Texas A&M University | Department of Computer Science  gcube

Library for Exact Solid Modeling

Home |
Examples | Download | References | Numerical Perturbation 

What's new?

08/12/04 The Web pages are renewed.
04/17/06 References are added.
04/17/06 The release Ver. 0.3 is uploaded. [Download]

[Jump to the Top of this Page]


ESOLID, developed primarily at the University of North Carolina, is a geometric solid modeler. Given a CSG-tree of primitive solids, ESOLID performs exact boundary evaluation and returns the B-rep of resulting solid. ESOLID uses exact computation in order to eliminate numerical errors. ESOLID assumes that all the solids are in general position.

Robust Geometric Solid Modeling

A significant problem can arise when performing boundary evaluation. This is the problem of numerical error. Slight inaccuracies in the representations or the computations can cause numerical errors, which can lead to seriously incorrect output or program crashes. Even seemingly minor errors can eventually compound until they are serious enough to cause these problems. These problems with numerical error are magnified considerably when dealing with curved surfaces.

Exact Computation for Algebraic Numbers

The approach that we chose to follow has been to use exact computation. Exact computation means that all numerical values are determined to a precision guaranteed to make any decision based on that number correct. The drawback to exact computation is that it can be extremely slow (many orders of magnitude). The goal of our work has been to create a boundary evaluation program that uses exact computation (thus eliminating all numerical errors) that runs on low-degree curved solids (i.e. the types found in most CSG models), at speeds no more than two orders of magnitude slower than a comparable inexact program.


ESOLID assumes that all the solids are in general position. Thus, ESOLID may fail or crash on the input with degeneracies.

[Jump to the Top of this Page]


[Jump to the Top of this Page]

Copyright 2003-2006 - ESOLID Project, All Rights Reserved

This material is based upon work supported by the National Science Foundation under Grant No. 0138446. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.