The power transmission efficiency of continuously variable transmissions (CVTs) based on the pushing metal belt is acknowledged to be lower than that of discrete ratio alternatives. This tends to negate the potential fuel economy benefits that are obtained by improved engine/load matching with a CVT. This series of three papers details an investigation into the loss mechanisms that occur within the belt drive as a first step to obtaining improvements in efficiency. This third paper follows on from two previous papers in which an analysis was performed modelling the torque losses that occur due to relative motion between the bands and segments of the belt, and between the pulleys and the belt due to pulley deflection effects. It describes additional experimental work, measuring the belt-slip speed tangentially about both of the pulleys in the variator. Additional loss models are proposed beyond those discussed in Parts 1 and 2 to describe the belt-slip phenomena, based on existing theory proposed by others. The analysis produced in this paper is validated against a range of experimental data and additionally through its close interaction with the torque-loss and torque-force distribution models proposed in Parts 1 and 2. The work takes into account new findings in other research and changes in the design of the current metal V-belt.

The use of high-speed tracked vehicles results in a number of accidents each year, some of which cause the vehicle to roll-over. Rollover caused by striking a low object, e.g. a kerb or tree stump (tripped), is normally thought to be the case but video evidence exists which shows that roll-over is initiated on a flat horizontal surface while the vehicle undergoes a 90°turn (untripped). This work looks at the initiation of roll-over in a tracked vehicle by increasing the lateral coefficient of friction with load transfer. A four-degree-of-freedom vehicle-and-track model is developed to investigate the problem. This work shows that the initiation of roll-over is possibly caused by increasing the lateral coefficient of friction between the track and the ground.

The power transmission efficiency of continuously variable transmissions (CVTs) based on the pushing metal belt is acknowledged to be lower than that of discrete ratio alternatives. This tends to negate the potential fuel economy benefits that are obtained by improved engine/load matching with a CVT. This series of three papers details an investigation into the loss mechanisms that occur within the belt drive as a first step to obtaining improvements in efficiency. Experimental work has been undertaken to investigate the no-load and low-load torque losses associated with a pushing metal V-belt CVT. This first paper describes a new analysis of the principal torque losses occurring in the metal belt CVT due to relative motion occurring between the belt segments and bands. The work takes into account new findings in other research and changes in the design of the metal V-belt. The torque loss model proposed in this paper is supported by experimental data from several different test procedures. A number of additional torque loss mechanisms, due to pulley deflections, are described in Part 2 of the series. The findings from this current paper support an analysis of belt-slip losses, which is described in detail in Part 3.

The power transmission efficiency of continuously variable transmissions (CVTs) based on the pushing metal belt is acknowledged to be lower than that of discrete ratio alternatives. This tends to negate the potential fuel economy benefits that are obtained by improved engine/load matching with a CVT. This series of three papers details an investigation into the loss mechanisms that occur within the belt drive as a first step to obtaining improvements in efficiency. This second part follows on from an initial paper in which an analysis was performed of the losses that occur due to relative motion between the bands and segments of the belt. Additional experimental work has been performed indicating that a significant deflection occurs in the pulleys of the variator. Further torque-loss models are proposed in addition to that discussed in Part 1, representing a smaller but still significant torque loss associated with the belt. The work takes into account new findings in other research and changes in the design of the metal V-belt. The third paper in this series develops a number of models to predict belt-slip losses in the variator system, based on force distribution models developed in Part 1.