MODELING WATER FOR COMPUTER-GRAPHICS

Water motion is a phenomenon the full complexity of which has yet to b e realized in simulation. Although a good deal of work has been done i n modeling water waves, there remain interesting behaviours which have not been captured in any existing model. The variety of phenomena rea lizable by water is extensive, including waves, foam, bubbles, and spr ay. The present work is motivated by the complexity and range of pheno mena possible as well as by the frequency with which such phenomena ar e encountered. Water motion has been extensively studied from a physic al perspective. A computational model should take advantage of past wo rk in this area. Computer scientists have incorporated some physics in to models of water; however, currently existing models are incomplete. This paper considers the strengths and shortcomings of previous attem pts to model water. The work involves the development of an extension and refinement of an existing microscopic model of fluid, and a formal ization of some of the results of using this model. Some results of th e model have been compared with the predictions due to hydrodynamic th eory. Other results have been presented in the context of the situatio n being simulated, and the discussion hinges on observations of the re al situation rather than on theory. In brief, the model displays great potential for simulation of surface waves, spray, and interactions be tween solid objects and fluid. The experimental work is largely able t o substantiate claims as to the accuracy of the model at simulating th e underlying physics. (C) 1997 Elsevier Science Ltd. All rights reserv ed.