ABSTRACT
Chapter 15 describes three-degree-of-freedom methods for haptic display of the interaction of a point and a virtual object, such as the one introduced by Zilles and Salisbury [Zilles and Salisbury 95]. Three-DOF rendering methods are effective for single-point interaction, but designing similarly effective methods for object-object interaction becomes a remarkable challenge, due to the high computational requirements. The approach of Zilles and Salisbury for 3-DOF rendering presents two main benefits: (1) a nonpenetrating simulation of the motion of the point as it slides on the surface of the obstacles; (2) a constraint-based computation of the force applied to the user, which results in a force orthogonal to the constraints. These features are highly desirable, in that non-interpenetration of virtual objects is known to increase their perceived stiffness [Srinivasan et al. 96], and that an incorrect orientation of the force has been shown to perturb the perceived orientation of the virtual surfaces [Sachtler et al. 00].
