Topology Atlas Invited Contributions 9 (2004) no. 1, 12 pp.

arXiv:math/0402446v1 [math.GN] 26 Feb 2004

COMPUTATIONAL TOPOLOGY FOR REGULAR CLOSED SETS

(WITHIN THE I-TANGO PROJECT)

T.J. PETERS, J. BISCEGLIO, D.R. FERGUSON, C.M. HOFFMANN, T. MAEKAWA,

N.M. PATRIKALAKIS, T. SAKKALIS, AND N.F. STEWART

Abstract. The Boolean algebra of regular closed sets is prominent in topol-

ogy, particularly as a dual for the Stone-Čech compactification. This algebra
is also central for the theory of geometric computation, as a representation
for combinatorial operations on geometric sets. However, the issue of com-
putational approximation introduces unresolved subtleties that do not occur
within “pure” topology. One major effort towards reconciling this mathe-
matical theory with computational practice is our ongoing I-TANGO project.
The acronym I-TANGO is an abbreviation for “Intersections—Topology, Ac-
curacy and Numerics for Geometric Objects”. The long-range goals and initial
progress of the I-TANGO team in development of computational topology are
presented.

1. Introduction and brief literature review

Throughout this paper, all sets considered will be assumed to be subsets of R3 ,
with its usual topology. The Boolean algebra of regular closed sets in R3 will be
denoted as R(R3 ). Furthermore, any regular closed set considered will be assumed
to be compact. Any surfaces and curves considered will be assumed to be compact
2-manifolds and 1-manifolds, respectively. All neighborhoods will be assumed to
be open subsets of R3 .
The theoretical role for R(R3 ) was introduced into geometric computing to cor-
rect the unexpected output seen from combinatorial operations on geometric sets
[27]. For instance, consider the two dimensional illustration shown in Figure 1. The
original operands of A and B are indicated in Figure 1(a). The unexpected output
is shown in Figure 1(b), where the expected result would have been what is shown
in Figure 1(c).
The phenomenon shown in Figure 1(b) was informally described as “dangling
edges” [33]. The formalism that was proposed to eliminate this behavior was that
geometric combinatorial algorithms should accept only regular closed sets as input
and then execute the Boolean operations of meet, join and complementation on
these operands, thereby creating only regular closed sets as output [32]. The intent

Date: November 2003.

Partial funding for all authors was from NSF grant DMS-0138098. Authors Peters and Sakkalis
were also partially supported by NSF grant CCR 0226504. All statements in this publication are
the responsibility of the authors, not of any of these indicated funding sources.
1
2 I-TANGO PROJECT

A B
(a) Original sets

A−B A−B
(b) Unexpected result (c) Expected result

Figure 1. Subtraction of two sets

was to eliminate “dangling edges” and, in principle, this should have been suffi-
cient1. However, each operand also has a geometric representation that depends
upon the approximation methods used to compute the results. This additional
subtlety raises issues in both theory and computation.
For this short article, only a brief literature review will be presented. An ear-
lier survey on topology in computer-aided geometric design [25] is recommended as
introductory material for topologists. The texts [15, 24] discuss the integration of
computational geometry, shape modeling and topology. The subject of computa-
tional topology is still a nascent and emerging sub-discipline. This article focuses
upon the authors’ particular perspective in its development. The first use of the
terminology “computational topology” appears to have been in the dissertation of
M. Mäntylä [20]. Further contemporary views can be gained from the following
web sites [6, 9, 34].
Additional perspective can be gained by understanding the broader context in
which topology has already been successfully applied to computer science. We
mention two particularly notable successes. The first is the use of non-Hausdorff
topology by Kopperman, Meyer, Kong, Rosenfeld, Smyth and Herman in digital
topology for computer graphics and image processing. A good overview is readily
available [18]. Similarly, the work of Mislove, Reed and Roscoe on domains explores
variants of limits for fundamental algorithmic and programming language studies,
continuing the expressive power and the broad applicability of the language of
topology in denotational semantics and concurrent programming. The monograph
[26] is recommended as an introduction.

2. Theory versus computation

One elegant computational representation for the combinatorial operators is to
assign each object a symbol and then to indicate operations in a tree referencing
those symbols. For instance, such a tree structure could be as depicted in Figure 2.
At this level of abstraction, the mathematical theory and the computational
representation are completely consistent, and this representation became known as
Constructive Solid Geometry (CSG). Difficulties arose in instantiating the basic
geometric information that is represented by the operands at the leaf nodes and,
sometimes, in computing geometric representations at the internal nodes of the
1The subtraction operation between two sets, shown as A − B in Figure 1, is not specifically a
Boolean operation. However, the use of A − B should be understood to be conveniently shortened
notation equivalent to the operations A ∧ B ′ , where B ′ represents the standard Boolean operation
of complementation on the operand B.
 COMPUTATIONAL TOPOLOGY FOR REGULAR CLOSED SETS 3

∨❦
✠ ❅ ❘
❅
∧❦ −❦
✁☛✁ ❆❆❯ ✁☛✁ ❆❆❯
A B C D

Figure 2. Tree for (A ∧ B) ∨ (C − D).

tree. In CSG, the leaf nodes are restricted to a small set of specific geometric
objects, known as primitives. A typical collection of primitives might consist of
spheres, parallelepipeds, tori and right circular cylinders. The critical geometric
algorithm underlying each Boolean operation is the pairwise intersection between
the operands.
As the boundary of each of these primitives can be represented by linear or
quadratic polynomials, the needed intersection between each pair of primitives was
relatively simple and numerically stable, for most cases considered, although spe-
cific intersections could be problematical. For instance, suppose two cylinders of
identical radius and height were created and then positioned so that the bottom
of one cylinder was co-incident with the top of the other cylinder. This special
case was specifically considered in most intersection algorithms and could usually
be processed without problem. However, if one then rotated the top cylinder a
fraction of a degree about its axis (so that the planar co-incidence remained in-
tact) many software systems would fail to produce any output for this problem,
sometimes even causing a catastrophic program failure. This particular problem
became a celebrated test case and most systems developed ad hoc methods to solve
this cylindrical intersection problem. Yet, this was just avoiding the more serious
issue of the fragile theoretical foundations for many intersection algorithms. People
using CSG systems became sensitive to their limitations and continued to use them
effectively by avoiding these challenging circumstances, although the work-arounds
were often tedious to execute.
The imperative, largely initiated by the aerospace and automotive industries,
to model objects using polynomials of much higher degree than quadratic created
a movement away from CSG systems. The alternative format was to represent
compact elements of R(R3 ) by their boundaries, and this became known as the
“boundary representation” approach, or “B-rep” for short. This has become the
dominant mode today. Again, within this clean conceptual overview, the realities of
computation pose some subtle problems. In most industrial practice, the modeling
paradigm was further restricted so that the boundary of an object was a 2-manifold
without boundary. However, it was difficult to create computer modeling tools that
could globally define 2-manifolds without boundary, though there existed excellent
tools for creating subsets of these 2-manifolds. For example, computational tools for
creating splines were becoming prevalent. Again, in principle, if each such spline
subset was created with its boundaries, then the subsets could be joined along
shared boundary elements to form a topological complex [14] for the bounding
2-manifold without boundary.
The inherent computational difficulty was to separately create two spline patches,
each being a manifold with boundary, so that the corresponding boundary curves
4 I-TANGO PROJECT

were identical and could be exactly shared between the patches. In some situations,
algorithms for fitting spline patches were used successfully. In other cases, patches
have been slightly enlarged and intersected so as to obtain improved fits. Indeed,
such intersections are well-defined in pure mathematics, but, again, approximation
in computation poses subtle variations from that theory, as described in the next
section on pairwise surface intersection.

3. Subtleties of pairwise spline surface intersection

It is well-known that unwanted gaps between spline surfaces or self-intersections
within intended manifolds often appear as unwanted artifacts of various imple-
mented intersection algorithms [10]. The mismatch between approximate geometry
and exact topology has historically caused reliability problems in graphics, CAD,
and engineering analyses—drawing the attention of both academia and industry.
The severity of the problem increases with the complexity of the geometric data
represented, both from high-degree nonlinearity and from the intricate interdepen-
dence of shape elements that should, but do not, fit together according to the
specified topological adjacency information.
The conceptual view of these joining operations is illustrated in Figure 3, with an
intersection curve2 denoted as c. But this picture of c only exemplifies the idealized,
exact intersection curve. For practical computations, an approximation of c is often
created [11] and, in many systems, an intersection curve will be approximated twice;
once within the parametric domain of one of the intersecting surfaces and then again
within the parametric domain of the other. These approximations are labeled as
c1 and c2 in Figure 3. The spline functions from [0, 1]2 into R3 then also rely upon
algorithmic evaluation of these approximated intersection curves, as indicated by F
and G in Figure 3. It is virtually certain that those evaluations will not be exactly
equal in R3 .
The mismatch between concept and reality depicted in Figure 3 creates ambigu-
ity, as the intersection representation is sometimes considered as a unique set, from
the symbolic topological view, and at other times as two approximating sets, from
the geometric view.

4. Specific progress
To resolve this ambiguity, we are investigating richer representations. We intro-
duce a neighborhood of the true, but unknown, intersection set. This neighborhood
is created from newly determined rigorous upper bounds on the error incurred dur-
ing efficient intersection approximations. To date, it has been convenient to create
these neighborhoods as tubular neighborhoods [13], but broader generalizations
seem to be possible.
The role of topological equivalence. Before introducing these bounding neigh-
borhoods, we discuss the meaning of topological equivalence between an object and
its approximations. We have proposed the use of ambient isotopy for this topolog-
ical equivalence versus the more traditional equivalence by homeomorphism, as is
explained more fully in our publications [2, 3, 4, 28, 29, 31]. Intuitively, two closed
curves will not be ambient isotopic if they form different knots. Figure 4 shows
2We focus on the generic case of an intersection curve, although isolated points and co-incident
areas can also arise, with similar complications.
 COMPUTATIONAL TOPOLOGY FOR REGULAR CLOSED SETS 5

Concept

F
c c
1

Reality

G
c
2

Figure 3. Joining operations for geometric objects

two simple homeomorphic space curves, where the piecewise linear (PL) curve is
an approximation of the smooth curve. However, these curves are not ambient iso-
topic, because they depict different knots3, with the smooth curve illustrating the
simplest knot, known as the unknot. In the right half of Figure 4 the z coordinates
of some vertices are specifically indicated to emphasize the knot crossings in R3
(All other end points have z = 0). All end points of the line segments in the PL
approximation are also points on the original curve. Having this knotted curve as
an approximant to the original unknot would be undesirable as output from a curve
approximation algorithm, particularly for applications in graphics and engineering
simulations. Similar pathologies can happen in approximating 2-manifolds, both
with and without boundary, but results [2, 4, 19, 28, 29] summarized here can pre-
vent these difficulties by appropriately constraining the approximations produced.
In response to the example of Figure 4, a theorem was published that provided
for ambient isotopic PL approximations of 1-manifolds [19]. The proof utilizes “pipe
surfaces” from classical differential geometry [21] to build an appropriately small
tubular neighborhood such that if the PL approximant is constrained to lie within
the constructed neighborhood, then the PL approximant is ambient isotopic to the
original curve. While the techniques of tubular neighborhoods are well known in
differential topology, the relevant theorems usually state only the existence of an
ambient isotopy. Additional work in applied mathematics was needed to comple-
ment the theorems with constructive formulations from which to obtain effective
procedures and algorithms. Such extensions were successfully accomplished for
both non-singular compact, orientable 1-manifolds and 2-manifolds (with or with-
out boundary) [1, 2, 19, 28, 29] by I-TANGO team members (and their co-authors).
We used geometric characteristics to compute a specific upper bound on the size of
a tubular neighborhood and to then specify a particular isotopy.
3The different knot classifications of 0 and 4m are indicated in Figure 4.
1 1
6 I-TANGO PROJECT

z = -2 z = -1
z = 100

z = -1

z=1 m
0 4
1 1

Figure 4. Nonequivalent knots

The importance of this topological equivalence class extends beyond the mani-
folds described to the geometric models created as compact subsets of R(R3 ). Con-
sider that a well-defined topological complex could be created from 2-manifolds
with boundary if the difficulties along the intersection boundaries could be solved.
However, the non-uniqueness of the geometric representation of the intersection sets
seems to pose an intractable problem to creation of a single, well-defined topological
complex. Our alternative approach is to find a neighborhood within which it can be
proven that the true intersection curve lies. Since the construction of these bound-
ing neighborhoods is dependant upon the specific intersection algorithm used, some
further details of the two intersection algorithms used within the I-TANGO project
are presented. While the neighborhood construction details will vary with the spe-
cific intersection algorithm chosen, the following two neighborhood constructions
were chosen both for their carefully defined error bounds and for their potential for
generalization.

Error bounds for topology from Taylor’s theorem. First, we present the
Grandine-Klein (GK) intersection algorithm [12]. Referring to Figure 3, we note
that the GK algorithm bases its error bounds on well-established numerical tech-
niques in differential algebraic equations (DAE). While these DAE techniques pro-
vide rigorous error bounds, these bounds are expressed within the parameter space
[0, 1]2 , which serves as the domain of the spline functions (indicated as F and G
in Figure 3). The code implementing the GK algorithm then has an interface that
allows the user to specify an upper bound ǫ for this error in parameter space and
the algorithm provides guarantees for meeting this error bound. However, the typ-
ical end user is often unaware of the role of this parametric domain, so selection
of this parametric space error bound has often relied upon heuristics. It would be
more convenient for the user to be able to specify an error bound within R3 . One
accomplishment within the I-TANGO project has been to demonstrate a mathe-
matical relation [22] between the error bounds in R3 and [0, 1]2 , following from a
straightforward application of Taylor’s Theorem in two dimensions [8, p. 200]. The
conversion between these error bounds has been implemented in a pre-processing
interface to the GK algorithm and this new interface has been tested to be reliable,
efficient and user-friendly.
 COMPUTATIONAL TOPOLOGY FOR REGULAR CLOSED SETS 7

Using the notation from Figure 3 for the spline function F , Taylor’s Theorem
provides a bound on the error of F evaluated at a particular point (u, v) versus
F evaluated at a point (u0 , v0 ), where (u, v) and (u0 , v0 ) are within a sufficiently
small neighborhood. This sufficiently small neighborhood will have radius given by
the value in the parametric domain [0, 1]2 which was denoted as ǫ in the previous
paragraph. Then it follows [22], with k · k being any convenient vector norm, that
kF (u, v) − F (u0 , v0 )k ≤ ǫM
for any M satisfying
∂F ∗ ∗ ∂F ∗ ∗
(u , v ) + (u , v ) ≤ M,
∂u ∂v
for some point [u∗ , v ∗ ] on the line segment joining [u, v] and [u1 , v1 ].
For the single spline F , let γ(F ) be an upper bound for the acceptable error in
R3 between the true intersection curve c and one of its approximants F (c1 ). In
order to guarantee that this error is sufficiently small, it is sufficient
ǫM ≤ γ(F ),
where an upper bound for M can be computed by using any standard technique
for obtaining the maximums of the partials ∂F ∂F
∂u and ∂v . For G, a similar relation
4
between γ(G) and ǫ exists .
Then it is clear that a neighborhood can be defined that contains the true in-
tersection curve c and both of its approximants. Let F (c1 ) denote the image of c1
under F and similarly, let G(c2 ) denote the image of c2 under G. Let Nγ(F ) (F (c1 ))
be a tubular neighborhood of radius γ(F ) about F (c1 ), where c1 has been gener-
ated from the GK intersector to satisfy the inequality presented in the previous
paragraph. Similarly, define Nγ(G) (G(c2 )). Then, let
N (c) = Nγ(F ) (F (c1 )) ∪ Nγ(G)(G(c2 )).
It is clear that N (c) is a neighborhood of c, which contains both of its approx-
imants, F (c1 ) and G(c2 ). However, there is both a theoretical and computational
limitation to this approach.
• There is no theoretical guarantee that either approximant is topologically
equivalent to c, and
• Any practical computation of N (c) would depend upon an accurate com-
putation of the set Nγ(F ) (F (c1 )) ∩ Nγ(G) (G(c2 )), which is likely to be as
difficult as the original computation of F ∩ G.
While the above bounds are often quite acceptable in practice to compute a rea-
sonable approximant, further research has been completed into alternate methods
to give guarantees of topological equivalence within a computationally acceptable
neighborhood of the intersection set, as reported in the next subsection.
Integrating error bounds and topology via interval solids. Recent work
by Sakkalis, Shen and Patrikalakis [30] emphasized that the numeric input to any
intersection algorithm has an initial approximation in the co-ordinates used to
represent points in R3 , leading to their use of interval arithmetic [24]. The basic
idea behind interval arithmetic is that any operation on a real value v is replaced
4This error bound assumed that the error due to algorithmic truncation within the numerical
DAE methods dominated any other computational errors.
8 I-TANGO PROJECT

by an operation of an interval of the form [a, b], where a, b ∈ R and a < v < b. The
result of any such interval operation is an interval, which is guaranteed to contain
the true result of the operation on v. This led naturally to the concept of an interval
solid and some of its fundamental topological and geometric properties were then
proven, as summarized below.
Throughout this section, a box is a rectangular, closed parallelepiped in R3 with
positive volume, whose edges are parallel to the co-ordinate axes5. Let F be a
non-empty, compact, connected 2-manifold without boundary. Then the Jordan
Surface Separation Theorem asserts that the complement of F in R3 has precisely
two connected components, FI , FO ; we may assume that FI is bounded and FO is
unbounded. Let also B = {bj , j ∈ J} be a finite collection of boxes that satisfies
the following conditions:
C1: {Int(bj ), j ∈ J} is a cover of F .
C2: Each member b of B intersects F generically; that is, b∩F is a non-empty
closed disk that separates b into two (closed) balls, Bb+ and Bb− , with Bb+ ,
(Bb− ) lying in FI ∪ F (FO ∪ F ), respectively.
C3: For any bi , bj ∈ B, let bij = bi ∩ bj . If Int(bi ) ∩ Int(bj ) 6= ∅, then bij is
also a box which satisfies C2.
Notice that condition C2 indicates that every b ∈ B intersects F in a natural
way (see Figure 5).

+ F bj
Bb
− b
Bb bi

b ij
F
b F
Condition C2 Condition C3

Figure 5. 2D versions of conditions C2 and C3

The following result summarizes several previously appearing results, where a

solid is defined to be a non-empty compact, regular closed subset of R3 .
Theorem 4.1S([30, Corollary 2.1, p. 165]). If F is connected and B satisfies C1–
C3, then F ∪ j∈J bj is a solid.
Bisceglio, Peters and Sakkalis [28, 29] have recently given sufficient conditions to
show when the boundary of an interval solid is ambient isotopic to the well-formed
solid that it is approximating, as described in the following theorem. For a positive
number δ, define the open set
F (δ) = {x ∈ R3 | D(x, F ) < δ},
where D(x, F ) = inf{d(x, y) | y ∈ F }, with d being the usual Euclidean metric in
R3 .
5Enclosures other than boxes are quite possible and this is a subject of active research.
 COMPUTATIONAL TOPOLOGY FOR REGULAR CLOSED SETS 9

Theorem 4.2. Let F be a connected 2-manifold without boundary. For each ǫ > 0,
there exists δ, with 0 < δ < ρ so that whenever a family of boxes B satisfies
S B, b is a subset of F (δ) (Please see Figure 6)
conditions C1–C3, and for each b of
then, for S = F ∪ FI and S B = S ∪ j∈J bj , the sets F and ∂S B are ǫ-isotopic with
compact support. Hence, they are also ambient isotopic.
The quoted theorem depends upon results from Bing’s book on PL topology [7,
p. 214], and related literature [17], as is explained in full [28, 29]. The proof shows
that normals to F do not intersect within the constructed tubular neighborhood,
as is illustrated by the depiction of its planar cross-section in Figure 6.

2γ

Figure 6. 2D version of proper subset condition

If the boxes containing the true intersection curve can be made sufficiently small
so that each such box fits inside F (ρ), then the resultant intersection neighborhood
will contain an object that is both close to the true solid and is ambient isotopic
to it. Considerable success in meeting these constraints has already been achieved
[24] when two splines intersect transversally, while very recent progress from the I-
TANGO team for more subtle spline intersection pairs is now under review [23]. As
a further note on integration, results on root computations from interval arithmetic
[24] are used to provide estimates of initial starting points for the GK algorithm.

Work in Progress. The previous discussion emphasizes results from the I-TANGO
project that have already appeared in the literature, whereas the remarks in this
section are intended to provide some indication of related results that are expected
to appear soon.
Recall that the result of the GK algorithm will be a set of spline patches that
do not fit together precisely along the approximation of their boundary. Recent
work by Andersson, Stewart and Zidani [5] uses the Whitney Extension Theorem to
formulate a conceptual model of how imperfectly fitting patches might be perturbed
to form a topological complex which is the boundary of a non-empty compact
element of R(R3 ). Furthermore, the proposed process extends the patches under
Lipschitz mappings so that the resulting element of R(R3 ) can be shown to lie
within rigorous error bounds of the given geometric input data. Our initial results
on ambient isotopic approximations of 2-manifolds have followed the prevailing
simplification of considering only those manifolds without boundaries [2, 28, 29], but
our recent results have been for the more technically challenging cases of 2-manifolds
with boundaries [1]. Our team is also completing work on spline intersections with
multiple roots [23], to improve the rather loose error bounds known to-date.
10 I-TANGO PROJECT

Interestingly, the interval solids were not initially intended as a means to in-
tegrate issues of topological equivalence and approximations in intersection algo-
rithms. Yet it is a hallmark of our project that these concepts are merging, as
discussed above. Note, further, that the intervals include both approximation er-
rors from truncation of numerical processes within intersection algorithms as well
as the approximations that arise by using a finite set of floating point numbers in
computation as an approximation of the reals. How floating point approximations
effect error bounds for intersection computations is another important issue this
project considers. Attention has focused upon the impact of floating point arith-
metic on polynomial computations near multiple roots [16]. This is an important
issue because the intersection algorithms have to search for roots of the system of
polynomial equations representing the intersecting surfaces and exact arithmetic
computations have not been shown to be practical.

5. Conclusions and future work

Topology and computer science are finding common interest in the emerging area
of computational topology. Various branches of pure topology (point-set topology,
differential topology, low-dimensional topology, . . .) can make important contribu-
tions to establishing the appropriate theoretical foundations. A fundamentally new
perspective arises from the role that computational approximations should play
in the reformulation of central topological concepts. The I-TANGO project is a
research effort concentrating upon these issues specifically with respect to surface
intersections within R3 . Considerable progress has already been made, but many
questions also remain open, as is summarized in this article.

References
[1] Abe, K. et al, Shape preserving isotopic approximations of 2-manifolds, Preprint, October
31, 2003.
[2] Amenta, N., Peters, T. J., and Russell, A. C., Computational topology: ambient isotopic
approximation of 2-manifolds, invited paper, Theoretical Computer Science, 305, 3–15, 2003.
[3] Andersson, L-E., S. M. Dorney, Peters, T. J., Stewart, N. F., Polyhedral perturbations that
preserve topological form, Computer Aided Geometric Design, 12 785–799, 1995.
[4] Andersson, L-E., Peters, T. J., Stewart, N. F., Equivalence of topological form for curvilinear
geometric objects, International Journal of Computational Geometry and Applications, 10
(6), 609–622, 2000.
[5] Andersson, L.-E., Stewart, N. F. and Zidani, M., Error analysis for operations in solid
modeling in the presence of uncertainty. Preprint, June 25, 2003.
[6] Bern, M. et al, Emerging Challenges in Computational Topology, Workshop Report, National
Science Foundation, June, 1999, http://xxx.lanl.gov/abs/cs/9909001 .
[7] R. H. Bing, The Geometric Topology of 3-Manifolds. American Mathematical Society, Prov-
idence, RI, 1983.
[8] Buck, R. C., Advanced Calculus, 3rd Edition, McGraw-Hill, 1978.
[9] Edelsbrunner, H., Home page. http://www.cs.duke.edu/~edels
[10] Farouki, R., Closing the gap between CAD model and downstream application, SIAM News,
(32) 5, June 1999.
[11] Grandine, T. A., Applications of contouring, SIAM Review, 42 (2), 297–316, 2000.
[12] Grandine, T. A. and Klein, F. W. IV, A new approach to the surface intersection problem,
Computer Aided Geometric Design, 14, 111–134, 1997.
[13] M. W. Hirsch, Differential Topology, Springer-Verlag, New York, 1976.
[14] Hocking, J. G. and Young, G. S., Topology , Addison-Wesley, Reading, MA, 1961.
[15] Hoffmann, C. M., Geometric and Solid Modeling, An Introduction, Morgan Kaufmann, San
Mateo, CA, 1989.
 COMPUTATIONAL TOPOLOGY FOR REGULAR CLOSED SETS 11

[16] Hoffmann, C. M., Park, G., Simard, J.-R., and Stewart, N. F., Residual iteration and accurate
polynomial evaluation for shape-interrogation applications, Preprint, 2003.
[17] J. Kister, Small isotopies in euclidean spaces and 3-manifolds, Bull. Amer. Math. Soc., 65
371–373, 1959.
[18] Kong, T. Y., Kopperman, R., and Meyer, P. R., A Topological Approach to Digital Topology,
Amer. Math. Monthly, 98 (1991), 901-917.
[19] Maekawa, T., Patrikalakis, N. M., Sakkalis, T. and Yu, G., Analysis and applications of pipe
surfaces, Computer Aided Geometric Design, 15 (5), 437-458, May, 1998.
[20] Mäntylä, M., Computational Topology: A Study on Topological Manipulations and Inter-
rogations in Computer Graphics and Geometric Modeling, Acta Polytechnica Scandinav-
ica, Mathematics and Computer Science Series 37, Finnish Academy of Technical Sciences,
Helsinki, 1983.
[21] G. Monge. Application de l’Analyse à la Geometrie, Bachelier, Paris, 1850.
[22] Mow, C., Peters, T. J., Stewart, N. F., Specifying Useful Error Bounds for Geometry Tools:
An Intersector Exemplar, Computer Aided Geometric Design, 20 (5), 2003, 247–251.
[23] Mukundan, H., Ko, K. H., Maekawa, T., Sakkalis, T., and Patrikalakis, N. M., Tracing surface
intersections with a validated ODE system solver, MIT Design Laboratory Memorandum 03-
06, Cambridge, MA, October 2003.
[24] Patrikalakis, N. M. and Maekawa, T., Shape Interrogation for Computer Aided Design and
Manufacturing, (Mathematics and Visualization), Springer-Verlag, New York, 2002.
[25] Peters, T. J., Rosen, D. W., and Dorney, S. M., The diversity of topological applications
within computer aided geometric design, Annals of the New York Academy of Science, 728,
198–209, 1994.
[26] Reed, G. M., Roscoe, A.W., and Wachter, R.F., (eds.). Topology and Category Theory in
Computer Science, Oxford University Press, Oxford, 1991.
[27] Requicha, A. A. G., Representations for rigid solids: theory, method and systems, Lecture
Notes in Computer Science, ACM Computing Surveys, 12 (4), Springer-Verlag, 437–464,
1980.
[28] Sakkalis, T. and Peters, T. J., Ambient isotopic approximations for surface reconstruction
and interval solids, Proceedings of the Eighth ACM Symposium on Solid Modeling and
Applications, 176–184, June, 2003.
[29] Sakkalis, T., Peters, T. J. and Bisceglio, J., Application of ambient isotopy to surface ap-
proximation and interval solids, invited paper, to appear, CAD.
[30] Sakkalis, T., Shen, G. and Patrikalakis, N. M., Topological and geometric properties of inter-
val solid models, Graphical Models, 63 (2), 163–175, May, 2001.
[31] Stewart, N. F. Sufficient condition for correct topological form in tolerance specification,
Computer Aided Design 25 (1), 39–48, 1993.
[32] Tilove, R. B., and Requicha, A. A. A. G., Closure of Boolean Operations on Geometric
Entities, CAD 12 (5), 1980.
[33] Voelcker, H. B. and Requicha, A. A. A. G., Geometric modelling of mechanical parts and
processes, IEEE Computer, 10 (12), 48–57, 1977.
[34] Zomorodian, A., Homepage for graduate course Introduction to Computational Topology.
http://graphics.stanford.edu/courses/cs468-02-fall/
12 I-TANGO PROJECT

T.J. Peters, University of Connecticut, Storrs, CT 06269-3155 USA

E-mail address: [email protected].

J. Bisceglio, Bluesky Studios, Inc.

D.R. Ferguson, The Boeing Company, Seattle, Washington

E-mail address: [email protected]

C.M. Hoffmann, Purdue University, East Lafayette, IN

E-mail address: [email protected]

T. Maekawa, Yokohama National University, Yokohama 240-8501 Japan & MIT

E-mail address: [email protected]

N.M. Patrikalakis, Massachusetts Institute of Technology, Cambridge, MA

E-mail address: [email protected]

T. Sakkalis, Agricultural University of Athens, Athens, Greece & MIT

E-mail address: [email protected]

N.F. Stewart, Université de Montréal, Montréal, Canada

E-mail address: [email protected]