Research & Development
|
There are many interesting things
that could be further developed in the CAD world. Some of the below things
have been part of EU projects, small projects with universities and “home projects”,
i.e. work has been done, and still is being done, in the spare time as a
hobby. Some examples: New Mathematics and Optimization Techniques for
surface definition The available mathematics (like
NURBS) is not suited for surface definition. Free Form Surface Elements that
are of a high quality should be used to define Surface Areas. Optimization Techniques should be used to
put Surface Areas and Surfaces together. Measurements on deformed surfaces, Deviation
Objects The CAD model is a nominal
(perfect) model that never exists in reality. Some deformations (from
gravitation, temperature, etc) are expected and not any errors of the
product. Measured data has to be modified for such natural “nuisance errors”. One way to this is to modify the physical
product itself. Use jacks to make an aircraft wing more like the nominal CAD
model, make a special tool for the measurement of a thin sheet metal part,
etc. But such corrections can be made
mathematically. The real errors should also not be just a table of deviations
or a colored picture that represents the deviations. The result should be a
Deviation Object (Model) that easily can be used in the CAD system for
further analysis and to define corrective actions. Big Point Clouds Some types of measurement
instruments generate large amounts of point coordinate data. Special database
techniques must be used in order to be able to handle such large amounts of
data efficiently. It is however for many (most) applications not at all necessary
to store and use all this point data. There are smarter ways to do it. Generic Models Product data that is stored in
computers are mainly Results Data, and not Method Data. Information about
analysis that were made, alternative solutions that were analyzed and
criteria that were used in the design process are seldom in any computer
model. This information is in the heads of engineers and perhaps on paper.
But it is this Method Data that is of interest when a similar product shall
be developed! Method Data can and should be
stored in the form of Generic (Computer) Models! |
|