aero and automotive engines and engineers

  by Richard Hodgson

This 1993/4 article first appeared in the Rolls-Royce Heritage Trust's (Coventry Branch) Magazine "The Sphinx".  Please use your Web Browser's "Back" button if you want to return to this article after visiting any hyperlink which are shown in bold dark blue.  Please click here to go to the list of articles on this site's home page.

Aero Engine Experimental Department Report No. 353

Report by S. Allen of Work Carried Out on a 16 H.P. Car Engine

A copy of this revealing experimental report has been found by [the late] Mervyn Cutler of Tiger two-speed supercharger and Sapphire fame and to whom I would like to express my thanks.   A copy is in the Branch's library [1993].  I would also like to thank [the late] John Densham, Sid Allen's assistant from 1938 - 1942 and the Car Division's Experimental Engineer from the late 40's to the mid-50's, for his kind help.

The "all-new" 16hp cars, the Hurricane, Lancaster, and later Typhoon, were announced in May 1945 and were the result of work carried out during 1939-40(/1?), and from 1942 to 1945.  Most of the car work after 1940 was carried out at Crackley Hall by the three pre-war car musketeers: Allard, Thornett and Holister.  The 16hp's were the first cars made by any British company after the war that were not merely pre-war production models slightly brought up to date.  Though well received by the motoring press, the car division and elements of the motor trade were only too well aware that, with their essentially pre-war 1991cc engine, the otherwise splendid 16hp's would be considered underpowered in the brave new post-war world.  This indeed proved to be the case, as recorded by Donald Bastow on page 189 of W. O. Bentley - Engineer (Haynes 1978):

Armstrong Siddeley's immediate post-war car, which was in production very quickly, had a 2 litre six cylinder push rod engine and three body styles of which the roomiest was the Lancaster Saloon, a four door, 4-5 seater. It was generally accepted in the industry that the engine was not powerful enough for the size and weight of car.

W.O. Bentley and Bastow had been appointed to design a 3 litre ohc engine (and car) in the late 1940's after the failure of Thornett's 2.6 litre aluminium unit which, as Bastow states in his book, was designed to rectify the above state of affairs.  The 2.6, called the "2.5" by ASM, and the fabulous, massively oversquare, all-aluminium 4.5 litre ohc 4 main bearing V8 Bristol Siddeley Special Projects Division (this division was part of the aero-engine company) design of November 1959 (not the proposed V8 Star Sapphire derivation) will be the subject of later articles.

The car division were all too aware of the 16hp engine problem by late 1945, but were unable to improve output without a fundamental redesign.  Realising the potential seriousness of the situation in the mid-to-long term and knowing all too well that an all-new piston engine can rarely be designed and developed in a short time by a minuscule team as then existed, the obvious answer was for the highly experienced aero-side to consider the situation.  John Densham recalls that the rumours (and later the appearance) of the new Jaguar ohc engine acted as the main catalyst for action.  Sid Allen, best known for his remarkable work on vaporising combustion chambers and rockets, had been in charge of, inter alia, Deerhound MK.III single cylinder testing.   He had also played a part in the design of the cancelled 24 cylinder Wolfhound.   Allen was asked as a matter of some urgency towards the end of 1945 to investigate the 16hp engine.  The brief to Allen stated that:

"... recommendations should be made as to what steps could be taken to improve its performance."

He and his team investigated engine No. 160042.  Leading particulars were:

        6 cylinder 65mm bore x 100mm stroke

        Compression ratio:  7.0 : 1

        Inlet valve opens:  10 BTC, closes: 54 ABC

        Exhaust valve opens:  48 BBC, closes: 16 ATC

        Inlet and Exhaust lift:  0.343"

        Inlet diameter:  1.2"

        Exhaust diameter:  1.125"

        Valve seating:  45

        Depth of inlet seat:  0.05"

        Depth of exhaust seat:  0.075"

        Ignition timing set:  3 ATDC

        Filtrate Lubricating Oil; "Pool" petrol (nil lead content)

At full throttle and 4400rpm, the engine actually developed:

        70hp with the test bed exhaust system

        65hp with the standard Burgess silencer without flexible tail pipe

        65hp with a Cheswick and Wright silencer without flexible tail pipe

        61.5hp with a Cheswick and Wright silencer with a flexible tail pipe

        60.5hp with the standard Burgess set-up with flexible tail pipe (EN.54261)

After testing the engine both on the test bed and in a car, Allen came to the conclusion that both the indicated and brake horse powers developed by the engine were:

"considered to be lower than those which may be reasonably expected from an efficient engine of this type".

His first discovery, and probably the most startling, was that the mechanical efficiency rapidly decreased at higher speeds.  Up to 2400rpm, efficiency was uniform at 85%, dropping to 80% at 3700rpm and to about 72.5% [sic] at 4400rpm.  He noted that just the mechanical losses at 4400rpm amounted to 25hp!  He tactfully stated that he had no data on mechanical losses to be expected in a car engine, but added that the generally accepted figure for a naturally aspirated liquid cooled in-line aero-engine was about 90%.  In his textbook "The Elements of Motor Vehicle Design", 2nd edition, Oxford University Press, 1935, Donkin gave a figure of 80 to 90% for then modern engines.  Allen stated that the power absorbed by the fan at 4400rpm was "negligible".   [See also article on the RAC h.p. formula].   He added that further work was required to discover the cause of the mechanical losses.  It is now thought that much energy was lost as friction in the cylinders and to a certain extent in the main bearings - the crank set-up on the 16hp not being the "ideal" 6 cylinder 4 main bearing arrangement, unlike the Sapphire.

It was also discovered that there was a low indicated mean effective pressure "due to low volumetric efficiency [about 81% on induction pipe conditions between 1600-4400rpm, but rising rapidly below 1600rpm] and possibly due to inefficient combustion under certain conditions".  He concluded the low volumetric efficiency was due to a reduction in pressure due to losses in the carburettor, induction pipe and induction manifold, and due to a reduction in density of the mixture by heating in the induction pipe manifold.  Allen considered that by properly designing the induction system, separating the manifold from the cylinder head, putting the exhaust and inlet manifolds on opposite sides of the cylinder head and using larger valves [entailing a decrease of the stroke/bore ratio]: "we may hope to halve the mixture temperature rise and to halve the pressure loss".  He also identified the valve timings as potentially having an effect on the heating of the incoming charge, as unsuitable exhaust timings could increase the weight of the residual exhaust gas.

As to combustion, the fact that indicated thermal efficiency based on both air and fuel consumption increased as rpm was increased pointed to either unsuitable ignition timing or bad design of combustion space.  Correct timing and an efficient combustion space should lead to no appreciable difference in indicated thermal efficiency over the running range.  He recommended the adoption of hemi-spherical combustion spaces, and further experimentation on timing, using as a starting point two sets of timings:

        Inlet opens 10 early, closes 54 late
        Exhaust opens 58 early, closes 16 late

        Inlet opens 10 early, closes 59 late
        Exhaust opens 48 early, closes 26 late

The standard exhaust fitted to the 16hp with flexible tail pipe caused a very serious back pressure at higher engine speeds, causing a yet further appreciable loss of power.   Allen suggested it was merely a matter of experiment to find a combination of silencer and tail pipe that gave the minimum of resistance with the necessary noise level.   The output figures quoted above speak for themselves - there being a 9.5hp difference between the test-bed exhaust and the standard set-up.  The following acceleration figures from the report are of interest:

        Speed Range       (mph)        Time for Acceleration (secs)               Exhaust System

        Speedometer     Actual         With Wind      Against Wind

         65 - 75            (59-69)             11.5                      18.1              }     Standard set-up with
         70 - 80            (64-73)             17                         >34               }     flexible tail pipe.

         65 - 75            (59-69)             11.7                      15.8              }     Standard set-up without
         70 - 80            (64-73)             16                         21.4              }     flexible tail pipe

Even the results for the standard silencer set-up without the flexible pipe could be said to explain John Densham's famous reference to: "Those gutless wonders!"

Of course, some of Allen's suggested changes would have entailed an alteration to the fundamental design - a time consuming, expensive, process.  These would have been the adoption of hemi-spherical combustion spaces, placing the exhaust and inlet manifolds on opposite sides and a decrease in the stroke/bore ratio to permit the use of larger diameter valves to increase volumetric efficiency.  Echoes of the Sapphire engine ....

When Allen's report arrived on 31st January 1946, it only confirmed the car division's fears.   A new engine was required.  Thornett immediately set about designing a 2593.7cc near square 6 cylinder 7 main bearing wet-linered ohv aluminium engine referred to at the beginning of this article.  It was ready for the test bed within a year.   Despite heroic efforts by the car experimental department, that engine remained a "turkey" and was eventually dropped, to be replaced by the Bastow/Bentley 3 litre ohc engine (and chassis).  Due to "certain problems" [not the least being financial considerations], the 3 litre was itself replaced by the well-known ohv Sapphire engine (and chassis) of which the car division was especially proud.  In the meantime, the 16hp’s lack of power was attacked as follows:

(1)  Fuel Injection [sic]. By July 1947, extensive tests were being carried out on various fuel injection schemes, including direct injection;  and then,

(2)  When it was discovered that fuel injection was not the cure for the 16hp's ills, the well-known 18hp variant was introduced as a stop-gap until the A.S. Bentley was due to be ready.  Even the 18hp was not considered a very satisfactory stop-gap, and John Densham (who had been a junior draughtsman on the V12 Lagonda, one of the "supercars" of the 1930's) designed a decidedly sporty high lift high overlap camshaft in August 1948 for test in EX 2.3 - 2, but the basic problems with the engine remained.  As a matter of record, the Densham camshaft gave an increase of some 7 hp at 4300rpm, 4.5hp at 4000 rpm, 2.5hp at 3750 rpm, and less than 2hp at 3000rpm.  Below 2250 rpm, the standard 18hp camshaft produced more power.


Richard Hodgson 1993, 2000 - may not be reproduced in whole or in part without permission