Wikipedia:Reference desk/Archives/Science/2017 July 30

Science desk
< July 29	<< Jun | July | Aug >>	July 31 >

Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

July 30

Steam train

What wheel arrangement is best for a steam train whose main task is bringing minerals from mines in the mountains to the mainline in the valley (the line over which it operates is NOT a narrow-gauge or rack railroad, but does have at least some tight curves and steep grades)? The parenthetical phrase was added after the first response. Would that be the same 4-8-4 arrangement which is best for heavy mainline trains, or would some other arrangement work better? 2601:646:8E01:7E0B:298A:91F:6885:FC (talk) 00:34, 30 July 2017 (UTC)

You really have to mention a lot more factors. For example, if you have lots of tight curves, an articulated locomotive with two sets of drivers (e.g. a Mallet locomotive, with wheel arrangements like 2-6-6-2) has an advantage over locomotives like a 2-10-4, while if you just have a lot of straight stretches, that specific advantage disappears. Also bear in mind that many such railroads were narrow-gauge railways; such routes typically were built on a lighter strength than full-size railroads (they were narrow because of size restrictions that also would prevent the use of heavier trains), so wheel arrangements with lots of wheels might be a very bad idea simply because such locomotives would be a good deal heavier. Nyttend (talk) 01:20, 30 July 2017 (UTC)

Clarification: the line is NOT a narrow-gauge or rack railroad, but does have tight curves and steep grades; also, the trains are very heavy (as is often the case with mineral traffic). 2601:646:8E01:7E0B:298A:91F:6885:FC (talk) 01:51, 30 July 2017 (UTC)

Thank you for the clarification. Side note, if you're in a really tight environment and can't build heavy lines, I suppose you just have to run lots of smaller, lighter trains to extract the heavy minerals. You need to read up on the history of the Chesapeake and Ohio Railway in West Virginia and southeastern Kentucky, as they had to handle precisely the kind of situation you describe. (Here is a short discussion of the subject, and this article would probably be really useful to a magazine subscriber.) For a quick glance, go to 37°52′48″N 81°59′12″W and look around: this is their huge Peach Creek coal marshalling yard near Logan, West Virginia, out of which they sent coal trains of 160 loaded cars out of the mountains, and a quick look around the area will show you the topography they had to handle. In most of this region, mines were served by Mallets (primarily 2-6-6-2), with 2-8-8-2s for full-length trains before the development of the 2-6-6-6. The railroad's smaller locomotives and larger non-articulated locomotives were generally used in less mountainous regions (for example, their 2-10-4s were used for faster freight), although late in the steam era, they employed 2-8-4s in the mountains. Nyttend (talk) 02:23, 30 July 2017 (UTC)

PS, this book, about locomotives of the C&O 1900-1965, would be particularly helpful to you (my parents have a copy, which I frequently read when growing up), although as your WHOIS suggests that you're in California, it's likely to be a good deal harder to obtain than if you were here in Virginia. Contact your local library about interlibrary loan, if it looks interesting. Nyttend (talk) 02:27, 30 July 2017 (UTC)

Assuming 1930s technology, probably some big Garratt design, like a (2′D1′)(1′D2′). Garratts were commonly used for such tasks in many parts of the world, although for some reason they were never very popular in Europe and North America. More recently, one wouldn't consider using a steam train. Around 1890 one would expect a 2′D1′ or something similar.

Note that in your case the loaded trains are going downhill and the empty ones uphill, so you may not need a very powerful locomotive. Electric trains using regenerative braking could turn your railway into a power generator instead of power consumer. After 1920 this may be the best option.

For your mainline trains, you have less tight curves to deal with, less severe gradients and you may want more speed to mix with express passenger traffic. The same Garratt could do the job, but you may consider a non-articulated design instead, like a 2′E1′. Also note that steam locomotive design was often more an art than a science. PiusImpavidus (talk) 08:59, 30 July 2017 (UTC)

• "It depends". The best wheel arrangement is an 0-4-0, because it's simplest and it puts all the loco's weight onto the drivers. Everything else beyond this is a complication.

You probably need more power, so that means more weight. That needs more wheels to keep individual axle load from overwhelming the track, so it becomes an 0-6-0 or an 0-8-0. This is a mineral railway, so it may involve hill climbs (even going "downwards" it might need to cross a ridge), the track could be lightly laid, cheaply maintained and in poor condition. Tank locomotives might be favoured because they keep useful weight over the wheels, giving better adhesion.

The loading gauge could become an issue in the UK, so fitting a large enough boiler in place could be an issue, especially with a tank locomotive. UK minerals were hauled by the Hunslet Austerity 0-6-0ST which was a small, powerful loco, but the size of the boiler limited its speed and sustained output. Heavy loads could be drawn by working multiple of these together.

If the rear of the loco becomes heavier, owing to a larger firebox or more bunker capacity (thus greater range), there may be a need for a carrying axle behind the firebox. This gives an 0-6-2T arrangement, such as the South Wales types, which were another stalwart of UK coal working.

There will be two cylinders, as the speed and budget is against any more. Early British practice was for inside cylinders, but eventually they are too large and move outside. This gives a 2-8-0 layout, such as the GWR 2800 and GWR 4200 Classes. These were a tender and tank locomotive version of originally the same engines, the tender version with greater capacity and range, the later tank version having detail changes and a slightly shorter boiler to allow tighter curves.

As bunker capacity again increases, there may be a need for a 2-8-2T class, as was found with the GWR 7200 Class - converted from some of the earlier 2-8-0T.

British distances were short, so there was little need for the large tender locos with high fuel and water capacity used elsewhere, or for the huge boilers to supply steam across such a prolonged climb as the Rockies or Andes. So British locos never grew much beyond the 2-10-0 arrangement of the last large mineral locos and the BR Standard Class 9F. This was such a sophisticated design that it was also one of the fastest express passenger locos of BR, owing to the sophistication of its suspension and wheels. Still a single axle front pony truck was enough to carry the front cylinder weight and to guide it at speed, without requiring a 4-wheel bogie. The rear, owing to the small wheels and the shallow firebox, could be carried over the driving wheels and so not needing a rear bogie or truck at all.

In general, mineral locos were slow and so 4 wheel bogies were avoided in favour of 2 wheel trucks. This kept more weight on the drivers, where it was usual for adhesion. There were of course exceptions.

Articulated locos were primarily needed when either a narrow gauge limited the width of a locomotive and required it to be extended lengthways, or when the boiler or grate capacity needed to be very large. The Garratt design in particular places the boiler in isolation, with neither wheels, cylinders or water tanks around it, allowing it to expand as much as possible. This was often used when a wide grate was needed to fire with poor quality fuels. The Mallet, as favoured in the US, had the advantage of compound working, but it is harder to fit a large boiler in there too. Only the large US loading gauge really allowed this for very large locos - most other countries went for the Garratt. For narrow gauge, where the loading gauge can be far larger than the track gauge. the Meyer or Kitson-Meyer layout has advantages and is better articulated than the Mallet, so was popular in South America.

There is far more to this question than could be answered here. Andy Dingley (talk) 12:21, 30 July 2017 (UTC)

Thanks! So, a Mallet or a big 2-8-0 or 2-10-0, right? 2601:646:8E01:7E0B:FCEE:E7DF:4CAD:C2E5 (talk) 04:24, 2 August 2017 (UTC)

A Mallet or a big 2′D, 2′E or something similar, or a Garratt, it's difficult to say. In any case a design with many small driving wheels (for high force, low speed, whilst staying withing axle load limits), but not a very long rigid wheel base (for sharp curves). That's why an articulated locomotive would be likely.

It depends on the curves, gradient, lauding gauge, permitted axle load, train weight, fuel quality, whether it's a new line or you want new locomotives on an existing line, the length of the line (how much water and coal do you need for one trip?), what your drivers and maintenance workers are familiar with, whether you want a few standard classes or a design optimised for this particular line and whatever I forgot. Mallets are great for compound locomotives, which are more fuel efficient, but difficult to drive and relatively heavy. Simple designs are cheaper on maintenance, but may be less fuel efficient and may require straighter and stronger track. Some operators used standard locomotives and even more standard components by the 1890s, others tried to optimise for every line even in the 1930s.

Furthermore, you didn't specify a time period. In 1840 the 1A1 arrangement was everywhere. Mallet locomotives were introduced in the 1880s, Garratts in 1909. By 1930, the end of steam was looming, although that took quite a while. No mineral railway in Switzerland would use a Garratt, because by the time those were available, they wouldn't use steam at all. But in Zimbabwe they were extensively used for minerals. PiusImpavidus (talk) 12:00, 3 August 2017 (UTC)

Identifying a butterfly

Hi all, I photographed this butterfly in the Butterfly House in the Melbourne Zoo on 16 February 2015. Can anyone help with identifying it? The rest of the butterflies there were native to Australia, so I would suppose this one is too. Thanks, --SuperJew (talk) 13:28, 30 July 2017 (UTC)

I see you've changed the name of the pic to Common Eggfly (Hypolimnas bolina). Does this mean this Q is now resolved ? StuRat (talk) 21:08, 30 July 2017 (UTC)

I notice that the photo is a very good match for those at the second (of the three) external links in the Hypolimnas bolina article. {The poster formerly known as 87.81.230.195} 90.204.182.36 (talk) 04:45, 31 July 2017 (UTC)

Because of mimicry, identifying butterflies by wing pattern can be chancy. It would be better to go by taxonomic characters like genitalia to be sure, since there is a risk that otherwise Wikipedia starts collecting images that aren't right until everyone believes them. That said, the only thing I found on PubMed, which is not doing well here, about Hypolimnas mimicry was from the 1800s [1] and still locked up by Science ... the notion of troubling Sci-Hub for a no-pictures article using probably obsolete taxonomic terms from ancient history isn't getting me excited today. But a Google search directs me to our own article that says it mimics species of Euploea in western Australia; another paper here. Wnt (talk) 12:23, 31 July 2017 (UTC)

Butterfly genitalia ? Will it hold still while I whip out my microscope ? :-) StuRat (talk) 21:08, 31 July 2017 (UTC)

Outdoor thermometer time delay

In case of an outdoor thermometer for air temperature how much time elapses until measurement after a particular temperature actually sets? Is there a significant delay and if yes, are mercury-based thermometers inferior or superior to other types? Brandmeister^talk 14:56, 30 July 2017 (UTC)

The delay will depend mainly on the design of the enclosure, the windspeed, and the rate of change of temperature, but it should normally be measured in seconds not minutes. The Volumetric heat capacity of mercury is not significantly different from that of ethanol, but perhaps non-fluid thermometers might respond slightly more quickly? Dbfirs 15:10, 30 July 2017 (UTC)

What would you define as a "significant delay"? A second? A minute? An hour? ==Phil Holmes (talk) 16:02, 30 July 2017 (UTC)

Around a minute or more, also due to occasional rapid weather changes. Brandmeister^talk 16:08, 30 July 2017 (UTC)

In case you mean the type that has an outdoor sensor and display, and then broadcasts the results to an inside receiver for display there, this type may have more of a delay for the inside unit, as to save on the battery they may only send this info periodically. (A smart system will send the data less often, if there's no change.) StuRat (talk) 16:16, 30 July 2017 (UTC)

The idea of the delay is hysteresis, and googling for [thermometer hysteresis] gives lots of information about standards and different sensor designs. DMacks (talk) 22:26, 30 July 2017 (UTC)

Influence and borrowed content

How much from the scientific content of Luca Pacioli's work Summa de arithmetica, geometria, proportioni... is taken from his predecessors going back to Fibonaci and even earlier?(Thanks!)--82.137.13.229 (talk) 15:49, 30 July 2017 (UTC)

I think "most of it". We have an article about Summa de arithmetica, geometria. Proportioni et proportionalita (Venice 1494) which was a school textbook containing a synthesis of the mathematical knowledge of Luca Pacioli's time. It added to the tradition of Liber Abaci (1202) in which Fibonacci c. 1175 – c. 1250 advocated the use of the digits 0–9, and of place value (until this time Europe used Roman Numerals). But the only original work that Luca Pacioli is known for is publishing the Double-entry bookkeeping system and a "Rule of 72" for estimating an investment's doubling time; Pacioli certainly used earlier sources such as works of Piero della Francesca c. 1415 - 1492 and drew on ancient sources such as:

• Babylonian numerals were the first known positional numeral system.
• The mathematics of Archimedes c. 287 – c. 212 BC, possibly in Latin translations by Gerard of Cremona c. 1114–1187
• The invention of Geometry is attributed to Euclid's Elements circa 300 BC.

For general reading see History of mathematics. Blooteuth (talk) 22:20, 30 July 2017 (UTC)

Minimum half-life below which a nuclide isn't a nuclide?

I seem to remember reading somewhere that a species composed of protons and neutrons had to last for some minimum duration (IIRC, on the order of 10^-14 or 10^-15 seconds) to attain some physical property or other that was considered to define an atomic nucleus, and thus by definition a nuclide with a half-life shorter than that could not by definition exist. Is that right? What is the lower bound? What are some of the half-lives of transient species in nuclear reactions that decay too quickly to qualify as nuclides? NeonMerlin 20:12, 30 July 2017 (UTC)

I'm not sure how to find a sourced answer, but this makes sense. The strong force has a range of 1 to 3 femtometers in an atomic nucleus, i.e. 1E-15. Therefore, if a nucleon moving at near-lightspeed cannot move ~ 1 fm before the nucleus decays, you could not have a meaningful binding together of the nucleus. light is 3E8 m/s, so this back of the envelope calculation comes up with maybe 3E-24 seconds for a minimum half-life. But alas, this is too much speculation, not enough answer. Wnt (talk) 20:19, 30 July 2017 (UTC)

Think the above answer is pretty good and likewise I can't find good references either. To my simplistic mind, I can only conceptualize thus: That when we are considering such short time intervals, then physics appear to advance not smoothly anymore, but in incremental steps. It is as if energy has not had time (in these very short intervals) to sort out, if it it is to become a photon or particle (or combination of) or of which type and flavour. This is why CERN run hundreds of thousands of experiments to statistically explore and validate the main pathways, thus cutting out the noise of random events. So, I would agree, that in this time period, a nuclide can not be said to exist as it it properties can not be measured, until it escapes into the realm where it can be objectively observed. Aspro (talk) 21:54, 30 July 2017 (UTC)

The shortest half-life at List of radioactive isotopes by half-life is ⁷H with t_1/2=2.3x10^–23 s, so the lower bound for "actually considered an isotope" is at least that short. DMacks (talk) 22:24, 30 July 2017 (UTC)

Mind you, ⁷H is not actually bound, and immediately leaks out a neutron. If you look at articles like isotopes of hydrogen you will see that many of these are assigned an energy instead of a half-life in the standard tables to let you know that they are not particle-bound: you can get the half-life from there by simple calculation. Double sharp (talk) 23:39, 30 July 2017 (UTC)

The ref for ⁷H (doi:10.1103/PhysRevLett.90.082501) says "The evidence for existence of the ⁷H state near the t+4n threshold was obtained", but I don't have access to the full article or have enough background (ha!) to know what exactly that means. It seems like it means there is some definition of being bound enough to be considered a nuclide and that this nuclide meets the definition.

The ⁷H isotope's half-life in our table is annotated as "not purely derived from experimental data, but at least partly from systematic trends". But the ⁵H half-life is apparently directly measured (neither annotated as trend-based nor listed as an energy), a with other elements' nuclides with half-lives in the 10^–22 s order of magnitude. DMacks (talk) 13:38, 31 July 2017 (UTC)

t+4n means a tritium nucleus plus four neutrons, which is what this loose conglomeration of one proton jostled between six neutrons would evidently much rather be. It is "bound" in the sense that if you plot the energy levels of ⁷H and t+4n the nucleus cannot immediately run down the potential well, but in practice the binding energy is so low that it makes close to no difference. (From the chemist's point of view, it would not count anyway since a nucleus would need about 10⁻¹⁴ s to get a stable electron cloud, but this is more physics than chemistry. If we were talking about superheavy nuclides then it would be a different story.) Double sharp (talk) 13:48, 31 July 2017 (UTC)

Thanks! DMacks (talk) 02:21, 1 August 2017 (UTC)

I think I read somewhere that a nuclide has to last long enough for the nucleons to arrange themselves into energy shells before it's actually considered a nuclide (before then, it's just an amorphous blob of protons and neutrons, known as a "compound nucleus"; several compound nuclei with Z>118 have been created, but none of these have formed actual arranged nuclides as of yet), but I can't recall precisely where. Whoop whoop pull up ^{Bitching Betty | Averted crashes} 23:41, 31 July 2017 (UTC)

Good lead! compound nucleus#Compound nuclear reactions says something about 10^–19 . DMacks (talk) 02:21, 1 August 2017 (UTC)

Turning orange juice into an alcoholic drink

Can it be turned into an alcoholic drink? Uncle dan is home (talk) 20:44, 30 July 2017 (UTC)

Of course. Just add vodka. --Trovatore (talk) 22:52, 30 July 2017 (UTC)

Almost any liquid which contains fermentable sugars and who's pH does not inhibit the growth of yeast can produce alcohol. This is not to be confused with orange liqueurs. There, the orange just serves as a flavourent. Aspro (talk) 21:04, 30 July 2017 (UTC)

There's some info at Fruit_wine#Orange_wine (and we have an article on pruno too). ---Sluzzelin talk 21:12, 30 July 2017 (UTC)

These are good article references, but there's a surprising apparent contradiction: orange wine is said to be difficult to make commercially because the pH is too low and there are bacteria that spoil the process, yet it can be made in prison using bread yeast? In any case, I can't think of any insurmountable obstacle to coming up with a strain of organism, perhaps a yeast, that could do this fermentation at low pH. Anyone? New industry just waiting for someone to claim it! And ... heck, doesn't it seem like old oranges you wait too long to eat get an ethanoley taste spontaneously? I wonder if that's a start... Wnt (talk) 01:03, 31 July 2017 (UTC)

Just because it's possible doesn't mean it's commercially viable. Orange juice is relatively expensive, relative to apple juice and grains, so unless there was some advantage to using it, the cost disadvantage would make it impractical. StuRat (talk) 01:13, 31 July 2017 (UTC)

Why isn't it widely commercially available? Uncle dan is home (talk) 21:38, 30 July 2017 (UTC)

Ph is too low. Raising the pH for fermentation destroys the essence of orange. Whist adding orange essence afterwards to a liqueur doesn’t. One doesn’t make wines from grapefruit, limes and lemons for the same reason. So why attempt to make orange wines? Aspro (talk) 15:41, 31 July 2017 (UTC) 22:29, 30 July 2017 (UTC)

It would be possible with very ripe oranges. In some varieties, as they ripen, the pH rises above that of grapes, but, as stated above, very ripe oranges don't taste very much of orange. Dbfirs 06:23, 31 July 2017 (UTC)

I think I once ate some fermented cantaloupe cubes accdentally. I think they were fermented, because they had a burning sensation like ethanol. Wild animals can eat fermented fruits and get "drunk". 50.4.236.254 (talk) 21:25, 30 July 2017 (UTC)

I can be turned into an alcoholic drink, i tried and succeed. That's quite easy, actually. It wasn't bad, but not good enough to do it again on purpose. There is a wikiHow to Ferment Fruit, it applies to orange as any other.

I bet someone did try to make and sell alcohol out of orange juice, it seems not successful enough to be notorious, only god knows why. Maybe because commercially it makes lot more sense to sale the juice and use the zest to flavor a cheaper and better alcohol, that to turn juice into a not-so-good sort of beer/wine.

Gem fr (talk) 14:45, 31 July 2017 (UTC)

Note: the WikiHow link above is about ferment the fruit itself and not the juice. Aspro (talk) 16:00, 31 July 2017 (UTC)

Fresh orange juice has on occasion accidentally fermented in my fridge. The result is a pleasantly effervescent drink, not unlike Buck's Fizz (OR). --catslash (talk) 00:54, 1 August 2017 (UTC)