Steve Williamson once said that there
are analysts and then there are synthesysts. The analysts are the ones who
model and simulate physical systems and the synthesysts are the ones who take
all the various engineering knowledge needed and design things.
Machine design is an applied
engineering skill, it is a synthesis of many skills. Machine designers are different
than machine theorists. There are people who comfortably occupy both spaces;
but by and large, the analysts reside in academia while the synthesysts, the
designers, reside in industry. There are a number of good reasons for that.
The people who are analysts think and
teach in terms of equations, physics, and mathematics. The people who are
synthesists think in those terms as well, except they think about them in different
ways. The best practice for the university setting does not align itself with
the best practice necessary for design, the pedagogy that we adopt for the
universities is that the knowledge is presented in a very logical linear progression,
based upon what we know about the subject. We start at the basics and we slowly
move into the more complex subjects as the student develops more ability and
knowledge.
The problem with machine design
specifically is that not only is it multidisciplinary, but the design problem involves
many coupled systems happening at the same time. Mechanical system design,
thermal design, drive design, manufacturing system constraints, cost constraints,
performance constraints, and the interaction between them all affect each part
of the design, so there is a never-ending iteration loops amongst all of them. This
is not to say that motor design is
unique in that aspect; this is why many complex applied engineering problem
solving skills are not taught in the universities: the linear pedagogy does not
lend itself to the kind of procedures that we need to apply in engineering. Machine
design is the fellow who is keeping all the plates spinning at the same time,
you need to have and eye on everything. Of course, that is assuming that the
designer has the overall picture of what they're trying to do. Therefore, it is
called the synthesis part of engineering: you are synthesizing all the
divergent and diverse knowledge.
To answer the question of why the universities
don’t have people teaching machine design, the best teachers are in industry,
practicing their art. In order to effectively teach machine design, the teachers,
are themselves well versed with the problem and are well versed in the holistic
vision of the design problem. The normal academic pedagogy is not well aligned
with the synthesis of the many knowledges inherent in the design problem. The
usual academic pedagogy is much more aligned with the analysis portion of the
puzzle, that is the crux of why we cannot just throw anyone to the wolves and
have them have them teach a holistic mind frame of design with sequential
reasoning and procedures. There was a time when people used the linear pedagogy
to teach machine design, go to some older references in machines from the 1940’s
and 1950’s and you will find design sheets, where the designer has to fill in
the sheets sequentially. This is to keep the designers from forgetting all the
variables they need to calculate for their design. The design equations are
mostly algebraic and assumes linearity. The procedure deals with the
nonlinearity by assuming linear performance, in addition, these are in the days
before variable frequencies, so the designs are assumed to be operating at one
operating point. Fortunately, computational power has increased exponentially,
and we are able to calculate the design variables through numerical simulation,
which gives us the ability to make design decisions quicker and with greater precision.
In so doing, however, it allows us to incorporate more of the constraints from the
other coupled systems into consideration at a much earlier point in the design
process than we have been able to before the computational power, which means
that the up front design is much more completely able to meet the designs and
constraints.
Teaching machine design is not just
teaching the science of mathematics, physics, chemistry, and other academic
subjects. It is also the passing on of experiential knowledge, there aren’t
that many people around who can do justice to the task.
