Prev: Maximum ammount of local variables in cg shaders ?
Next: Online Exams for Certification, Free Practice Exams, Study Material, Dumps
From: ChrisQ on 3 Nov 2009 10:40
Robert Myers wrote:

>>
> Are you now, or have you ever been, a rocket scientist?
>
> Most engineering that goes wrong is because most things are under-
> engineered: the design was supported by inadequate analysis.
> Sometimes the analytical capability doesn't exist (the science isn't
> there) and sometimes people are just unwilling to spend the money.
>
> I'm not now nor have I ever been a solid state physicist, but the
> black art of device engineering seems to result from a lack of
> adequate science, so that much of it is inspired guesswork and cut and
> try. As far as I know, computational and theoretical device physics
> aren't in very good shape, which makes the progress that does take
> place even more amazing.
>
> Rocket engineering is also a lot of inspired guesswork and cut and
> try, sometimes with unwitting test pilots aboard the under-engineered
> vehicle. Rocket science could be in much better shape, but no one
> wants to spend the money.
>
> Robert.


Real world engineering is always a devils compromise between cost,
timescales, performance and available technology, while science builds
the theory that says something is possible.

Engineering is the process of translating abstract concepts into
physical reality. The science is only the starting point on the way to
building things that work. If you have ever done any discreet component
analog rf circuit design, for example, you would find that all the math
in the world won't describe a physical layout and circuit values that
are optimal. You use the science to model and calculate approximate
circuit values. add in previous experience and some assumptions, build
it, then fine tune layout and values to optimise the performance. There
are just too many variables to describe some things in any reasonable
mathematical way. Seems to me that a lot of engineering is like that.

I guess the philosophical point is that while engineering is constrained
by scientific limits, the creative part is what reconciles those limits
with the original vision of a working solution. The vision comes from
where ?. One could argue that that's part of the science as well, though
it can be more intuitive than analytical :-)...

Regards,

Chris

From: Robert Myers on 3 Nov 2009 16:21
On Nov 3, 10:40 am, ChrisQ <m...(a)devnull.com> wrote:

> Real world engineering is always a devils compromise between cost,
> timescales, performance and available technology, while science builds
> the theory that says something is possible.
>
> Engineering is the process of translating abstract concepts into
> physical reality. The science is only the starting point on the way to
> building things that work. If you have ever done any discreet component
> analog rf circuit design, for example, you would find that all the math
> in the world won't describe a physical layout and circuit values that
> are optimal. You use the science to model and calculate approximate
> circuit values. add in previous experience and some assumptions, build
> it, then fine tune layout and values to optimise the performance.  There
> are just too many variables to describe some things in any reasonable
> mathematical way. Seems to me that a lot of engineering is like that.
>
> I guess the philosophical point is that while engineering is constrained
> by scientific limits, the creative part is what reconciles those limits
> with the original vision of a working solution. The vision comes from
> where ?. One could argue that that's part of the science as well, though
> it can be more intuitive than analytical :-)...

I don't know how I got into this. If I see smoke rising from an
argument over definitions, I usually turn the car around and drive the
other way. ;-)

There are some important issues, although, in the end, arbitrary
labels are arbitrary labels.

Just as there are many different kinds of people, there are different
kinds of scientists and engineers with all kinds of different ways of
pursuing their vocations.

The characterizations of "rocket science" that got me into this are so
naive and wrong-headed that I just couldn't not say anything. Whether
you want to call it rocket science or rocket engineering, believe me,
it is not easy, as the US quickly learned as it tried to hurl into
space satellites that the Russians sneered at as "grapefruit." In the
process, US rocket scientists (actually, the German rocket scientists
we had captured) learned the hard way just how difficult it is to put
something even into into sub-orbital flight that might span
continents, never mind actually to put something into orbit around the
earth, and there were many spectacular failures to drive the point
home. Whether the US succeeded or failed was a very big deal, and the
stakes were much higher than most at the time appreciated. If anyone
had snotty thoughts or attitudes about "rocket science," they would
have kept them to themselves. That was the error of "duck and cover"
drills in grade school.

In such a context, all philosophical and dictionary considerations
seem silly. Whatever talent was available was put to work, and there
was a serious shortage of talent that was up to the task. Now, of
course, you could just shop it all out to India, but this is a
different world, because everyone else internalized the lesson that
the United States learned while the United States is now in the
process of aggressively forgetting so that the smartest people can use
their talents stealing from everyone else.

In *that* context, is there anyone here who deserves to be sneered
at? The departmental labels don't matter.

Both science and engineering should and do concern themselves with
identifying and asking questions worth exploring. That's a hugely
important skill and one that can't really be taught. Whether it's
science of engineering, someone in Cal-Trans has been making too many
assumptions and asking too few questions unrelated to money or
votes.

If you persist in trying to make class distinctions among people with
technical skills, you simply increase the likelihood that some hack
will be able to make mistakes without adult supervision. The nation
cares more about Michelle Obama's wardrobe than it does about the
competence of the technical people who supposedly take care of the
nation's infrastructure. There's not much glamor in fatigue or
fracture, but, believe me, it's *hard*. There is science,
engineering, mathematics, intuition, and just plain dumb luck (or the
lack of it, as in the San Francisco-Oakland Bay Bridge incident).
There is also theory and experiment, under whatever category you put
them.

Both science and engineering are constrained. You can make design
choices in science just as you can in engineering. That is
particularly true if you are a modeler. Neither scientist nor
engineer works with a free hand, and just exactly what the constraints
are varies wildly. Both scientist and engineers vary wildly as to how
they think, whether abstractly or in very concrete terms. Even those
who think concretely (and they can be found in both science and
engineering) have huge variations in what they think of as concrete.
Among those who think abstractly, examples of which can be found in
both science and engineering, also bring wildly differing cognitive
tools to whatever they do.

When you try to insist on distinctions, that potential benefit (many
different eyes and minds) becomes a huge liability. If everyone
thinks pretty much the same way, it might make for a more peaceful
workplace, but it very likely does not result in better science or
engineering, and you can't tell how someone thinks by looking at a
C.V. (degrees and job experience).

I remember Stephen Weinberg stopping himself in the middle of a
lecture hall peroration to comment, as if to no one, that he was
pontificating. He was, and I didn't understand a word of it. I'll
stop.

Robert.










From: Robert Myers on 3 Nov 2009 16:54
On Nov 3, 8:58 am, n...(a)cam.ac.uk wrote:
> In article <30b13890-50c6-4b47-8ed4-99769f847...(a)o10g2000yqa.googlegroups..com>,
> Robert Myers  <rbmyers...(a)gmail.com> wrote:
>
>
>
> >Nick works at the
> >University of Cambridge, where the University Press publishes the
> >Journal of Fluid Mechanics and there is a long and impressive history
> >of non-trivial contributions to the field.  He really should have
> >known better.
>
> Oh, yes, I know that - our Engineering department is very highly
> regarded for its work in that area.  Now, what is it that I should
> have known better? :-)

Fluid mechanics, in whatever department it is taught, is a scientific
discipline. I believe that Harvard teaches it in the "Engineering
Science" department.

Robert.
From: Terje Mathisen on 4 Nov 2009 01:38
Robert Myers wrote:
> Fluid mechanics, in whatever department it is taught, is a scientific
> discipline. I believe that Harvard teaches it in the "Engineering
> Science" department.

Robert, isn't the interesting/relevant part of Fluid Mechanics the fact
that almost all interesting flow regimes are impossible to solve
exactly, so even the science part of it becomes a set of simplified
equations for various hopefully interesting subsets?

If so, then even those calling it science are doing engineering imho, in
the form of finding mathematical simplifications that give useful results?

I guess I'm saying that the boundary between science and engineering is
rather fluid.

Terje

--
- <Terje.Mathisen at tmsw.no>
"almost all programming can be viewed as an exercise in caching"
From: Robert Myers on 4 Nov 2009 02:42
On Nov 4, 1:38 am, Terje Mathisen <Terje.Mathi...(a)tmsw.no> wrote:
> Robert Myers wrote:
> > Fluid mechanics, in whatever department it is taught, is a scientific
> > discipline.  I believe that Harvard teaches it in the "Engineering
> > Science" department.
>
> Robert, isn't the interesting/relevant part of Fluid Mechanics the fact
> that almost all interesting flow regimes are impossible to solve
> exactly, so even the science part of it becomes a set of simplified
> equations for various hopefully interesting subsets?
>
> If so, then even those calling it science are doing engineering imho, in
> the form of finding mathematical simplifications that give useful results?
>
> I guess I'm saying that the boundary between science and engineering is
> rather fluid.
>

I was trying to make that argument among others in my longer post.
Every branch of science has pure and applied incarnations.

Engineering is the odd beast because it invents rather than studies,
but every branch of applied science finds its way into some kind of
engineering. A natural scientist does not have the luxury of
designing the object of his interest: the natural world is a given.
When a scientist invents a new abstraction, though, he's doing
something very much akin to what an engineer does: he designs
something that is more or less useful for an intended purpose, and
more or less ex nihilo, even if while standing on the shoulders of
giants.

Science v. engineering isn't all that helpful a distinction for those
of us who live in the world of applications, but there are
incarnations of fluid mechanics that are most definitely pure science--
certainly if you allow mathematics to be a science. If it's not a
science because mathematics isn't a science, then it's a branch of
mathematics. It has nothing whatever to do with the mechanical
engineering, aero engineering, or chemical engineering departments
when you are doing, say, astrophysical fluid dynamics and the
equations are still the same, maybe only with more phases and/or the
MHD approximation. Engineers and scientists both wrestle with
radiative transfer.

Judicious approximation and idealization are practically hallmarks of
good science. No planet orbits the sun in an ellipse, but it's a
helpful idealization. For back of the envelope estimates, both
scientists and engineers will be happy to ignore the eccentricity of
the orbit entirely and put the sun or the earth at the center of a
circle and even to ignore the fact that neither the earth nor the sun
is an inertial frame, even though almost no geophysical fluid
mechanics can ignore the rotation of the earth.

The equations of fluid mechanics are the first three moments of the
Boltzmann equation. The infinite hierachy is closed at the first
moment for incompressible flow by adding viscosity, which is a form
of what field theorists call renormalization. If you do turbulence
theory, you get a similar hierarchy with more elaborate methods of
attempting to renormalize, none of which seem as natural or to work as
well as viscosity (but no one knows why the low order closure of the
Navier Stokes equations seems to work so well). To me, all these
equations look so much alike that the distinctions people make seem
artificial. You can attempt Feynman diagram methods for turbulent
flow, but they turn out to be just a low Reynolds number expansion.
If I learn a trick in one field, why shouldn't I get mileage out of it
wherever a similar circumstance occurs?

The idea that it's all so hard that you might as well just throw it
onto the computer and stop losing sleep may well have seized the
imagination of managers who don't like uncertainty, but the most
popular codes make some really crude approximations that couldn't be
fixed even if anyone really knew how to do satisfactory turbulence
modeling, and someone whose insight goes deeper than the input
necessary to run a code is necessary to see when those problems are
leading to unrealistic results. Is that science or engineering? I
think it's science. The design decisions you make based on the
predictions and whatever uncertainty is inherent in them is
engineering. I've done both. Am I a scientist or an engineer?

Yes, computer architecture is hard, and there is the inevitable
tension between those who idealize and those who have to make it
work. Fluid mechanics is a teensy branch of nineteenth century
physics, but, at that, the field could never be mastered in its
entirety by anyone I've ever known. And it's just one piece of
"rocket science."

Robert.


First  |  Prev  |  Next  |  Last
Pages: 3 4 5 6 7 8 9 10 11 12 13 14
Prev: Maximum ammount of local variables in cg shaders ?
Next: Online Exams for Certification, Free Practice Exams, Study Material, Dumps