Lagonda Rapier Valve Operation

Article on Rapier Valve Operation by John Macdonald

Have you ever wondered why the valve operation on the Rapier is so unusual, using as it does a finger follower with a convex face? This arrangement makes it very difficult to design a conventional cam profile which will give good results and there are differing opinions about how this should be achieved.
More recently I have had time to study this matter and some interesting facts have come to light. It is my understanding that the engine designer, Tim Ashcroft, was regarded as a bit of an expert on engine valve gear so there was probably a good reason for the method chosen – this has turned out to be the case.
By studying the valve operation on racing engines of the period it soon became clear that the curved finger follower was a feature on some of the most successful engines of the period and there is a very good reason for this design.

It is well known that camshaft design is always a compromise. A camshaft designed for low speed torque is unlikely to be ideal at high revolutions, similarly a camshaft designed for maximum power will need to run efficiently at higher revolutions and is unlikely to produce good low speed torque. Some modern manufactures, such as Honda and Alfa Romeo have used forms of variable valve timing to try to achieve the best of both worlds but this is complex and only of marginal worth. A much simpler system was known in the early 1930’s which increased the useable revolution band.

One has to bear in mind that at the time the Rapier was conceived it was fashionable to have a car which would pull strongly from low revolutions, the Lagonda M45 at the time was very highly regarded because of its immense low speed torque but it had a modest rev limit. The Rapier, being a small engine could be endowed with a decent amount of low speed torque to the detriment of overall power but Ashcroft wanted it to rev freely and to achieve both was going to require some clever camshaft design.
The usual way to achieve low pulling power with a camshaft is to limit both the time that each valve is open and the overlap but this stifles the gas flow at high revs where longer duration and overlap are required. With a conventional camshaft profile the valve is lifted from the seat and gradually accelerated to maximum lift. The initial valve aperture is quite small and it may well take around 100 crankshaft degrees to be fully open. On a high speed engine therefore the valve duration has to be quite long so that the total valve aperture can pass the necessary volume of gas, however at low speed this long duration is a disadvantage.

If it was possible to fully open and close the valves very quickly without overstressing the valve gear this would allow for a relatively short duration and narrow overlap to enhance the torque but with a bigger total port area to enable high speed operation. The way this was achieved in the early 1930’s by such manufacturers as Bugatti (Tipo 57), Fiat (402.403), Frontenac (in 1920), Mercedes Benz (M25) etc on their racing engines was by the use of a finger follower with a convex face thereby combining a relatively light follower with the ability to use a re-entry or hollow-flank cam profile. Other successful manufactures such as Maserati and, indeed the Lagonda V12 used a convex follower although not with a finger follower.

The hollow flank cam is the means by which the aforementioned manufacturers tried to achieve good torque as well as good power.
For those who are not familiar with the meaning of “torque” and “power” may I be permitted a few lines to make clear what these mean.
TORQUE:      A measure of how much a force acting on an object causes that object to rotate. Imagine using a socket and bar to tighten a nut. If you exert a pull of 50 lbs on a 2 ft bar the torque is 100lbs ft. (50 lbs x 2 ft). The same torque can be applied by pulling 25 lbs on a 4 ft bar (25 x 4)
POWER:         We are referring here to “Brake Horsepower” (bhp). In simple terms it is the torque measured on a brake (dynamometer) and multiplied by the speed.

The formula is  P = T x N
                                5250                            where P = BHP
                                                                               T = Torque in lbs ft
                                                                               N = Revolutions/minute

So 50 lbs ft @ 1,000 rpm = 9.52 BHP
and 50 lbs ft @ 5,000 rpm = 47.62 BHP

i.e. more revs, means more power.

It follows therefore that BHP is not a very good comparison for different engines or states of tune as even low torque engines can appear powerful if very high speeds are used.

The way that the hollow flank cam works is actually quite clever – the follower fits into the hollow flank allowing the valve to remain closed for longer but then open fully very quickly and remain fully open almost to the point of closure when it is then rapidly lowered back to closed as the follower settles into the hollow flank on the other side of the cam. This allows a relatively short duration and overlap for low speed torque combined with a larger total valve aperture to pass the necessary volume of gas for high speed power. Whilst the rate of lift is much greater than with a conventional cam the initial phase of lift is actually more gentle. This applies also to the final closing so the effect on the valve gear, even at high speed, is not as serious as may be imagined. Indeed, the Honda CB 450 motorcycle engine used a convex finger follower in the 1970’s which was safe to 12,000 rpm.

There is one final clue to the theory that the valve gear was designed for a hollow flank cam profile and this is the fact that the cylinder heads were designed for camshaft dampers which would damp out the snatching which is inherent with this form of cam profile.

So that is the theory, now I suppose you are wondering if this will work in practice. With this in mind I set about producing a suitable profile for test purposes and with expert professional help was able to produce a pair of camshafts for testing. The profile chosen was a little wilder than would have been considered in the early 1930’s because with modern technology everyone expects more performance and there is a need to be able to match modern cars on the road. We may have sacrificed the low end torque a small amount to achieve a better top end but testing on the dynamometer would prove the point.

The first test engine was already fitted with Kent competition camshafts which we had timed to get best power for the circuit but on a basically standard engine. A power/torque curve was recorded on the existing camshafts which were then removed and replaced with the hollow flank test cams. Without changing anything else the car was re-tested and the power/torque curves recorded. The results surprised even us. The torque at 2,500 rpm was roughly doubled resulting in a power increase of that amount but this was largely because the Kent cams were sacrificing low end torque for top end power. However the increase in torque from 3,500 rpm to 6000 rpm was averaging more than a 35% improvement resulting in a power improvement to 6,000 rpm of similar proportions.

The second engine tested was also largely standard specification with new connecting rods for safety. The camshafts were original later type but with the exhaust profile ground onto the inlet. These had been set up to give best torque for road use. Again we changed only the camshafts. This time the torque produced up to 2,500 rpm was pretty much identical but whereas the original torque peaked at 2,600 rpm and then fell away steeply the torque from the new cams climbed from 2,600 rpm to peak some 25% better at 3,300 rpm and declined only marginally to 6,000 rpm before falling away. The result was a much more powerful engine without any loss of low end torque.

These results would tend to show that this is most probably what Tim Ashcroft had in mind but he was perhaps not given the finances or the time to make the camshafts as intended. Having proved the theory I believe that less radical cams of this type would have provided even stronger low end torque as well as the ability to rev to 5,500 rpm as originally intended.

It is interesting that the published power output for this engine was 45 bhp at 4,500 rpm but an early “Light Car” road test found the performance to be disappointing. Of the many more or less standard engines we have tested over the years none has come anywhere near this figure but even with our slightly wild hollow flank cams this figure is readily available. In our case, however, the maximum power is considerably more and peaks at around 6,000 rpm.