Wikipedia:Reference desk/Archives/Science/2016 June 3

Science desk
< June 2	<< May | June | Jul >>	June 4 >

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.

June 3

Hole size vs cost

When drilling a hole in metal, the smaller the hole the faster/less costly it is, but this only applies up to a point. When the hole size gets too small the drill bit becomes too small and thus is prone to breakage. Approximately what is the hole size at this inflection point?

Basically I'm asking what size hole is the cheapest to drill in metal. I'm interested in stainless steel in particularly. Johnson&Johnson&Son (talk) 08:05, 3 June 2016 (UTC)

Sorry, forgot to mention that the hole depth would be 5mm. Johnson&Johnson&Son (talk) 09:48, 3 June 2016 (UTC)

Check this out. Lots of good tips about holes in general but point 14 is relevant i.e. "Deep, narrow holes with length to diameter ratios of larger than three should be avoided." 196.213.35.146 (talk) 10:16, 3 June 2016 (UTC)

Thanks! "length to diameter ratios of larger than three should be avoided" means anything that's smaller than ~1.6mm should avoided for a 5mm depth. But what about sizes above that? Are holes bigger than 1.6mm, let's say 2mm, cheaper or more expensive than 1.6mm holes? Johnson&Johnson&Son (talk) 10:25, 3 June 2016 (UTC)

Without more information about the commercial environment in which this/these hole/holes is/are to be drilled, there can be no single/satisfying answer to the question. For example, if a hobbyist is occasionally to drill one of these holes in a piece of stainless he will find one answer to the question. If a small workshop is to drill these holes a couple of dozen times they will find a different answer. If a factory uses automatic machine tools to do it as mass production they will find a third answer; and if an ultra-high production factory dedicated to drilling and fabricating components in stainless steel does it they will find yet another answer. Dolphin (t) 11:48, 3 June 2016 (UTC)

Quite true. Knowing the purpose to the hole is also important. If you are going to tap a thread you need to consider the pitch and diameter of the stud/bolt going into the hole to get the maximum strength over a 5mm depth. So it's not about the cheapest hole but the strongest join. 196.213.35.146 (talk) 11:57, 3 June 2016 (UTC)

Let's say it's a factory drilling the holes with a vertical machine center then. Johnson&Johnson&Son (talk) 12:29, 3 June 2016 (UTC)

So there must be another bound to the ratio, right? Bits that are too thin for their length have problems, but on the other end, bits that are very wide can never drill a clean shallow hole well in a single pass, due pitch of drill blade and the the taper of the tip. So for 5 mm depth, it must also more difficult/expensive to drill e.g. a 200 mm diameter hole than a 2 mm diameter hole (using conventional metal drill bits). I don't know what the cost/ease graph looks like between these points. I suspect that a lot of the answers are "it depends", and you might not find any good general rules, just empirical data for specific situations. For further searching - perhaps unsurprisingly the depth/diameter ratio is known as the aspect ratio of the hole. There has been much research on how to drill thin deep holes (e.g. [1] [2]). It seems to overcome the limits you have suggested, lasers and ultrasound and other things are used rather than conventional drill bits. Those papers are not directly relevant to this question, but many of the references contained will be. Here [3] [4] are a couple of other articles that might help you with the literature and search terms. SemanticMantis (talk) 13:09, 3 June 2016 (UTC)

Thanks a lot!Johnson&Johnson&Son (talk) 01:44, 4 June 2016 (UTC)

In what real situation do you only want to drill holes, but you don't care what the size of them is? 86.171.42.219 (talk) 03:26, 7 June 2016 (UTC)

Blood tests

Are there any blood tests that would involve boiling the sample taken? In other words, something that would literally make my blood boil? Lugnuts ^{Dick Laurent is dead} 12:48, 3 June 2016 (UTC)

A Blood ethanol test done by Distillation of a blood sample. This finds the level of ethanol in the blood of an intoxicated person for both medical and legal purposes, and may be ordered as an alternative to breath testing. AllBestFaith (talk) 15:50, 3 June 2016 (UTC)

Thanks! Lugnuts ^{Dick Laurent is dead} 18:03, 3 June 2016 (UTC)

Oshchepkov's device

I've stumbled upon a story whose author purpotedly describes how in late 1970s Pavel K. Oshchepkov obtained electric energy out of thin air. According to this (translated from Moskovsky Komsomolets): "Pavel Kondratyevich showed me a bulb from the flashlight with two wires wielded to it, their other ends were hanging freely in the air. And the bulb was glowing. [The inventor] said that the wire is made from a sly allow of copper and aluminum, the left end is purely copper, the right end is pure aluminium. Throughout the entire length, from left to right the copper content gradually decreases, while the aluminium content gradually increases. The electron is forced to run with an acceleration from aluminium to copper. The average speed of the electron in copper is higher than in aluminium. Thus the electrons run. Hence, there's an electric current, with no violation of the thermodynamics laws". Is this ever workable? Brandmeister^talk 14:27, 3 June 2016 (UTC)

The description sounds exaggerated, if not entirely fictionalized.

It is possible to use dissimilar metals to build a thermoelectric generator. Those can be built into a real device, and they can extract energy if you put heat energy into the system. The description above may have been corrupted (by hearsay) from a real demonstration of some similar device, or the device's capabilities might have been intentionally exaggerated for the purposes of promoting some scammy pseudoscience.

There are a lot of neat thermodynamic properties at any junction between dissimilar metals; more typically, at least one material is a semiconductor; but none of these neat properties permit perpetual extraction of energy. Aluminum/copper isn't a fantastic pairing either, at least not if you're aiming for a big effect - though those metals can be used to build a simple thermostat! To a casual observer, the thermoelectric effect may seem like creation of energy, but if you study it carefully, you see that it's just energy conversion. It is unlikely that you could use a bimetallic junction or a thermocouple to produce enough energy to power a light bulb - certainly not with the wire leads exposed to ordinary room temperature.

Nimur (talk) 15:27, 3 June 2016 (UTC)

Ok so room temp is out. But maybe the heat of your hand would work? Or a hot cup of tea? Or are we talking more like a bonfire? I don't know how to get a back-of-the-envelope estimate for converting heat in BTU to electrical watts in this manner. Also in the late 70s the story is surely about incandescent bulbs but there are some rather bright and low-watt LEDs out there now. It sounds to me like these principles could be used to make a hot beverage run a small LED for a classroom demo, and that sounds fun :) SemanticMantis (talk) 16:39, 3 June 2016 (UTC)

No comment on the conservation of energy issues with which the original question seems rampant, but as far as your unit-conversion question, we don't need backs of envelopes any more when we've got the magnificent units program available!

$ units
586 units, 56 prefixes
You have: btu/hour
You want: watt
    * 0.29307222
    / 3.4121282

So if you have an energy source that's capable of generating one BTU per hour, multiply by 0.29 to get its power-delivery rate in watts. (Assuming 100% efficiency, of course.) —Steve Summit (talk) 17:15, 3 June 2016 (UTC)

[Oh, and, sorry, I guess that was OR. But I can cite a reference!]

Haha thanks for the math plus the tip on the Units program :) I guess I was thinking about a total conversion for a specific yet hypothetical device, including efficiency estimates, but this is at least a start. SemanticMantis (talk) 19:07, 3 June 2016 (UTC)

Does house current induce amnesia?

I just took a moment to read the news just now and came across the case of some poor sap who got carjacked and forced into the stolen car. His captor soon realized the man had "seen his face", and came up with the most drastic way to fix that, for which now I think he is facing the death penalty. My first thought, needless to say, was gee, that was stupid, why didn't he just rip the cord out of a table lamp and give him a couple of minutes of electroconvulsive therapy? Then I had to look it up... is house current suitable for ECT side effects? While modern forms claim to be slightly more refined, ECT opponents describe it as basically house current perhaps hopped up to 150V or so. [5] On the other hand, the level and permanency of the amnesia is prone to dispute. [6] It's not clear to me whether that is a triumph of psychiatric restraint or just that electric current doesn't do as much as people think. So... is there any empirical data? Do people exposed to electrical accidents -- or incidents -- involving heads and house current (whether 110 or 220V, I won't be picky) report lasting amnesia from the event? Also, it is my impression that current across the head shouldn't stop the heart, but would it kill another way? Wnt (talk) 18:00, 3 June 2016 (UTC)

This would seem to fall into the same category as inducing amnesia by hitting somebody in the head with a frying pan. Sure, it's possible either might work, but the chances of doing so, without causing permanent damage, is quite remote. StuRat (talk) 19:13, 3 June 2016 (UTC)

According to the ECT article, "Retrograde amnesia occurs to some extent in almost all ECT recipients...", which is different from a frying pan. And there is a limit to how many times you can hit someone with a frying pan without killing him but I don't actually know if there is a limit to how much you can abuse him with ECT, if his future mental condition is not an issue for you. (Avoiding permanent damage would not be a priority for the culprits here; avoiding a murder charge would be) Wnt (talk) 20:06, 3 June 2016 (UTC)

could you post a link to the news story? Asmrulz (talk) 19:27, 3 June 2016 (UTC)

I'd prefer to keep it out of play;. Last time I asked a hypothetical question about a living person here I got a lot of threats and policy-brandishing, and I'd rather keep things on the level of pure mad science. :) Besides, these "he saw our faces, what do we do" stories all end the same way, and not the way I would have thought of. Wnt (talk) 20:06, 3 June 2016 (UTC)

We need to talk about Kevin: physician holds the hand of a patient in the air

In the movie We need to talk about Kevin, a physician holds the hand of a patient in the air for a sec during a diagnostic. Then he lets it loose. It was implied that if the patient maintains his hand in the air, that would be a sign of autism. If the patient left his hand fall, then that implies normality.

Does this diagnostic procedure exist? Or is this just one of those things only found in movies? --Llaanngg (talk) 23:58, 3 June 2016 (UTC)

I'm not aware of one, though similar tests are used to determine whether a patient is truly unconscious (you attempt to drop their hand in their face; if it moves away, it is assumed they're conscious on some level) and some Googling suggests that there are types of seizures or movement disorders that might be diagnosed that way (in fact, tying in to the unconsciousness test mentioned earlier, see here for example). Epilepsy is more common in autistic folks than the general populace and they may exhibit stiffening seizures, but that's the closest link I've found (so far anyway). Oh! There's also this (possibly better cite here), but that doesn't seem to be the same kind of thing you're describing. Matt Deres (talk) 02:42, 4 June 2016 (UTC)

I'm almost used to seeing "loose" when the writer means "lose"; the reverse error is a new one on me. I assume it's just a typo, though. --Trovatore (talk) 02:44, 4 June 2016 (UTC)

If anything, trying to take their hand in the first place might tell you more. Many on the autism spectrum would pull their hand back. StuRat (talk) 02:53, 4 June 2016 (UTC)

Yes, some people on the autism spectrum are hypersensitivity to touching. But do they get testing this way? Llaanngg (talk) 16:54, 4 June 2016 (UTC)

Not sure, but since autism lacks a specific medical test, behavioral tests like this might well be used. StuRat (talk) 15:38, 7 June 2016 (UTC)