Cad Guidebook: A Basic Manual for Understanding and Improving Computer-Aided Design (40 page)

For better or worse, many CAD system vendors have tried to make 3-D CAD
systems as fundamental and simplistic as possible. This has certainly expanded
the reach of the software, but this unfortunately insulates the user from grappling
with a number of issues that should not be ignored. These issues often relate to

180 Chapter 7

making a good mathematical model that unleashes the underlying power of the
3-D CAD system (instead of just making a model that looks good at the moment).

What designers really need to do is make models that can automatically
store what they meant or their design intentions. They also need to know how the
3-D CAD modeling works so that they can debug the model when it can’t comply
with the desired design intent. In other words, it is important that users gain a
basic understanding of how the modeling software really works (as opposed to
just mimicking the models from an on-line tutorial). In order to gain this deeper
understanding of the software, there are three misconceptions that most users
need to overcome. The remainder of this section attempts to address these mis-
conceptions.

7.3.1 “You Can’t Do 3-D Without 2-D First”

It may come as a surprise to some users that good 3-D models almost always
require a good 2-D foundation. So, time spent understanding 2-D drawings or
planar mathematics is not a waste of time. And, although some 3-D CAD systems
have the option to create arbitrary 3-D surfaces without any 2-D geometry (using
points, lines, splines, etc. in space), building part models in this fashion is often
not productive. Instead, the 3-D CAD user needs to figure out what 2-D geometry
can be used to build up the 3-D model. This often relates to finding various planes
within the imagined part that can be used as a place to sketch 2-D geometry.

In Figure 7.10, one can see the underlying 2-D entities (lines, dimensions,
etc.) that make up a particular section or feature of a 3-D model. The plane on
which these 2-D entities are created should not be thought of as a drawing, how-
ever. The 3-D CAD system will refer to it as a sketch or sketch plane or sketch
pad or sketching datum. The sketch is more informal than a drawing since it is
not expected to be seen by anyone besides the designer (unlike a drawing that is
sent to manufacturing, vendors, customers, etc.). However, the sketch is more
formal than a drawing in a mathematical sense. The geometric entities on the
sketch plane are “intelligent.” The 4 lines that form a rectangle on a 2-D CAD
system are really just 4 separate entities with no formal relationship to each other.
In the 3-D CAD system, however, 4 lines that are created for the rectangle shown
in Figure 7.10 will have relationships to each other. The 3-D CAD system knows
that the end points of the lines are connected. This allows the 3-D CAD system to
create 4 faces based on the 4 lines and know whether the 4 faces touch or share
an edge.

Beyond the relationships between the various entities on the sketch plane,
the sketch plane usually also offers the ability to parameterize dimensions. This
means that mathematical formulae can be programmed into the sketch plane so
that one dimension drives another dimension. For example, a hole that is sup-
posed to be 25% of the distance between two edges, can be programmed to be

3-D CAD 181

FIGURE
7.10

An example of the 2-D foundation for a 3-D feature.

located at 0.25 times the dimension between the edges. Then, when the edges
move, the hole will automatically reorient itself to be at the correct location. This
is generally what is meant by making a parametric or intelligent model, and this
is an example of capturing design intent. Although this concept can be expanded
to include parametric relationships between many of the features of a single part
model, or even to the relationships between totally different parts models (such as
in an assembly model), one should first master this parametric concept at the 2-D
level within the context of the sketch plane, and then build on the possibilities
offered in the full 3-D part modeling and/or assembly modeling.

So, as one learns to master 3-D CAD, one should not be mislead into think-
ing that mastering 2-D geometry is no longer needed. Indeed, understanding 2-D
mathematics may be even more important now.

7.3.2 “Solid Models Are Really Just Special

Surface Models”

As mentioned earlier, there are different types of 3-D models. For instance, a sur-
face model is a model that is open (refer to Figure 7.3). The faces or surfaces of
the model do not completely enclose a volume of space. It is also basically a
physical impossibility. No physical part is going to have infinitely thin walls like
the surface model.

182 Chapter 7

The more common type of model is the solid model. The solid model is
physically realistic since it does have a volume associated with it. However, it is
really a misconception to think that they are different kinds of models at all. In
fact, the solid model is a surface model. It is merely a special case of surface
model where all the surfaces meet at mutually shared edges (within a tolerance or
limit). For the solid model, it is said that the surfaces are completely stitched, and
the surfaces become faces of the solid model.

There are a number of advantages to understanding the nature of the
surface model within the solid model. First, there are solid models that are best
made by starting with a surface model. In this case, an open part is created that
has the basic shape needed (with no thickness applied). Then, an amount of
thickness is added all over the surfaces. This is usually referred to as shelling a
part. Figure 7.11 shows an example of this type of part; notice that there is now
wall thickness.

Another case of using the surface nature of solid modeling to one’s advan-
tage is making a normal solid model, then making it open, and then making the
open part solid again (refer to Figure 7.12). In this case, one surface was removed
from the solid model. This exposed the internal surfaces. These surfaces can then
be used as the basis for a part that can then be shelled to regain a solid model.

Finally, understanding the underlying surfaces in solid models helps one
correct or debug part models that are not working as desired. For instance, when
one sketches a 2-D shape to be extruded into a solid (such as a rectangle to be

FIGURE
7.11

An example of a shelled surface model.

3-D CAD 183

FIGURE
7.12
inal model.

An example of making a solid part after deleting a surface from the orig-

extruded into a block), but one does not make sure that the lines of the 2-D shape
do not connect at their ends, the CAD system may wind up making a surface
model instead of a solid model. As shown in Figure 7.13, a small gap between the
edges of the surfaces is left where the original 2-D lines did not connect. Some-
times it is not clear that this has occurred (particularly for small features on a
part). One way to correct the problem is to have the CAD system look for free or
unstitched edges of surfaces.

So, as one learns to master 3-D CAD, one needs to realize that the solid
model is really a special case of a surface model.

7.3.3 “You Are a Historian, Not a Sculptor”

The next concept, called part history, is is probably the most difficult concept for
new users of 3-D CAD systems to appreciate. This is probably because there is a
natural tendency to view the 3-D model creation as an exercise in sculpting. New
users often create a basic shape and then cut away segments of that shape until it
has the desired final shape. Although this may be an appropriate modeling tech-
nique in some cases, there are usually much better approaches. These better tech-
niques are usually superior in their ability to permit fast and easy modifying or
tweaking the part over the design iteration process.

Instead of viewing modeling as sculpting, users should consider part mod-
eling as creating history. The 3-D CAD system does not electronically store tril-

184 Chapter 7

FIGURE
7.13

An example of a surface model created by mistake.

lions of little specks that the user carves away (as if sculpting with 3-D pixels).
Instead, the 3-D CAD system records or tracks or captures the steps that a user
has followed during the creation of the 3-D part model (whether it is a surface
model or a solid model). Recording these steps does not really take up much
space on the computer system (as would the 3-D pixels in the sculpting idea).
The history of steps is also a concise means of storing the intelligence of the
mathematical model of the part model. It also retains a logical approach to how a
part has been modeled. The designer starts with the some big first step that gives
the basic geometry, and then one adds, removes, or builds upon that base to get
the desired level of detail for the final 3-D part model.

Although there are some 3-D CAD systems that do not use this history-
based modeling method, the vast majority of systems and users do use this
method in one way or another. CAD systems may refer to this recording of steps
as part history, model history, history tree, model tree, feature list, etc. They all
mean the same thing—the user’s sketching and modeling process for the part is
stored in some sort of sequential format. Figure 7.14 shows some steps in the
formation of a part. The steps proceed from the top of the list to the bottom.
These are steps that would be used to create the part in Figure 7.1.

This approach to 3-D modeling might also be considered features-based
modeling. This is because each of the major steps in the part history may be
called features. A boss protrusion sticking out of a part would be a feature; a par-
ticular hole drilled through the side of the boss would be a feature, etc., and each
of these features would be found as a step in the part history. However, although

3-D CAD 185

FIGURE
7.14

An example of part history.

all 3-D models should have features, some systems may not track this history as
fully as others. So it seems to be more accurate to refer to what most 3-D CAD
systems are doing as history-based, not just features-based.

As with the other misconceptions that are trying to be dispelled here, it
turns out that there are significant advantages to the user if they understand what
the history really means. First of all, the terminology associated with part model-
ing makes more sense. The 3-D part models can be rolled back, replayed, up-
dated, re-generated, stepped through, etc. All these terms are referring to
activities relevant to the part history. Refer to Table 7.2 for a simple explanation
of some of these terms.

An extremely important concept that needs to be understood with respect to
history-based modeling is that the order of steps is usually very important to
making a good 3-D part model. One can not just pick any order to follow for the
steps. For instance, if one wants to create a hole in a part using a specific flat
planar surface as the sketch plane, then that surface must exist at the moment that
one needs to make the hole. If the planar surface turns into a curved surface by
some other previous step in the history, then the flat plane is lost. So, the hole is
probably going to be “messed up” in some way. By the same token, once the hole
is created in the history, then it often does not matter if the plane is destroyed in a
later step in the history. What is important is that the plane existed at the appro-
priate time.

186 Chapter 7

TABLE
7.2

Part History Terminology

Terminology Basic Explanation

Feature A step in the history of a part.

Replay or regenerate Replaying a history-based model refers to having the CAD