ENGINEERING AND THE COMPUTER
| Pages | 43-45 |
| Date | 01 January 1981 |
| Published date | 01 January 1981 |
| DOI | https://doi.org/10.1108/eb057175 |
| Subject Matter | Economics,Information & knowledge management,Management science & operations |
ENGINEERING
AND
THE
COMPUTER
Our article on a new approach to CAD/CAM in mechanical engineering has
been written by Dr. Brian Walker who, after graduating from Queen's
University Belfast, went on to research in the use of computer based control
systems, leading to his Ph.D in 1968. Since 1959 he has been with Ferranti
and is now head of software and services development, and Product Manager
for Ferranti Cetec Graphics' systems for engineering applications.
In the fifteen years or so that have
elapsed since SKETCHPAD showed
that computer techniques could be used
to manipulate shapes
as
well
as
numbers
and words, there have been recurring
predictions that this new technology
would quickly result in increased
efficiency and economy in engineering
design and manufacture. These expec-
tations have not in fact been matched
by results, particularly in the field of
mechanical engineering.
In some application
areas,
such as the
manufacture of printed circuits and
integrated circuits, reasonably wide use
is now being made of CAD techniques
to improve methods of design, design
analysis and production. A number of
systems are available offering benefits in
cost and lead time. Although similar
claims are made for equivalent systems
applicable to mechanical engineering
design and manufacture, no significant
penetration of the industry has been
achieved by any of the equipment
currently available.
Why is this so? Are there any useful
lessons to be learned by comparing a
relatively successful application area
such
as
printed circuit manufacture with
mechanical engineering?
The most obvious difference is
that printed circuits are essentially
two-dimensional, whereas (with a few
exceptions) mechanical engineering is
concerned with three dimensional
objects. Secondly, the analysis of an
electronic circuit can usually be made
without a knowledge of the geometry
of the printed circuit layout. There are
notable exceptions in the design of high
frequency circuits, but by far the major
proportion of electronic equipment
operates at frequencies low enough to
make circuit layout insignificant from
the point of view of circuit design.
Adequate analysis of a component
or sub-assembly in mechanical engin-
eering, on the other hand, requires full
knowledge of material, topology and
geometry.
The third important difference is
that the electronics industry thrives on
a range of standard components. Users
have a broadly similar component
library, whereas the users of computer
techniques in mechanical engineering
need widely differing and often totally
user-generated component libraries.
Finally, the end product of greatest
importance in electronic equipment
manufacture is the printed circuit,
virtually a drawing in
itself.
A
relatively
small production data "fan-out" is
necessary for a small range of drilling,
component insertion, and test equip-
ment. Conversely in mechanical
engineering, while a 20-drawing may
well be a valuable output from the
design process, the completed product
involves a wide range of manufacturing
methods, tools, and processes between
the drawing and the delivery of the
assembled product.
What the User Needs
Taking these major differences in order,
better use of CAD/CAM techniques in
mechanical engineering could be made
if:
—
the whole range of user-generated
and bought out components could be
held in a company-wide data base.
Until the designer has access to both
geometric and non-geometric data
for each component or assembly, the
use of computer based CAD/CAM
techniques can only be piecemeal.
—
the information generated at design
and analysis stages could be used
directly to produce appropriate data,
not only for drafting, but for manu-
facturing processes and management
information fan-out such as stock-
control, order input processing, and
cost analysis.
These attributes of a comprehensive
design and manufacturing system
require a new approach to 30-part
modelling, data base management, and
the output of manufacturing and
management information. To match
the needs of the mechanical engineering
industry and the preferred working
methods of draughtsmen and designers,
the whole system must be turnkey, and
access from terminals and graphic work
stations which do not demand a pro-
found change in the attitude of
the
user.
Viewed against these requirements,
previous CAD systems offer, at best, to
the mechanical engineering business
a patchy solution to its CAD/CAM
needs.
However, the major part of the
work relating to all the requirements
had already been done at various un-
connected centres in the UK. All that
was needed was the integration of
available software and hardware into a
linked system.
Total Objectives of
CAD/CAM System
To review the requirements, it has been
concluded that the mechanical engineer-
JANUARY 1981 43
To continue reading
Request your trial