Theoretical and experimental research indicates that radial loads have a significant influence on the value of belt-on-idler rolling resistances. Computational models discussed in literature use the notion of unit rolling resistance, i.e. rolling resistance per unit length of the idler. The total value of the rolling resistance of belt on a single idler is determined by integrating unit rolling resistance with respect to the length of the contact zone between the belt and the idler. This procedure requires the knowledge of normal load distribution along the contact zone between the belt and the idler. Loads acting on the idler set have been the object of both theoretical analyses and laboratory tests. Literature mentions several models which describe the distribution of normal loads along the contact zone between the belt and the idler set (Krause & Hettler, 1974; Lodewijks, 1996; Gładysiewicz, 2003; Jennings, 2014). Numerous experimental tests (Gładysiewicz & Kisielewski, 2017; Król, 2017; Król & Zombroń, 2012) demonstrated that the resultant normal loads acting on idlers are approximate to the loads calculated in theoretical models. If the resultant normal load is known, it is possible to assume the distribution of loads acting along the contact zone between the belt and the idler. This paper analyzes various hypothetical load distributions calculated for both the center idler roll and for the side idler roll. It also presents the results of calculations of belt rolling resistances for the analyzed distributions. In addition, it presents the results of calculations with allowance for load distribution along the generating line of the idler.

In the paper, the authors present the solution aimed at increasing reliability of the conveyor units. The analysis of technological and operational defects of conveyor rollers is presented. The changes in manufacturing technology have been proposed, which allowed for avoiding welding and provided the required level of tightness.

Computer simulation of the motion of air in the labyrinth seal of the roller was conducted to determine the numerical parameters of possible airflows. It is proved that the airflow is present in the gap of the labyrinth seal due to the roller rotation. It is shown that the reason for the penetration of abrasive particles through the labyrinth seal after stopping is decompression, which occurred as a result of temperature change and push out of airflows during rotation. It is also suggested that the number of stops during the operation should be taken into account when determining the durability of rollers. Practical recommendations are given for preventing the penetration of abrasive particles during conveyor stops and the need for combined seals. The results can be used for the construction of roller conveyor belts in any industry.