With the VSCC quite rightly opposed to newly-built “old” cars and the rules governing historic vehicles due for revision, it is interesting to look at how an accomplished engineer tackled the task of recreating a very significant 1904 car whose chassis had been scrapped as long ago as 1908.
I refer to the 15-litre Napier L48 “Samson” which Bob Chamberlain has revived in Australia, and my information is based on data supplied by Chamberlain himself. It is important to make it clear from the outset that he has never claimed that “Samson” is the original car; rather, having come across its original engine, he has recreated the world’s first successful six-cylinder racing car to house it.
Designed by AJ Rowledge (who was later responsible for the famous Napier “Lion” aero-engine before moving on to RollsRoyce), “Samson’s” career has already been well documented (Motor Sport, November 1982 and Anthony Heal’s 1951 articles). Suffice it to say that in SF Edge’s hands it lost its 1908 Brooklands Match Race against Nazzaro’s Fiat “Mephistopheles” when the crankshaft broke in its replacement 20-litre engine, but was very successful in speed-trials here and in America, and won the 1908 Montagu Cup race at Brooklands, where it took records at up to 119.34 mph. It was regarded as the fastest British racing car of its day.
When the chassis was scrapped, both L48’s engines went into racing boats, the first being used by the Cornwall brothers in Australia to win the 1914 and 1915 Griffiths Cup races. They later installed a Sturtevant aero-engine in Nautilus 2 before going off to the war in Europe, and the 15-litre Napier power-unit lay disused in their pottery works until Bob Chamberlain heard of it, and acquired it.
Realising what a prize he had, Chamberlain decided to build a chassis for it. He had comprehensive facilities available in his own engineering works, but being over 70 years of age he couldn’t hang about! So thorough was he that he travelled to London to inspect the original Napier drawings in the Science Museum, and to the USA to measure and photograph the two similar four-cylinder racing Napiers then in the Harrah Museum.
Rowledge had had access to the 70hp Panhard, the most successful racing car of its time, when designing L48, but his biggest problem had been to make the pioneer six-cylinder engine as large as possible while keeping overall weight to the 1000kg limit imposed by the regulations of the time. Weight-saving is evident throughout, even to internal parts, and there was space only for a two-speed gearbox. Even floorboards were dispensed with in 1903-04.
One of the problems facing those who restore, or rebody, historic cars is to decide which period the rebuild should cover. The larger engine had been installed before the Match Race in 1908 (and the boat it was later used in apparently sank), so L48 was made to its last specification with the 15-litre power-unit (but with the two road-racing bucket-seats, whereas at Brooklands the petrol tank had been placed alongside the driver’s seat).
In spite of Bob’s fine workshops and foundries, no press of sufficient length was available for rolling the side-members, but welds in the most highly-stressed part proved satisfactory. Similarly, nobody in Melbourne could produce the tapered-section spring leaves, since the necessary tapering rollers no longer exist, but Australia’s biggest springmaker came to the rescue, its shop superintendant himself being the only man who could forge them by the old-fashioned method.
Meanwhile, Australia was scoured for three sets of six early Rudge-Whitworth wheel centres, and four 23/60 Vauxhall wheels had to be sacrificed to obtain mint-condition rims. To obviate the messy and time-consuming lob of changing ratios which Napier’s had had to endure (dismantling the axle and then precision-shimming it in the rebuild), three complete rear-axle assemblies were made up.
Chamberlain’s work was complicated by the fact that L48 had been a one-off, using no parts from other Napiers. Its original flat-fronted radiator had been replaced in 1908 with cooling tubes along the sides of the bonnet, which came to a point at the front; it is said that Edge demanded this not to increase cooling efficiency but to make the car more distinctive! These tubes made the engine inaccessible, so Bob made separate tanks for them. The four small brass tanks at the ends of the tubes were all different, but a leading metalworking firm made replicas from the drawings without difficulty.
Axle-ratios of 1.85:1 and 3.0:1 (for rallies) and 1.56:1 were made up, the latter originally used with gearbox ratios of 1:1 and 1.866:1. Dunlop supplied 880x 120 tyres like the Clinchers used at Brooklands, and 6.00 x 21s and Firestone 6.50 x 20 tyres were also used. The only other main areas in which the new chassis departed from the original were larger brake drums on the back wheels (11in x 21/4in like the transmission-brake drum, instead of the ridiculously small 8in x 11/2in dictated by the 1000kg weight-limit), the use of modern bolt-sizes instead of the unavailable early Whitworth standard bolts with large nuts, and a 21/2in increase in length to accommodate a new engine vibration-damper.
The engine was as removed from the boat more than 60 years earlier, but was in reasonably good order; it was found to turn over freely and all cylinders had even compression. The water jackets were badly cracked, but the seven-bearing crankshaft, machined from the solid, showed no such flaws even though it had been reground several times, and could still be used.
Bear in mind that Napier had had no prior knowledge of torsional vibration (Edge’s famous “power rattle”) in a six-cylinder engine, and that Rowledge probably lightened the crankshaft to further reduce overall weight.
Chamberlain pays tribute to the quality of the alloy castings — good foundry work for 1904—and notes that Napier must have had a precision cam-grinder. Incidentally, the purpose of using six cylinders had been to ease tyre strain through smoother torque, but the flexible running prompted the ebullient Edge to forecast that production cars would soon be able to dispense with gearboxes! No attempt was made to use materials to original specification in the rebuild (these were unknown for the chassis anyway, and its new side-members are of ‘3/16ths-thick 1045 steel). The only usable 1904 parts were the crankcase, crankshaft, three of the inlet manifold branch-pipes, one timing-pinion, the ignition-drive skew gears, a thrust-race, some bolts, and the water-pump.
It was known that L48 had design failings, and some of these could be eliminated in the rebuild. It would never idle, a fault blamed on the Napier carburettor, though the weak valve-springs were also a likely cause, since the actual valve-timing was suited to a good idle. The 32 lb exhaust-valve springs allowed the valves to lift and gas to re-enter the cylinders and the cam-followers to lag when this ioe engine with push-rod inlets was running at high speeds. It is believed that it had to be restarted at least 17 times (one report says 25 times) during the 1905 Gordon Bennett race. So 110 lb exhaust-valve springs and 60 lb inlets were fitted and a larger-diameter camshaft made, with lift more suited to the engine (Rowledge having believed low-lift was suitable with Napier’s odd multi-opening flat-seated valves).
Napier cylinder-design was also unusual, and four “pots” had failed in the car, one more in the boat; since they were unusable, new cylinders were made, to a revised design. The crankshaft might fail if it vibrated, so the aforementioned Lanchester-type torsional damper was fitted.
The original pistons were of cast-iron, weighed 11 lb, and had 3/4in-wide rings; compression-ratio was 3.6:1, which in the boat was increased to 4.0:1 by machining the heads, allowing half a ring to protrude above the bore at tdc! New 7 lb alloy domed-head pistons were made, giving 4.6:1 cr. New chains and sprockets now drive the water-pump. The total-loss lubrication system was replaced with a dry-sump system incorporating one pressure and three scavenge-pumps, with provision for priming to 75 lb/sq in, before starting up. The forged tubular con-rods were of 23 tons tensile strength, using Napier’s racing bearing-caps which were only about half the width of the journal, giving a large dip area but, as is now known, a broken oil film so new 70-ton rods were manufactured.
Twin carburettors were used in the boat, but a single World War Two-type 27/16 Bendix-Stromberg is now fitted. The original Napier trembler-coil ignition system has been replaced with a dual system using “off-the-shelf” Lucas parts but with the coils in the correct dash-mounted box. There is a choice of modern make-and-break or trembler coil, and two plugs per cylinder. An original-type starting handle is fitted, but a Ki-gass aids firing up and, as a substitute for two strong brave “swingers”, an Eclipse Bendix aircraft-starter with 130:1 reduction-gear is available. Although a new clutch-stop was made up, Bob says “the 1907 one probably worked better”.
The rebuilt engine was first started up in July 1982, using sae-30 Shell Super M oil. The carburettor required no adjustment, but ignition-advance was increased from 20° to 40° btdc. The Napier was soon making demo runs in Australia, but its tyres gave trouble, creeping on the rims even at low speeds, as security bolts had been eschewed.
Seats and tanks are 11/2in further back than in 1904, the wheelbase longer at 9ft 2in to take the crank-damper. With 35° ignition-advance, a rolling-road test gave an estimated output of 240 bhp at 2250 rpm.
In 1983 the famous Napier came to Donington and was demonstrated at Brooklands. Then it ran at Shelsley Walsh (recording 50.4 seconds) and at Colerne (30.67 sec for the standing-start kilometre, driven by Tony Gaze) using the 1.85 axle and 880 front, 7.00×21 rear tyres. Whether or not you approve of the modern reconstruction of old cars, you must concede that this is the recreation of the decade. Modifications made were in keeping with the ethics of a highly-experienced engineer intent on providing a habitat for a decidedly historic engine, and had the task not been undertaken there would now be no 1904 Napier L48. WB