Physically Based Animation - Computer Animation: Algorithms and Techniques - page 198

CHAPTER

7

Physically Based Animation

Animators are often more concerned with the general quality of motion than with precisely controlling
the position and orientation of each object in each frame. Such is the case with physically based anima-
tion. In physically based animation , forces are typically used to maintain relationships among geometric
elements. Note that these forces may or may not be physically accurate. Animation is not necessarily
concerned with accuracy, but it is concerned with believability, which is sometimes referred to as being
physically realistic . Some forces may not relate to physics at all—they may model constraints that an
animator may wish to enforce on the motion, such as directing a ball to go toward making contact with
a bat. For the purposes of this chapter, the use of forces to control motion will be referred to as physically
based animation whether or not the forces are actually trying to model some aspect of physics.
When modeling motion that is produced by an underlying mechanism, such as the principles of
physics, a decision has to be made concerning the level at which to model the process. For example,
wrinkles in a cloth can be modeled by trying to characterize common types of wrinkles and where they
typically occur on a piece of cloth. On the other hand, the physics (stresses and strains) of the individual
threads of cloth could be modeled in sufficient detail to give rise to naturally occurring wrinkles.
Modeling the surface characteristics of the process, as in the former case, is usually computationally
less expensive and easier to program, but it lacks flexibility. The only effects produced are the ones
explicitly modeled. Modeling the physics of the underlying process usually incurs a higher computa-
tional expense but is more flexible—the inputs to the underlying process can usually produce a greater
range of surface characteristics under a wider range of conditions. There are often several different
levels at which a process can be modeled. In the case of the cloth example, a triangular mesh can
be used to model the cloth as a sheet of material (which is what is usually done). The quest for the
animation programmer is to provide a procedure that does as little computation as possible while still
providing the desired effects and controls for the animator.
The advantage to the animator, when physical models are used, is that (s)he is relieved of low-level
specifications of motions and only needs to be concerned with specifying high-level relationships or
qualities of the motion. For our example of cloth, instead of specifying where wrinkles might go and the
movement of individual vertices, the animator would merely be able to set parameters that indicate
the type and thickness of the cloth material. For this, the animator trades off having specific control
of the animation—wrinkles will occur wherever the process model produces them.
For the most part, this chapter is concerned with dynamic control —specifying the underlying forces
that are then used to produce movement. First, a review of basic physics is given, followed by two
common examples of physical simulation used in animation: spring meshes and particle systems. Rigid
body dynamics and the use of constraints are then discussed.