Rolling resistance of inflated tires is a factor that contributes to the total load and fuel consumption of a vehicle. Therefore, models of rolling resistance is an important area within computer simulations of vehicles used to predict fuel consumption and emissions. In these applications the coefficient of rolling resistance is usually described as a function of velocity. We have earlier shown that this is not a satisfactory solution [1, 2].

In this paper it is demonstrated that the temperature of the tires is a dominating factor for rolling resistance in real driving. The tires typically start at ambient temperature and are then warmed up by the heat generated in the tire. As the temperature increases the rolling resistance decreases (to some limit). After a long period (2 hours for truck tires) of driving at constant conditions, a stationary temperature (and rolling resistance) is reached. In real driving there are usually also other (faster) changes in conditions, like a speed change, that affects the rolling resistance as described in the traditional velocity models.

These aspects are demonstrated and lead to a proposal of a new model for rolling resistance. The new model is based on the relation between rolling resistance and tire temperature at stationarity. A differential equation for the tire temperature uses the relation between momentary temperature and stationary temperature as input. A speed dependent term is used to model fast changes of speed. In this way both fast and slow phenomena can be described.

The new model is in good agreement with results published by others. Further, in our own experiments with a heavy truck the new model has shown a very good ability to calculate the true dynamic rolling resistance.