Kemp K4


Name Kemp K4 Designer Harold Kemp
Bore 5/8" (15.88 mm) Stroke 7/8" (22.23 mm)
Type Compression Ignition Capacity 0.268 cuin (4.4cc)
Production run Low Country of Origin England
Photo by Graham Podd, Eric Offen Year of manufacture 1947-48

 

Background

The K4 was the first commercially produced engine made by Harold Kemp in his workshop at 7 Bank Street, Gravesend, Kent. The engine was expensive to produce and was only offered for a small time, but Harold continued with smaller engines producing 1cc and 0.2 cc diesels in small production runs. This latter engine proved very popular, being the smallest British production diesel of the time. Harold sold the Kemp Engine business in 1948, with "K" Model Engineering Co Ltd taking over at the same premises. The Kemp 4.4 was dropped from the line, but the 0.2cc diesel was retained and revised into the K Hawk. For more details on Harold Kemp, see the Harold Kemp Tribute page.

We are fortunate to have scans of a rather tattered original drawing for the K4 which has permitted us to redraw the engine under CAD, taking the usual liberties to simplify construction for modern times, while retaining the outward looks of the original. The General Arrangement at the head of this page shows the re-drawn engine, although the details that follow reflect the original design. This picture shows the title block and materials call-out table, possibly drafted by Harold Kemp himself.

Construction

The Kemp K4 was based on a number of intricate die castings, initially aluminium, later magnesium. Internally is was a conventional side-port, "long-stroke" design, popular at the time. A simple tubular liner slid into the bored case to sit on a lip at the bottom of the case bore. This was held seated by the head casting which was secured by four 6BA screws. The crankshaft was nickel steel, drilled axially to reduce weight, and hardened. This rode in two separate cast-iron bushings pressed into either end of the magnesium nose casting which was very well secured to the main casting by six 6BA screws. Details are sketchy regarding the prop drive washer. The front of the crankshaft was a simple step from the 3/8" diameter journal to 5/16", the latter section being threaded 5/16-40 TPI—an unusually fine thread for a crankshaft!

The connecting rod was called out as case hardened mild steel, or 3% nickel steel, with a hardened big end. The big end was reamered 5/16" in diameter; the little end, 3/16". The piston was a light weight, two piece assembly. A steel gudgeon pin carrier was topped with a stub threaded 1/4 BSF. This threaded (tightly) into the hardened steel piston sleeve and was lightly peened to assure they stayed together. A wise decision seeing the thickness of the piston crown was a mere 1/16". The 1/4" BSF thread has a not so fine pitch of 26 TPI, so that crown would have had a mere 1.6 threads; provided that is that the die could run right down to the shoulder of the gudgeon yoke! I suspect it was the riveting, not the screwing that counted. The compression screw was a conventional L-bar with a 1/4-40 TPI thread—which I'd have thought a better thread choice for the yoke pin, but I'm sure Harold had a good reason for his design choice. The contra-piston (charmingly termed the "obdurator" on the plans) was unhardened mild steel. Curiously, it featured very thin 0.037" walls, similar to the DCO contra method favoured amongst the Motor Boys today. To complete the compression adjustment department, the magnesium head casting was drilled 13/32" diameter for a pressed-in, threaded, mild-steel bush.

To complete the engine, a cast magnesium backplate was very well secured using the same six bolt pattern as the main journal casting. This must have been faced at the rear, as the cast magnesium fuel tank replicated the bolt pattern and was secured as a simple but-joint against the backplate by the 6BA screws. The venturi was yet another magnesium X shaped casting, drilled through #18. The rear face had a one-sided flange like that often seen on Westbury carburettors for tapping to take a swinging shutter type choke plate, although none was fitted. The vertical sections were threaded 1/4-40 on the outside for the needle valve thimble and a brass fuel pickup tube/nut assembly. The venturi threaded into the crankcase using a 5/16-40 thread. No jam-nut was fitted, so air leaks would be likely unless the thread was very carefully cut. The fine thread used would also assist in aligning the fuel pick-up with the hole in the tank.

On the factory drawings, the needle valve is shown as comprising a needle pointed length of 0.1" diameter silver-steel, threaded 7BA on the blunt end to screw into a brass carrier where it was secured with lock-nut (the major diameter of 7BA is 0.098", so this makes perfect sense). Production engines seem to have abandoned this expensive design in favor of soldering the needle in place. The needle seat and jet were formed by a length of 3/32" OD copper tube pressed into the lower part of the venturi casting so as to project slightly into the carburettor bore. A note on the plan states (in part) "Close end to suit needle".

Evidence exists through the example in Eric Offen's collection of a more interesting assembly to provide a fuel cut-out, as was popular for free-flight work. This may have been an optional but is certainly was complicated and intricate! It comprises a spring loaded needle with an over-center, cam operated lever, drilled to take an actuating wire, or thread. As can be seen in this modern reproduction, with the actuating arm vertical, the cam action of the lower part of the arm draws the needle upwards, compressing the spring and hopefully locking the needle into the thimble so it will not move vertically, while still being somewhat free to pivot. In this state, the needle thimble will move the needle up and down in the usual manner, although the needle itself can rotate in the carrier—an important factor as we shall see.

British timers of the day were mostly of the type that gave a pull when activated—be they of the Auto-Knipps, or ED clockwork variety, or of the air-hydraulic type. Regardless, they would generally result in a line providing a pull towards the rear of the airframe which could be connected to the hole in the top of the cam arm. To function, the cam arm needs to be oriented so that the slot is aligned fore-aft. When the timer activates and pulls the line, the arm can swing backwards. As it passes over-center, the spring will assist, snapping the needle down onto the jet seat, cutting off the fuel supply. Tricky and fiddly to make, assemble, and adjust. Plus as a bonus, air leaks would be almost inevitable making repeatable, stable needle settings uncertain. But what the hell? It may be complex and not eminently practical, but the uniqueness has a certain charm to it that appeals to the non-engineer in me (while the other ninety-eight percent cringes in horror ).

Conclusion

The K4 enjoyed a rather brief existence, and magnesium castings do not weather well, seeking as Bert has said, always to return to the sea from hence they came. So the K4 is a rare bird today, but the little Kemp 0.2cc engines are still around, as are enough of the 'K' Engineering products to make them prized items in many collections.

There is some question as to the part, if any, that Harold Kemp played in the design and development of these later 'K' series engines, having sold the business in 1948. Certainly, his career as a engineer took a different, but still highly productive direction, but that was not until 1952, by which time the 'K' series had come and gone. Sadly, Harold passed away in 1975, so a definitive answer is difficult. But notice that the circled 'K' emblem visible at the foot of the K4 drawing sheet and cast into the bypass of the K4 itself was retained by the new company as the logo cast into their engines, suggesting some carry-over. Regardless, we can state with certainly that the the K4 and the 0.2cc Kemp were his designs, placing him amongst the pioneers of the British diesel.

 

Ruler

 

  Model Engine News Home

Please submit all questions and comments to enquiries@modelenginenews.org