Units (geekery)

[lederhosen]

My co-worker and I are physicist/mathematicians in a workplace made up mostly of biologists and other squishy scientists. As such, we end up helping people with maths & physics problems a lot, and in the course of this we keep discovering more and more painful examples of how little some alleged scientists understand the disciplines they're working with, or even the basic common language of those disciplines.

One of the people we help a lot is B. She's fairly bright, but she has trouble grasping the difference between 'force' and 'pressure', and we have a lot of conversations with her where we explain this and she goes away and then comes back confused. A large part of this seems to be that in the interim, she's been reading papers that abuse these concepts so badly that they drain knowledge like some sort of steroid-bloated head mosquito.

This was brought home to us very painfully recently when B. came in trying to compare results which she'd obtained by hard work and diligent supervised-by-us calculation with somebody else's published data that came with a calibration table.

The data in question was apparently intended to measure pressure. The units they were using were identified merely as 'g/s'. We spent hours on this trying to figure out what this could possibly mean; taken literally, that's "grams per second", which ordinarily would refer to some sort of bulk flow rate. Since nobody in vision sciences except the two of us and our immediate boss seems to understand the distinction between mass and force, we thought it quite likely that the 'g' might really mean "grams-force", i.e. one gram multiplied by one Earth gravity. But even with that substitution, it made no sense whatsoever.

Eventually, B. wrote to the guy who produced this data, and asked him what it meant.



We were right about the 'g'. As for the 's', it stood for 'square centimeter'.

*twitch*

*twitchtwitch*

ARGH!

Comments

[shadow-5tails.livejournal.com]

*head->desk*

I think the bit that I can't get over is that someone published this travesty? Don't things normally get reviewed before publication?

[lederhosen.livejournal.com]

I think it was published as documentation for equipment or some such, so not peer-reviewed. I'd hate to think something like that survived review; it certainly didn't survive mine ;-)

[shadow-5tails.livejournal.com]

Even then, surely, you get someone to proofread things? Someone who understands what the hell you're going on about?

Gah.

[lederhosen.livejournal.com]

One would think, wouldn't one?

[shadow-5tails.livejournal.com]

Well, one does think "Gah." That's why one said it.

[quatranoctal.livejournal.com]

Well it looks like someone didn't think. Why do so many otherwise intelligent people make the foolish assumption that the person they're interacting with is actually an intelligent, sensible human being, and so are constantly disappointed?

[wingedkami.livejournal.com]

I guessed square metre. Ouch.

Also I seem to have forgotten the difference between stress and strain, and what their units are, and am now googling Hook's Law in an attempt to remind myself.

[lederhosen.livejournal.com]

That's one of the things that pisses me off - even if you figure out that the 's' might stand for 'squared unit of length' (which is what it *has* to be, if the 'g' is a gram-force), the abbreviation still leaves no way to tell *what* unit of length. The centimeter-gram-second system is the bane of my existence, I hate it worse than imperial units.

Stress and strain are closely related quantities. Strain is unitless - it refers to the stress-induced deformation of an object proportional to the size of an object. (If you take a 1m cube and squish it so it bulges out to 1.01m, the strain is 0.01.)

Stress is the associated internal force per unit area, measured in pascals (same as pressure). If I take a wire of cross-sectional area 1E-6 m², and use that to support a mass of 1 kg (i.e. a force of about 10 newtons), the associated strain is 10/1E-6 = 1E7 pascals. (Assuming it's distributed evenly through the cross-section of the wire, which it probably isn't.)

Note that both stress and strain are direction-dependent quantities.

[wingedkami.livejournal.com]

Ahh, physics joy. :-) I managed to remember what stress is, but couldn't get further than 'something to do with elongation' for strain. Now I must resist the temptation to go and experiment on my housemate's slinky.

I first encountered the centimeter-gram-second system in my first stellar physics lecture. My lecturer made some comment along the lines of, "When faced with some of the largest distances ever measured, scientists have for some reason decided to measure them in centimetres." I don't think she liked it very much either. We stuck with SI for the rest of the course.

[tenner.livejournal.com]

Wow. Just... wow.

[silmaril.livejournal.com]

I had to stare at it for a while, then thought "grams-force per square"? Then I thought "nah, can't be..." And then I thought, I must be right, or lederhosen wouldn't be so itchy.

And of course I was right.

The pain.

I have an excuse to be right, though: In semiconductor device design there is a convenience unit of "ohms per square". That comes from the fact that if the material thickness is the same, a piece of the same material 50 micron by 50 micron will have (1) the same DC resistance as the same material 1 micron by 1 micron, so you can just define some sort of resistance-per-square and use it to calculate the resistances of the structures you lay out on silicon. It's a convenience thing.

But your folks---square what!?


[1] Almost, void where prohibited, some exceptions apply.