Viscosity Perception of Virtual Fluids rendered by a Hand Exoskeleton

Abstract

Enabling haptic interaction with non-solid materials, such as liquids or sediments, could expand possibilities for exploration of virtual or remote environments, which would e.g enable training divers and astronauts in simulators. To allow application of natural investigation procedures in such scenarios, haptic interfaces with several degrees of freedom (DOF) are necessary, which allow the interaction with a variety of solid objects as well as different surrounding mediums. The goal of this work is to develop an algorithm for a high-DOF hand exoskeleton as haptic interface connected to five points on the human hand, that enables the perception of virtual fluids by rendering the fluids' prominent proprioceptive characteristic (viscosity). To allow simultaneous rendering of virtual solid objects of varying stiffness, a high update rate should be maintained. To quantify human perception of the rendered fluid, two user studies are carried out. The first investigates the ability to perceive fluids of low viscosity such as water, while the second deals with the discrimination ability for higher viscous virtual fluids. For virtual fluids with low viscosity, it is found that a linear relationship exists between the rendered and perceived viscosity with a scaling factor of 2. Fluids with high viscosity (> 10 Pa s) can be discriminated well, achieving similar values for the Weber fractions (w = 0.3) as are found in real interactions with fluids. The results of both experiments prove that properties of fluids rendered using simplified models to allow high update frequencies (833 Hz) can still be discriminated by human users.