Development of the QUB 250cc twin cylinder

It was in June 1967 that Gordon Blair, M Johnston and their associates at Queens University Belfast conceived a new type of inlet valve for a two-stroke engine. A 250cc water cooled twin cylinder engine with this rotary sleeve valve was designed and built from scratch and first road tested in 1969.

The rotary sleeve valve was described by Blair as a “disc valve translated into cylindrical terms”. See the following diagrams for details of the rotary sleeve valve. The valve (A) rotated on two bearings (D and E). Gas sealing between the cylinder, sleeve and the open ends is done by using labyrinth type seals (G). The running clearance between the stationary surfaces and the valve is 0.003”. The drive to the valve comes from a toothed rubber belt from the crankshaft (2:1 from the crank).

One advantage of this type of valve over a two-stroke disc valve engine is that the sleeve valve engine can be much narrower. A twin cylinder disc valve engine has a valve and carburettor on both ends of the crankshaft. Fitting of the primary drive to the gearbox and ignition system also becomes more difficult with a disc valve engine. For example the primary drive will come out between the two cylinders, making the whole engine package wider. The extra engine width limits chassis design and can have a negative effect on machine handling. The width of the QUB engine was 13”, which compares favourably to the 250cc units from Yamaha and MZ of the period (20” and 22” wide respectively).

The team at QUB did acknowledge some of the limitations involved with the use of their rotary sleeve valve. The first of these was that in order to obtain compact valve dimensions to achieve port timings of 40° abdc opening, 70° atdc closing, and approximately 50° from the fully closed to fully open position, the inlet port could not be a round hole, otherwise a valve with a diameter of 6” would be required! Instead a rectangular inlet tract was utilised with dimensions of 2.5” wide and 0.6” high in conjunction with a 2.5” diameter valve in order to achieve the required port timings. The addition of a conventional carburettor to the inlet port would produce a step change in section from round to rectangular, and have the associated large coefficient of discharge losses. Therefore the QUB team designed a “low pressure fuel injector with a barrel throttle valve”. Fuel was supplied from a varying section spindle set into G in the diagram. QUB also suggested that there would be a weight increase penalty of approximately 10 lbs due to the valve, inlet tract and fuel/air metering system.

Looking back with hindsight it can be seen that the inlet tract design has some other limitations. The use of the rotary valve means that the inlet tract is relatively long, which is less than desirable on a high performance engine. The inlet tract has also has a poor pathway into the crankcases. The port follows a ‘blind path’ to a wall before being expected to turn down into the cases.

For the port plan shown in the diagram QUB used an exhaust port timing of 201° and transfer port timing of 142°. These timings were for a designed 120 lbf/in² bmep at 11,000RPM resulting in a power output of 50 bhp.

The cylinder head of the engine ensured the squish area covered 50% of the piston area, at a clearence of 0.020”. The trapped compression ratio of the engine was 7.6:1, whilst the actual compression ratio was 15:1. The engine was water cooled and two separate cylinder heads cast out of RR50 and sealed by a single aluminium cover with an O-ring.  

A Hepolite piston with a single dykes ring was used in the QUB 250cc. The single L-shaped piston ring was made from nodular iron. A 12mm gudgeon pin was utilised with an INA roller bearing in the connecting rod little end. The pistons run in cast iron liners with a 0.006” clearance. The liners are shrunk into the alloy water jacket and then honed to size with a surface roughness of 12-15µin c.l.a. The bore was 57mm and the stroke is48mm, giving a bore/stroke ratio of 1.19.

The steel I section connecting rod had the small and big end eyes nitride to a depth of 0.010” to Rockwell 60c for the needle roller bearings. The big end was a 'Durkopp' cage from Ransome, Hoffman, Pollard, SK 22 x28 x 13 FV.  

The original crankshaft for the engine had the crankpins and mainshafts held in with Loctite, as opposed to being pressed together. For alignment a jig was used to hold the crankshaft assembly whilst the joining compound set. During early service, chattering was noticed at the crankpins. The crank design was then modified to incorporate interference fits for the crankpins. The crankpins were pressed into position with an interference of 0.0035-0.004”. The mainshaft joint was retained as a Loctite joint. Double row needle roller main bearings were used, with a ball bearing on the timing side to stop end float. The crankcases split horizontally.

The water pump is driven by a rubber belt at 3.5:1 from the rotary valve (7:1 from the crankshaft). And has a flow rate of 10 gal/min. A thermostat controlled the block temperature to 88°C. An oil pump was driven off the water pump which delivers oil to the intake tract on top of the big end bearings. Castrol R40 oil was used and also mixed in the fuel to make a petroil mix of 25:1. A Lucas transistorized ignition system was used.

The engine was coupled to a 5 sped Albion gearbox via a Morse “Hy-Vo” chain and mounted in a frame made by Colin Seeley. The total weight of the machine was 265lbs.

The engine was run up on a Heenan and Froude DPXO dynamometer. The ignition timing was optimised at 11,000 RPM and found to be 21.5-22°BTDC. There was a very steep rise in power from the engine from 9,000 RPM and a sudden cut-off at 11,250 RPM. Peak power was found to be 45 BHP, but it was expected that with some development 10% more would be achieved.

In 1969 the machine was raced on two occasions, but it broke down due to small end bearing failures. The best performance on these occasions was a 92 mph of the Ulster GP circuit at Dundrod.