The Kemp 2cc Trio:

The K Falcon, Kestrel, and Tornado

by Adrian Duncan

---

In our companion article on the Kemp Hawk, we've covered the history of the British model engine manufacturing concern founded in 1946 by Harold Kemp and continued up to late 1950 under the name of the K Model Engineering Co. We've also taken an in-depth look at most of the engines which were produced by this manufacturer. To complete our coverage of the range, it's now time to look at the series of engines in the 2 cc (0.12 cuin) category which together constitute one of the best-remembered product lines of the K Model Engineering Co, but first, some acknowledgements.

Acknowledgements

As always, a study like this cannot be carried through to an authoritative conclusion by any individual working alone, and I wish to extend my very sincere thanks to my good mate David Owen for his invaluable assistance in researching serial numbers and for providing images of engines of which I do not possess examples. Similarly, I'm most grateful to fellow Kemp/K enthusiast Paul Rossiter of Rochester, England, for his help in clarifying a number of matters in connection with this study.

Special thanks go to our long-suffering Editor Ron Chernich for his efforts in extracting advertisements from his extensive collection of old Aeromodeller magazines. In the absence of any previous in-depth commentary on the history of the K engines, the advertisements constitute one of our most authoritative sources of information. But Ron does far more than this—the work required to set up these pages and the associated images and links represent a very considerable task indeed. We should all be grateful to Ron for his seemingly tireless efforts in this regard—I know I am!

Further much-valued assistance came from long-time Kemp/K user Brian Cox, who was able to add significantly to our knowledge of the serial numbers associated with the Kemp and K engines. Finally, it would be difficult to overstate my gratitude to my friend and colleague Alan Strutt for generously making available fine examples of both the later variant of the K Kestrel and the rare K Tornado glow model for inclusion in this study. Without assistance of this nature, it would not be possible to make any claim to authority for these articles.

Now, before we turn our focus onto the series of engines in the 2 cc category which were manufactured by the K Model Engineering Co, it may be helpful to new readers to present a brief summary of the chain of events leading up to their introduction. Readers requiring the full story are advised to read the background section of our companion article on the 0.2 cc Hawk. Conversely, those not requiring such information are invited to skip the following section altogether. It's your choice ...

Background

The range of model aero engines produced under the K in a circle trademark had its origins in 1946, when Harold Kemp moved to Gravesend, Kent to establish a small precision engineering concern called Kemp Engines. The possibility has been suggested that Kemp may previously have had a hand in the design and/or construction of the K6 spark ignition engine marketed in 1946-47 by Eddie Keil, but this is very far from being an established fact. The K6 is undeniably very different in style from any other product definitely attributed to Kemp, which seems to throw considerable doubt upon the attribution. Still, it remains a possibility.

The new company established premises at 7 Bank Street in Gravesend, a few doors from Harold Kemp's new residential address at 3 Bank Street. The business commenced operation on a very small scale, initially employing only three full-time workers (including Harold Kemp himself) plus one part-time assistant. The first definitely-attributed Kemp product was a 4.4 cc sideport diesel called the K4, which was released to the market in early 1947. A few 8.8 cc twin cylinder diesels were also manufactured using components from the 4.4 cc single-cylinder model. These large engines were soon joined by a 1 cc crankshaft front rotary valve (FRV) diesel.

In early 1948 the company released perhaps its most famous model, the cute little 0.2 cc Kemp Hawk Mk I. Shortly thereafter, the FRV 1 cc model was replaced by a simple sideport 1 cc design based very much on the Hawk. This was named the Eagle, in keeping with Kemp's evolving "bird of prey" nomenclature.

All of these engines were relatively expensive and production figures were quite small by comparison with those achieved by the more prominent British manufacturers of the period such as ED, Mills Bros and International Model Aircraft (Frog). However, the range acquired a good reputation for being well-made and dependable, and both the Hawk and the Eagle appear to have sold as fast as the tiny company could make them. It must soon have become clear to Harold Kemp that the potential for increased sales was considerable. The key to tapping this potential lay in expanding production capacity and thereby also reducing unit production costs.

Accordingly, Kemp went looking for additional investors, and he found one in the form of the well-known modeller Len "Stoo" Steward. The result was the reorganization of the Kemp Engines concern into a new company trading as the K Model Engineering Co from a different location on Darnley Street in Gravesend. It's presently unclear whether or not Harold Kemp retained a connection with this company, but it's entirely possible that he did so. He certainly remained in residence at 3 Bank Street in Gravesend until at least late 1951, well after the closure of the K Model Engineering Co.

The new company seems to have come into existence in around July of 1948, since it was mentioned in the July 29, 1948 issue of "Model Engineer" and its first paid advertisement appeared in the August 1948 issue of Aeromodeller. That advertisement focused on the 0.2 cc Hawk and the 1 cc Eagle sideport models which remained in production at this time—the 4.4 cc diesel had been dropped immediately upon Steward joining the company, if not before.

Steward was a leading member of the West Essex club and a prominent participant in control-line stunt contests at the time in question. With this background, it should come as no surprise to learn that his influence over the company's affairs was very much directed towards the development of engines suitable for stunt use. This made complete business sense, since control-line stunt was probably the fastest-growing branch of the aeromodelling hobby in England at the time; how times change.

In order of their appearance on the market, the first new products of the K Model Engineering Co were an updated version of the sideport Hawk (c. September 1948), a 5 cc FRV diesel called the Vulture (October 1948) and a Mk II version of the 1 cc Eagle featuring FRV induction (December 1948). The Eagle Mk II was also offered in kit form, a not-uncommon marketing strategy during the cash-starved period which the British economy underwent followed the conclusion of WW2.

It's clear that the 5 cc Vulture was specifically designed for control-line stunt use and was intended to be the "flagship" model of the new company. However, as chronicled in our companion article on the Vulture series, it got off to a very shaky start from which it never appears to have recovered. There was nothing fundamentally wrong with the design—it was more a case of handling difficulties (which were manageable but required patience to overcome) coupled with certain components initially not quite being up to the job in terms of strength.

The company made changes to address the latter issues, substantially improving both the engine's durability and performance in the process. However, their efforts appear to have been insufficient to revive the fortunes of the Vulture in commercial terms. They had evidently anticipated a strong demand for this engine and had set their staffing levels and production quotas accordingly. The fact that the anticipated buyers did not materialize left them with a growing inventory of unsold engines. Since they had already borne the cost of manufacturing these units, a considerable amount of their capital was now tied up in unsold inventory—hardly a healthy situation from a fiscal standpoint. From this point until its final demise, the company appears to have operated under the restraint of what amounted to a set of financial handcuffs.

This was a great pity, because a well set-up late-model Vulture is actually a perfectly useable and durable engine in the hands of a capable operator and does indeed make a very good stunt diesel by the standards of its day. As a Vulture user himself, Len Steward was doubtless well aware of this and he seems to have decided that one way to demonstrate the basic soundness of the Vulture layout and also restore the company's credibility was to release a smaller version of the engine in a more "popular" displacement category which would benefit in design terms from the lessons learned with the Vulture and would be easier to handle. Moreover, such an engine would have broader sales appeal given the related facts that it could be used with smaller and less expensive models and would be far easier on the fuel budget—an important factor in early post-war Britain.

The resulting new model was designated the 2 cc K Falcon, and it launched the series of engines which form the main subject of the present article. The Falcon was announced in an advertisement placed by the K Model Engineering Co in the April 1949 issue of Aeromodeller. It was in essence a downsized version of the internally-strengthened K Vulture Mk II which had first appeared in around February of 1949. Once again, the Falcon's primary intended use was for control line stunt.

At this point, some readers may be asking themselves—"Why 2 cc?" No doubt a somewhat odd displacement from a present-day standpoint, but this is a matter which has to be considered in a late 1940's context as opposed to a twenty-first century one. At the time in question, the "magic" displacement of 2.5 cc (0.15 cuin) had not yet been adopted by the FAI for International competition, and accordingly no particular displacement really had much of an edge over another. The only issues were what size of model was desired and whether or not a given engine could do the job of flying that model to the standard required.

Engines of 2 cc (0.12 cuin) displacement or thereabouts were in fact quite common in Britain at around this time. Aerol Engineering had started out in 1947 with their 2 cc Gremlin and Hurricane models before replacing them in the latter part of 1948 with the hugely successful Elfin 1.8, and ED had enjoyed considerable sales and competition success with their 2 cc "stovepipe" models, one of which powered the 1948 Gold Trophy control-line stunt winner. International Model Aircraft were represented by their Frog 180 diesel, and Rawlings also had a 1.8 cc model on the market. Other makers such as Dyne, Majesco and Airstar also offered models of around the 2 cc displacement mark.

Seen in this 1949 context, there was actually nothing at all extraordinary about the K Model Engineering Co releasing what was intended to be a "popular" new model in the 2 cc displacement category. It was only later, when the 2.5 cc limit was formally adopted by the FAI and the British 1/2A displacement limit was set at 1.5 cc, that the attention of manufacturers became progressively more focused on engines built to the new competition limits and the 2 cc displacement category became something of an orphan, lying between competition classes.

Now that the background to its introduction has been made clear, let's take a closer look at the Falcon ...

The K Falcon

As noted earlier, the K Falcon was introduced in April 1949. In most important respects, it was simply a scaled-down version of the companion Vulture Mk II which had first appeared in early 1949. It featured the same ultra-simple construction in which no screws were used, all the parts being of the screw-in or screw-on type. The entire engine consisted of only 17 individual components, and three of these (the prop nut, prop washer and spraybar nut) were off-the-shelf items that did not have to be manufactured.

It was perhaps the Falcon's misfortune to arrive on the scene at a time when it faced formidable opposition from a well-established competitor in the shape of the Elfin 1.8 cc radial mount model. This ground-breaking design had appeared in the latter part of 1948 as a replacement for the earlier 2 cc Aerol Hurricane and was taking the British modelling world by storm at the time. The Falcon shared a number of features with its rival, including the radial cylinder porting and the use of updraft FRV induction. It's hard to say whether this would have enhanced or diminished the marketplace view of the engine, but it would certainly have made direct comparisons inevitable.

Regardless, the Falcon was an individual design in a number of ways and was quite distinct in design terms from its famed rival despite the superficial similarities. The nominal bore and stroke of the Falcon were 0.520" (13.21 mm) and 0.562" (14.27 mm) respectively for a displacement of 1.96 cc (0.119 cuin). This made it a long-stroke design, in contrast to the short-stroke arrangement of the parent Vulture. The stroke was in fact identical to that of the Elfin 1.8, but the extra 0.015" bore of the Falcon gave it a slight displacement advantage as well as bringing it closer to "square" bore/stroke dimensions.

To offset this, the Falcon was the clear loser in term of weight, tipping the scales at 4.3 ounces, just over an ounce heavier than the 3.25 ounce Elfin. But the simplicity of its construction allowed the makers to offer the engine for sale at a very competitive price of �2 19s 6d, which presented a notable contrast with the �3 19s 6d price tag of the competing Elfin. So despite its weight disadvantage, the Falcon appeared on paper to offer quite serious competition to its Liverpool rival.

Looking now at the engine's construction, we find ourselves largely repeating our description of the Vulture which appears elsewhere. The case-hardened steel cylinder was of basically tubular form—the step down in outside diameter just above the exhaust ports which was a feature of the larger Vulture was omitted in the smaller model. The upper portion of the cylinder was externally threaded to accommodate the screw-on machined alloy cooling jacket. The lower end below the ports featured a base flange which served to locate the cylinder vertically on top of the crankcase. A thin gasket between the cylinder base flange and the upper crankcase ensured a good seal.

As with the Vulture, there was no "spigot" length below this flange to locate the cylinder laterally on the crankcase. Instead, the cylinder was both secured and laterally located by an internally-threaded locking ring which bore on the top of the cylinder flange and screwed onto an externally-threaded portion of the upper crankcase. This threaded ring quickly became known colloquially as the "dog collar", a highly descriptive term which has stuck to this day!

The dog collar on the Falcon was modelled very closely upon the matching item used on the contemporary Mk II Vulture. It was a light alloy item having an upward tubular extension to a point just below the exhaust ports. This extension carried a set of shallow turned cooling fins. Given their location, these could have had little practical effect and were doubtless more for show than anything else.

The upper cylinder was equipped with a conventional contra piston made of hardened steel. This is normally a problematic choice of materials for use in a steel cylinder since in most cases a steel contra-piston freezes solid in a steel bore after the engine warms up a bit. Some of the K 2 cc engines do exhibit this behaviour, and a cast iron contra piston would undoubtedly have been a better choice. If you ever have occasion to fit a new contra piston to one of these engines, use cast iron!

The fully machined screw-on aluminium alloy cylinder jacket featured a smooth head in contrast to the finned head of the Mk II Vulture. This configuration was shortly to appear on the Mk III version of the Vulture as well. The omission of head fins was doubtless a cost-cutting measure. The jacket featured integrally-machined barrel fins of conventional pattern.

Unusually, the thread for the compression screw was not formed in the centre of the head—instead, the cylinder was topped by a hard brass disc with a central spigot which engaged with an oversized hole in the centre of the head. This brass disc was sandwiched between the underside of the head and the top of the cylinder when the jacket was screwed down. The thread for the comp screw was formed in the spigot rather than in the head itself. These arrangements were carried over directly from the Vulture.

The presence of this component doubtless contributed to the relatively high weight of the Falcon and also made the engine somewhat taller than it would otherwise have been. However, it did have the advantage of permitting a worn comp screw thread to be rectified simply by replacing the brass insert rather than the entire cylinder jacket.

Turning now to the cylinder porting, this too possessed some distinctive features which were also carried over directly from the Vulture. The exhaust ports were essentially conventional for a radially-ported engine, consisting of a ring of four sawn ports separated by relatively thin pillars. However, the transfer arrangements were rather more unusual. The actual transfer port itself consisted of a continuous annular channel machined around the entire circumference of the inner cylinder wall just below the exhaust port ring—an example of true 360 degree porting! This annular channel was fed from the lower crankcase by six shallow bypass flutes machined internally at a uniform spacing into the cylinder wall below the channel.

This system did not provide much in the way of directional control of the incoming mixture, and in the case of the larger Vulture it clearly resulted in substantial losses of unburned mixture through the exhaust ports. It's reasonable to expect that the Falcon would be similarly afflicted, and indeed a well-trained "diesel nose" can detect the unburned fuel in the exhaust of a well-tuned Falcon or its successor.

The conical-topped piston was again of hardened steel and was used in conjunction with a hardened steel connecting rod. The rod was of quite substantial proportions—the designers had clearly learned from their sad early experiences with the Vulture! Unusually once again for a British engine, the piston and rod were connected by a ball-and-socket joint rather than the more usual gudgeon pin arrangement. This system had in fact been introduced to British model diesel manufacture in the former Kemp Eagle Mk I sideport diesel and had also been applied to the Vulture. It was thus something of a Kemp/K trade-mark feature by this time.

The use of a ball-and-socket small end was probably a good move in view of the bypass arrangements used—unless it was well secured, a conventional gudgeon pin might well have tended to foul the bore at the top of the bypass flutes. The downside was that since both components were hardened after installation, replacement of the rod was impossible—if that requirement arose, one had to replace both piston and rod, and that would mean a rebore or a new cylinder into the bargain. The system also prevented the re-swaging of the ball-and-socket joint if it became loose—any attempt to do so would likely result in a fracture of the hardened socket. Be warned ...

The hardened steel crankshaft was of conventional form and appeared to be adequately sized for the job, in contrast to the matching component on the Vulture. The crankweb was of the full disc type but lacked the substantial counterbalance which was a feature of the companion 5 cc Mk II and Mk III Vultures. The shaft ran in a well-fitted steel bushing which was pressed into an alloy housing cast integrally with the main crankcase. A round crankshaft induction port was used in conjunction with a venturi base of similar shape formed in the bushing.

The Falcon designer must surely have been anxious to avoid the fragility issues which had bedevilled the working components of the original Vulture. Crankshaft journal diameter was a nominal 5/16" (0.312"), only 1/16" less than the diameter of the same component in the 5 cc Vulture shaft and a quite generous figure for a 2 cc diesel. The shaft in a checked example actually measured up at 0.315".

So far so good, but the central hole for the passage of incoming mixture had a checked diameter of an equally generous 0.200", resulting in a crankshaft wall thickness of only 0.058"—roughly the same as that of the Mk I Vulture shaft which had given so much trouble. However, the fact that the K 2 cc engines are not noted for suffering crankshaft failures implies that this must have provided sufficient strength given the fact that the Falcon had a 60% smaller displacement than the Vulture and was moreover a long-stroke design. The crankpin had a very sensible diameter of 0.1875"—no shortage of strength there!

The steel prop driver was secured to the shaft by being pressed onto a short splined section of shaft forward of the locating shoulder. A standard prop nut and prop washer were used to secure the prop against the driver. Somewhat unusually, the thread used was 0 BA! If a need for a replacement prop-nut ever arises, this is almost exactly equivalent to a standard 6 mm x 1.0 mm metric thread.

Looking now at the main casting, the 28 TPI threads on the upper part of the case onto which the dog collar screwed were apparently formed using a large Coventry die-head. This forced the designer to keep the rest of the casting below the level of these threads so that nothing would foul the die-head. This was likely a significant factor in the choice of updraft induction since it placed the venturi on the opposite side of the main casting from the cylinder retention threads.

The venturi itself had a nominal bore of 0.1875" (4.76 mm). This may appear a little skinny for a 2 cc engine, but it was used in conjunction with a spraybar which was thinned across its working length to a thickness of only 0.110" (2.80 mm). These components combined to give good gas velocities and hence excellent suction, a real asset for a diesel intended for stunt use. The needle was quite conventional, utilizing the standard split thimble setup for tension. The spraybar thread was 5 BA.

The main bearing housing on the Falcon was of generous proportions and was quite sturdily braced at four locations at the point where it met the front of the crankcase. However, the mounting lugs remained something of a weak point with the Falcon, as they had been with the Mk I and Mk II versions of the Vulture. They were somewhat thinner than ideal for an engine of this size and were in fact less substantial than those on many engines of significantly lesser displacement. Moreover, the Falcon did not have the benefit of the centrally-located stiffening webs or the forward extensions which had been features of the Vulture.

Overall, it has to be said that the lugs on the K 2 cc engines appear somewhat marginal for the job, and cases with damaged lugs are not infrequently encountered. When flying these engines myself, I have always used small plates of 1/8" aluminium alloy which are shaped like the lugs in plan view and have holes drilled through them at the appropriate spacing. These act in effect as clamps to hold the lugs down on the mounting beams in the model. The result is that the loading is distributed over the entire mounting lug area rather than being concentrated at the mounting hole locations as it would be if the mounting bolts acted directly on the upper lug surfaces. This approach makes breakage far less likely and also prevents the generation of mounting marks on the lug surfaces.

Returning to our description, the backplate was a simple screw-in item turned from solid aluminium alloy bar stock. It sealed to the rear of the case with a gasket. An additional radial mount backplate was supplied with the Falcon as an accessory. Radial mounting was enjoying something of a vogue in Britain at the time, largely one suspects as a result of the great success of the early Frog engines and the Elfin 1.8 cc model. The K Model Engineering Co hedged their bets by featuring beam mounting as standard with the option of radial mounting if desired. This was another factor which might have been expected to give the Falcon something of a marketplace edge over the radial-only Elfin.

The engines all carried serial numbers which were stamped on the right-hand side of the case (looking forward in the direction of flight) beneath the mounting lug on that side. Apart from the serial number, the only other identification to appear on the engine was the company's well-known "K in a circle" trade mark. This appeared on both sides of the crankcase cast in relief above the mounting lugs.

Thanks to the splendid cooperation of our colleagues acknowledged at the outset, we are fortunate enough to have access to a highly representative sample of serial numbers for the Falcon. It's evident from the numbers recorded to date that the Falcon received its own distinct serial number sequence beginning with engine number 6001—all authenticated numbers reported so far are within the 6xxx range and beyond. The earlier Eagle Mk II FRV model had been numbered in the K5xxx series but had been discontinued soon after the introduction of the Falcon and never approached numbers in the K6xxx range. Serial numbers for the other contemporary K models were nowhere near the 6000 mark, nor likely to get there, and it's pretty clear that the intent was to keep the numbers on the Falcons quite distinct from those of other models in the range.

At present, the lowest Falcon serial number of which we are aware is 6045, and the highest is my own 7462. This seems to confirm that perhaps 1500 examples of the Falcon were produced—the total is unlikely to be much if any higher. Such a figure is certainly consistent with the fact that the Falcon is significantly less common today than its successor, the "K" Kestrel which we will consider in a following section of this article.

Prospective buyers who care about such things should be aware that there are a number of claimed Falcons out there which are actually conversions of the subsequent K Kestrel model (see below). The Falcon and Kestrel were more or less identical designs with a few largely cosmetic changes, and converting one to the other is perfectly straightforward if one has (or can make) the necessary parts. The giveaway is the serial number—an authentic Falcon will have a number in the 6000 plus range, whereas Kestrel serial numbers are far lower and have a K prefix. Converted Kestrel number K349 which recently sold at auction was an example—it was offered as an original Falcon but is in fact either a conversion of an early Kestrel or a repair to an original Falcon using a Kestrel case.

The Falcon remained on the market for some 10 months, making its final advertised appearance (as far as we are presently aware) in the February 1950 issue of Aeromodeller. By that time, the K 2 cc diesels had demonstrated good potential in the control-line stunt field, although sadly this had not translated into increased sales activity for the company. However, the engine did attract the attention of the resident engine tester for the rival magazine Model Aircraft (almost certainly Peter Chinn). Let's see how the engine fared in his hands ...

The K Falcon On Test

The Falcon made its one and only published test appearance in the January 1950 issue of Model Aircraft. For reasons which remain unclear, the magazine did not publish the name of the individual responsible for these tests at the time in question or for many years thereafter. However, there's a large body of circumstantial evidence to support the conclusion that the tester was Peter Chinn, who contributed regular articles on model engine development to the magazine. To our knowledge, no-one has ever presented a credible challenge to this conclusion, and accordingly we shall henceforth maintain the assumption that Chinn was the author of this test. If anyone can prove us wrong, let's see your evidence!

Chinn (assuming it was he) reportedly had two examples of the Falcon available for test, both obtained directly from the manufacturers. However, one of these had evidently been assembled very much on the tight side and was reluctant to perform with any great authority. The fact that all of the working components were made of hardened steel meant that a tightly-fitted engine would take many hours of running to free up, an expensive process in terms of fuel costs and one which the average modeller would quite reasonably not expect to have to face.

Chinn's other example was evidently fitted considerably less tightly and performed far better than its companion right out of the box. It was this engine that was used as the basis for the major content of the report. It was given a one-hour running-in period and was then tested in the usual manner.

There is an implication in the above findings that there was considerable variation between different examples of the Falcon. If this was a common issue, there's little doubt that some buyers must have been quite disappointed, and this would scarcely have furthered Len Steward's objective of restoring consumer confidence in his company's products. We'll be recalling this factor in a later section of this article.

Returning to the test report, Chinn reported that the Falcon was "not quite so easy to start as some engines", a factor which may well be connected with the type of porting used. The 360 degree transfer port coupled with the radial exhaust porting meant that there was nowhere for a prime to "pool" in the cylinder when the engine was mounted in an upright configuration, as it would be during a typical test. Despite this, a prime is more or less essential for these engines when mounted in an upright position because fuel drawn from the tank by choking tends to drip out of the downdraft intake by gravity rather than being absorbed by the engine. Starting is actually much improved when the engine is mounted in the sidewinder configuration, as it often would be in a stunt model.

Another factor which could potentially affect starting is the width of the transfer port (the entire circumference of the bore). Such a wide port would tend to result in reduced transfer gas velocities, a characteristic which is not conducive to good starting. The larger Vulture suffered noticeably from the effects of both factors noted above.

Having made the above statement, Chinn softened the blow by stating that "neither Falcon presented any operational difficulties under test". He found that once the correct needle and compression settings had been established, little further adjustment was required between starting and running settings. Present-day experience bears this out completely.

An issue encountered with one of the two engines tested by Chinn was the loosening of the dog collar. This is a typical issue with the K engines and is more of a characteristic than a real problem—it's mostly a consequence of the sealing gasket compressing over time, and the assembly does stabilize eventually. All that's necessary is to check periodically to ensure that the collar remains snug. But please preserve the knurled surface of the dog collar by protecting the jaws of the tool or using a strap wrench when doing this!

It's interesting to compare the power output figures obtained by Chinn with those recorded by Lawrence Sparey in his test of the competing Elfin 1.8 which had been published in the July 1949 issue of Aeromodeller. Sparey reported a figure of 0.1138 BHP @ 12,100 rpm for the Elfin as opposed to the 0.121 BHP @ 9,600 rpm recorded for the better of Chinn's two Falcons.

In comparing the two sets of results, it's worth recalling that Chinn consistently tended to find more power at higher revs than either Sparey or his Aeromodeller successor, Ron Warring, when testing the same engine. This must be down to variations in the testing and engine management procedures. It's likely that Chinn would have extracted a somewhat better performance from the Elfin than Sparey was able to do.

Regardless, on this showing the Falcon far out-torqued the Elfin, producing greater peak power at substantially lower rpm. It would seem that the Falcon would happily swing a larger and hence more efficient prop than the Elfin. Chinn used an 8x8 airscrew for flight testing, and this is probably a good choice for stunt applications with this very torquey low-revving engine.

A useful feature of Chinn's reports at the time was the occasional addition of information based upon actual flight testing. The Falcon was one of the engines tested in this very practical manner. Chinn reported that the better of his two Falcons was mounted in a small control-line stunt model of 150 square inch wing area that had originally been built for "a well-known Class "A" diesel of rather smaller capacity". The weight was increased by less than one ounce to 9-1/2 ounces for a still quite practical wing loading of 6.27 ounces/100 sqin.

Chinn reported that using an 8x8 airscrew and flying on 40 foot lines, the model's speed increased from less than 40 mph to almost 52 mph, and the "feel" of the model during manoeuvres was greatly improved. He was clearly quite impressed by the Falcon's airborne performance in this application.

As a final note, it's interesting to observe that the Falcon peaked at only marginally greater rpm than the 8,900 rpm peak reported by Sparey in his January 1949 Aeromodeller test of the Mk I Vulture. The fact that the change in scale of the basically common design did not substantially affect the speed at which peak power was achieved suggests that the unique porting used on the two K models was a limiting factor in terms of peaking speeds.

This was the only test of a K product that was destined to appear in Model Aircraft. Somewhat unusually, the rival magazine Aeromodeller never published a test of the Falcon or either of its successors.

The K Kestrel

In May of 1949 the K Model Engineering Co began to advertise both the Mk II and MK. III versions of the 5 cc Vulture simultaneously. As discussed in our companion article on the Vulture, this was apparently a strategy intended to help the company liquidate a rather large inventory of unsold Mk II Vultures which they had on hand at this time. The two versions were destined to be offered side by side in the range until the early part of 1950, when supplies of the Mk II Vulture were finally exhausted.

The application of this dual-variant marketing strategy to the Vulture seems to have got the company thinking about a similar approach to other models in their range. In June 1949 they announced a Mk II "Special" variant of the recently-introduced FRV 0.2 cc Hawk Mk II, although it's presently unclear whether or not this actually reached production. They then turned their attention to the Falcon ...

The consequence of their deliberations was the August 1949 introduction of a second variant of the 2 cc series which had been initiated with the Falcon. This was the well-known Kestrel diesel. It was clearly aimed squarely at the "popular" market, selling for the rather startlingly low price of only �2 5s 0d and thus undercutting the Falcon (and the competition) by a substantial margin.

The serial numbering sequence was re-started at 1 when the Kestrel was introduced. To further distinguish it from its Falcon predecessor, the letter K was attached to the serial number as a prefix. There is thus no possibility of confusion between cases originally used in Falcon and Kestrel engines—engines with serial numbers in the 6000-plus range which lack the K prefix but otherwise appear to be Kestrels are collector-assembled hybrids made up from a collection of parts, including a Falcon case. They do exist, as do "Falcons" with Kestrel cases...check those serial numbers!

The commonly-held view of the Kestrel is that it was a somewhat less carefully-built version of the Falcon. This may be a rather simplified view of the matter, as we shall see, but it is undoubtedly correct to say that the Kestrel was to all intents and purposes the same engine with very minor alterations. Indeed, the only immediately-apparent difference between the two models was the dog collar! The component used on the Falcon had three barrel fins above the knurled ring itself, while the shorter collar used on the Kestrel had only one.

A more careful examination will also reveal that the cooling jacket on the Kestrel had a slightly more rounded head profile than the component used on the Falcon, but that is of little real consequence. The number of fins remained unchanged and the weight of the engine was very little affected as well, being only fractionally less at 4.2 ounces.

In all other respects, the two engines were apparently identical, so there is no need to indulge in a detailed description here. A few minor points are worthy of mention, however. It appears that production of the short-lived FRV 1 cc Eagle Mk II model had ceased by May of 1949—certainly, the engine no longer appeared in the company's advertising as of that date. This left the company with a few extra parts for the Eagle model, one of which was the L-shaped compression screw. It seems that a decision was taken to use up these extra components on the Kestrel production line, since a number of the earlier examples of that model use L-shaped compression screws—my own illustrated K201 is an example and engine number K320 is another. Later they switched to the more conventional T-shaped item which appears on most (but not all) examples of the Kestrel.

The other changes which appeared midway through the production cycle were a switch from steel to aluminium alloy for the prop driver and from brass to aluminium alloy for the head insert. A similar change to the prop driver material had previously appeared on the companion 5 cc Vulture Mk III. The motivation for these changes, which seem to have occurred at around engine number K1600 in the case of the Kestrel, was likely weight reduction. However, it was not completely successful in the case of the prop driver at least, since the alloy driver had a greater tendency to strip on the rather skimpy splines on the crankshaft and leave the driver free to turn on the shaft.

The radial mount backplate which had been introduced with the Falcon continued to be offered for the Kestrel. Since the two engines were essentially identical, the same component was an equally good match for both models.

A number of owners have reported that the average Kestrel seems to be set up a little more loosely than the typical Falcon. It's presumably this factor that has led to the previously-expressed view that the Kestrel was a less carefully-built version of the Falcon, and there certainly seems to be some validity in this view.

But this then poses the very clear question—why concurrently offer what was in effect the same engine in two distinct and competing guises which were built to differing standards? This is a matter which deserves our fullest consideration.

After much thought, we do have two possible explanations to offer. However, we must emphasize at the outset that neither interpretation may be viewed as established—they merely explain the observed facts, as any scenario must necessarily do. We fully accept that there may be other interpretations that do so just as well—however, we can't think of them! All that we can do is offer our two plausible scenarios and allow our readers to judge for themselves ...

Looking at our first scenario, readers of our companion article on the Hawk (in which the history of the company is set out in detail) will recall that the Kemp manufacturing operation was almost certainly established using surplus wartime machine tooling obtained at low cost following the cessation of hostilities. In general, this equipment was very well-used, or to put it more bluntly, totally knackered! Not only that, but its continued use following the war could only have made matters worse ...The maintenance of consistent tolerances using such well-worn equipment must always have presented a challenge, especially when there was a concurrent need to keep costs down.

It thus appears possible that the advent of the Kestrel may have represented the manufacturer's response to the challenge of maintaining consistently close production tolerances at a reasonable cost using the clapped-out machine tooling which was all that was available. We have the evidence of the Model Aircraft test report summarized earlier to confirm that there was considerable variation in fits between different examples of the Falcon. If nothing else, customer feedback would have ensured that the company became aware of this quite early on.

Although this cannot of course be proved, one plausible explanation of the concurrent marketing of the Falcon and Kestrel is that the company came to accept this situation and recognized that while the only effective fix would be to upgrade their machine tooling, the non-availability of funds precluded this. After all, a good portion of their capital was by then tied up in their inventory of unsold Vultures, as noted earlier.

So they decided instead to accept the situation and deal with it by in effect "high-grading" the components as they came off the line. The closest-fitting components went into the Falcons and the less precise components were earmarked for Kestrels. This would relieve the manufacturer of the need to take drastic steps to maintain consistent tolerances in the face of the major handicap represented by the continued use of clapped-out equipment which they could not afford to replace.

We repeat—the above scenario cannot be proved! It merely represents one possibility among others. It also embodies one serious objection in that it implies that the company was willing to take on the challenge of concurrently producing two versions of essentially the same engine, and to differing standards at that. No subscriber to the KISS principle (Keep It Simple, Stupid!) would entertain such convolutions for a moment!

OK, so much for scenario number one—a bit of a doubtful starter! Can we propose something that seems a little more rational? Fortunately, the answer is a resounding YES—there is another possible interpretation which definitely bears consideration and in my personal view is far more likely. This is based in part upon the established history of the companion Vulture, which has been covered in detail elsewhere.

Readers of the Vulture article will recall that although production of the Mk II Vulture appears to have ended by May of 1949 in favour of the Mk III model, unsold examples of the Mk II version continued to be offered well into 1950 alongside the Mk III. We submit that the Falcon may have followed a similar path in relation to the Kestrel.

The company no doubt started with the assumption that the Falcon with its low price and "popular" displacement category would sell like hot cakes, and production levels were likely established accordingly. However, it seems entirely possible that sales of the Falcon were adversely impacted by the problems encountered with the earlier Vulture, which had likely in turn affected the marketplace view of the company itself. The fact that the Falcon looked pretty much like a downscaled Vulture may actually have contributed to sales resistance.

We saw earlier that the Falcon first appeared in April of 1949, and we noted that the Kestrel was introduced in August of 1949. We also referred to the fact that surviving serial numbers imply the production of perhaps 1500 examples of the Falcon in total. Based on the above considerations, the possibility cannot be discounted that all of the Falcons were manufactured prior to August of 1948 and that the Kestrel was a replacement for the Falcon rather than an addition to the 2 cc range. The implied production rate of some 300 Falcons per month (including the pre-release stock build-up in March) was well within the demonstrated production capacity of the K Model Engineering Co.

There is some supporting evidence for this in the form of Chinn's test report summarized earlier. Chinn returned the tighter of his two test Falcons to the manufacturers and was told that it was from "an initial batch which were assembled somewhat "tighter" than current models of which the second engine was an example" (our emphasis). Interesting! What was Chinn doing in late 1949 testing an engine recently "obtained from the manufacturers for test" that was from the "initial batch" of April 1949? Chinn unfortunately did not record the serial numbers of his test engines, but the inference here is that as late as the end of 1949 the manufacturers still had engines on hand from the "initial batch" known to have been in existence by April of 1949!

Suddenly, things begin to fall into place! It appears (if this scenario is correct) that the K Model Engineering Co came out of the starting blocks at a good clip and began the production of the Falcon with high hopes of good sales figures. Monthly production figures in the 300 range may well have been achieved or approached, as they undoubtedly were with other models in the K range.

However, the anticipated sales boom did not materialize. Worse yet, there were ongoing difficulties in consistently maintaining the desired close tolerances given the increasingly tired machinery which the company had no alternative other than to continue using. So by August of 1949 they had manufactured 1500 examples of the Falcon, many of which remained unsold. This would have made an existing adverse financial situation even worse—now they had substantial amounts of capital tied up in unsold inventory of both the Vulture and the Falcon!

What to do? Well, first accept the fact that the tolerances would have to be increased to accommodate the limits of the available equipment and establish more practicable limits for the ongoing production of the Falcon. Secondly, dress the revised Falcon up in slightly different clothes and promote it as a new model. The introduction of the "less carefully made" Kestrel was the outcome of these deliberations.

But they still had a good stock of unsold Falcons, just as they had with the Mk II Vulture at the time when the Mk III Vulture was introduced. The answer? Follow the pattern already established with the Vulture by continuing to sell the two models side by side and hoping in this way to liquidate the asset represented by the unsold Falcons.

A final piece of evidence which I personally find quite persuasive in support of this scenario is the fact that the alloy prop driver and head insert which were used on later versions of the Kestrel appear never to have been used on the Falcon. If the two models were produced concurrently, one would undoubtedly expect to find later examples of the Falcon also featuring the alloy components.

We repeat that neither of the above scenarios may be taken as established fact. We only claim that they do at least conform to the evidence as it appears today, as any scenario must do. Personally, I'm inclined to accept the second of these two scenarios as being by far the more likely. To me, it makes a lot more sense to envision the company simply revising its production tolerances and carrying on with a repackaged model than to see them accepting the challenge of producing two distinct variants side by side. However, readers are invited to make up their own minds. Any additional evidence or alternative fact-based interpretations will be most welcome.

In terms of actual production numbers for the Kestrel, we are once again fortunate in having access to a highly representative selection of confirmed serial numbers from a range of sources. At present, the lowest number of our acquaintance is K 201 and the highest is K1954. It thus appears that they started from K 1 and that some 2000 examples of the Kestrel were manufactured in total. This explains why it is undoubtedly the most commonly-encountered member of the K 2 cc family today. As noted earlier, a number of examples have been converted by collectors into "Falcons", but this does not affect the validity of the above comments regarding the serial numbers.

At present, we don't know exactly when production of the Kestrel ceased. The engine appeared in the final advertisement placed by the K Model Engineering Co in the November 1950 issue of Aeromodeller, but it seems almost certain that the company had ceased production long before that date and was simply liquidating its remaining assets. We probably wouldn't be far wrong in assuming that production of all models had ceased by mid 1950, and quite possibly earlier.

If this is correct, the Kestrel was in production for only 10 months at most. Even so, this implies an average monthly production figure of at least 200 engines, and it was almost certainly higher at the beginning, tailing off as the company management recognized the inevitable and commenced the sad process of winding up and progressively handing out employee redundancies. It seems probable that the monthly production of some 300 units which appears to have been approached with the Falcon was continued for a time with the Kestrel before the company entered its wind-up phase.

The K Tornado

As of early 1950, the K Model Engineering Co was clearly in deep trouble. Largely as a result of the doubtless well-publicised problems with the Vulture and perhaps quality control issues with the Falcon, the company's position in the marketplace had become seriously eroded, as witness the efforts that had been ongoing since mid-1949 to liquidate unsold stocks of several models which were no longer in production. Despite these efforts, as of early 1950 the company still held sizeable stocks of several of their discontinued models, including the Falcon. These represented a significant amount of capital which was unavailable to the company and was if anything depreciating. Time perhaps for desperate measures.

Glow-plug ignition had been enjoying something of a vogue in Britain during 1949, and the K Model Engineering Co had not stood aside from this, offering a glow conversion for the 5 cc Vulture at a cost of 10 shillings. As the company entered the new decade, a decision was evidently made that perhaps some of their remaining stock of Falcons could be liquidated very simply and cheaply by dressing them up in glow-plug guise.

The result was the February 1950 announcement of the final K model that was destined to be offered by the company, the 1.96 cc Tornado glow-plug engine. The evidence overwhelmingly suggests that this was simply an existing Falcon that had been converted to glow-plug ignition. The only differences lay in the treatment of the upper cylinder.

The Tornado was based upon the same crankcase, crankshaft and piston/cylinder unit that had previously appeared on both the Falcon and Kestrel models. However, the added length of the cylinder above the working bore to accommodate the contra-piston was of course not required for glow operation, and the cylinder was consequently shortened at its upper end to suit. It was presumably this requirement that dissuaded the manufacturers from offering screw-on glow conversions for the Falcon and Kestrel.

The contra piston and compression screw were naturally dispensed with. In their place, the engine featured a revised brass head insert which accommodated a glow-plug instead of the comp screw. A close examination of the head insert used in the Tornado reveals that it appears to be nothing more than a standard Falcon comp screw insert which has been re-threaded 1/4-32 to accept a long-reach glow-plug and has been re-contoured a little on the underside to match the piston! A reconfigured light alloy screw-on cooling jacket was used to retain this modified head insert, which sealed to the ground upper surface of the shortened cylinder with a gasket. The compression ratio resulting from this arrangement was a rather marginal 6 to 1, based on direct volumetric measurements and assuming the use of a long-reach plug.

The cooling jacket used on the Tornado appears at first glance to be a standard Falcon unit which has simply been re-machined to reduce the external diameter of the fins, change the head contour and make the bore conform to the shorter cylinder. However, a close examination suggests that it was in fact a separate purpose-made component. The head of the Falcon above the top cooling groove appears to have insufficient thickness to accommodate the stepped contour and flat top of the Tornado's cooling jacket. But whether or not it was a modified Falcon component, the production of the required revised cooling jacket was a very simple matter, involving only simple turning and tapping.

In all other respects, the Tornado was identical to its Falcon ancestor apart from being slightly lighter (as one would expect) at 3.7 ounces. Accordingly, there is no need to repeat the foregoing description here. However, we hope that the preceding comments have made it quite clear that the conversion of an existing Falcon to a Tornado was a very simple matter indeed. All that was necessary was to temporarily remove the cylinder, discard the contra-piston and comp screw, shorten the upper cylinder (apparently by grinding of the hardened component), modify the existing brass head button to accept a glow-plug and produce a modified cooling jacket to suit; all very simple tasks. Then re-assemble with a head gasket, and there you are! There are doubtless a few collector-converted examples of the Falcon (and even the Kestrel!) out there, masquerading as Tornados! Check the serial numbers ...

Our earlier claim that the Tornado was simply a glow conversion of previously-existing examples of the Falcon rather than an all-new production is supported by a considerable body of evidence, which I myself find highly persuasive. Let's see if I can convince you ...

Firstly, the Tornado never received the alloy prop driver which was used on the later examples of the Kestrel. Like all known authentic examples of the Falcon, it retained the steel item which had been used at the commencement of production of the K 2 cc models. If the Tornado was in fact a new production as of February 1950, we would undoubtedly expect to find the alloy prop driver being used. There seems to be no logic whatsoever behind the notion that that the company would have reverted to the older and heavier design on a new product built from scratch.

In addition, the fact that the head insert in the Tornado was made of brass instead of light alloy strongly implies the use of a modified Falcon component. By the time of the Tornado's appearance the companion Kestrel undoubtedly featured an aluminium alloy insert, and it's hard to understand why the company would revert to brass for their new model unless they were re-using existing components in the manner described above.

Moreover, the fact that the Tornado used a multi-finned Falcon "dog collar" to secure the cylinder is surely significant. At this stage the company was trying to keep their prices at rock bottom to attract customers, and maximum use would undoubtedly be made of existing components rather than making new ones. Moreover, if new components were to be produced, they would surely be of the simplest pattern possible. The fact that the manufacturers used the more complex Falcon dog collars for the Tornado strongly implies that those components were already there to be used ...

Further evidence that the Tornado was simply a converted Falcon taken from existing unsold inventory is to be found in the serial numbers. Although at present we only have two confirmed serial numbers for the rather elusive Tornado, these are quite illuminating. The numbers in question are supplied by my own engine and the example illustrated in Mike Clanford's well-known A-Z Pictorial book. These engines are respectively numbered K 7391 and K 7445.

The key point here is the fact that the numbers on the Tornado appear to be consecutive with those from the Falcon apart from the addition of the letter K. This observation is completely consistent with our working hypothesis that the Tornados were in fact glow conversions of a batch of previously-manufactured and previously-numbered Falcons which the company had been unable to sell. They merely did the glow conversions (a very simple production task, as we've shown above), added the letter prefix to the existing serial number to confirm that the engine was sold as a glow model and hoped for the best yet again!

Viewed in light of these observations, it appears that the usual designation of the Tornado as "the final K Model Engineering Co design to enter production" is in all probability quite misleading. In fact, all of the evidence summarized above supports the notion that it wasn't a new production at all—it was simply a re-packaging of a group of existing engines. Its introduction appears to be yet another case of the company taking a few remaining examples of a discontinued model that they were having trouble selling, dressing it up in new clothes, presenting it as a new model and hoping that the buyers would appear.

The other matter that is clarified by the above view of the situation is the frequently-expressed notion that the Tornado was a glow derivative of the Falcon or the Kestrel. In fact, the evidence suggests that it was nothing of the sort—it was a glow conversion of remaining stocks of the Falcon.

The simplicity of the conversion coupled with the fact that the Falcons were evidently not selling anyway allowed the manufacturer to offer the Tornado at the very competitive price of only �2 9s 6d, only 4s 6d more than the Kestrel but 10 shillings less than the asking price of the Falcon. Clearly this was another case of the company attempting to cut its losses...

The previously-noted February 1950 advertisement in which the Tornado first appeared also marked the swan-song of the Falcon in advertisements placed by the K Model Engineering Co The same advertisement made reference to a 2.5 cc model called the Falcon Mk II, but there is currently no evidence in the form of known surviving examples that this model ever reached the production stage. It may have been more of a "feeler" than anything else. It's actually by no means impossible that virtually all K production apart perhaps from the Tornado glow conversions had ceased at this point.

The fact that the original 1.96 cc Falcon disappeared from view at about this time creates a strong impression that as of March 1950 the company had liquidated their entire remaining stock of Falcons either by selling them or by rebuilding them into Tornados in the hope that the switch to glow-plug ignition might attract a few more buyers.

If this was the plan, it didn't succeed. For one thing, the initial flirtation period between British modellers and glow-plug ignition had virtually run its course by this time, at least as far as the smaller displacement categories were concerned. In those categories, the diesel now reigned supreme in Britain, as it was to do for the next 15 years or so. In large part this was due to the approach taken by the majority of British manufacturers (like the K Model Engineering Co) of simply converting an existing diesel model to glow-plug ignition. Example after example went to show that such conversions were almost never successful in that the glow-plug version signally failed to match the performance of its diesel progenitor. With the notable exception of ETA, British manufacturers were seemingly slow to learn that a successful glow-plug engine had to be designed as such from the ground up.

The Tornado was no exception to this—it fell well short of matching the performance of the Falcon and Kestrel diesel models. Word of this evidently got around fast, and the Tornado seems to have attracted relatively few customers. Unsold examples were still being offered by the company in their final advertisement which was placed in the November 1950 issue of Aeromodeller, all of ten months after the engine's introduction.

On the basis of the very limited serial number data-base presently available for the very elusive Tornado, it's impossible to say with confidence just how many Falcons received the Tornado glow treatment. We saw earlier that the highest known serial number for an original Falcon is 7462. The fact that the next-lowest number is 7074 leaves a significant gap without any Falcon serial numbers. The only two numbers for authentic Tornados in or possession are K 7391 and K 7445, which fall squarely within the gap in the Falcon sequence. It appears from this that perhaps some 300 high-numbered Falcons may have been set aside for receiving the Tornado glow conversion treatment but the factory never got around to converting all of them before manufacturing operations ceased. This would completely explain the existence of Falcon number 7462.

In summary, the implication of the numbers known to date is that perhaps 1500 examples of the Falcon were originally manufactured but only some 1250 of these were actually sold as Falcons—the other 250 or so eventually went out the door or into the scrap bin as Tornados.

Whatever the actual numbers, it's clear that the number of Falcons converted to Tornados was pretty small. The Tornado is certainly by far the least common member of the K 2 cc clan today.

Kestrel vs Tornado—a Head-to-head Comparison

I'm a long-time Kestrel owner and have had many happy hours using this well-mannered and dependable little engine. Just look after those mounting lugs by installing it properly and keeping it out of the ground, and it'll run forever for you!

Thanks to the kindness of my valued friend Alan Strutt, I'm also lucky enough to have a nice example of the Tornado on hand for test purposes. What an opportunity! Let's put them on the bench head to head and see how they fare.

In deference to the Tornado's rather low compression ratio, I fitted the hottest long-reach plug that I could find and also elected to use an "oily" fuel having a 10% nitro content. I realize that nitromethane was pretty much in the unobtanium category in Britain as of 1950 and was prohibitively expensive even if one was lucky enough to find a source. Therefore, few British glow-plug users at the time would have been in a position to use anything other than a straight methanol/castor mix. However, I wanted to give the engine a fair chance, hence my decision to use a bit of nitro.

Having reviewed the Model Aircraft test of the Falcon, I had a pretty good idea of the speed range on which to focus and selected my airscrews accordingly. These are not high-speed engines, and the very low compression ratio of the Tornado glow model looked to me like a factor which would limit the operating speed range of that unit too. Even so, I stayed away from the larger props with the Tornado given the fact that ignition timing control is pretty iffy with glow-plug ignition and I didn't want to over-stress the working components through the common glow-plug problem of pre-ignition at low speeds.

As an old hand at running Kestrels, I anticipated no testing difficulties there, nor was I disappointed. My mate David Owen has characterized the dear old Kessie as "a little darling of an engine", and I've never seen a more apt description! A small prime, a few flicks, and away it goes, every time! Puts its similarly-designed larger relative (the Vulture) to shame in this regard. It's also very responsive to the controls and hence extremely easy to set. Extracting the propeller/rpm figures was a real pleasure!

I actually tried two examples of the Kestrel during the course of these tests. They yielded very comparable figures, although the more "experienced" of the two tended to be 100 or so rpm up on its companion across the board. This kind of discrepancy is well within the range of variation to be expected with series-produced engines having had different levels of use. I've used the results from the slightly faster engine for the purposes of this comparison.

The Tornado glow model upheld the family honour in one respect—it was extremely easy to start. Like most glow motors, it needed a healthy prime but was very responsive once the plug was lit up. It was of course far easier to flick over, and in fact required only a quite lazy flick to start. It would have made a very good beginner's engine.

The low compression ratio did manifest itself once a start was achieved—the engine needed to warm up for a few seconds before the plug could be disconnected without stopping the motor. Once this was done, it was quite easy to find a needle setting at which the engine ran very smoothly with no trace of sagging. Needle control was in fact very positive without being overly sensitive—the setting for optimum performance covered about a quarter-turn range.

At the optimum setting, the engine ran very smoothly indeed without missing a beat. A further manifestation of the low compression ratio was the fact that once two-stroking, it tended to pick up speed quite noticeably as it reached maximum temperature, which of course causes the ignition timing to advance a little. For me, this is further confirmation that the 6 to 1 compression ratio was perhaps a bit lower than ideal for the best ignition timing. A switch to 7 to 1 or even a little higher might have worked wonders.

So how do the two models compare in performance terms? Here it has to be admitted that the glow Tornado confirmed its reputation by failing to approach the performance of its diesel relative, even with the help of the 10% nitro fuel which I used. At the lower speeds it was no contest, although the glow did make a bit of a run at the diesel at the higher speeds. However, the majority of British modellers at the time had yet to appreciate the need to use smaller props on glow-plug motors and consequently tended to over-prop their glow engines. Many contemporary users would have fallen headlong into this trap and would likely have been quite disappointed as a result..

The propeller/rpm figures obtained are shown in the accompanying table. The cited BHP figures cannot be taken as precise since they are merely implied based on the established power coefficients for the props in question. However, these coefficients do generally give figures which fall well within credible limits, and they undoubtedly form a valid basis for direct comparisons given the fact that the exact same props were used for both engines.

The figures shown in the above table plot together to form reasonably coherent power curves for each of the two engines. It's immediately apparent even without looking at such plots that the Tornado falls well short of matching the performance of the Kestrel. The figures depicted above seem to imply that both engines achieve their peak power delivery at around 9,600 rpm, just as reported by Peter Chinn.

The "experienced" example of the Kestrel which yielded the above figures appears to have slightly exceeded the performance of Chinn's example of the Falcon—the implied peak is around 0.135 BHP or thereabouts. By contrast, the Tornado seems to max out at around 0.110 BHP, albeit at more or less the same speed as the diesel. This latter point seems to offer further confirmation that it was the porting that limited top-end performance in these engines. In the case of the Tornado, a slightly higher compression ratio might have worked wonders by optimising the ignition timing for a slightly higher speed. As it is, more nitro would doubtless reduce the performance gap somewhat.

Still, the Tornado isn't completely buried by the diesel—it would doubtless do a creditable job of flying a model of similar size, and the weight reduction would certainly be a benefit. In fact, the 10% lower weight of the Tornado does much to minimize the difference in the power-to-weight ratios for the two models. And there's no doubt that it's a real pussy-cat to handle—I certainly enjoyed testing it!

Conclusion

The K 2 cc series represents the final displacement category into which the K Model Engineering Co threw its hat. It's likely that this series was initially seen as the road to redemption from the reverses suffered by the company as a result of the problems encountered with the Vulture. Sadly, their introduction proved insufficient to save the company in the long term—the damage had been done, and there was no recovering from the consequent erosion of the company's capital.

This is a great pity, because the K 2 cc diesel models were (and are!) excellent little engines with few of the vices of their big brother and an entirely satisfactory performance for their intended purpose. Even the glow Tornado has its charms too. The engines may be run with complete confidence and will give any owner many hours of enjoyment. So if you get a chance to acquire one of these nice little units, don't hesitate; I never have!