Wikipedia:Featured picture candidates/delist/Translational Motion

Image:Translational_motion.gif

The randomized, or *thermal*, energy of particles such as atoms and molecules gives a substance its temperature. For ideal monatomic gases this means kinetic energy from translations. Here, the size of helium atoms relative to their spacing is shown to scale under 136 atmospheres of pressure. These room-temperature atoms have a certain, average speed (slowed down here two trillion fold). At any given instant however, a particular helium atom may be moving much faster than average while another may be nearly motionless. Five atoms are colored red to facilitate following their motions.

Reason: Image is not accurate. These atoms are represented as moving in a two dimensional square instead of in a three dimensional box. Because we live in a 3-D universe, this image can't be an accurate representation of what helium atoms do at room temperature and 136 atmospheres. In the real world, they could pass in front of and behind each other, and my suspicion is that the image misrepresents many things like Mean free path and Collision frequency because of this. It looks good but it's factually incorrect.
Nominator: Flying Jazz

Delist — Flying Jazz 23:58, 11 August 2007 (UTC)[reply]
Keep Granted, we live in a 3-D universe but there is a dearth of 3-D computer monitors. Accordingly, concepts like the Maxwell–Boltzmann distribution are best conveyed in 2-D. Clearly, atoms don’t live in 0.062-nm-thick windows; this much is just too obvious. The point of the animation is to demonstrate how random elastic collisions result—at any particular instant—with a certain portion of atoms moving quickly while others are moving slowly. To this extent, the animation engine is spot-on. It’s algorithm precisely recreates rebound kinetics of elastic collisions, and—if you tally all the velocities over time and plot them on a histogram—generates a Maxwell-Boltzmann distribution bell curve. Greg L (my talk) 03:29, 12 August 2007 (UTC)[reply]
Keep if only for all the work that must have been put into so precisely simulating the subject. Also I find it fascinating that the atoms are so close to each other in a gas at only 136 atmospheres.. in other words it's a valuable image, and interesting image- not the best featured but still it deserved the original promotion. keep --⁪froth^t 04:23, 12 August 2007 (UTC)[reply]
Keep and I recommend the nominator read up on scientific model. ~ trialsanderrors 06:12, 12 August 2007 (UTC)[reply]

I've taken your recommendation and read that article. As a model of elastic collisions and the Maxwell-Boltzmann distribution, this is a fantastic animation worthy of featured article status. However, the caption describes it as a model of the translational motion of helium atoms at room temperature and 136 atm. To Froth and myself and lots of other people, that supposed correspondence to physical reality is the fascinating part of the animation, but the animation is not a good model of that situation. With a different caption reflecting what is actually being shown, I'd withdraw my delist nomination. Flying Jazz 11:32, 12 August 2007 (UTC)[reply]

Keep. While a three-dimensional model might perhaps be more encyclopedic, there are a number of points against it. Firstly, it's my feeling that it would be far more difficult for the layman to follow, or even understand. The size of such an image would also be a factor - most simulations I've seen have been pretty hefty files. As I recall in the nomination of this image, even just anti-aliasing the particles shown would quadruple filesize. I do agree with the nominator that the model is incomplete, and worse, that it doesn't state its limitations either in the caption or on the description page. However, barring any proof to the contrary, I'll stick with my earlier Support vote, since I feel this is likely to be the best representation of temperature at this scale that we're going to see uploaded to Wikipedia. GeeJo ^(t)⁄_(c) • 11:07, 12 August 2007 (UTC)[reply]

Yeah that's a good point- we certainly need a good picture somewhere showing little balls bouncing around to demonstrate the idea of average motion making up temperature. And this is pretty much as good as that could get --⁪froth^t 17:05, 12 August 2007 (UTC)[reply]

Delist Accuracy first, above everything else. A pseudo-3D image could be made by drawing an outline of a cube, with the particles changing size and apparent motion to illustrate depth. I'm not saying such an image would be easy to make, but the filesize would be comparable and it would be decidedly more accurate. Matt Deres 11:28, 12 August 2007 (UTC)[reply]

If there's ever a proper way of doing this without resorting to a 2D projection, then I'd wholeheartedly agree, but there's no svg animation yet and gif wouldn't work since without the balls being textured (with a gradient perhaps instead of a solid color) there would be no illusion of depth- just balls getting bigger and smaller inside a projection of a cube --⁪froth^t 17:08, 12 August 2007 (UTC)[reply]

It would depend on how detailed and realistic you wanted the animation. A GIF is essentially nothing more than a flipbook (though you can play with the filesize and other optimizations to a limited extent); if you want the particles to fade or darken, shrink or grow, you just need to draw them that way. When dots of differing shades and sizes passed over one another, there would be an illusion of depth. Even with keeping some particles red for demonstration purposes, you'd still be well within the 256 colour limit. I'm not sure if I've properly understood your objection, though. Matt Deres 17:27, 12 August 2007 (UTC)[reply]

Technical background, from Greg L: Computers aren't infinitely fast and the broadband connections don’t have infinite bandwidth. Tradeoffs must necessarily be made in animations, otherwise their file size rapidly gets out of hand and file download times become bothersome. Doing a 3-D animation necessarily requires shading. Four bits per pixels looks like crap so once you head down that path, you really need at least eight. You’ll notice that the edges of the above balls don’t have anti-aliased edges. I purposely used only pure red, blue, black, and white in this animation so all the color content could be described using only two bits of data per pixel. Anti-aliasing the edges would have doubled or quadrupled the size of the file! This also explains why the five tracking balls are all red instead of a mix of colors: bigger file size. File size is especially important for the thermodynamic temperature article because it features three animations plus eleven other graphics. This animation has 371 frames, which is a lot. Doing so provides a nice long viewing period before it loops. In turn, this leads to another trade-off: frame rate. This animation runs near the edge of what is considered to be fluid motion: between 16.4 to 18.2 frames per second. This is the frame rate of Super 8 movie film. The interframe delay is set at 50 ms. All computers wait the required 50 ms while displaying a frame. After that wait, most computers devote between about 5 and 11 ms to actually process the next frame. This totals between 55 to 61 ms per frame (18.2 to 16.4 frames/second).

No one in their right mind could possibly think that a 2-D representation of this phenomenon is a perfectly accurate representation of what really occurs in 3-D, nor does any caption so suggest. And who would want to watch a 3-D animation for a period of time necessary to witness very many collisions? The real issue surrounding this vote shouldn’t be the technical limitations that pertain to all GIF-based animations; it should be whether constraining the animation to 2-D is a valid way to demonstrate the rebound kinetics of elastic collisions and the Maxwell-Boltzmann distribution bell curve. If you consider the mathematics of the issue, one can perfectly model the 3-D physics of rebound kinetics in 2-D (like steel balls in a pinball game). The bottom-line issue should be this: does the animation effectively demonstrate how, in perfectly random elastic collisions “a particular helium atom may be moving much faster than average while another may be nearly motionless.” Was there a serious flaw in this underlying premiss that warrants delisting? I don’t think there is any flaw in the premiss. Greg L (my talk) 19:40, 12 August 2007 (UTC)[reply]

P.S.: The speed of the “helium atoms” is quite accurately represented as two-trillion times slower than at room temperature. The “disks with the radius of atoms” (they also happen to have color, which atoms don’t really have) move an average of 7.16 pixels per frame. Given that the atoms are 11 pixels in diameter and have an actual diameter of 62 pm, this is 40.3 pm of movement per frame. Notice that the speed is independent of size as displayed on any particular computer monitor; displayed diameter and displayed movement per frame scale proportionally. After going through all the frame-rate issues, this works out to beween 1.852 trillion and 2.055 trillion times slower than at room temperature. I think this is also part of what makes the animation interesting to me and others: seeing how quickly atoms move and knowing they really move two trillion times faster. Greg L (my talk) 21:44, 14 August 2007 (UTC)[reply]

OK. Now with the dimensions of the interior of the square, you can calculate the surface pressure shown by the animation and report that in the caption instead of the incorrect 136 atm value. Flying Jazz 13:09, 17 August 2007 (UTC) [reply]

When the caption says "atoms" instead of "disks with the radius of atoms" and says "atmospheres" (a 3-D pressure) instead of "Newtons per meter" (or other 2-D units of pressure) then many people in their right mind will think this is an animation that depicts atoms at a certain number of atmospheres. The physics of elastic collisions can also be modeled in 1-D, and you can get Maxwell-Boltzmann distribution curves that way too. But line segments bouncing into each other on a line are just as much not-atoms as disks bouncing into each other in a plane. Either the animation should be altered to match the caption or, if that is too technically difficult, the caption should be altered to match the current animation. Great animations (with proper units of pressure) of 1-D, 2-D, and really pretty 3-D situations can be made with the Java software at [1]. Flying Jazz 21:58, 12 August 2007 (UTC)[reply]

Keep. Obviously the caption can easily be rewritten for clarity, but that's no reason to delist the picture. The whole complaint about 2D versus 3D is completely bogus. This is a teaching tool, not an accurate representation of reality. — BRIAN0918 • 2007-08-17 19:54Z

Kept MER-C 04:23, 18 August 2007 (UTC)[reply]