Simulation and Haptic Rendering of the Interaction with Diverse Fluid Media

Abstract

Virtual reality (VR) technologies are committed to the development of solutions that enable the long-standing vision of creating immersive synthetic worlds that transcend the boundaries of reality. Although VR is at the peak of its history thanks to the confluence of various technologies developed over decades, the diversity of interactions that may be portrayed in VR experiences remains limited. While interaction with solid objects and deformable bodies has attracted a great attention from researchers, other interesting media, such as fluids, have been largely ignored. In the particular case of fluids, this is primarily due to a combination of two factors. First, interactive fluid simulation methods are incapable of conveying materials other than inviscid fluids. This severely limits the capacity to replicate fascinating everyday materials such as honey, whipped cream, paint, or clay. Second, conventional haptic devices struggle to provide compelling contact with fluid media, particularly in applications requiring direct manipulation. In this thesis, we investigate strategies to overcome the limitations of the current state of the art in order to enable physical contact with rich and complex virtual phenomena such as fluids. This is accomplished in two ways. First, we present a highly efficient constraint-based method for viscoelastic fluid modeling. Our approach is motivated by a constitutive model for polymeric fluids, which enables the portrayal of a wide variety of materials under a single formulation. Second, we present two methods for depicting tactile interaction with such media by leveraging on the AM and STM control metaphors commonly employed in ultrasonic haptics. We approach the device actuation as a numerical optimization problem, finding the control variables that best reproduce the pressures arising from virtual interactions. Furthermore, we incorporate knowledge of the technical and perceptual constraints of both control metaphors to maximize the efficacy of our solutions. To conclude, we demonstrate the applicability of the presented approaches, combining them to address the challenge of virtual simulation of clay interaction. As a result, our method enables unprecedented degree of realism in natural manipulation of materials exhibiting extreme viscoplastic behavior.