The commonly adopted arrangement, whereby the roll inertia is made equal to that of the vehicle, gives a response to such disturbances as are induced by driveline oscillations, i.e. judder, that are significantly removed from on-road behaviour.
For reasonably accurate simulation of phenomena such as judder, roll inertia should be at least five times vehicle inertia.
Electronically simulated inertia is not effective in this instance: actual mass is necessary.
• The test vehicle should be anchored as lightly and flexibly as possible: this is not an insignificant requirement, since it is desirable that the natural frequency of the vehicle on the restraint should be at the lower end of the range of frequencies, typically 5–10 Hz, that are of interest.
It is strongly recommended that investigations of vehicle behaviour under conditions of driveline oscillation should proceed with caution if it is intended to involve running on a chassis dynamometer.