Introduction
On Scott's 1910–1913 expedition to Antarctica three newly developed motor sledges where taken to assist with haulage of supplies. One fell through the ice shortly after arrival and the other two were plagued by mechanical failures and were abandoned on route to the South Pole. The failure of the big ends was the primary mechanical problem and this was attributed to overheating of the air-cooled engine and inadequate lubrication producing excessive mechanical stresses (Day Reference Day1912; Dibbern Reference Dibbern1976; Anon. 1991). The intense dryness of the air combined with the positioning of the cooling fan meant that only the first two of the four cylinders of the engines were cooled effectively. The constant over-heating of the third and fourth cylinders was suggested to have resulted in the excessive oil consumption of the engine and hence inadequate lubrication (Day Reference Day1912).
One of the more minor issues was that the engines were difficult to start due to poor volatilisation of the petrol. As a consequence, either at the start of the day or after a break to let the engine cool, a blow-torch was required to heat the carburettor for up to half an hour. A wooden box was constructed around the motor to keep it warm while starting may also have disrupted the cooling airflow and contributed to the overheating and subsequent failure.
Fuel analysis
The fuel for the motor-sledges was donated by Shell and retrieved from a petrol canning factory in Melbourne, Australia (Wilson Reference Wilson1972: 50). Fuel from metal tins in wooden crates labelled ‘Shell Motor Spirit’ at a fuel-dump besides Scott's hut has been sampled and analysed for composition and geochemical properties (Gardner and others Reference Gardner, Glavinceski, Moon and McDill1984; Dougherty Reference Dougherty1985; Volk and others Reference Volk, McIntyre, Batts and George2005). It was a straight run-gasoline produced from southeastern Asian crude and is consistent with fuel production at that time (Encyclopædia Britannica 1911). The simulated distillation by gas-chromatography indicated the fuel had a boiling range of 2–176 °C with 22% n-alkanes, 37% iso-alkanes, 15% cycloalkanes and 7% aromatics by weight. Dougherty (Reference Dougherty1985) notes that the composition would be markedly different from modern fuels as the initial boiling point is about 20 °C higher than for modern fuels, because it was produced before modern refinery processes such as catalytic cracking, alkylation, naphtha hydrofining, reforming and hydrocracking processes were available. A modern Australian gasoline, for example, will contain approximately 12% n-alkanes, 32.4% iso-alkanes, 6.4% cycloalkanes, 42.9% aromatics and 6.3% olefins. Dougherty (Reference Dougherty1985) goes on to suggest that an octane rating should be determined so that it can be compared to the compression ratio of the engine. The fuel was re-analysed to obtain an octane number for comparison with published engine designs (Anon. 1910) and to consider its suitability for the task in the Antarctic. Details of the fuel used are given in Volk and others (Reference Volk, McIntyre, Batts and George2005). 100 mL of fuel was analysed by Intertek, Caleb Brett, Australia. Bulk hydrocarbon group-type composition (PONA) and a research octane number (RON) were calculated from data for individual compounds obtained using gas-chromatography (ASTM 2001). The fuel has a density of 0.73 g/cm3 (15 °C) and a vapour pressure of 24 kPa (37.78 °C). The average molecular weight was 96.2 and the elemental composition was Carbon = 85.2%, Hydrogen = 14.8% and Oxygen = 0%. Eighty peaks were identified (99.2%) ranging from C5 to C11 and the RON was calculated to be 65. The density is normal for a fuel of that time which ranged from 0.70–0.74 g/cm3 (Encyclopædia Britannica 1911). The vapour pressure is lower than for modern gasolines that have a range from 50–80 kPa. However, although the fuel was stored under Antarctic conditions some loss of volatile components may have taken place. The RON of the fuel is normal for a fuel of this composition and for that time, although considerably lower than that of modern fuels that are typically around 85–95. The text with the engine diagrams indicates that the valves were mounted on the side of the engine, indicating that it was a flathead or side-valve configuration (Anon. 1910). A major disadvantage of flathead engines is that the exhaust follows a more complicated path through the block to leave the engine leading to overheating under sustained heavy use. No cylinder volume is given and from the diagrams it is not possible to calculate one. Therefore, no compression ratio can be calculated. No piston or cylinder damage has been reported, suggesting that no pre-ignition or detonation occurred and that the compression ratio was correct for the fuel. In conclusion the information presented here suggests that for the period the fuel was probably suitable for the purpose of powering the motor sledges and did not contribute to any major failings. Volatility may have been low, but not such that the motor could not be operated, and the octane rating appears suitable for the engine. Fuel composition is therefore not likely to have added significant stresses or contributed to failure of the big ends. These findings support the explanation that inadequate engine design was the primary cause of the failings.
Acknowledgements
The authors would like to thank CSIRO petroleum for funding this work and Steve Dibbern for his helpful comments.
