Chapter 1
Introduction to Computer Aided Design Drafting

Computer Aided Design packages employ several different ways of representing the geometry of objects. Most packages have two different ways in which the objects may be displayed. They may be displayed as Wire Frames or as Solids.

Wire frame displays are easier to work with during model construction since the display is regenerated much more quickly than a shades solid. It is important for the creative process that display regeneration is as quick as possible, so that the thought process is not unduly interrupted by having to wait. CADKEY employs wire frame model construction using lines, arcs, conic sections, and splines for defining the edges and shape of the objects. The lines, arcs, circles, conic sections, and splines are referred to as entities. The entities can be thought of as the building blocks we use for defining the shape of the objects we want to design. We will look at the entities and their definitions and attributes in detail later in the chapter.

SOLIDS MODELLING:

Solids modelling can be approached in two different ways; Constructive Solid Geometry (c.s.g.); and Boundary Representation (B-rep). CADKEY's solids modeller uses the latter approach as can be deduced from the above description of the mode of wire frame construction.

Constructive Solid Geometry:

In constructive solid geometry one uses a set of solids primitives on which one performs boolean operations to create the desired geometry. The standard solids primitives are such objects as blocks, spheres, cylinders, cones etc. (see figure 1.1).
Most c.s.g. software uses wire frame display, with hidden line removal, of the solids during model construction in order to speed up screen redraw. Shaded solids are generally only used for final presentation purposes.

Constructive solid geometry is efficient since it requires relatively little storage space in order to fully describe an object. For instance a block can be described by its primitive type, its length, depth, height, location in space, and material density.

The main problem with pure c.s.g. software is that it is difficult or impossible to define complex shapes like automobile bodies, aircraft and ship hulls.

Boundary Representation:

Boundary representation solids modelling is ideally suited for modelling complex shapes. The boundary of the body
is defined as a surface. The interior and exterior of the body is defined by the sign of the surface normal i.e. the software will have certain surface construction conventions which force us to clearly indicate which side is intended as the interior, and which the exterior. The software uses this to define the positive and negative direction for the surface normals.
The surface is often defined by B-spline surface patches (see Appendix C for an introduction to mathematical surface
modelling techniques). One way that this can be implemented in practice is for the user to define a series of sections through the body. These sections are displayed using B-spline curves (see Appendix B for an introduction to B-spline and other curve fitting techniques). When one is content with the shape of the sections, the surface can be generated by performing a "skinning"
operation i.e. taking the complete set of points defining all the sections and fitting a surface through them. This surface is divided into surface patches as illustrated in the wire frame display of a ship's hull shown in figure 1.3.
B-rep solids modelling requires much more storage space than c.s.g. since it takes a large number of points to define each surface. This means that one will reach the practical limitation on the size of the model much more quickly with boundary representation than with constructive solid geometry. The practical limiting factors are imposed by the hardware available e.g. processor speed, RAM, and storage space.

One can perform boolean operations on solids generated using B-rep, in the same way as with constructive solid geometry. Again there are practical limitations to the complexity of the model that can be handled.

Software that combines boundary representation with constructive solid geometry would give one the best of both worlds. Also, at the current rate of growth of affordable computing capacity the practical limitations on model size are constantly being pushed out further and further.

CADKEY's modelling techniques:

CADKEY uses a wire frame modelling technique, and CADKEY Solids uses boundary representation. Solids takes the wire frame model and fits planes and complex surfaces through the edges and sections. More will follow on the operation of the solids modeller in a later chapter. The remainder of this chapter will deal with some of the basic principles behind the operation of modelling in CADKEY itself.

The basic building "blocks" of CADKEY are called entities. These entities come in different types just like the primitives mentioned in connection with Constructive Solid Geometry. Cadkey has the following basic entity types and sub-types:
Line
Arc
Circle
Point
Polyline
Fillet
Conic
Polygon
Spline
These entities are all lines or curves of one form or another which have what we refer to as attributes. The attributes are such things as line type (see figure 1.4), line width (see figure 1.5), line colour, pen number (pen # refers to selecting plotter pens without using line colour to specify the pen to be used). Polygon and closed polyline entities have an additional attribute. They can be outlined or filled (see figure 1.6).
In addition to the entities that we use for constructing the objects we have another group of entities called detail entities. These entities are:
The text used in details has the following attributes:
The dimensions used have the following attributes:
The arrows have the following attributes:
The witness lines have the following attributes
The leader lines have the following attributes:

The leader line options are shown in figure 1.10.

The cross-hatching has the X-HATCH Type as its basic attribute (i.e. brick, steel etc.). In addition we have the following two attributes:

Figure 1.11 shows the X-HATCH Types available.

The above outline is intended to give you an idea of the thinking behind the CADKEY program. It will give you a better concept of how the program functions, and how you may change things that are not the way you would like them to be. This applies to both your model and presentation drawings. For more detail on the program menus and sub-menussee Appendix D. All these principles will become clearer to you when you start using the program.
If we look at the way in which the entities are defined we get an idea of how model information is stored. Lines for
example have to have the following information stored:

LINE,pt1(x,y,z),pt2(x,y,z),L-type,L-width, L-colour, [PEN #].

In other words, we need to store the entity type, the minimum geometrical information (in this case the x, y, and z coordinates of the line end points, since all points in between can be reconstructed by using the mathematical equation for a line in 3-space), plus the entity attributes (i.e. line type, line width, line colour, and an optional specification of pen number. The way in which the information about each entity type may be stored can be seen from the structure specified for the CAKEY Advanced Design Language (CADL). See chapter 16 for more details on CADL.

End of Chapter 1