Experimental Physics & Engineering

Question

I know that most of us are "paper theorists," but I think we need to remember that experimental physics is physics too! We have several tags for experimental physics (with tag excerpts):

• experimental-technology

  Use this tag for questions pertaining to the limits, management, and operation of equipment necessary to experimental physics. This tag is not intended for "does this thing I heard about actually work" type questions.

• experimental-physics

  for questions about design, process, data, or analysis of experiments and observations.

• experimental-technique

  filling in the unspecified details of elegant-sounding descriptions of experimental methods.

I think some of us are a little quick to close questions that seem to be engineering (e.g., How can I make a dry dilution refrigerator quiet? that was closed and even downvoted--and also the cause for this Meta post). This merely begs the question:

What objectively separates engineering questions from experimental physics ones?

---

see also

• Experimental technology questions: on topic?
• Does the Area 51 proposal "Experimental or Applied Physics" duplicate this site?
• Should "How do I solve this experimental problem?" questions be on-topic
• What is engineering and what is experimental design? (of which this might be a possible duplicate?)

Answer 1

What objectively separates engineering questions from experimental physics ones?

Nothing.

I will argue that trying to formulate such a distinction is not the correct way to attack the problem of deciding what questions should and shouldn't be considered on-topic for this site. A better approach is to ask whether or not physics or engineering flavor is more appropriate in the answer.

It's a question of answers

Suppose I want to build a waveguide resonator with low dielectric loss. I could post the following question to either Physics SE or Electrical Engineering SE:

How can I design the geometry of a waveguide resonator so that it has the minimal dielectric loss?

If I pose this question to an electrical engineer, I will likely get design rules, possibly in the form of immediately useful formulae. If I pose it to a physicist, I may learn something about the physics of electromagnetic fields near dielectric boundaries, diverging fields at conductor corners, and maybe even something about conformal mapping. Heck, if I pose it to the right person I might learn about numerical recipes to solve electrostatics problems.

This example illustrates that the appropriate place for a question can depend on what kind of answer the person posing question wants. It would be wise to keep this in mind when judging the appropriateness of questions on this site. Before knee-jerk voting to close a question, we must ask ourselves if the asker may have had a reason to post it here instead of somewhere else.

Who is reading your question?

In our capacity as experimental physicists, my colleagues and I spend >99% of our time building equipment, fixing equipment, and programming a computer. These are all activities which could be classified as "engineering", yet we identify as physicists. I frequent Physics SE and not Electrical Engineering SE. This is not without reason: I am far more qualified to answer questions here than I am on Electrical Engineering and, even when I'm asking an engineeringy question, I would rather get the benefit of people who work in similar environments and with similar equipment as myself.

The point here is that another important aspect to deciding whether a question should be categorized as "engineering" is the readership of the site. If I ask a question about a machine, the Physics Lab Quizzwopper 2000X, nobody on Electrical Engineering SE will ever have heard of it. So, while a theorist might find a question about equipment off-topic, an experimentalist will not, because the experimentalist wants the answer which only a physicists can give.

Audiences aren't well delineated

Consider the following question:

How can I use the symmetries of physical system XYZ in order to vectorize my simulation program in computer memory for optimal performance?

There is no way to objectively categorize this question as "physics" or "engineering". All we can say is that the person most like to give a useful answer to this question is someone who has solved a similar problem in the past. This could be a "physicist" who had to learn some programming or an "engineer" who had to solve a physics problem. In this case, it's probably reasonable to let the question sit wherever the asker put it, and if it doesn't get any attention, that person can move it somewhere else and try again.

P.S. I have avoided proposing an objective criterion by which to decide whether a given question is appropriate for this site, as that issue is not strictly within the scope of the original question.

Answer 2

I try to think of "what kind of person is most likely to have the answer to this question"? If the answer is an experimental physicist, or a physicist, then I am very loathe to close any question, no matter how "engineering" it is.

There was a question recently where some poor user asked why she was getting essentially a link error in building a piece of software to do scattering amplitude calculations. The software was written by CERN and the question was closed as being "software engineering". I really thought we'd done the user a disservice: she linked the home page for the software (so she'd more than likely sought an answer from the right sources), and, from what she was doing, clearly had the intelligence to know well in advance the answer that she would have gotten from Stack Exchange. "You're missing a library", or someone hasn't defined a symbol in extant libraries (highly unlikely in released software!) are the only two possible correct answers. Well, DUH, I think such an answer would simply insult the user's intelligence - if she can wield Mellin-Barnes integrals she would surely know this. I notice she hasn't been active since and I don't blame her. Someone of her obvious knowledge was likely to be an asset to Physics SE, so we lost out in closing this question. What of course she was looking for was someone with experience with that particular piece of software, or someone who could point her towards other resources. That person, unquestionably, had to be a physicist.

I hasten to add that, for me, the engineering-out screening is important: I come here to think about fundamental problems and not be bothered by engineering problems. The latter are pretty much my day job. I come here to get away from them. I'm not disparaging them, but I'm not lucky enough (at this time) to be working in physics, so when I'm not at work I want to be either spending time with my children, or working on the mathematics and physics that I love.

I also appreciate John Rennie's reason for closing: we want the site to have use to not simply one person. But it's conceivable that other people want to use this particular software. Also, one would have to wonder how many times such a question would be asked. Allowing the odd question like this might help the physics community a little more.

Answer 3

I am an experimental physicist (not an engineer) and a Physics teacher. More often than not, I am denigrated for this in real life, as one theoretical physicist stated:

Experimental, and especially physics teachers are the lowest on the research totem pole

This is rubbish, utter rubbish, insulting and says more about the one said it than his target.

Experimental Physics = Theoretical Physics = Physics

A bit of background, I have a PhD in Physics (not Experimental Physics etc), and am published internationally. Due to circumstances currently beyond my control, I teach Physics to pay the bills, and my experiments use equipment that is often self made or borrowed, and are performed at a local lookout, or in the school science lab or on my kitchen bench.

Nothing as grand as CERN, my research is based on what I can get from the hardware and electronics store and put together and calibrate with the borrowed 'standard' equipment. It is often referred to as "Backyard Physics" - but is no less valid.

To put it succinctly:

To effectively make useful homemade tools for use in experimental 'backyard' physics, physics principles must be understood and applied as the major underlying principles for the functionality (and as I have found in publishing work, the validity) of the tool. I fear that questions that are of this nature will just be closed as 'engineering' akin to 'toolmaking' - however, a question about the application of physics principles for an aspect of the tool's use is not engineering, but is applied physics.

For example, in my question Microtops sunphotometer sun target optics, I am clearly asking about the optics principles involved (applied physics) - not asking about how to build a sun target (engineering).