A Comper Swift Project

CONTACT's picture

by John Illsley, Chapter 322

Origins of the design

Nick Comper was an RAF officer who went on design a series of light aircraft and set up the Comper Aircraft Company. Of his designs (Kite, Mouse, Streak, Swift) only the  CLA-7 Swift  (first flown in 1930)  went into production and 34 were built  between 1930 and 1933 before the company went into insolvency. By the standards of the time, this made the Swift one of the more successful British light aircraft of the period, keeping in mind that this was during the Great Depression. Intended as a small sporting and racing aircraft, the Swift made its mark in pre and post-war air races in Britain, but also became associated with some long distance record-breaking flights and attempts. By the end of this year, there will probably be four original and two replica Swifts flying, spread across the UK, Spain and Australia.

 Some Specs

empty weight: 320 kg
max weight: 480 kg
cruise speed: 115 mph
max speed: 130 mph
stall speed: 45 mph

The Swift is a small aircraft with a wingspan of  only 24 feet and an overall length of just over 18 feet. Most of the structure is of Spruce with a few lengths of Ash in the centre fuselage longerons. Although often described as being an ‘all wooden’ airframe, it derives much of its immense strength from the use of metal plates and brackets at almost every point where significant structural pieces of timber come together. These in turn are frequently cross braced with metal rods. I have often half jokingly said that Nick Comper obviously had little faith in glued wood  for load-bearing elements,  because it is only in part of the rear fuselage that there is a significant area of pure glued structure evident. The use of hundreds of metal fittings is what makes the aircraft so complex. Indeed there can be few aircraft from that era of a comparable size that are this complicated. Of course, it is yet another example of British over-engineering, with all its pros and cons. One definite advantage of the airframe design was that it lent itself to much development in terms of engines with almost no changes to the airframe. The prototype flew with a 40hp ABC Scorpion engine and Swifts  variants followed  that used the 50hp Salmson; 75 (and later 90hp) Pobjoy engines and eventually a racing version that used the 130hp DH Gipsy Major engine!


Why choose a Swift?

Even given my well known passion for vintage aircraft, this may seem like an unusual choice of project. I have several reasons for wanting to build and fly one of these.  The design and the unique radial engine have always appealed to me, particularly when the airframe is completed with wheel spats, as many of the air racing examples were in the 1930s through to the 1950s. That is a entirely subjective reason and I don't expect too many in the flying fraternity would agree with     my taste!

Victor SmithA second motivation was to try and undertake something that would help to add another unique pre -WW2 type to the SA register. As I have mentioned before, there is a dearth of aircraft from SA's aviation heritage that have been preserved, let alone kept flying in the country. I keep hoping (perhaps in vain) that by completing projects that fall into this category, others with greater resources than I will follow the lead in plugging gaps in important types that we have lost entirely.

The third reason is the inspiration provided by Victor Smith, who in 1932 flew a Swift in an attempt on the England to Cape Town speed record of Amy Mollison (nee Johnson). Unfortunately he ran into high winds on the last leg, started to run out of fuel and had to do a precautionary landing at night near Van Rynsdorp by the light of a parachute flare! This is all related in his wonderfully readable book ‘Open Cockpit Over Africa’. My intention is to complete my Swift in the same colour scheme as Smith's Swift. That well travelled example was unfortunately lost on a subsequent northbound flight when he abandoned the plane in the Western Sahara (where some parts may still survive). Remarkably, the Swift which Victor Smith used had already set a speed record from England to Australia, flown by CA Butler in a little over nine days. This Swift (registered G-ABRE) had been modified with the addition of a second fuel tank (to give a range of 1,000 miles), further instruments and navigation lights. With the weight pushed to over the maximum AUW, Butler had to fly in carpet slippers! His aircraft had flown over 22,000 miles by the time he sold it to Smith who registered it as ZS-ABU. Given that my project will in a sense be a reincarnation of the most well travelled of Swifts, there may be potential for a short book on the plane and its pilots – watch this space!

How do you see in front of you?

I have never shown anybody a picture of a Swift without that question being asked! Admittedly, it does seem as though a pilot must be flying blind as far as the forward view is concerned because the wing is mounted on a faired pylon which is forward of the cockpit. However this tapers upwards to a narrow point at the wing junction and although it forces the pilot to look either side, it by no means blocks out the entire view ahead. Furthermore, because the wing is about eye level, there is almost no blind spot from that surface, which is the case with most light aircraft, even today.

Finding an engine

Pobjoy engine shortly after arrival The heart of any replica project is the correct engine and in this case it is such a  characteristic feature of the aircraft that would be impossible to substitute with an alternative without losing much of the charm of the original Swift. The engine is a Pobjoy radial with either the original R-model (of 75hp) or the later Niagara (90hp) being suitable for the Swift. The Pobjoy is unusual in that it is a geared radial which means that the prop shaft is offset from the crankcase, giving it a very distinctive appearance as the gearbox dominates the front of the engine. The engine was one of the earliest to achieve its torque through high revs, although its 3,300rpm seems paltry by comparison to today's Rotax engines. Nevertheless, the Pobjoy was for many years the engine with the best power-to-weight ratio in its category. The gearbox also imparts a unique sound to the engine over and above that generated by seven cylinders pounding away. Viewed alongside any of the radials we are accustomed to in SA, the engine appears to be positively tiny in dimensions. When reduced to components, it is very evident that the workings are rather more delicate than any of your Pratt and Whitney radials (Victor Smith described the engine as having been made by Swiss watch makers). Apart from an almost total absence of any original spares, this  small size does add somewhat to the challenge of rebuilding the fiddly little engine.

The search for a suitable example of a Pobjoy took a number of years, even with the help of web sites and printed publications. I traced several down in the UK, but inevitably the cost in Rands made it prohibitive. One in France was missing a few cylinders and the ones in Australia were all part of current projects (for a Swift and a Short Scion). Eventually I had a response from the USA that looked feasible. Several Pobjoys were exported to the USA in the 1930s and were utilised mainly in autogyros and in very small racing aircraft. One of the latter  designs was the Pobjoy Special and, when commissioned to build a replica of this racer in the 1990s, Bill Turner of Repeat Aircraft acquired two Pobjoys for the purpose. (Those of you who have been to Oshkosh may have seen this replica in the Whittman hangar at Pioneer Field).   His son offered me the spare engine at a reasonable price and I shipped it to SA. Seemingly I am never destined to acquire vintage aircraft engines that are complete and this one was missing the magnetos; prop hub and exhaust ring.

Reproduction or replica?

In the world of replica vintage aircraft, there is much debate as to what qualifies as a true replica, accurate in every last detail and what constitutes a reproduction that approximates the original in general appearance and being a full scale (as opposed to 3/5 or 5/8 the size of the original) aircraft. The American approach has more often than not fallen into the latter category, for example often substituting welded fuselage frames for wooden. This certainly speeds up building time and provides for a more long lasting structure, generally without the external appearance being compromised.

Being something of a purist, I wanted to stick closely to the original design but with some pragmatism applied for certain ‘under the skin’ aspects. Apart from the correct engine the project will also utilise period instruments, magneto switches and a Sutton harness. In the airframe I have decided to make one major deviation from the original design and that is in not building the aircraft with folding wings. This was a feature much beloved of British aircraft companies in the 1920s and 30s, but it adds considerably to the complexity of the Swift by way of the folding mechanism, hinges and aileron controls. Consequently I am going to dispense with wing folding which will simplify construction and achieve some weight saving but without detracting from the lines of the aircraft.


Tracing or manufacturing all of the parts missing from the Pobjoy engine has taken some years, but these are now virtually all in hand. A meeting with Dr Roy Fox, an Australian vintage aircraft collector at Oshkosh in 2008, led to him kindly donating a prop hub and the exhaust collector pipes. Some British Swift owners have donated a BTH magneto and distributor rotor arms. I have also bought a spare oil pump from Australia and a pair of BTH M-1 magnetos from the UK. Simms couplings, the vernier couplings used on the magneto drives are still being made and a pair of these was recently bought from a company in England. Still missing are distributor caps which will probably have to be made from a pattern or a drawing.

Pobjoy Niagara engine stripped downThe engine has been completely stripped down and although in good condition, with no corrosion evident, it will be rebuilt with new bearings, piston rings, valves and valve springs.     The small valves seem to have been a weak point of these engines and so modern equivalents will be made using CNC technology.

All the wing ribs have been made up from spruce and birch ply, using the Arthur Orde Hume type of jig which utilises offset rotating wheels to lock the rib components in place prior to gluing . A total of six jigs were needed for the various full; three-quarter;  wing taper and nose ribs that go into the Swift wings. With all the rib components pre-cut, a type of production cycle can be set up whereby, for each gluing session, ribs are either being started in jigs or coming out of the jig for attachment of the remaining gussets on the opposite side of the rib.

After several years of having the original factory drawings for metal fittings  gradually converted into Autocad format, I was able to take all of the digital format drawings to a company to have the fittings cut out of 4130 steel using computer-controlled laser cutting machines. A company in Britain went through the same exercise a decade ago and can offer a full set of Comper Swift flat packs of all the metal fittings. However, given the weakness of the Rand to the Pound Sterling, this has never been an option for me and I went the route of replicating all the parts locally. Apart from the huge cost saving, it does mean that in the unlikely event that there is anybody else in the EAA fraternity who wishes to build a replica of one of the Swift variants, I am in a position to assist with metal fittings. Seeing the components laid out on the "nesting" drawings for each sheet of steel is a reminder of the fact that there are several hundred fittings that have to be built into the airframe and that this is no small undertaking.

The way forward

Wing Rib JigsThe spruce needed to complete the project has just been ordered from Aircraft Spruce (who else?) and will be shipped shortly. With the Rand-Dollar exchange rate being what it is, I have to be careful not to leave any invoices lying around lest I be smitten by the wife for excessive outlay on this hobby!

I will probably complete the wings and tail surfaces first.      The rudder and fin (the only surfaces made from welded tubular steel) can be made relatively easily and should provide a morale boost – parts that at least look like they belong on  an aircraft!

The approach to each major sub assembly will be to assemble a kit of parts and then proceed into assembly. The way the type was designed and manufactured originally lends itself to this approach and even allows construction of sub assemblies in a small work area. The fuselage, for example, consists of a front, middle and rear fuselage, each of which can be built independently.

Wing rib production line


Bending profiles for fin and rudder

All in all, there are years of fun ahead – as a fellow EAA 322 member commented on hearing about this project, "You don't choose the simple ones, do you?"