From the Rolls-Royce experimental archive: a quarter of a million communications from Rolls-Royce, 1906 to 1960's. Documents from the Sir Henry Royce Memorial Foundation (SHRMF).
Technical analysis of wheel fight, comparing Dubonnet and wishbone suspension types and steering mechanisms.
| Identifier | ExFiles\Box 170\3\ img068 | |
| Date | 14th March 1936 guessed | |
| -2- (4) Of the two types of wheel fight the high frequency type is by far the more violent, and the more difficult to damp out by steering friction. In tuning the steering it appears better to err on the low frequency side so long as this can be done without getting a "rubbery" steering. (5) Dubonnet suspensions have a great advantage in the fact that the mass swinging about the kingpins is approximately doubled. This gives them an inherent tendency to the low frequency type of fight which is easily damped out. Another probable advantage is that there is more flexibility between the wheels and the steering rods, which also tends to low frequency of wheel flap. (6) Wishbone suspensions and conventional suspensions both tend to the high frequency type. Typical examples of "softening devices" in the steering mechanism of such cars are :- Shimmy shackle. Lateral flexibility of front spring shackles. Opposed springs in drag link. Bell crank in wishbone suspensions. A bell crank being notoriously difficult to make rigid. Mercedes device of rubber mount for lower wishbones, etc. (7) Chief fault of cross-steering on wishbone suspensions is its greater stiffness as compared with the bell-crank method. (8) Best method of trial tuning of steerings is by means of "hairpin" type of pitman arm in which the stiffness of the pitman arm can be adjusted through a wide range in a few seconds by shifting or adding clamps. (9) Accuracy in the up and down geometry of the steering layout; i.e. change of direction of front wheels in conventionals, or change of toe-in on wishbone suspensions, often blamed for wheel fight, appears to have no connection whatever with wheel fight within any range of errors which have been tested or which would be tolerable in practice. | ||