Another look at valve multiples

Browse pages
Current page

1

Current page

2

Current page

3

Current page

4

Current page

5

Current page

6

Current page

7

Current page

8

Current page

9

Current page

10

Current page

11

Current page

12

Current page

13

Current page

14

Current page

15

Current page

16

Current page

17

Current page

18

Current page

19

Current page

20

Current page

21

Current page

22

Current page

23

Current page

24

Current page

25

Current page

26

Current page

27

Current page

28

Current page

29

Current page

30

Current page

31

Current page

32

Current page

33

Current page

34

Current page

35

Current page

36

Current page

37

Current page

38

Current page

39

Current page

40

Current page

41

Current page

42

Current page

43

Current page

44

Current page

45

Current page

46

Current page

47

Current page

48

Current page

49

Current page

50

Current page

51

Current page

52

Current page

53

Current page

54

Current page

55

Current page

56

Current page

57

Current page

58

Current page

59

Current page

60

Current page

61

Current page

62

Current page

63

Current page

64

Current page

65

Current page

66

Current page

67

Current page

68

Current page

69

Current page

70

Current page

71

Current page

72

Current page

73

Current page

74

Current page

75

Current page

76

Current page

77

Current page

78

Current page

79

Current page

80

Current page

81

Current page

82

Current page

83

Current page

84

Current page

85

Current page

86

Current page

87

Current page

88

Current page

89

Current page

90

Current page

91

Current page

92

Current page

93

Current page

94

Current page

95

Current page

96

Current page

97

Current page

98

Current page

99

Current page

100

Current page

101

Current page

102

Current page

103

Current page

104

Current page

105

Current page

106

Current page

107

Current page

108

Current page

109

Current page

110

Current page

111

Current page

112

Current page

113

Current page

114

Current page

115

Current page

116

A few more observations are perhaps justified under this heading, used for one of last month’s articles, in view of the return by a number of leading makers of high-performance cars to multiple valves. The 1912 GP Peugeot is often quoted as having set the trend in four valves per cylinder but, in fact, this advance was more bound up with pioneering twin overhead camshafts, because four valves for each cylinder had been known before Peugeot’s well-known Racing Department had need of them.

For instance, and without suggesting that these were necessarily the very first multi-valve cars, the significant Prince Henry Benz of 1910 used such an arrangement, in conjunction with push-rod valve gear, and by 1913 the road-going 100 hp Isotta-Fraschini also had four valves in each cylinder head, prodded from an overhead camshaft and rockers. So far as three valves per cylinder figure, adopted today by Honda for their new production cars, Ettore Bugatti may have used them first but we should not overlook the employment of one inlet valve and two exhaust valves per “pot” for the pre-1914 Mercedes Ninety. Mercedes used four valves per cylinder to dominate the classic 1914 French Grand Prix, this layout being commonplace for racing engines that year, found, in fact, on 11 out of 14 makes entered for the French GP, that is to say, used by Mercedes, Peugeot, Sunbeam, Nagant, Delage, Schneider, Opel, Fiat (which was the order they finished the race), Aida, Nazzaro, and Vauxhall (all of which failed to finish the race).

The only breakaways in the race from this then universal use of multi valves were Fiat and Aquila-Italiana, who both preferred a single inlet and exhaust valve per cylinder, thereby pre-empting post-war practice, and Piccard-Pictet, who scorned poppets altogether, preferring sleeve valves. As there was but one Aquila-Italiana running, and two each of Piccard-Pictet and Nagants, whereas all the other makes were represented in the GP by teams of three cars each (except for Mercedes, who had put in five, to make certain of victory) it will be seen that in this great 1914 race multiple valves were all but universal.

It is interesting that, after the war, when such an arrangement was scarcely justified on grounds of increased reliability and when supercharging soon removed any clear excuse for employing multiple valves, Mercedes-Benz was still using four valves per cylinder at 60° included angle in a pent-roof combustion chamber on their all-conquering W125 and W154 Grand Prix cars of 1937 and 1939, having retained the layout after 1914 for others of its racing power units, including the W25B GP engines, and is now reverting to the system for the more sporting versions of its new 190 model. This favouring of such a layout instead of the by-then accepted two valves per cylinder in a hemispherical head is interesting — the ploy of four valves in a hemispherical head presented valve-operating complications, although the Rudge-Whitworth motorcycle engineers overcame this for the famed radial-valve Rudge, its single-cylinder lessening, however, the valve-actuating complexities.

Technical advancement killed off the multiple-valve head rather than any race regulations governing engine design, because, apart from rules relating to stock car or sports car racing, where certain stipulations had to be adhered to, I can think of no control of design made on pure racing car engines, apart from dimensional requirements, the banning of superchargers for certain sizes of power unit, and those which fuel-consumption restrictions may automatically have imposed — apart, that is, from recent controversy over exhaust-pipe configurations and water-injection. That is, with the one exception, of Indianapolis.

The sole departure from unfettered racing-engine design applies to the American Indianapolis 500 Mile Race before the war, as I related in the article on “The Ford V8 as a Racing Car”, in the November 1983 issue of Motor Sport. Among the various rules aimed at encouraging “stock-block” engines alongside the out-and-out racing cars in Indy’s one big annual race was the banning in 1930 of more than two poppet valves per cylinder. Apparently this ruling was debated very thoroughly by the Technical Committee of the AAA which was acting in an advisory capacity to the Indianapolis Motor Speedway Association. Although the ban excluded some fast semi-stock European cars like the Bentley and Bugatti and would apparently, according to the American Motor magazine, give a 16-cylinder engine a marked power advantage over an eight-cylinder engine of equal swept volume, the rule was nevertheless introduced, as discouraging the building of special racing engines. Maybe the American Conunittees did not expect serious entries of Bentley and Bugatti cars anyway, after the lone effort by the former in 1922 and the pathetic performance of the latter as a five-car team in 1923. They certainly knew that the recently-introduced new American car with four valves per cylinder. the Model-J Duesenberg, was outside the engine size limit of 366 cu in by a matter of 66 cu in — presumably no one visualised this engine being linered-down or otherwise reduced in capacity for the race. The question that arises is why, at this stage of engine development, the AAA Technical Committee expected multi-valve racing engines to be built for the Indy 500 when every successful Indy car for many years had not used more than one inlet and one exhaust valve per cylinder?

Even more odd is the fact that a year later, for the 1931 Indy 500 Mile Race, former restrictions on numbers of valve, and carburetters were removed / amended. It was said at the time that multiple valves were permitted again in order to attract entries of stripped European sports cars. If this were true the aim was moribund, the only multi-valvers entered for this 1931 race being the Buehrig cars with eight-cylinder Duesenberg engines, two Empire Free State Millers, a couple of Stutz presumably of kind that had harried the Bentleys for a time at Le Mans in 1929 and the Curtiss-powered DeBiase, plus an old Mercedes. Indeed, the vast majority of the 1931 entries had only two valves per cylinder, and it was rumoured that Ralph de Palma’s Miller would rely on Wehr rotary valves. So it was apparent that in the USA as in Europe multi-valves were out-dated, as Coatalen had delighted in telling W O Bentley some time previously.

After qualifying, a full field of the permitted 40 cars started in the 1931 500, whereas under the 1930 restrictions the number had been two short. So it is difficult to understand what the ruling over the number of valves an entrant could use was about. The new Indy regulations were otherwise virtually unaltered, two-seater bodies and riding mechanics still being required, fuel capacity limited but the larger engines still being permitted, and superchargers again barred to all except two-stroke engines, which were represented by the blown 16-cylinder Durays. The only previously-entered foreign cars of recent years that the changed rules for 1930 would have eliminated were a Bugatti Special with straight-eight 24-valve engine that had failed to qualify in 1928, and the supercharged Amilcar Six and 1927-type GP Deluge which ranin 1929, Chiron in the Delage finishing seventh at 87.7 mph. None of these returned to the fray in 1931.

Here I would digress to remark that in the aforesaid article about racing Ford V8s I quoted the 1930 Indy rules but omitted to say these were changed, as described above, m 1931. It was, of course, the latter ruling that enabled the 10 Fords built for the 1935 race to use four single-choke or two dual-choke carburetters, as the rules by then permitted two carburetters, or a single dual carburetter, per two cylinders. whereas in 1930 only two carburetters per engine, regardless of how many cylinders it had, had been allowed. — WB