Wikipedia:Reference desk/Archives/Science/2009 December 23

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.

December 23

fiberglass rot

When fiberglass rots I assume the resin is what rots and not the silicon dioxide fiber. By what organism does this rot occur? 71.100.6.206 (talk) 00:10, 23 December 2009 (UTC)

It could just be some chemical degradation. Light, oxidation, there are a host of ways that something can "break down" that don't involve organisms. --Jayron32 00:13, 23 December 2009 (UTC)

Fiberglass is just glass, there is no resin. Do you mean Fiber reinforced plastic? Ariel. (talk) 02:38, 23 December 2009 (UTC)

Ariel, Did you even read the article you linked? Here is the second sentence "It is used as a reinforcing agent for many polymer products; the resulting composite material, properly known as fiber-reinforced polymer (FRP) or glass-reinforced plastic (GRP), is called "fiberglass" in popular usage." APL (talk)

And the pink stuff that they sell as insulation is called what? It's called fiberglass, and it actually is just glass, with no resin or anything else. At least call the other stuff fiberglass panels/boards, etc. Ariel. (talk) 05:25, 23 December 2009 (UTC)

You are both right - according to Wiktionary the word has two distinct meanings:

1. silica based glass extruded into fibers that possess a length at least 1000 times greater than their width.
2. a composite material made from fine fibres of spun glass held together with resin. (Also called Glass-Reinforced Plastic.)

In (1) there is nothing but glass - and that's what you find in roof insulation and optical fibre. In (2) there is resin holding the fibreglass filaments together - but the resulting material is still called fibreglass...it's the stuff they make boat hulls from, for example. When type (2) fibreglass degrades it's generally because UV light has degraded the epoxy. The glass fibres are not degraded - so type (1) lasts pretty much forever (at least on "human timescales"). There are various coatings they can put on the material to reduce the UV breakdown - but it seems that none of them are perfect. SteveBaker (talk) 12:53, 23 December 2009 (UTC)

IMHO I agree with APL here. Ariel is right that fiber-reinforced polymer is a better name but our article itself makes clear early on it's commonly called fibreglass. If Ariel read the fibreglass article and/or understood this already then I would suggest a better answer would have been "fibreglass properly only refers to stuff that is just glass, you're probably thinking of fibre reinforced polymer which is common called fibreglass but which has resin in between the fibres and comes in boards and panels" or something of that sort since that's far more helpful to the OP and question answerers Nil Einne (talk) 14:10, 23 December 2009 (UTC)

But why correct the person at all when they are communicating effectively and properly using the common name of a substance? With the mention of "Resin" it was 100% clear to everyone what was meant. It's not as if the linked articles answered the question. APL (talk) 01:06, 24 December 2009 (UTC)

In my experience the fluffy insulation is called 'glass fibre', while 'fibreglass' means the bonded stuff. --ColinFine (talk) 17:38, 24 December 2009 (UTC)

Getting a bit too OT so taking to talk page Nil Einne (talk) 12:10, 25 December 2009 (UTC)

UV degradation of the resin is a big part. Store in shade. Polypipe Wrangler (talk) 12:23, 26 December 2009 (UTC)

Per-unit-cost energy content of different battery sizes

Of the several common battery sizes (AA, AAA, C, & D), which of them tends to give you the most energy on a per unit cost basis? --173.49.11.197 (talk) 00:19, 23 December 2009 (UTC)

Usually larger sizes are more cost effective, but D size is rare(ish), so may be more expensive. Battery (electricity)#Life of primary batteries and List of battery sizes has a chart with capacities, so you can do your own math. Be aware that rechargeables are different (but should have the capacity listed on the package), and that some battery manufacturers cheat, and put a rechargeable AA inside a C and D shell. Ariel. (talk) 02:43, 23 December 2009 (UTC)

Icosahedrons or Dodecahedrons?

If perfect (or nearly perfect) spheres are crushed together with equal pressure from all directions (similar to what one might expect when lowering a weighted bag of ping-pong balls or balloons into the depths of the ocean) will their interfacing surfaces form Icosahedrons or Dodecahedrons? 71.100.6.206 (talk) 00:32, 23 December 2009 (UTC)

How many spheres? How are they situated in the bag? What's the motivation and context for this question? ~~ Dr Dec (Talk) ~~ 00:34, 23 December 2009 (UTC)

See Sphere packing. You won't get either as they can't be arranged in a packing. People have actually tried the experiment with balls randomly thrown into a box. Dmcq (talk) 00:41, 23 December 2009 (UTC)

I was under the impression that the spheres got squashed by the pressure and lost their shape. This isn't the same as sphere packing where the spheres remain rigid. ~~ Dr Dec (Talk) ~~ 00:44, 23 December 2009 (UTC)

No, but it sounds like a related problem. You need to start with a sphere packing and then crush the spheres together. What you end up with will depend on what sphere packing you started with. --Tango (talk) 01:12, 23 December 2009 (UTC)

You can get most any shape you want. You can get cubes, if the spheres start out in that pattern. Perhaps you are asking what is the tessellation with the greatest ratio of volume to surface area. Which is another way of saying you want the best sphere packing. Ariel. (talk) 01:11, 23 December 2009 (UTC)

If you drop spheres into a container randomly you will most likely get sphere close packing. The polyhedron packing that most closely corresponds to this is a truncated octahedron honeycomb, so I am guessing that is what crushing under pressure will get you. To get a cube honeycomb you would have to carefully pack the spheres in a cubic honeycomb before crushing - which is quite difficult as the spheres will not naturally stay lying in that configuration. SpinningSpark 01:36, 23 December 2009 (UTC)

Truncated octahedrons are a space-filling geometry, but I don't know if you could get them by squishing spheres together. Nimur (talk) 01:53, 23 December 2009 (UTC)

The results of circles so treated in 2 dimensions is a honeycomb as stated. I'm merely looking for the same results in 3 dimensions using spheres starting with a regular lattice. 71.100.6.206 (talk) 02:06, 23 December 2009 (UTC)

No, it's not. If you start with your circles in a square grid, and squish evenly on the sides, you will get squares, not honeycombs. And you got an answer: truncated octahedron. BTW circle packing is not 100% definitively proven, so there is no definitive answer to your question. (Assuming that your real question is "what space-filling shape maximizes volume to surface area". Since your actual question has no answer.) Ariel. (talk) 02:30, 23 December 2009 (UTC)

Oooh, this paper is interesting. The guy has actually done the experiment with soap bubbles and gets the conclusion that the resulting honecomb is irrational. SpinningSpark 03:06, 23 December 2009 (UTC)

That paper is about randomly placed spheres. I think the OP wants a regular close packing. Dauto (talk) 03:53, 23 December 2009 (UTC)

The shape the OP is looking for is the Rhombic dodecahedral honeycomb. Dauto (talk) 03:47, 23 December 2009 (UTC)

Correct on both counts... Thanks Dauto! 71.100.6.206 (talk) 04:37, 23 December 2009 (UTC)

Sorry I thought the sphere packing article would mention that or link to it, I'll go and put in a link. Dmcq (talk) 11:10, 23 December 2009 (UTC)

I like rhombic dodecahedra, particularly because of the disconcerting way they look like cubes from most angles. [1]

[2] [3] Felis cheshiri (talk) 15:51, 23 December 2009 (UTC)

The comparative values of specific health behaviours

I'm aware of the standard advice regarding healthy behaviour (for example, don't smoke, stay slim, take exercise, avoid saturated fats, eat plenty of fruit and vegetables). But I'm wondering if the comparative worth of these different behaviours has been published anywhere? Knowing that, for example, that staying slim was worth 100 times more than eating X amount of fruit&veg every day would motivate weight-loss, and increase dedication to the most valuable behaviour. 92.29.68.169 (talk) 13:39, 23 December 2009 (UTC)

I don't think you'll find definite numbers like that, partly because there is some disagreement and partly because all of these behaviours exist on a continuum and interact with each other, your genes, and other aspects of your environment. For example, cutting down on salt would have a minimal impact on most people's health, but for a few people it would drastically increase their lifespan because they are sensitive to salt. And there is no procedure to identify these people before they are talking to a doctor about their serious health problems. 86.176.191.243 (talk) 14:26, 23 December 2009 (UTC)

Official government advice (5 A Day) seems to disagree with this notion that we are all different. Felis cheshiri (talk) 16:11, 23 December 2009 (UTC)

Of course, the article that you linked notes that these nominally-similar programs offer different advice in different countries. Do Americans actually need eight times more fruit than Britons? It seems unlikely. 'Official government advice' is going to be based on some combination of best guesses, old data, received wisdom, studies in other countries, expert consultion, powerful lobby groups, typographical errors, and political biases. The advice that makes it into the final pamphlet is going to be based largely on the practices which, in the aggregate, tend to produce an overall improvement without doing significant harm to any individuals.

For example, recommending a moderate intake of sodium (salt) isn't goign to be harmful for anyone, and it can significantly reduce hypertension (and associated morbidity) in individuals who are particularly sensitive to sodium. The net result is a (small) overall increase in average health, though most of that improvement is going to be due to major gains in a smaller fraction people. Similarly, an individual who has a family history of type 2 diabetes is likely to derive a greater potential benefit (delay or evasion of disease progression) than the average person from exercise and a healthy diet — but exercise and diet improvements will probably offer at least some benefit to nearly everyone. TenOfAllTrades(talk) 16:39, 23 December 2009 (UTC)

The Wikipedia article about the difference in UK and US portion sizes is misleading: if you read the source site, the UK "one tablespoon" portion size only refers to dried fruit, for other types of fruit and vegetables the portion sizes are similar to those in the US. I have now edited the article to remove the misleading information. 92.29.51.193 (talk) 20:22, 23 December 2009 (UTC)

I have added the curious fact about dried fruit after your edit, in a non-misleading way. Felis cheshiri (talk) 20:53, 23 December 2009 (UTC)

I have now looked at the souces given in the article: I quote from what I wrote on the discussion page - "The article says the US and UK portion sizes are eight times different. Not true if you look at the souces given: the US sources do not mention dried fruit except for "1 small box (1/4 cup) of raisins", and the UK source mentions "1 tablespoon" of dried fruit. Tablespoons, different from dessert spoons, are big things so the portion size is about the same. Also, it unimportant and insignificant if the portion sizes (for dried fruit) are different.". Please remove the misleading information. Update: the confusion may be do to differences in US-english and UK-english: in the UK a tablespoon is a very large spoon. The Wikipedia tablespoon article appears to be about what we in the UK would call a dessert spoon. See the discussion page to that article. A heaped tablespoon (in UK-english) would probably be much more than the two fluid ounces referred to as the US dried-fruit portion size. 78.149.147.198 (talk) 23:13, 23 December 2009 (UTC)

Alright, I bow to your accuracy. :) Felis cheshiri (talk) 01:04, 24 December 2009 (UTC)

I respectfully disagree, and have said so on the tablespoon talk page. I have supported my assertion with a reference from Delia Smith and some OR based on a 30 year old UK tablespoon. Felis was right. 86.176.191.243 (talk) 01:18, 24 December 2009 (UTC)

From my personal experience of UK tablespoons and UK dessert spoons, I have to disagree with you. A tablespoon was much larger than a dessert spoon. The size was not standardised, so you must have come across a smaller one. Tablespoons are now rare in the UK due to changes in cooking and eating styles. 78.146.200.137 (talk) 11:04, 24 December 2009 (UTC)

Turns out the UK's 5 A Day recommendations are based on (among other sources) "the Leatherhead report", summary available here: [4] which states that (in this context, at least) a tablespoon = 15ml. If that's comparable to a quarter of a US cup, that gives us a 60ml cup rather than (minimum) 240ml, yielding a nice factor of 4 disparity. Felis cheshiri (talk) 13:47, 24 December 2009 (UTC)

You have made the same mistake again. The "tablespoon" (UK-english) only refers to dried fruit, not fruit and veg in general. As described in tedious detail on the discussion page to the five-a-day article, the US portion size for dried fruit is very similar. Note that raisins are a type of dried fruit. 92.24.73.139 (talk) 15:01, 27 December 2009 (UTC)

As the previous response notes, there are significant and not-fully-understood genetic factors which can make some individuals more susceptible than others to particular habits. It's important to remember that any particular numbers related to loss of life expectancy are aggregate statistics, and that any individual's response to a given lifestyle choice will ultimately depend on their own genetics, the interplay of assorted environmental and lifestyle factors, and sheer dumb luck.

That said, there's a brief summary of a couple of large studies which mostly focused on obesity here. Most studies find an 'optimum' body mass index somewhere between 22 and 25. Being mildly overweight (BMI up to about 30) doesn't tend to have huge penalties, particularly as one grows older. (I have seen speculation that a modest amount of excess fat provides a 'cushioning' effect during periods of illness when one might not tend to consume enough calories, particularly among the elderly.) Being significantly overweight when you're a young adult hurts you more than growing overweight as you age. BMI between 30 and 35 costs you (statistically speaking) 2 to 4 years. BMI of 40 to 45 knocks down median survival by 8 to 10 years; smoking carries roughly the same 8- to 10-year penalty. Being underweight (BMI less than about 20) also is linked to excess mortality.

That said, the apparent power of BMI as a predictor should not be taken to mean that merely being somewhat overweight is a cause of excess mortality. Factors such as poor diet or a lack of exercise can cause both disease and obesity. A 2007 study of seniors (discussed here) found that obese individuals (BMI from 30 to 35) who were 'fit' (based on a treadmill test) had a significantly reduced mortality compared to 'unfit' people of normal weight.

I don't have comparable numbers for nutrition. The challenge there is that while weight is easy to quantify and track (and often recorded regularly and consistently by physicians and patients) it is much more difficult to get accurate data on everything that a large cohort of individuals eats. TenOfAllTrades(talk) 15:22, 23 December 2009 (UTC)

Density of heavy water

I was wondering why the density of pure heavy water is not 2g per cc as it contains one proton and one neutron per atom and as the neutron is about the same mass as a proton and I assume the atomic radius of water and heavy water molecule the same. —Preceding unsigned comment added by 41.212.208.234 (talk) 14:35, 23 December 2009 (UTC)

Heavy water contains one proton and one neutron per atom of hydrogen. However, water is not pure hydrogen. The mass of the oxygen does not change, and so the mass of the molecule does not double. — Lomn 14:46, 23 December 2009 (UTC)

(After edit conflict) You have forgotten the 16 protons and neutrons in the oxygen atom. A molecule of D₂O contains 20 nucleons whereas H₂O contains 18, so you would expect density of D₂O to be about 20/18 = 1.11 g/ml, which is approximately correct. Gandalf61 (talk) 14:48, 23 December 2009 (UTC)

It's also possible to get water with a heavy isotope of oxygen (up to 2 extra neutrons) to make even heavier water. I believe it's used in some medical tests. Ariel. (talk) 18:34, 23 December 2009 (UTC)

(ec) Normal water has two hydrogen atoms (one proton each) and an oxygen atom (8 protons and 8 neutrons). Heavy water has one normal hydrogen atom, one deuterium atom (one proton and one neutron) and one oxygen atom. That means you have a total of 19 nucleons compared to 18. So the density ought to be about $1{\frac {1}{18}}$ grams per cc. Of course, actual heavy water is a mixture of H₂O, HDO and D₂O. --Tango (talk) 14:51, 23 December 2009 (UTC)

A minor quibble about terminology. 'Heavy water' should not be assumed to mean a 1:1 ratio of protium (regular 'light' hydrogen) and deuterium. In my experience, heavy water almost always means pure deuterium oxide (that's all you get if you look for 'heavy water' in the Sigma catalog), although IUPAC seems to be more flexible, accepting any "Water containing a significant fraction (up to 100 per cent) of deuterium in the form of D₂O or HDO"[5]. TenOfAllTrades(talk) 15:34, 23 December 2009 (UTC)

Indeed, the terminology is not consistently used. It depends on context. Some uses of heavy water need almost pure D2O, others just need a higher than usual ratio of D to H. What people call "heavy water" will depend on what use they are putting it to. If the OP meant pure D2O then the density should be changed to $1{\frac {1}{9}}$g/cc. --Tango (talk) 15:38, 23 December 2009 (UTC)

Totally interjecting here, but I never expected an extra two neutrons per molecule would make such a difference as .11 g/ml. Does this mean that heavy water would sink to the bottom of a tank of regular H₂O? Ks0stm ^(T•C•G) 15:23, 23 December 2009 (UTC)

No, because the two liquids are not immiscible - the heavy water molecules will dynamically combine with the ordinary water molecules to create a homogenous mixture of H₂O, HDO and D₂O. But our heavy water article does say that heavy water ice will sink when placed in ordinary water. Gandalf61 (talk) 15:31, 23 December 2009 (UTC)

There will be some separation with a concentration of heavy water at the botton slightly higher than at the top. But the difference will be small at 1 g gravity. Higher concentrations can be achieved with centrifuges. Dauto (talk) 17:07, 23 December 2009 (UTC)

Average time of sunlight worldwide

If we ignore issues such as topography, is there anywhere in the world in which the sun is not above the horizon for about 2,190 of the 4,380 hours in a non-leap year? I'm pretty sure that the answer would be no, since places on the Equator have equal amounts of sunlight every day, and anywhere else in the world, it seems that the amount of daylight in any given day and the amount of night exactly six months away would be equal, but I could be forgetting to account for something. Nyttend (talk) 22:21, 23 December 2009 (UTC)

If we ignore topography then I think that is right. It should all average out. If we do count topography then it can be very different (eg. see Viganella). --Tango (talk) 22:53, 23 December 2009 (UTC)

You are asking if there is anyplace where there is an imbalance in day vs night, as measured over a year? I would have to say yes for two reasons.

1: Because the sun is a point source, but darkness covers everything else - they two are not symmetrical (it's not half light, half dark). Let's assume the earth was not tilted - the poles would never get any light. The poles are tilted of course, but the intensity of their light (add up all the light energy they receive in a year) is less than the total intensity of light in the equator - but the "intensity" of darkness is the same for both. This is not what you are asking though. You are asking if you can simply see the sun or not.

2: Let's assume the earth rotates on it's axis once per year. There would of course be places that get no light. Now lets assume twice per year, then simply because of the tilt of the earth, some places would get more than others because they happened to be tilted away from the sun during their moment in the light. Obviously the earth rotates faster than that - but it's not an integer number of rotations per year. So there are parts of the earth that get just a bit less sun than others because they happened be tiled away from the sun during their moment. The difference is probably small, but you seem to be looking for an exact answer. If you average over eons, it probably works out, but not averaged over a year. Ariel. (talk) 23:08, 23 December 2009 (UTC)

1:The sun is not a point source. It is infact larger than the earth.

2:True but completely negligible and probabily not the kind of precision the OP is looking for. His arbitrary exclusion of leap years shows he is not interested in those minor details. Other effects are more important such as refraction in the atmosphere. Dauto (talk) 23:42, 23 December 2009 (UTC)

The simple answer is yes, and this can be seen without considering atmospheric effects or the fact that the Sun is not a point source. Just remember that from the equinox in March until the one in September the Sun is continuously above the horizon as seen from the North Pole, and for the rest of the year it is continuously above the horizon as soon from the South Pole. But because the Earth moves faster in its orbit when closer to the Sun, and it's closest in January and farthest in July, the two equinoxes are not equally spaced -- the North Pole always gets more time in sunlight. Here's are the equinox dates for the years 2000-2020 from the United States Naval Observatory web site (all times are UTC), and I've added a column showing the elapsed time between them in days. Note that it is always longer than half a year (which is about 182.62 days).

 2000  Mar 20 07:35  Sept 22 17:28  186.41
 2001  Mar 20 13:31  Sept 22 23:04  186.40
 2002  Mar 20 19:16  Sept 23 04:55  186.40
 2003  Mar 21 01:00  Sept 23 10:47  186.41
 2004  Mar 20 06:49  Sept 22 16:30  186.40
 2005  Mar 20 12:33  Sept 22 22:23  186.41
 2006  Mar 20 18:26  Sept 23 04:03  186.40
 2007  Mar 21 00:07  Sept 23 09:51  186.41
 2008  Mar 20 05:48  Sept 22 15:44  186.41
 2009  Mar 20 11:44  Sept 22 21:19  186.40
 2010  Mar 20 17:32  Sept 23 03:09  186.40
 2011  Mar 20 23:21  Sept 23 09:05  186.41
 2012  Mar 20 05:14  Sept 22 14:49  186.40
 2013  Mar 20 11:02  Sept 22 20:44  186.40
 2014  Mar 20 16:57  Sept 23 02:29  186.40
 2015  Mar 20 22:45  Sept 23 08:21  186.40
 2016  Mar 20 04:30  Sept 22 14:21  186.41
 2017  Mar 20 10:29  Sept 22 20:02  186.40
 2018  Mar 20 16:15  Sept 23 01:54  186.40
 2019  Mar 20 21:58  Sept 23 07:50  186.41
 2020  Mar 20 03:50  Sept 22 13:31  186.40

In general, this reason means that places in the Northern Hemisphere will get more time of daylight than in the Southern Hemisphere. However, the effect is offset, and I believe more than offset, by the variation in distance of the Sun. --Anonymous, 23:51 UTC, December 23, 2009.

(ec) Well, we can also consider whether the sun is up for 2,202 of the 4,404 hours in a leap year; my point was simply exactly half a year. I was wondering if perhaps refraction would contribute, but I've never quite understood — how could this make the day seem longer or shorter? If it makes sunset seem to be later than it would if we had no atmosphere, how is it that sunrise doesn't appear to be several minutes later as well? Refraction doesn't cover the subject; I'll welcome a link to an article that does. Nyttend (talk) 23:55, 23 December 2009 (UTC)

The difraction makes the sun's light curve a little bit around the earth making the sun rise earlier and set later. Dauto (talk) 00:38, 24 December 2009 (UTC)

Since the sun is much bigger than the earth, can we not also say that more than half a sphere will be illuminated - so something like 181° rather than 180° - therefore the sun will always be showing above 2,190 hours  Ronhjones  ^(Talk) 00:11, 24 December 2009 (UTC)

I should think that the finite angular size of the sun as seen from earth should be the more important contributor. Sun has an angular diameter of about half a degree. So, even though the time for which one could see the sun at the equator if it were a point source is about 12 hours per day, this finite size increases the time by 2 minutes on the equinox day and even more on the other days. This effect becomes even more important closer to the poles (If the earth's rotation axis were perpendicular to the ecliptic, people at the poles could have seen the sun throughout the year). So what I believe is that the total time for which one can see the sun at any point on the earth is more than 50% when a year is considered -- Raziman T V (talk) 05:29, 24 December 2009 (UTC)