PEDESTRIAN AND BICYCLE-FRIENDLY ROUNDABOUTS; DILEMMA OF COMFORT AND SAFETY

This paper addresses the circulatory speed of motorized traffic on roundabouts. After all, for the safety of pedestrians and cyclists, the difference in speed between cars and bicycles at a conflict point is very important: a reduction in collision speed from 30 mph (48 km/h) to 20 mph (32 km/h) means that the risk of fatal injury is reduced from 45% to 5% (a factor of 9). The speed through roundabouts is determined by the vehicle path curvature. On single-lane roundabouts, an increase in the vehicle path curvature results in a reduction of vehicular accidents. On multilane roundabouts, however, increasing the vehicle path curvature can result in a higher potential for sideswipe collisions. On double-lane roundabouts, designers are faced with a dilemma: accepting a higher number of sideswipe collisions involving motorized traffic (when they increase vehicle path curvature) or accepting serious accidents involving pedestrians and cyclists (when they decrease vehicle path curvature). The turbo-roundabout offers a solution to this dilemma. This kind of roundabout is based on important principles applied to single-lane roundabouts: (1) no weaving traffic on the roundabout; and (2) dealing with conflict points by means of slow speeds. Addressed second are the right-of-way regulations for cyclists and pedestrians in which cyclists are usually given priority in the Netherlands. But in the case of roundabouts, this leads to a situation in which either safety or convenience is diminished. In attempts to resolve this dilemma, Dutch guidelines recommend that within built-up areas, cyclists on the cycle track going around the roundabout be given right-of-way (for convenience) but that outside of built-up areas (and when another design is applied), they should not be given right-of-way (for reasons of safety). Research findings are discussed. It is concluded that further research is needed to demonstrate the degree to which roundabouts that give cyclists the right-of-way decrease their safety, even when given the best roundabout design possible. Finally, this contribution devotes attention to the designing of cycle crossings for crossing two double lanes. For pedestrians, a width of 3 meters of the splitter island (or reservation) is sufficient to anticipate motorized traffic satisfactorily. The conclusion is that the higher speed of the cyclists in comparison with that of pedestrians places additional demands on the geometric design for creating sufficient anticipation time.