1

Modern engineering designs often combine several diferent kinds of modeling techniques. Me-
chanical designs today frequently come to life as mathematical solid models instead of as 2D draw-
ings. Tese solid models frequently must work with other kinds of engineering software representing
entities such as wiring diagrams and component layouts. Datafow programming techniques are
often the means of building such connections and expressing this information.
A solid model represents a shape as a three-dimensional object having mass properties. Solid
models are useful in several ways. For example, it is easier for nontechnical personnel to understand
3D renderings than to grasp two-dimensional drawings that consist of orthographic projections, aux-
iliary projections and cross
sections.
Solid modeling soft-
ware may use any of several
methods to represent model
information. Feature-based,
parametric, and so-called
direct or explicit tools
let designers push and pull
models as if they were made
of clay.
Frequently, solid models
are useful because their
geometry can represent
not only the parts being
designed but the intent of
the designer. An example of
design intent might include
keeping two part faces paral-
lel no matter how other part
features change, or maintain-
ing the same mathematical
relationship between a parts
length and width. No matter
what dimensions the de-
signer types in while build-
ing the model, the software
ensures the part defnition is
Basics of
Solid Modeling
Geometric solid models are the preferred way of
defning manufactured parts and assemblies. In
recent years these techniques have taken a role in
characterizing control cabinets and wiring.
Contents
Basics of Solid
Modeling 1
Surface modeling
3
Building blocks
for solids 4
Electrical design
models 4
Presented by
Feature-based models
let designers defne
features pertaining
to geometry as
well as to steps in
downstream analysis
and manufacturing.
Parametric modeling
is the term used to
describe the capturing
of design operations
as they take place,
as well as the editing
that takes place on the
design.
2
E
l
e
c
t
r
i
c
a
l

d
e
s
i
g
n


m
o
d
e
l
s

4
B
u
i
l
d
i
n
g

b
l
o
c
k
s


f
o
r

s
o
l
i
d
s

4
S
u
r
f
a
c
e


m
o
d
e
l
i
n
g

3
B
a
s
i
c
s

o
f


S
o
l
i
d

M
o
d
e
l
i
n
g

1
is a constraint. It will further assume that it should preserve
this relationship throughout any model changes so the part
maintains its original design intent.
Another advantage of the direct-modeling approach is
that the set of active constraints can change dramatically
depending on events during the construction of the model.
For example, the system may capture constraints to support a
modifcation that can contradict constraints captured during
an earlier design iteration. Tis can save time during model
creation compared to the parametric approach, where it is
the user who must change the constraints. Te process of
changing constraints can be tedious and complex.
Conversely, the beneft of using a parametric modeler is
that the CAD system neednt guess at the constraints because
the user spells them out. But the predictability of the CAD
model then depends on the CAD user being skillful enough
to avoid difculties that can arise because of too many or
poorly defned constraints. Tere can also be issues when
one designer must modify a parametric model that another
designer created. Sometimes the complexity of the constraints
can make design intent difcult to grasp.
High-level model quality problems can arise in all feature-
based modelers. Typical problems include unintentional
interactions among features. Tese interactions typically
take the form of small cracks, knife edges, voids, and similar
artifacts between features. Tese efects not only cause
problems in analysis software such as FEA, but also get worse
composed of individual features that describe
how the geometry is supposed to behave.
Early solid modelers were not based on
features. To put a hole through a part, for ex-
ample, the designer might defne a simple cyl-
inder having the diameter of the desired hole
and which was long enough to go through the
part. Designers would then tell the software
to perform a Boolean diference operation
between the part and the cylinder. Te result:
a hole in the part having the diameter of the
cylinder.
Te problem with this scheme comes if
the part dimensions change. Suppose, for ex-
ample, the designer later modifes the part and
makes it thicker. If the designer didnt happen
to make the cylinder long enough to extend
through the new, thicker part, the result is the
model of a blind hole. In this case, the model
captured the geometry the designer specifed,
but it did not capture the design intent of the
designer.
A designer working in feature-based
software, on the other hand, would approach
the through-hole diferently. Te designer
would defne a feature called a through-hole
such that no matter what the dimension of the part, the
hole extends all the way through it. In other words, once
the topology of the design has been called out in terms of
features, any changes to the design always keep these features
operational unless the designer specifes otherwise. Typically,
the software prompts the user for inputs during the defnition
of the feature. Tese may include positional constraints,
algebraic defnitions, and other factors.
A related type of solid modeling scheme is parametric
modeling. A parametric modeler defnes the part model in
terms of parameters. A simple parameter might be expressed
as an equation such as, (diameter) = 3 x (depth), or (width) =
2 x sqrt(length). Parameters can also establish links between
parts as in an assembly model. An example might be a part
position with respect to a reference plane on another part.
Tus parametric CAD systems represent each geometrical
and dimensional constraint in terms of a relationship among
two or more entities. When the designer changes a parameter
(dimension), the CAD system propagates the change
throughout the entire solid model while maintaining the
other constraint relationships.
Another type of solid modeler is a direct modeler. Here
the CAD system creates constraints on the fy; the user
creates no constraints. Te advantage is that this approach
more easily handles high-level changes to the solid model.
For example, the CAD system might note that two nearby
planar faces are parallel to each other. It may assume that this
Many solid modelers include primitive and boundary representations.
In the primitive approach, the user combines elementary shapes in
building-block fashion to create a new shape. Boolean logic commands,
such as union, difference, and intersection, aid in forming new shapes.
With boundary defnitions, 2D surfaces get swept through space to trace
out volumes. Most packages provide several types of sweeps to help
create shapes.
3
E
l
e
c
t
r
i
c
a
l

d
e
s
i
g
n


m
o
d
e
l
s

4
B
u
i
l
d
i
n
g

b
l
o
c
k
s


f
o
r

s
o
l
i
d
s

4
S
u
r
f
a
c
e


m
o
d
e
l
i
n
g

3
B
a
s
i
c
s

o
f


S
o
l
i
d

M
o
d
e
l
i
n
g

1
the designer to change a feature and then propagate it to all
instances in the model. If the designer changes a bolt pattern,
for example, the fexible modeler can put the new pattern in
all locations where the original resided.
Surface modeling
Prior to the advent of solid modeling, computerized
geometry models frequently took the form of wire frames.
Wire-frame models represent 3D part shapes with inter-
connected line elements. Wire frames are the simplest 3D
geometric representation, though not necessarily the easiest
to create. Some modeling programs still use a wire frame
data structure. Te benefts are that wire-frame models
use little computer time and memory and provide precise
information about the location of surface discontinuities on
the part.
Wire frames, though, contain no information about
the surfaces themselves nor do they diferentiate between
the inside and outside of objects. Tus, wire frames can be
ambiguous in representing complex physical structures and
often leave much interpretation to users.
Wire-frame models are created by specifying points and
lines in space. One commonly used approach to creating a
wire frame model divides the computer screen into sections
showing various model views. Designers draw lines to create
top, bottom, side, isometric, and other views of the model.
Designers need not manually draw each line in a wire frame.
Rather, the CAD package constructs the lines based on user-
specifed points and commands chosen
from an instruction menu.
It is important to remember, however,
that not all models that look like 3D wire
frames are wire-frame models. Some soft-
ware lets users build isometric models that
appear to have Z-axis depth, but in reality
do not. Tis software is usually called
2-D software.
Although wire-frame models are the
simplest form of geometric model, the
term is sometimes associated with both
surface and solid modeling. Surface mod-
els defne the outside part model precisely
and help produce NC machining instruc-
tions where the defnition of the structure
boundaries is critical. However, surface
models represent only an envelope of
part geometry, even though tools such as
automated hidden-line removal make the
model appear as a solid.
Surface models, in turn, are created
by connecting various types of surface ele-
ments to user-specifed lines. Typical CAD
surface elements include planes, tabulated
if they pass over to another brand of CAD system where
feature history is lost.
Unfortunately, parametric models imported into a
diferent CAD system may come across without parameters,
features and design intent, even if the new CAD system
supports parametric modeling. Tus users of CAD systems
relying on parametric feature-based approaches generally must
remove and recreate diferent pieces of geometry that they
want to change through parameters and features.
Some direct modelers can work with previously created
parametric models. Te resulting models and their changes
remain parametric and feature-based. Another advance is
the development of modelers that will associatively update
imported solid models when the original model changes in a
diferent CAD application.
Some modelers have whats called a fexible modeling
extension which works directly on geometry. It comes in
handy where a designer needs to make change on imported
geometry that has come in with no features, or on a model
created elsewhere and there is no time to fgure out how the
model was built up. Flexible modelers include a facility for
recognizing patterns and symmetry within the model as a
means of adding intelligence. A user can pick the geometry
to be modifed, for example, and drag it to new place. Te
system then reattaches it to the model, with the automatic
addition of rounds if need be.
If the designer imports a fle, a fexible modeler can
add intelligence to it through such measures as allowing
Surfaces available for geometric modeling range from a simple planes
to complex sculptured surface. These surface usually are represented
as a set of ruled lines. The computer program recognizes these lines as
continuous surfaces. Users select surfaces types from a menu to model
individual details or fully envelope parts.
4
E
l
e
c
t
r
i
c
a
l

d
e
s
i
g
n


m
o
d
e
l
s

4
B
u
i
l
d
i
n
g

b
l
o
c
k
s


f
o
r

s
o
l
i
d
s

4
S
u
r
f
a
c
e


m
o
d
e
l
i
n
g

3
B
a
s
i
c
s

o
f


S
o
l
i
d

M
o
d
e
l
i
n
g

1
tation. Tere are many kinds of sculptured
surfaces, including curve-mesh, freeform,
B-spline, and cubic patch surfaces. Curves
need not even be parallel. Te two curve
families intersect one another in crisscross
fashion, creating a network of intercon-
necting patches.
Building blocks for solids
Solid models can be constructed
from successive combinations of simple
geometric operations with primitives or
with boundary defnitions. Te primitive
approach lets elementary shapes such as
blocks and cylinders be combined in a
building-block fashion. Users position the
primitives and then create a new shape
with the proper Boolean command. With
boundary defnitions, two-dimensional
surfaces are swept through space to trace
out volumes. A linear sweep translates the
surfaces in a straight line to produce an
extruded volume. A rotational sweep pro-
duces a part with axial symmetry, while a
compound sweep moves a surface through
a specifed curve to generate a more com-
plex solid.
Each of these construction methods
is good at handling a particular class of
shapes. Most industrial parts, for example,
consist of planar, cylindrical, or other sim-
ple shapes and are readily modeled with
primitives. But components with complex
contours such as automobile exhaust manifolds and turbine
blades are more easily modeled with boundary defnitions.
Electrical design models
CAD can be used to characterize entities such as electri-
cal panels and cabinets for controls. A CAD program looks
at a control cabinet as an entity described by connections
between blocks representing physical functions such as
relays, terminal blocks, and circuit breakers. One kind of
CAD for electrical cabinets typically takes the form of stand-
alone programs for drawing 2D schematics. Such programs
generally also handle automation for PLC (programmable
logic controller) wiring, terminal blocks, reporting, and so
on.
Tere are facilities for bringing 2D schematic electrical
design data into the 3D model. Once in the 3D model, the
user can place components like motor starters, DIN rails,
wiring ducts, and similar entities in enclosures. Tere are also
routing routines that handle such tasks as proposing alternate
ways to route wires in 3D, handling component spacing, and
cylinders, ruled surfaces, and surfaces of revolution along
with sweep, fllet, and sculptured surfaces. Of course, the
plane is the most basic surface type. Te software merely cre-
ates a fat surface between two user-specifed straight lines. A
tabulated cylinder is the projection of a free-form curve into
the third dimension. A ruled surface is produced between
two diferent edge curves. Te efect is a surface generated
by moving a straight line through space with the end points
resting on the edge curves.
A surface of revolution is created by revolving an arbi-
trary curve in a circle about an axis. Tis capability is useful
in modeling turned parts and parts with axial symmetry. Te
sweep surface is an extension of the surface of revolution.
Sweep curves, however, sweep an arbitrary curve through an-
other arbitrary curve instead of a circle. Te fllet surface is a
cylindrical surface connecting to other surfaces in a smooth
transition. Previously, this was a tedious and manual process.
But over time CAD has solved the problem with the precise
mathematical continuity required by many applications.
Sculpted surfaces are the most complex surface represen-
A typical CAD package might create a wire frame model from points the
user specifes.
5
E
l
e
c
t
r
i
c
a
l

d
e
s
i
g
n


m
o
d
e
l
s

4
B
u
i
l
d
i
n
g

b
l
o
c
k
s


f
o
r

s
o
l
i
d
s

4
S
u
r
f
a
c
e


m
o
d
e
l
i
n
g

3
B
a
s
i
c
s

o
f


S
o
l
i
d

M
o
d
e
l
i
n
g

1
Graphical data-fow gives the user access to black-box
processes only through their connections. Also, blackbox
processes can be viewed as reusable components that dont
know the name of other black boxes with which they
communicate. Users can reconnect diferent black boxes
endlessly to form diferent applications without having to
change any of the blackboxes internally.
In the case of designing the contents of a control
cabinet, for example, the CAD program might view each
individual component mounted on a DIN rail as a blackbox
process. Connections between the control cabinet compo-
nents dont afect their internal functions, only their states.
In IT lingo, the streams of data passing between black
boxes are called information packets, and the connections
through which they pass are bounded bufer connections.
Each process identifes its related connections by port
names, rather than directly. Typically, a connection engine
or scheduler routine relates port names to the real network
and drives the individual processes.
segregating low-voltage wires from
those carrying high-voltage. Other
tasks normally conducted in 3D
electrical models include checking for
clearances, planning wiring paths, cre-
ating harnesses, and so forth. Changes
typically get linked back to the 2D
drawing for documentation.
Some programs for modeling
such entities use a data-fow program-
ming approach as a representational
scheme. Te key to data fow is that it
is a handy way of defning networks
of blackbox processes. Tese pro-
cesses exchange data across predefned
connections. Data-fow programming gets its name from the
fact that application developers need only work with fows
of data through the connections rather than having to defne
a sequence of commands as with conventional sequential
procedure code.
Te frst widely used data-fow program was the spread-
sheet. Each cell in the spreadsheet can be considered a black-
box process. When any of those cells update, the frst cells
value automatically recalculates. One change can initiate a
lengthy chain of changes when one cell depends on another
cell which in turn depends on yet another, and so forth.
But data fow is not just for recalculating numeric values
as in spreadsheets. Te concept eventually expanded to let
drawn entities represent blackbox processes. Tus it can be
used to re-draw a picture as directed by mouse movements.
A graphical data-fow application becomes essentially a list
of connections which can be generated by a graphical tool.
Among the frst such graphical data fow programs to be-
come commercially available was the LabView program.
Modern CAD programs let
the user bring 2D schematic
electrical design data into
the 3D model. Once in the
3D model, the user can place
components like motor starters,
DIN rails, wiring ducts, and
similar entities in enclosures.
Tere are also routing routines
that handle such tasks as
proposing alternate ways to
route wires in 3D, handling
component spacing, and
segregating low-voltage wires
from those carrying high-
voltage.