Wikipedia:Reference desk/Archives/Science/2015 May 15

= May 15 =

What is a train's speed control "user interface" like?
What is the speed control "user interface device" of a train like? Is it more like the accelerator pedal of a car, which needs to be continuously operated to keep the vehicle going? Or is it more like the cruise control of a car, which once set will keep the vehicle going at a preset speed until further intervention? Or is it something entirely different? --98.114.98.58 (talk) 04:18, 15 May 2015 (UTC)


 * The controls are vaguely like those of an airplane (e.g., a hand-operated throttle). Here's a concise explanation with pictures: Short Brigade Harvester Boris (talk) 04:38, 15 May 2015 (UTC)


 * There isn't really much resemblance to airplane controls, although the throttle is indeed hand-operated on both vehicles.


 * The throttle on a locomotive is like the accelerator pedal in that it controls the engine power rather than directly setting the speed, but like the cruise control in that it stays at the same setting until moved. To ensure that the train doesn't run away if the driver falls unconscious, there is also typically either (1) some sort of deadman switch that must be continuously depressed, or (2) a vigilance device that requires the driver to take some positive action every 90 seconds or so or else the brakes will come on and the power shuts off.  The problem with 1 is that the driver may still depress it while unconscious.  The problem with 2 is that it may react 90 seconds too late.


 * The other primary control on a train is the brake, which is typically a separate control that also stays at the same setting until changed.


 * I'm using words like "typically" because there is variation not only between types of trains (diesel, electric, subway, etc.) but also between countries. For example, on this subway train there was a single control lever that was moved to one side to apply power and the other side to brake, while being pressed down to release the deadman switch. --174.88.135.200 (talk) 07:05, 15 May 2015 (UTC)

Somehow I had the notion that trains accelerated slowly, but according to this report from CBS, the train accelerated from 70 mph at T-65 seconds to 100 mph at T-16 seconds. That's a 30 mph acceleration in 49 seconds! Wnt (talk) 20:01, 15 May 2015 (UTC)


 * Slow acceleration means that more time is required each time that the train has to stop, and it means a train still accelerating to full speed may delay a following train that's already at full speed and closes in on it. Both of these are a big deal in a railway that's operating fast passenger trains. Freight trains are another matter. The wrecked train had an Amtrak Cities Sprinter electric locomotive, which weighs 97,766 kg and has a maximum power of 6,400 kilowatts (8,600 horsepower). It was a Northeast Regional train, so its 7 cars were probably Amfleet cars weighing about 53,000 kg each, for a total of about 470 metric tons.  Accelerating that mass from 70 mph (31.3 m/s) to 100 mph (44.7 m/s) in 49 seconds would require (470 &times; (44.7&sup2;&minus;31.3&sup2;)/2)/49 = about 4,900 kilowatts: well within the power of the locomotive. --174.88.135.200 (talk) 02:26, 16 May 2015 (UTC)


 * A further factor that might apply: the UK's Southern Railway had/has a lot of commuter lines serving London and therefore a lot of stations its commuter trains had to stop at. In order to minimise stopping delays and also energy lost in braking and accelerating, it built or re-built many of its stations on artificial humps: this meant that as each stopping train approached the station, the upslope helped to slow it down, and as it left the downslope helped to speed it up. (Through-only lines could be left "un-humped. The SR was also an early adopter of electric locomotion, which is more efficient than steam or diesel, and capable of greater acceleration.)
 * Since this principle was/is well known in railway engineering circles, it may well have also been adopted on the line involved in this recent tragedy. {The poster formerly known as 87.81.230.195} 212.95.237.92 (talk) 12:56, 18 May 2015 (UTC)


 * I don't have definitive data, but it doesn't make sense in this case. The accident was a few miles from 30th Street station, and they wouldn't have a continuous slope over that distance.  Also, the station is on low-lying land and the tracks there aren't elevated above ground level. --174.88.135.200 (talk) 05:29, 19 May 2015 (UTC)

FLUG
Yep, a FLUG as mentioned in Hainan Eastern Ring High-Speed Railway. What is it? Thanks. :) Anna Frodesiak (talk) 06:20, 15 May 2015 (UTC)


 * That was added in this edit by . Denis was annotating the entry for Sanya Railway Station as an airport stop ("flug" is German for "flight"; I think the "BS-" route templates were imported from the de:wikipedia). Perhaps Denis simply goofed and put FLUG into the infobox by mistake. -- Finlay McWalterᚠTalk 06:52, 15 May 2015 (UTC)


 * Thanks Finlay! Yes, you are most probably right (I do not remember exactly where I took the "flug" keyword from, but it is indeed possible that I got it from the German Wikipedia), as I wanted to specify that there was an airport in Sanya, namely Sanya Phoenix International Airport. Sorry for the goofing. -- Denis.arnaud (talk) 09:03, 15 May 2015 (UTC)


 * In the US, the word "flug" for flight might be recognized from the Red Bull Flugtag. StuRat (talk) 14:28, 15 May 2015 (UTC)

Many thanks, folks. :) Anna Frodesiak (talk) 22:36, 16 May 2015 (UTC)

Smallest number of cages in KenKen puzzle?
You can see a KenKen puzzle. It looks a bit like Sudoku, but it’s very different, because in KenKen ordinary arithmetic is the key, whereas Sudoku is pure logic. I am wondering what is the smallest number of cages that could be used to provide a single solution for a KenKen puzzle. I am not a mathematician so I am raising the question here rather than in the Maths desk. I usually play the larger 12X12 Calcudoku puzzles which are larger versions of KenKen, and also wonder how increasing the number of cells affects the minimum for cages. Myles325a (talk) 07:22, 15 May 2015 (UTC)


 * As I understand this question, the answer is obviously one. (Or am I missing something?)--Shantavira|feed me 07:50, 15 May 2015 (UTC)
 * Except in the trivial case of a 1x1 square, having only one cage would let you arrange the numbers within in multiple possible ways (e.g. if you had a 2x2 square with target 6+, it could be 1,2;2,1 or 2,1;1,2 (semicolons denoting a new line)). For a 2x2 grid, the answer is 2 (one cage of 3 squares, and one of 1 square), since there are two possible arrangements which differ by relfection, and the single cage or an arrangment of two 2x1 cages would not break the symmetry. Our article KenKen describes how the problems are set up, but importantly the grid is always an nxn square, filled with n copies of the numbers 1 to n.  The grid is divided into continuous regions (cages) which are set in such a way as to define the numbers in the cage, but not their order (exactly how this is done is probably not relevant for this problem).  No number can appear twice in the same row or column. MChesterMC (talk) 08:28, 15 May 2015 (UTC)


 * I don't see how you would break the reflection diagonally with less than three cages. The list of minimum numbers of cages depending on size nxn would certainly be a rather idiosyncratic discrete function; I imagine the mark of the true mathematician would be to make it continuous. :) Wnt (talk) 19:23, 15 May 2015 (UTC)
 * There is a reflection diagonally for the 2x2 case, but it doesn't end up in a distinct result - the only possible results are 1,2;2,1, and 2,1;1,2, both of which are the same when reflected diagonally. For n=3, you can't break more than two of diagonal symmetry, left/right reflection, and up/down reflection without a third cage (which is easy to show via brute force).  At n=4, you can break all those symmetries with just two cages, but the example I can find results in the top and bottom row being interchangable - numbering the squares 1 to 16, an L block covering 5,6,7 and 9 breaks the symmetries, but the row 1-4, and the row 11-16 are indistinguishable. MChesterMC (talk) 09:08, 18 May 2015 (UTC)
 * Wait, no there is an example for n=3 using two cages: using the numpad, one cage covers 1,2 and 4.  Using a, b and c to represent the numbers, set the problem so the smaller cage contains 1a and 2b's, the larger cage contains 3c's, 1b, and 2a's.  The c's must be on the diagonal line 7,5,3, the b's in the larger cage must be in 8,6, and the position of the remaining b is then completely defined.  It doesn't matter that the cage arrangement has diagonal symmetry, since the only possible solution also has diagonal symmetry, which makes it unique.  This solution can be generalised: for any nxn grid, split it into a triangle of side n, and a triangle of side n-1, and restrict so that the triangle of side n contains 1x1, 2x2, 3x3...nxn (where axb denotes a copies of digit b).  The only possible solution to fit n copies of a digit in the larger triangle is along the diagonal, at which point the only solution to fit n-1 copies of a digit in the remaining traingle is along the next diagonal, etc.
 * So to answer the OPs original question, for n=1, the minimum number of cages is 1, for every other n, the minimum number of cages is 2. Of course, that assumes it's possible to specify the operation in each cage sufficiently that there's only one set of numbers which could be in the cage - I'll leave that for someone with more mathematical knowledge than me! MChesterMC (talk) 09:24, 18 May 2015 (UTC)
 * I misunderstood the idea before, so I'll withdraw my comment. Wnt (talk) 20:11, 19 May 2015 (UTC)

What chemical properties make certain materials good insulators or good conductors of heat (or energy in general)?
On the molecular level, what are the electronic interactions or bonding arrangements or other chemical traits that make some elements/compounds/molecular structures better conductors (readily transmit energy) or better insulators (reluctant to transmit energy) than others?

I understand this depends on a huge number of factors, that a compound may transmit electricity very efficiently but another type of energy very poorly, and that certain molecules can lose conductive or insulative properties as their chemical environment changes, but it seems as if there must be a few (if not just one) overarching principle in chemistry that generally permits or most heavily influences the capacity or incapacity to transmit energy to other materials. Any ideas? Thanks! PJsg1011 (talk) 19:11, 15 May 2015 (UTC)


 * In metals, thermal conductivity is due to a conduction band of delocalized electrons that are free to wander throughout the metal, accepting energy from collisions here and releasing it there. See Wiedemann–Franz law.  According to thermal conductivity other materials depend more on phonons for energy transfer, i.e. vibrations moving through the material. Wnt (talk) 19:21, 15 May 2015 (UTC)
 * Other good articles to read on electrical conductivity would be metallic bonding and Fermi level and electronic band structure. -- Jayron 32 20:10, 15 May 2015 (UTC)


 * And note that it's not just chemical traits that have an effect. Physical arrangement can make a big difference, too.  For example, adding air bubbles to decrease the density of a material will lower it's thermal conductivity, even though the amount of material remains the same. StuRat (talk) 20:44, 15 May 2015 (UTC)