Cad Guidebook: A Basic Manual for Understanding and Improving Computer-Aided Design (51 page)

The problems for these constraints arise when the constraints can not be
resolved. In this case, the constraints conflict; or, they are inconsistent. If a bolt is
dimensioned to be aligned with a hole in a plate, but then that bolt is also dimen-
sioned from an edge of the plate and this dimension is not exactly the distance to
the hole, then the bolt is being tugged in two different directions. The constraint
solver tries to locate the bolt with respect to the edge and with respect to the hole,
but it can’t be done, so the constraints need to be changed. This is a simple exam-
ple, and one that is easy to correct. However, when there are dozens or hundreds
of instances involved, it can become difficult to figure out these problems, and
the solver is sometimes unable to give a clear indication of what instances are
actually causing the problem.

Table 10.2 lists a variety of assembly constraint types that may be provided
by the 3-D CAD system. This is not an exhaustive list of possibilities; there may
be many others available to the user. Usually, a set of standard or basic operations
is recommended for users (such as dimensioning between two planes or faces
whenever possible). Figure 10.6 shows an example of a standard dimension be-
tween two planar or flat faces of two part instances.

Note that the examples of possible assembly constraints listed in Table 10.2
are generally using geometric entities that are on the physical part (faces, edges,
vertices, etc.). However, 3-D part modeling often involves the use of extra
geometric entities that are contained in the part model, but would not be seen on
the physical part. Datum planes, reference geometry, and part coordinate systems
are examples of geometric entities that may be available for a part model. The
planes, lines, and points that are found in
this type of geometry would certainly
be other possibilities for use in the assembly constraints (indeed they can be
advantageous).

It is important to realize that some assembly constraints imply the presence
of other constraints. For example, the sample dimension shown in Figure 10.6
implies that the planar faces are parallel. It would not be possible to create such a

Assembly Modeling 253

TABLE
10.2

Some Typical Assembly Constraints

Constraint type Description

Dimension between
two planes

Dimension between
one plane and 1
straight edge

Dimension between
one plane and 1
point

Parallelism between
two planes
Parallelism between
two straight edges
Perpendicularity be-
tween 2 planes
Coincident and
colinear straight
edges

This constraint is probably the best constraint, since it is the
most similar to what a 2-D dimension implies in a drawing.
Keep in mind it is a linear dimension when the planes are par-
allel; another possibility would be an angular dimension be-
tween planes that are not parallel. This constraint can remove
one translation DOF and 2 rotational DOFs.

This constraint can be misleading since the part instance that
owns the edge would still be able to rotate about the edge.
This constraint can remove one translational DOF and one ro-
tational DOF.

This constraint will only control one point on the part instance
relative to another. The part instance with the point is free to
rotate in any direction about that point (sort of like a ball and
socket joint), and that is probably only used rarely. This con-
straint can remove one translational DOF.

This will keep two planar faces aligned, but they could be any
distance apart. This constraint can remove 2 rotational DOFs.
This constraint can remove one rotational DOF.

This constraint can remove two rotational DOFs.

Tangency between a

curved face and a

plane

“Grounding” or

“Rigidifying” one
part relative to
another
“Grounding” one
part relative to the
global origin

This constraint allows the two part instances to slide relative to
each other along the shared edge and to rotate about that edge
(similar to a hinge). This constraint can remove two transla-
tional DOFs and two rotational DOFs.

This constraint allows the part instance with the curved surface
to roll along the flat plane. This constraint can remove one
translational DOF and 1 rotational DOF.

This constraint makes two part instances behave as a single rigid
body, or they are locked together. This constraint can remove
all DOFs between the two instances: they cannot rotate rela-
tive to each other or translate relative to each other.
This constraint makes a part instance immovable. Its location in
the 3-D modeling space is completely fixed. This removes all
DOFs for that instance.

254 Chapter 10

FIGURE
10.6
two planar faces.

Example of a typical assembly constraint—a linear dimension between

linear dimension between these faces if they are not parallel (there would be
infinitely many distances between two faces that are at an angle relative to
each other).

Also, there are problems that can arise from the assembly constraints as the
part models are changed. Looking at Figure 10.6, if one of the planar faces partic-
ipating in the linear dimension were drastically altered to be totally rounded, the
dimension would not really be valid anymore. But, if the faces were just given a
new feature, such as a hole, then most likely the dimension would continue to be
valid. As the design evolves, it would be up to the designer to figure out what
ramifications result from the changes to the parts and how to reconstrain the parts
to preserve the integrity of the assembly model.

Finally, constraints such as dimensions can be used in the parametric sense.
In this case (as with part modeling), equations can be created that relate the val-
ues of one dimension to other dimensions. For instance, a dimension to locate
one part instance could be set to two times the dimension to another part instance.
Thus, when one part instance is moved, the other part instance moves automati-
cally according to the designer’s intention. This may also be referred to as having

Assembly Modeling 255

one assembly dimension drive other assembly dimensions. Keep in mind that
these equations (as well as all the constraint data) are another example of infor-
mation that would be stored in the assembly model based on the instances. That
is, the equations may apply to some instances of a part, but not others. The user
needs to determine how this data is tracked so that changes to the design can be
made in the correct assembly context.

10.4 ASSEMBLY MODEL CALCULATIONS

Once an assembly model is created, there are a number of important calculations
that can be performed using the assembly model. This would include volume,
weight (if material density is an attribute of the part models), surface area, inertial
properties (Ixx, Iyy, Izz, etc.), and center of mass or center of gravity (CG). The
meaning of these properties was explained in the chapter on part modeling.

The only real difference in these calculations versus part models is that the
properties need to be summed for the entire assembly model. The volume for an
assembly is really just the summation of all the part instance volumes. Assuming
all the part models had a density property associated with them, then the weight
of the assembly is just the summation of all the weights of the part instances. Of
course, the calculation becomes more complicated with the inertial properties
and the CG. Now the weight of each instance is not just added together, but con-
sideration is given for the location of the weights, or how the weights are distrib-
uted. Even though this is a more complex calculation, it is still a simple matter for
the CAD system to do this calculation. This is a very significant advantage over
any 2-D CAD systems. Calculating the inertial properties and CG for a large as-
sembly that includes a variety of manufactured parts could take quite a long time
without the assembly model.

Another calculation that can be done with the assembly model is interfer-
ence and minimum clearance. In this case, the 3-D CAD system looks at the var-
ious volumes of space occupied by the various part instances. If any part
instances are sharing the same volume of space, then an interference has oc-
curred, and the CAD system should be able to indicate this condition to the de-
signer. If part instances do not share volumes of space, then the CAD system may
be able to indicate how close they actually are to each other; this is the minimum
clearance calculation or just clearance. Figure 10.7 shows some parts that inter-
fere (the solid part interferes “into”
the translucent part). Figure 10.8 shows the
interference area highlighted for the designer by the CAD system. Obviously, this
can be very helpful in working with assemblies and preventing manufacturing
problems.

256 Chapter 10

FIGURE
10.7

An example of interference in an assembly model.

FIGURE
10.8

The interference volume is highlighted by the CAD system.

Assembly Modeling 257

10.5 ADVANCED ASSEMBLY MODEL

CAPABILITIES

Beyond the calculations that can be performed based on the assembly, there are
more advanced functions that are possible. This includes animations, part-to-part
associativity, and Bill of Material creation.

10.5.1 Animations

Animations are based on sequences of changing part instance positions. For ex-
ample, when a hydraulic cylinder is extended, the rod of the cylinder moves
along the axis of the cylinder. To show this movement, the rod part model needs
to be stored at a number of positions along that axis to show a smooth motion.

There are a couple of ways this can be handled by the CAD system. In one
case, there may be a dimensional constraint for the stroke of the cylinder. As this
dimension’s value is changed, the rod moves in and out. An animation of this
could be shown in real time as the user changes the stroke from one value to an-
other (the graphics adapter temporarily displaying the rod at intervals of the mo-
tion). In this case, the assembly model is showing the change all at once, and the
rod is simply left in a new location based on the new dimension value. This ani-
mation could be repeated by manually changing the dimension over and over.
This sort of animation is quick and easy, but is really temporary from a data man-
agement point of view. The animation is not saved as part of the assembly model.
Of course, some workstations may allow any motion created by the graphics
adapter to be saved in a video file; but this file cannot change automatically if the
parts or the assembly were changed or redesigned. It is really just a video clip.

Another way animation could be handled is to have the CAD system calcu-
late and store the location of the rod at the intervals of motion (the graphics
adapter not doing the calculations). In this case, the assembly model itself may
now store the dimensional values and the graphical information to show the rod at
various positions. Also, the user may be able to enter a starting and stopping value
for the dimension and the number of intervals desired. Thus the fidelity of the
animation can be controlled. Now the assembly model itself can save or store the
animation for the sequence, and it can be used again when part models or other
parts of the design are changed; it should reflect these changes appropriately.

In both these cases, the user is manually changing the assembly model to
show the animation. Another way animations can be created is by building a
mechanism or kinematics model. In this case, the assembly works with mecha-
nism-based constraints that mimic a physical system. For example, there may be
joint constraints (such as hinges or universal joints), ball-and-socket constraints,
cams and/or follower constraints, rigid body constraints, etc.. These constraints