has to be reduced by 550 lbs, to just under 4000 lbs. This performance is achieved using the British Civil Airworthiness Regulation speeds. Those speeds provide for a margin of 20 percent above the power-off stall speed for the take-off, and 30 percent above the power-off stall speed for the landing, when the aircraft is crossing the 50 foot screen height.

In due course, the term STOL (for Short Take-Off and Landing) came into use. Initially it was associated with an airfield length of 1000 feet over a 50 foot screen (Fig. 2). During the next twenty years, until the 1970’s, this definition crept up to 1500 feet as the Otter, Caribou, Twin Otter and Buffalo were developed, and their gross weights for STOL operation increased. In order to achieve this performance, the operator of these aircraft needs the approval of the appropriate civil or military airworthiness authority to use STOL technique. For the Caribou and Buffalo, the operator also has to abide by certain gross weight constraints.

What is STOL technique? deHavilland’s view rationalizes three important factors: flap settings, airspeed and consequences of an engine failure. Briefly, and depending on the particular aircraft model, STOL technique entails the use of a combination of increased flap deflection (beyond normal certification limits), and reduced landing and take-off speed margins above the power-off stall. Typically, for landing, full flap is used, the approach speed is reduced from 30 percent above the power-off stall speed to 15 to 20 percent above the power-off stall, and reverse thrust, if fitted, is used during the ground roll. For take-off, an additional 10 to 20 degrees of flap is used, the lift-off speed is at about 5 percent above the power-off stall, and initial climb is at 10 to 15 percent instead of 20 percent above the stall.

For the twin-engined aircraft, engine-out minimum control speed is not taken into account until the aircraft is established in the climb. STOL technique entails the acceptance that, in the event of an engine failure, the single-engined aircraft will be force landed straight ahead, under control (that is, not in a stalled condition), and at a relatively low speed, at which speed the occupants have a good chance for survival. For the twin-engined aircraft, there is a time interval during the take-off, of the order of 15 to 30 seconds after lift-off, during which, if an engine fails, the pilot has no option but to reduce power on the live engine to maintain directional control, and force landed more or less straight ahead, again under control and at a relatively low and survivable speed. In the absence of an engine failure during the lift-off and initial climb, the aircraft is promptly cleaned-up, minimum control speed is attained, and normal two-engined climb procedures, with single-engined safety apply. The gross weight constraints to which I referred earlier simply entail the limits for maximum take-off and landing weight, for some models of aircraft, when using STOL technique. For example, the Buffalo is limited to 41,000 lbs. for take-off and 39,100 lbs. for landing when using STOL technique, but can go up by 20 percent to 49,200 lbs. for take-off and 46,900 lbs. for landing when using the civil airworthiness technique.

Now back to the capricious definition of STOL. With the advent of the deHavilland Dash 7, an aircraft aimed at the commercial airline passenger market, it became expedient, at least in the Marketing and Sales Department of deHavilland, to speak of a 2000 foot field length as being STOL. However, this time STOL had to fully comply with civil airworthiness regulations. This meant providing for accelerate/stop,
continued take-off on 3 engines, and landing with the statutory field-length factor of 1.66, without reliance on reverse thrust. All of this was coupled to a 200 nautical mile stage length, with a 100 nautical mile alternate, and a load of 48 passengers.

The Dash 7 met the design requirement, and from then on, deHavilland stopped defining STOL.

So far, I’ve discussed STOL in the narrow context of field length; i.e. 1000 feet, 1500 feet, and 2000 feet. But to me, there are some other features in deHavilland products that are important if not vital for operation in some of the god-awful places that a STOL flying truck bush-plane can take you, and from which you may have to operate for extended periods of time, or get out of, with little or no outside support (which is nowadays called infrastructure). These features include rugged shock absorbing and energy dissipating undercarriage. The Buffalo main gear, when operating in the STOL mode, is designed to absorb a sink rate at touch-down of just over 12 feet per second as compared with 10 feet per second for a conventional undercarriage. The additional 2 and a bit per second may not sound like much, but it represents a 50 percent increase in landing loads and energy absorption. Furthermore, the gear is designed to cope with higher drag loads arising from landing and taxiing on rutted and soft surfaces (Fig. 3). Also, the nose gear and surrounding structure are designed to take the side-load generated by take-off power on one engine, if and when the nose-wheel drops into a pot-hole while taxiing on unprepared ground.

deHavilland’s STOL products include an overhead escape hatch, in case the aircraft goes through the ice on a frozen lake. The Beaver and Otter can be hand-crank started if the battery is dead. I'm told that the Beaver is the first aircraft in its class, anywhere, to have all metal control surfaces, thus eliminating the requirement for recurring fabric repairs and replacement. Both the Beaver and Otter have carried external stores which are too bulky to fit in the cabin, such as a canoe, or have to be air dropped. The Buffalo also has some other interesting design features, above and beyond what civil or otherwise conventional design requirements call for. It has three under-floor keel beams, designed to sustain a wheels-up landing. The intent was that, in such an unhappy event, after the dust had settled, one could jack the aircraft up, lower the gear, and fly back to a repair base. Another design feature of the Buffalo is the strength distribution of the rear fuselage.

If the high T-tail hits hangar structure while the aircraft is being put away, the damage should be confined to the tail surfaces, and not cause material distress in the rear fuselage, thus limiting the extent of repair.These are just some of the features that go into a successful STOL aircraft.

Other research and development projects carried on at deHavilland during the 1950’s and ‘60s included the installation of turboprop engines to the DHC-4 Caribou. At the behest of the US Navy, the standard P & W R2000 engines of 1450 SHP in the prototype Caribou, CF-KTK-X or RCAF 5303 were replaced by General Electric YT64 turboprop engines and Hamilton Standard 63E60 propellers for ground and flight testing, and qualification trials. The substantial extra power (I believe that at that stage the YT64 produced 2400 SHP), made for a very lively aircraft. This engine was subsequently chosen for the Buffalo. Another project, commissioned by the US Army had an Otter equipped with a 5-bladed propeller and exhaust mufflers was aimed at external noise reduction. As far as I know, nothing came of this project.

The most interesting and fruitful project was a programme, funded by the Defence Research Board that had four experimental phases to investigate the effects on airfield performance of:
- lift produced by propeller slipstream deflection,
- controllable, jet drag and thrust, and
- the pilot’s ability to cope with what might turn out to be extremely short field performance.

In the first phase a Otter was mounted on a high cradle on a test rig, similar to an earlier Caribou model test arrangement. The aircraft was mounted high enough to be largely out of ground effect, and was driven down the runway with its engine running, and measurements were taken. In the second phase (Fig. 4) still larger flaps were installed (“Batwing”), the empennage was redesigned and enlarged, and a sturdier landing gear with double the energy absorption capacity was installed. Flight tests were conducted and measurements taken. In the third phase, the standard flaps were re-installed, with modifications, and a General Electric J-85 jet engine was installed in the cabin in a steel “doghouse”. Air was supplied through a duct from an intake on the cabin roof. A tee-shaped jet pipe discharged out each side of the fuselage, aft of the cabin doors. Coupled rotating jet nozzles were controlled hydraulically (by the pilot) to modulate the jet efflux direction from full forward to full aft. A prototype Caribou nose wheel actuator and an electric power pack were utilized. The efflux direction was controlled by a lever beside the pilot’s seat. The lever mounted a twist grip to control the J-85 engine speed and hence thrust magnitude. The necessary additional engine instruments, fuel system, etc were installed. Again, the flight testing was successfully completed.

At about this time, Pratt & Whitney Canada (PWC) engaged deHavilland to install their new turboprop engine, the PT-6, in the nose of a Beechcraft Expeditor, and conduct limited and preliminary flight trials. deHavilland liked the engine very much and this led to a long association with PWC for the supply of engines to power dehavilland / Bombardier aircraft, which exists to present times. In the fourth and final phase of the experimental Otter programme the R1340 engine was removed and replaced with two wing-mounted PT6A-4 engines (Fig. 5). In this final configuration, landing performance was spectacular. The aircraft could approach at 3 to 4 times the steepness of a conventional technique, and land over the 50 foot screen in under 500 feet. I recall seeing design studies in which this concept was applied to the Caribou, but it was not to be. As a major benefit to deHavilland and Canada the prototype Twin Otter was developed from the results of this programme with 844 aircraft subsequently being produced and sold to all corners of the world. The Twin Otter -400 Series (Fig. 6) with upgraded engines, airframe improvements and a "glass cockpit", has just been developed from the final deHavilland -300 version design and put back into production by legacy manufacturer, Viking Air of Victoria, BC.

Click on any image below to enlarge it.

Fig. 1-DHC-2 Beaver with STOL Characteristics

Fig. 2 - Diagram of a Typical STOL Airfield

Fig. 3 - Buffalo Operating in STOL Mode


Fig. 4 - Experimental Otter Programme

Fig. 5 - Highly Modified Research Otter

Fig. 6 - Viking Air Twin Otter 400 Prototype

For further information on the heritage deHavilland Canada aircraft please view the following websites.

Historic Beaver DHC-2

Viking Air


The contents of this paper were originally presented to a meeting of the CAHS Toronto Chapter on February 6, 1996 and recorded by Gord McNulty in the chapter newsletter “Flypast”. All content has been edited and adapted for this website.