ABSTRACT
Geometric modeling is one of the most important features of a Monte Carlo algorithm, because it can signicantly impact areas of application, input preparation, degree of accuracy, and computation time. Throughout the years, groups have introduced different avors of geometric algorithms depending on their needs and limitations. Wang [2006] provides a good review of these algorithms. More commonly used approaches include:
• Combinatorial geometry, which forms a model by combining (using Boolean algebra) simple primitive objects, e.g., cuboids, cylinders, ellipsoids. This approach limits the user to modeling of simple or idealized objects.
