How are theory, experiment, and simulation related continued: a conversation

Hey, after the last post, I had a good conversation with a fellow graduate student/scientist from electrical engineering. With his permission, I wanted to post this discussion, since it illustrates nicely how theory and simulation interact within different fields (no pun intended this time, although he does work in applied electromagnetism/antenna design).

I enjoyed your blog post, and I'll post some thoughts on it later when I have more time!
I noticed that the title talks of simulations, but when you break research down into catagories, you instead use engineering.
(Simulations were lumped into theory.)
Granted, the theory/simulation/experiment breakdown isn't perfect or without overlap, but I personally think it's more accurate than putting engineering as a catagory.
"Engineering" to me is a process or tool, not a type of research

EE

Yeah. As I was going about it, I found it difficult to break down theory and simulations

Elizabeth Pogue

ahh, ok
Simulations use theory, that's true

EE

I think I might leave it as-is for the blog title. There are journals that just focus on instrumentation and engineering technologies and it is really distinct from the other two.

Elizabeth Pogue

Fair enough.
It hit me, I think why simulation and theory are different in my mind is that theory tends to build up more models and more complicated models
whereas simulations sometimes are built on repeated (and thus numerous and tedious) applications of first principles
This is a very broad generalization

EE

I think I know why we see different degrees of space between theory and simulation. For materials, you are already fairly close to first principles. You have theory that describes first principles but, that theory is often too tedious to totally do, so you end up doing simulations.

Elizabeth Pogue

The same is true for E&M: simulations usually are based on Maxwell's equations and derivations thereof

EE

True.

Elizabeth Pogue

Theory, such as transmission line theory, builds on that behavior to create larger, simpler models
There are usually assumptions or approximations built into theory, less so for first principles
It depends on the number and degree of approximations, of course
I think the difference I'm seeing could be between our fields
In my mind, theory is something that you can do on paper (possibly with a calculator) and get useful, practical results out
Whereas the theory you're talking about describes what happens, but may not always be directly and easily useful for predicting results or doing design
heck, even that's not always true
dunno
lol

EE

It is possible. If you look at something like density functional theory, you are making approximations about electron density inherent in the theory. It is based on quantum theory, but is generally still referred to as theory, not a simulation.

Elizabeth Pogue

Agreed
Simulation (or calculation) to me is number-crunching on more accurate models
Stuff you wouldn't do by hand

EE

That is why I was having difficulty clearly distinguishing them, other than saying, within theory, that simulations implement theory.
I agree with you on that.

Elizabeth Pogue

Yeah, that's a good point
simulation/calculation is a subset of theory
but given the number of researchers devoted to implementing such programs, it's really a realm of research distinct from theory
It builds off theory, but uses computer science (algorithms, data structures, etc) to practically model/simulate/calculate with said theory

EE

Yes. I changed it to say that generally simulations can't be done by hand, where theory can often be implemented by hand.

Elizabeth Pogue

Sure
And I'm sure physicists would nit-pick that definition too
Depends on your field, I guess!

EE