A reader writes:

I was watching an awesome Honda ad featuring the ghost of Ayrton Senna and his 1989 car...

...and I noticed people arguing in the comments on Jalopnik about Doppler effect, which I think you can hear in the video as the "car" goes past the camera.

Per your previous writing about "common sense" and concepts that "slither out of people's mental grasp", can a series of speakers set up around a racetrack and playing the sound of a car actually create the same Doppler effect as the actual car did?


No, they can't.

The Doppler effect happens when a moving object emits something, in this case sound waves. When each new wave is emitted in front of the sound source, it's closer to the previous wave than it would have been if the emitter were stationary. Behind the emitter, each new wave is a bit further from its predecessor than it would be if the emitter weren't moving.

This works for light as well, hence "redshift" and "blueshift".

We don't notice redshift or blueshift in everyday life because Doppler shift is a proportional effect, and the speed of light is so high that no light-emitter that humans normally deal with moves at an appreciable fraction of lightspeed relative to us. The speed of sound, however, is relatively low (about 340 metres per second close to standard temperature and pressure), and the human ear is quite sensitive to changes in pitch. So we can easily hear this effect on the sound of a car engine...

...or horn, when that car passes us at speed.

(My favourite example of car-horn Doppler shifting, which includes a lot of moderately comprehensible cursing, is this one.)

If you set up a bunch of speakers to imitate the sound of a passing car, none of them are moving, so there will be no Doppler shift from the point of view of a stationary observer. You could create the same effect by deliberately adding pitch shifts to the sound being played so that it sounds correct from a given listening location, but that'll make it sound wrong to listeners somewhere else. Doppler changes are caused by waves being bunched up and spread out by motion, and that just doesn't happen if neither listener not sound-emitter are moving. There's nothing about the order in which speakers play sounds that change what the sounds are.

(OK, there might be some interference effects audible at various listener locations. But that wouldn't sound Doppler-y.)

There actually would be Doppler effects if you were in your own car driving around the racetrack during the ghost-of-Senna performance, though. A moving listener creates Doppler shift in exactly the same way as a moving source:

Again, though, the pitch-shifts wouldn't sound right. They'd entirely depend on your speed relative to whatever stationary speakers are sounding at a given moment.

A related concept to this is the idea of the faster-than-light laser dot.

Consider flicking the dot of a laser pointer across, say, the face of the moon. (Presume you've got a laser that's well enough collimated that it still has a small dot at that distance.)

If the dot crosses the moon in, say, a hundredth of a second, and even if you ignore its curvature the moon is about 3,400 kilometres across, then that dot is going about 340,000 kilometres per second, which is faster than light. Address for delivery of Nobel Prize in Physics will be provided on request.

Unfortunately, and to the chagrin of a great many cats, a laser dot is not a "thing". It's just where photons happen to be falling and bouncing off at any given moment. Moving a dot faster than light is indeed perfectly theoretically possible, but you might as well give two blokes each a flashlight with an accurate timer built in, have them synchronise timers and then move a thousand kilometres apart, and then turn their flashlights on and off so that one light-pulse happens a thousandth of a second before the other. Presto, now a dot has moved at a million kilometres per second, more than three times the speed of light!

Except that doesn't mean anything, because that dot of light is not a thing moving faster than light. You could fill the space between those two flashlights with a trillion more flashlights timed to give a wonderfully smooth movement of the dot, but the dot would still not be a thing travelling faster than light. A spinning lawn sprinkler may have a contact point between droplets of water and the circumference of its spray pattern that goes round and round at a quite impressive speed, but that's just where the water hits the lawn, it's not an actual separate moving object.

(By the way, smart alecks, relativistic time dilation does not mean the flashlight timers would get significantly out of sync if the flashlight-carrier on one end got to his assigned location on foot, taking weeks, and the other got to his by rocket-sled at ten thousand kilometres per hour. At 10,000km/h your clock will tick slower than that of a stationary observer, but only by a factor of 1.0000000000429. The fastest object humanity has ever made is the Helios 2 probe, at 70,220 metres per second relative to the sun; it achieved a time dilation factor all the way up at 1.000000027!)

A further extension of this idea is to say, "OK, what if I've got a stick a million kilometres long, and I hold one end of it and spin it around my head in a circle in, say, five seconds? The circumference of a circle with radius one million kilometres is 6,283,185 kilometres, and the tip of the stick it will go all the way around that circumference in five seconds, which is 1,256,637 kilometres per second. The tip of the stick is a thing and not just a dot of light, so it's really going at that speed, which is 4.2 times the speed of light, NOW can I have my Nobel prize?"

No, you still can't.

Ignoring the obvious issues regarding the construction and inertia of a million-kilometre broomstick, there is no way for one end of an object to know what's happening to the other end at faster than the speed of light. Motion of the object occurs when the molecular bonds that hold it together are stretched and pull the molecules along, and there's nothing about those molecular bonds that causes them to influence each other faster than light. Otherwise you could make an instantaneous communication system by taking your very long magic broomstick and tapping on the end of it in Morse code or something.

So even if your very long stick were made of alien indestructium with an infinite tensile strength, spinning the middle of it round and round would just cause the whole thing to start wrapping up into a spiral. You could then try cracking it like a whip if you wanted, because you're Cowboy Galactus or something, but the other end of the object would still not travel faster than light, because no "information" within the object, in this case the information regarding the location and motion of its component particles, can travel faster than light either.

This seems bizarre, but again this is because we're talking about scales far larger than those on which humans normally operate. On the very large scale, nothing is particularly solid. If planets and stars and even galaxies run into each other, the energies involved may be unimaginably large, but all of the actual objects behave pretty much as if they were made of blancmange.

(Actually, in galaxy collisions, few to no actual collisions of the objects that make up the galaxies are likely to happen, because galaxies are mostly empty space.)

If Unicron were actually the size of even a small planet, no material that even theoretically exists in the universe would be stiff enough for him to be able to transform like his car-sized distant relatives. (Well, maybe if he's made of some kind of degenerate matter and has magical technology to prevent himself from collapsing into a black hole. Once you can cancel gravity, you might as well move information faster than light, too. It never seems to take a Transformer or Decepticon much more than twenty minutes to get to anywhere in the universe, after all.)

To reward anybody who managed to get to the end of this post, the ghost-of-Senna ad sounds pretty good, but the Shell-Ferrari one from a few years ago is much better:

(I think that version's the best one on YouTube in both resolution and sound. Aspect ratio's wrong, though.)

Psycho Science is a... sort of... regular feature here. Ask me your science questions, and I'll answer them. Probably.

And then commenters will, I hope, correct at least the most obvious flaws in my answer.

Zaps and bangs

A reader writes:

Hi, Dan! 

Though I didn't follow all of the details, I did enjoy your writing about electrocution and car batteries.

Do you know the odds of getting electrocuted if one is standing in a wet shower with wet skin using a cordless (battery-powered) sander? I don't know what kind of power I'll need to work on residential showers for hours at a time, but the electric chorded sander I WAS using (until I decided that I'm tired of risking electrocution) says it's a 120 Volt, 10 Amp model. 


There's probably no danger, but there could be some.

Cordless tools all run from low-voltage DC, although the voltage has risen in recent tools that use one or another flavour of rechargeable lithium battery. Higher voltage is better, from the tool-makers' point of view, because a given power from a higher voltage requires less current. This means thinner wires, less beefy switches, and generally speaking a cheaper, lighter tool with the same power.

Cordless tools are also, in general, significantly less powerful than corded versions. It's normal for corded drills and saws and sanders and such to draw peak power of at least several hundred watts. The ten-amp 120-volt rating of the sander you mention makes it a 1200-watt unit (so I presume you're talking about a belt sander, not an orbital one), though it'll only draw that much when it's working hard. You can expect even big heavy cordless tools to have no more than half the power rating of a similar corded tool.

Discovering exactly what that rating is can be difficult, partly because cordless tools can have a larger range between their "spinning freely doing nothing" and "working so hard it's barely turning at full power" power consumption than corded tools do. Mainly, though, cordless power ratings are harder to find because consumers think more watts are always better. So a cordless tool that costs three times as much as the wall-powered version, yet has a third the power rating, won't sell well, unless the manufacturer conceals that latter number.

I'm telling you all this just to explain my original wishy-washy "possibly dangerous" statement. If you're using a 12V tool then you probably won't be able to do yourself any electrical harm with it, even if you smash the thing on the wall until it breaks and then smack yourself in the chest with the pointy bits.

A 36-volt tool, on the other hand, is edging up toward the kind of voltage that actually can harm you, if only indirectly. (Direct harm: Current through your heart stops it, you die. Indirect harm: Current through some other part of your body causes you to spasm and dig a tool into yourself, fall off a ladder, flop out of the shower recess and smack your head on the toilet, et cetera. This sort of secondary injury following a non-fatal shock is a lot more common than injury or death caused directly by electricity.)

In the real world, even crappy bargain-basement cordless tools have enough plastic between you and the wiring battery terminals that no matter what voltage they run at, you pretty much have to make a specific and deliberate project out of killing yourself with one. Working in a wet environment is still dangerous, but only because it makes it easier to slip and then drill, saw or sand yourself instead of the workpiece.

Brand-name tools are generally safer still, and adding water to the situation may ruin the tool but is unlikely to hurt the user. Even the commonly-recognised-as-lethal "dropping a hair-dryer into your bath" situation is actually not terribly likely to kill you, though I don't recommend you try your luck.

If it's possible to electrocute yourself with cordless-tool gear in any way at all, here is I think your best chance of doing it without specifically running wires from the inside of the tool to nails driven into your chest. There are plenty of battery designs with exposed terminals of one kind or another, so suppose you eject the battery from the tool by accident, and then somehow grab that battery with both, wet, hands, so positive is touching one hand and negative is touching the other.

Even then, the resistance of human skin is way up in the tens of thousands of ohms - I found the resistance between two closely-spaced points on my tongue to be 70,000 ohms. So even with a 36-volt battery it'd be surprising if one whole milliamp managed to flow across your chest, and not all of that would go through your heart. I think you'd be an easy order of magnitude away from enough current through the heart for there to be any risk at all.

(I'm sorry to say that I'm not about to conduct heart-stopping experiments on myself. I have, however, previously zapped my arm for science.)

If both of your hands had bleeding cuts on them then 36 volts might be enough to at least give you a shock you could feel and it might have cardiac consequences, but this is really pushing it. And any sort of work gloves not made of chain-mail would erase the risk completely.

And, of course, back in the real world it continues to be downright difficult to actually touch the positive with one hand and the negative with the other. If you just grabbed both terminals of a 36-volt battery with one wet bleeding skinless lightly-salted hand then it'd sting like a bugger, but once again the only real health risk it'd present would be if the pain startled you enough that you then hurt yourself in some other way.

I won't be surprised if cordless-tool voltages rise further, though. There are already cordless mowers that run from 48-volt packs, for instance. So it's possible that a few years from now there'll be cordless tools running from voltages high enough to pose real electrocution risks.

It'll still be a lot less dangerous than it was in the olden days of corded tools, though, when casings were still commonly made of shiny cast aluminium. Then, the user's life was in the hands of the manufacturers and electricians who're meant to keep earth wires connected, and prevent live wires from touching the tool chassis.

With modern plastic casings and other construction improvements, even a theoretical 96-volt cordless tool is not likely to be an electrocution risk, even if you use it in the rain or, more realistically, get all hot and sweaty while working.

There's a lot of energy in a cordless-tool battery, though, and they definitely can hurt you if that energy is released very quickly because of, say, a short circuit...

...or severe over-charge...

...or physical damage...

The reason why drills and laptops and iPads aren't exploding all over the place is that the naturally excitable personality of lithium-ion technology, in particular, is kept calm by strong casings and protection circuitry ranging from simple fuses to smart current limiting:

If one of your cordless tools manages to puncture the battery of another, though, your life may still become quite exciting.

So I suppose I've allayed your fears of one kind of injury and then given you a new one to worry about.

There's no need to thank me.

Psycho Science is a... sort of... regular feature here. Ask me your science questions, and I'll answer them. Probably.

And then commenters will, I hope, correct at least the most obvious flaws in my answer.

My ever-vigilant Perpetual-Motion-Claims Patrol

A reader writes:

OK so this looks really cool:

Alabama-Based CMR Demos Programmable Magnets That Changes Polarity And Strength On A Whim

As long as they don't violate the laws of physics that is. What do you think, scam or not?


Correlated Magnetics Hoverfield demonstrator

I don't see any reason to think that Correlated Magnetics are doing anything snake-oil-y. They've just got a way of making single magnets which contain multiple differently magnetised regions.

Take the above "Hoverfield" demonstration unit gizmo, for instance...

...which you can buy for a hundred bucks if you like.

Clear acrylic and magnets doing odd things do indeed instantly start some red flags waving among those of us who're accustomed to independent thinkers like Steorn who, as you say, claim to break the laws o' physics but never quite manage to actually do it. But Correlated Magnetics are not in the physical-law-breaking business. You can make a less elegant contraption that does the same thing all of their products do, at home.

Take the "Hoverfield" thing, for instance. To make your own ugly version, all you need is a couple of large relatively weak magnets, plain ferrites, for instance, and two or more small strong rare-earth magnets.

Now glue the rare-earth magnets onto the faces of the ferrite magnets so that when the ferrites are facing and attracting each other, the rare-earth magnets are facing and repelling. The large ferrite magnets have a big, weak field; the small rare-earth magnets have a little, strong one. So the ferrites attract until they're close enough that the stronger but smaller repulsive force of the rare-earth magnets equals the attractive force of the ferrites, and provided your contraption prevents the magnets from slipping sideways and ruining the demonstration, it'll oscillate to stability with the magnets close, but not touching.

Correlated Magnetics can make a magnet array that does this, but looks like an ordinary single magnet. They produce various other one-piece arrays too. "Coded" magnets that have matching pseudo-random pole patterns so they lock together very strongly but only in one orientation, for instance, and other patterns that reduce the size of the field but strengthen its holding force. (This is also how rubber fridge magnets work - rub a couple of them over each other and you can easily feel the "ridges" of polarisation that make the weak magnetic material able to actually stick to a fridge.) Correlated Magnetics have a few other such creations.

Calling these magnets "programmable" is I think a bit of marketing puffery, since they've got nothing to do with true programmable matter. If these things are really "programmable", then so are those big Edmund-Scientific-type junk-fishing magnets that you can "turn off" with a handle (which moves the magnet inside away from the external field-concentrating pole piece that the junk sticks to).

There's no perpetual motion claims here, though. leave your pitchforks stuck in the haystack, and your torches unlit.

Psycho Science is a... sort of... regular feature here. Ask me your science questions, and I'll answer them. Probably.

And then commenters will, I hope, correct at least the most obvious flaws in my answer.

Raining rocks

A reader writes:

If you're up for it, I'd really love an article on why the size of that meteorite varies from 10 tons to 10,000, why the rarity of it is 1 in 5 years to 1 in 100, and why the explosion is everything from 1 to 500 Hiroshimas (I hadn't realized that was a new standard measure until today).


Right after everybody started goggling at YouTube videos of lights in the sky over Chelyabinsk, and blessing once again the everlasting source of comedy and horror that is the Russian dash-cam phenomenon, a lot of reporters were probably hitting the telephones. (Or just Twitter, now that having the first story is so much more important than having the first correct story.)

Anyway, early on there was an estimate allegedly from the Russian Academy of Sciences that said the meteoroid was only about ten tons. I don't know who at the Academy first said that to a journalist, since there doesn't seem to be an official press release or anything, but that mass estimate seems tied to them.

Since this estimate came before anybody had pulled together enough reports and readings to know for sure the size, velocity and explosion altitude of the rock, it's not surprising they were way off. I presume they just estimated a relatively low airburst of a relatively small rock, enough to do this sort of damage...

...without requiring the meteoroid to be a one-per-century size.

Around the same time, the European Space Agency similarly estimated the rock to be relatively small, with the caveat that they didn't yet have "precise information on the size, mass and composition of the object".

But then NASA estimated the rock was much bigger and heavier and blew up much higher. Since then, better readings have caused NASA to estimate it was a little larger again, putting it in the one-per-century category. I think this pretty conclusively overrules the early, low estimates.

This may also make it fortunate that this meteoroid came in at a grazing angle, and exploded so high up. I'm no expert on meteorite dynamics, but if the Chelyabinsk rock had managed to get down to ten kilometres or lower before it exploded, it would have Tunguska-ed the city, not just outshone the sun and then broken lots of windows.

This may have been impossible, especially if this meteoroid was one of the common stony types. Nickel-iron meteoroids are much rarer than stony ones, but also much more likely to make it to the surface without "exploding". The explosion effect when a meteoroid disintegrates in the atmosphere comes from the much greater surface area per mass of the fragments; they decelerate faster and heat up more, creating the kaboom. Whatever bits survive this process are generally small enough that they're only travelling at terminal velocity when they hit the ground.

I think gigantic dinosaur-killer meteorites can't help but make it to the surface largely intact, and ruin everybody's whole week. At the other end of the size chart are the tiny micrometeorites that drizzle down constantly, which anybody can harvest from the roof of a building.

I agree that using the Hiroshima bomb as an explosion-size yardstick is a bit silly, since it's not as if many people now living have a personal, visceral understanding of what that means. You might as well say something like "the meteoroid weighed more than 9000 tonnes, as much as fifteen thousand adult bluefin tuna".

I propose we use the PEPCON explosion instead.

The two biggest explosions there, which happened almost simultaneously, added up to only about 2.7 kilotons of TNT. The Chelyabinsk meteoroid explosion was, according to NASA's corrected estimate, close to 500 kilotons.

UPDATE: Phil Plait, the moderately famous actual astronomer who you should obviously have listened to about this stuff before you wasted minutes of your life on the above, has a couple of articles about the Russian meteor here and here.

Psycho Science is a... sort of... regular feature here. Ask me your science questions, and I'll answer them. Probably.

And then commenters will, I hope, correct at least the most obvious flaws in my answer.

12 AAs in the magazine, one in the chamber

A reader writes:

I've read that the problem with ray guns is that as an energy delivery system, pieces of high-speed lead propelled by a chemical reaction work much better than photons propelled by battery power.

If you could dump all of the energy out of, say, a AA battery really fast, though, could you get bullet levels of energy out of each battery?


Yes, you could.

Let's presume you're using nickel-metal-hydride AA batteries, which are somewhere between average-rifle-cartridge and average-pistol-cartridge in size. You can get a lot more current out of a NiMH or NiCd rechargeable than an alkaline or carbon-zinc battery, but, as you say, you still can't discharge them nearly fast enough for them to be useful replacements for firearm cartridges.

Even if you don't care whether the battery survives the experience, the biggest bang you can get out of a battery is the feeble "explosion" of a laptop battery. That may give you nasty burns if it happens literally on your lap, and shorted batteries have been responsible for the destruction of quite a few cargo planes, but batteries are no more than firecrackers compared with proper explosive devices.

Never mind that for now, though, let's just look at the energy content.

The most generally useful kind of NiMH cell is the "low self-discharge" type, which unlike the older kind of NiMH, do not go flat in a matter of weeks whether you use them or not. (Low-self-discharge cells are often sold as "pre-charged", or "ready to use".) LSD cells have lower capacity, though, so let's say we're using non-LSD cells with the absolute bleeding edge maximum capacity today available, which is about three amp-hours (3000 milliamp-hours).

1.2 volts (the standard NiMH or NiCd terminal voltage) times three amp-hours gives 3.6 watt-hours. A joule is a watt-second, there are 3600 seconds in an hour, so 3.6 watt-hours is 12,960 joules.

Firearm muzzle energy is often measured in foot-pounds, not joules, but I'll keep it all in SI units here. You also couldn't get the entire capacity of any electrical energy source into your beam or projectile, because no laser or mass-driver is 100% efficient, but I'll handwave that as well.

12,960 joules is a pretty darn respectable chunk of energy, way more than any handgun cartridge can manage. 9mm rounds top out around 500 joules of muzzle energy, .44 Magnum is a couple of thousand joules at most, and even the ludicrous .500 S&W Magnum is only around 4000 joules.

Rifle cartridges that qualify as "high-powered" seldom exceed 4000 joules. You have to start looking at exotic specialised sniper and large-game rounds, or heavy-machine-gun ammunition, before you get above ten thousand joules. The .50 BMG round easily beats 12,000 joules, and the more ludicrous kinds of elephant-gun double rifle roughly equal the battery's energy...

...as you'd bleeding well want them to, for this kind of recoil punishment.

But all of this is, again, just fantasy, because you can't dump the energy out of any kind of battery anywhere near fast enough to make it useful in a gun.

You can, however, dump the energy out of a capacitor in a very short period of time.

The highest-capacity "supercapacitors" can't be discharged in a tiny fraction of a second without damaging them; they're usable in a flashlight or for regenerative braking, but not for one lightning-strike discharge, as in a firearm.

Normal caps certainly can be discharged fast, though.

This cap bank could be used to power some kind of kinetic or ray-gun weapon - but it takes all of those huge beer-can capacitors to hold a mere 11.3 kilojoules, roughly the same as our one AA NiMH cell. The one killing the watermelon above is only 9270 joules, and its caps are huge.

I have one beer-can electrolytic cap of my own; it featured in this...

Ridiculous contraption

...extremely practical assemblage.

If it still had its full original capacity - which it doesn't - then fully charged to its 850-microfarad, 450-volt redline, it would hold 86 joules of energy.

You can get firearm cartridges that are that feeble, or even weaker, and I certainly wouldn't want to be shot with them. But I'd take them over a humble .38 Special or .22 Long Rifle any day.

The miserable performance-per-size of electricity-storing devices is why electromagnetic railguns are of moderate interest to navies...

...but not to armies.

It's also why the only really workable technology for a military laser gun (as opposed to lasers only used to temporarily or permanently blind the enemy) is the chemical laser.

Chemical lasers can be usefully powerful without requiring capacitor banks the size of a house. They are generally very unpleasant to be near, though, because they either run on, or produce, horrible toxic compounds.

Which is why plain old deflagrating gunpowder, propelling a piece of metal down a tube, remains the standard way to do unto others at a distance.

Oh - and if you've never watched Kaboom!, you really ought to.

UPDATE: On the subject of ludicrous electrical things, there's this piece I did on what a AAA battery composed of nothing but electrons would be like.

(It would not be kind to nearby spacetime.)

Psycho Science is a... sort of... regular feature here. Ask me your science questions, and I'll answer them. Probably.

And then commenters will, I hope, correct at least the most obvious flaws in my answer.

Falling from the friendly skies

How did Felix Baumgartner break the sound barrier by falling? I've always thought there was some kind of maximum velocity because of drag, even for someone trying to minimize it, in the vicinity of 200km/h.... Is this simply because the atmosphere is sparser up there? (Which would explain the bother about getting 42km off the ground when the max speed is reached after 40 sec.)

In the same vein, to what extent could an astronaut bail out of the ISS, Kursk-style?


(Everybody's seen the real video.)

The usually-quoted human terminal velocity - as you say, around 200 km/h for a skydiver in "star" pose, well over 300 km/h for a skydiver in a head-down pose with limbs tucked in - applies only to normal skydives, which don't start at such a high altitude that the divers even need supplementary oxygen, much less an actual pressure suit.

15,000 feet (about 4.6 kilometres) is a high jump altitude for a recreational skydiver. At that altitude atmospheric pressure is still above 50% of what it is at sea level. Unacclimated people won't be able to get much done at that pressure and will probably start feeling pretty miserable if they stay there for a long time, but if you're just sitting in a perfectly good aeroplane out of which you shortly intend to jump, it's not a huge problem.

5,000 feet (about 1.5 kilometres) is a much commoner skydiving altitude. At that altitude you've still got 80% of sea-level air pressure. The excitement of the impending jump will have much more effect on you, at that pressure, than the thinning of the air.

EDIT: As per ix's comment below, 13,000 feet is actually quite a common skydiving altitude for, as arkikol's comment explains, regulatory-loophole reasons.

(Katoomba, where I live, is about a kilometre above sea level, which is high for Australia; this country's pretty geologically inactive, so for a very long time erosion's been wearing the mountains down and nothing's been pushing them up. A thousand metres is still enough to drop atmospheric pressure to about 87% of that at sea level, though. I therefore get a very mild sort of altitude training any time I go to the shops, or take Alice the Wonder Dog, who needs more exercise than our friends who own her can quite manage to supply, for a walk.)

You need a pressure suit above the "Armstrong limit" (named for Harry George Armstrong, not Neil), which is the pressure where water boils at human body temperature. There is no way to survive for more than a minute or three above the Armstrong Limit, even if you've got pure oxygen to breathe.

The Armstrong limit is around 19.2 kilometres (about 63,000 feet above sea level, 2.2 Mount Everests), depending on the weather. Felix Baumgartner's Red Bull Stratos dive started from a little more than 39 kilometres above sea level.

At that altitude, the air pressure is about four thousandths of an atmosphere. That's 3.9 hectopascals, or 0.056 pounds per square inch. A home experimenter would be pretty pleased to own a mechanical vacuum pump able to pump down that low.

When the air is this tenuous, there is obviously not much air resistance to slow down a falling body. The terminal velocity of a skydiver (or a feather pillow, for that matter) will thus be far higher than it is for a human falling at normal skydiving altitudes.

The speed of sound in a gas, including air, depends on the gas's density, pressure and temperature. For the earth's atmosphere, this results in a rather odd variation of sound-speed with altitude, conveniently displayed in this graph I just ripped off from Wikipedia:

Speed of sound vs altitude

You can see that temperature is the major factor - the shape of the blue speed-of-sound line closely matches that of the red temperature line. This is because density and pressure decrease together with altitude, and cancel each other out.

You can also see, once again, that at 39 kilometres up where Baumgartner's dive started, there ain't much air left at all. The higher you go, the more perverse it therefore becomes to be concerned about the speed of sound at all, from the point of view of a skydiver.

Breaking the sound barrier at "normal" altitudes is a big deal. Even aircraft that only want to come vaguely close to the speed of sound, like jumbo jets, need special design features to prevent alarming things happening when they get above about Mach 0.75.

("Alarming things" include stuff like "the controls not working any more". Quite a lot of World War II airmen lost their lives when a power-dive pushed them fast enough that air was passing over certain parts of their aircraft at transonic speed. Some aircraft designs also helpfully went into a dive all by themselves if flown too fast.)

When the air's so thin that a paper plane would drop like a rock, though, all the same transonic shockwave stuff may be happening, but the forces involved are too feeble to worry about.

So yes, Baumgartner broke the speed of sound, but it wasn't that big a deal, because he was starting from so high up that he would probably have fallen at least a couple of hundred metres per second even if he'd opened his parachute the moment he jumped.

OK, on to bailing out from the International Space Station. This is problematic.

The ISS is in orbit, so if you jump out of it, you'll just be in orbit too. Whatever relative velocity you can give yourself with your legs will not be enough to make a significant difference. In order to actually fall into the atmosphere, you'll have to kill some of your orbital velocity with some sort of thruster - this is how spacecraft "de-orbit".

Let's presume you have a magical reactionless thruster doodad that lets you bring yourself to a halt relative to the surface of the earth directly beneath you, just as if you'd jumped out of a balloon that'd somehow made it to the ISS's altitude. Presumably you planned to further employ your reactionless lift belt or boots or whatever to float down majestically at whatever speed you wanted. But when you pressed the button to kill your orbital momentum, the device burned out, and now you're falling.

The ISS's low Earth orbit is about 400 kilometres above sea level. At that altitude, there's still enough of a trace of atmosphere to cause the ISS's orbit to decay by a couple of kilometres per month, so it requires frequent "reboosting" to stop it falling into the ocean ahead of schedule. From the point of view of someone who just told his fellow cosmonauts that he's just going outside and might be some time, though, it's a vacuum at 400 kilometres.

Low earth orbit is high enough that the Earth's gravity is somewhat attenuated, but only from about 9.8 metres per second squared to about 9.0.

So, starting at 400 kilometres and accelerating at nine metres per second per second, with both gravity and air density slowly rising as you fall. I don't know exactly how this'd work out, but I think that by the time you'd fallen 300 kilometres and passed the 100-kilometres arbitrary "start of space" altitude, you'd be falling at about 2.5 kilometres per second.

That's a pretty darn impressive speed, but it's much more manageable than actual orbital velocity. The ISS's orbital velocity is about seven kilometres per second; when the Shuttle Columbia broke up into flaming particles, it had managed to slow down to around six kilometres per second. Since energy increases with the square of the speed, an object travelling at seven kilometres per second that's trying to slow down has 7.8 times as much energy to get rid of as one travelling at 2.5 km/s.

2.5 km/s at a hundred kilometres altitude would probably be survivable, perhaps with some sort of ribbon parachute or similar drag device to bleed off speed steadily as the air got thicker.

But all of this is a bit silly, because it assumes that you've somehow managed to get rid of the several kilometres per second of your initial orbital velocity. That, there, is the big problem. If an orbiting spacecraft had enough reaction mass to kill its orbital velocity while it was still in space, it could then use wings or pop a gigantic parachute or three and sail down quite serenely, with no need for troublesome heat shields at all.

(This is why the Virgin SpaceShipOne and Two don't need heat shields. They're suborbital spaceplanes, not "real" spacecraft. They go very high by aircraft standards, then they fall back down again, never gaining or having to dispose of actual orbital velocity.)

Psycho Science is a regular feature here. Ask me your science questions, and I'll answer them. Probably.

And then commenters will, I hope, correct at least the most obvious flaws in my answer.

In New Zealand, the hobbits have them

I was just catching up on your Psycho Sciences (write more!), and read in the one about passive smoking that "very few [Australian] houses have basements", versus houses here in the USA, where we usually do have basements which, as you say, serve as convenient radon-accumulators for householders in need of a higher cancer risk.

Why is this? Is Big Radon Detector Business conspiring with architects to maintain demand?


Some of the differences in dwelling styles between countries are purely cultural, with no very logical reasons either way. Then there are obvious ones like "our houses are made of stone, 'cos there's no bloody wood for 500 miles". And then there are others that make only a very small amount of sense, like the almost complete absence of European-style heat-retaining technologies in Australian houses. This creates the peculiar situation that although a Sydney winter is not unlike a northern-European summer, Sydneysiders spend more time being cold in winter than Finns do, because Sydney houses are usually poorly heated, draughty, and often surprisingly poorly insulated too.

(This applies to Katoomba, where I live, as well. Katoomba winters are still a joke by countries-where-it-snows standards, but overnight temperatures around freezing point are still quite common in winter, as are tourists from elsewhere in Australia gazing in surprise at the ice on their car windscreens and wondering what on earth to do about it. Yet many houses here are built no differently from houses in much warmer beach towns, and their occupants suffer accordingly.)

Many house-design differences have a quite simple rational basis, though. Like, here in Australia it's easy to find houses with flat, or only gently inclined, roofs. In countries where it snows in winter, there's pretty strong selective pressure...

Carport snow collapse
(Image source: Flickr user HoundCat)

...against people who choose to live under a flat roof. Here, not so much, and a flat roof is simple and cheap to build.

The snow/no-snow roof design holds for most countries. In very hot areas, a flat-roofed house can also be built to let you sit out on the roof in the breeze of an evening.

The basements/no-basements thing has a rational basis, too, which once again holds for a large number of countries.

Digging a big hole under a house is time-consuming and expensive, and asking for trouble from water seepage. If you want that much extra space in the building, and are not living in some godforsaken wasteland that's shaved flat by tornadoes every ten years, it's faster and cheaper to just build a taller house.

Unless it gets cold enough in the winter for the ground to freeze to a significant depth, and then warm enough in the summer for it to thaw again.

If that's the case, then "frost heave" (which, coincidentally, is yet another thing Matthias Wandel has had to deal with) will slowly push anything sitting in the freezing and thawing earth up into the air.

Frost heave makes shallow house foundations a terrible idea. So you have to dig a hole to put the foundations in, and you might as well make that hole into a basement while you're at it.

Some basement-equipped houses are built in places where frost heave never happens, for the abovementioned cultural reasons. But you generally find them in cold-winter locations.

In Australia, you usually have to build to minimise the effects of heat, not cold. This has given rise to the famous underground houses of Coober Pedy, and the much more common "Queenslander" house style...

Queenslander house
(Image source: Flickr user Crazy House Capers)

...where the house proper sits as much as a whole extra storey high on "stumps", to catch the breeze and keep Queensland's trillion species of house- and/or human-eating arthropods that much further away.

Psycho Science is a regular feature here. Ask me your science questions, and I'll answer them. Probably.

And then commenters will, I hope, correct at least the most obvious flaws in my answer.

Great sound, shame about the cancer

A reader writes:

I came across your F5 speaker article...

Kit speakers

...and was very impressed with the information provided.

Is it true that glass fibre batts in speakers can cause mesothelioma? I noticed you linked to the Wikipedia mesothelioma page when discussing polyester and glass fibre batts with the word "carcinogenic".


A ported speaker with glass-fibre wadding inside (it's there to dampen internal resonances) will spit little bits of fibreglass out of its ports in normal use. The cancer risk from these is essentially nil, mainly because the amount of glass emitted is very small. But even larger glass-fibre exposures are generally less dangerous than asbestos exposure, and there's some debate about why this is.

Fibreglass and asbestos are mechanically, and somewhat chemically, similar. Glass-fibre, like window glass, is about three-quarters silicon dioxide (quartz), with the rest being additives, chiefly oxides of light metals, to reduce the glass's melting point and improve its strength and/or chemical properties. All forms of asbestos are essentially silicate minerals too, but with different elements mixed in with the silicon and oxygen.

The most common form of asbestos is the white "serpentine" kind, which is magnesium silicate. Blue and brown "amphibole" asbestos are closer to window-glass, being complicated sodium, magnesium and iron silicate minerals.

Asbestos is so particularly nasty (and useful) because its fibres can be very, very fine, routinely below twenty micrometres (or microns) in diameter, and even down to small fractions of a micrometre, versus around 100 micrometres for a human hair. These ultra-fine fibres are too small to even see, and bits of them can float around in the air waiting to be inhaled. This is why they tent whole buildings and put workers in moon-suits to do asbestos abatement; building materials that contain asbestos can be safe to be near, but as soon as you start busting those materials up, they can produce dangerous and invisible dust.

Fibreglass, also known as glass wool, is made in a similar way to fairy floss ("cotton candy", in the USA); extrusion of the molten material through tiny nozzles. The nozzle size determines the thickness of the filaments, so glass fibres can quite easily be made down to single-digit-micron thickness. As is the case for many other "whisker" materials, most of the desirable physical qualities of the fibres increase as the thickness drops. (This is explored in some detail in J.E. Gordon's classic "The New Science of Strong Materials, or Why You Don't Fall through the Floor", a book that I may not have mentioned on this site for as much as eight or nine minutes.)

Concrete dust and asbestos and glass fibres

This electron micrograph of dust from the wreckage of the World Trade Center (via the USGS) shows a thin glass fibre and a bundle of much thinner asbestos fibres.

(Nobody performed full asbestos abatement on the WTC towers while they were standing, because it would have been very expensive, the asbestos was largely safely bound up in building materials, and nobody expected the buildings to fall down.)

Glass fibres down in the single-digit-micron diameter range are a cancer risk, like asbestos, but glass fibre in general seems to be rather less carcinogenic than asbestos fibres of similar dimensions. Nobody's exactly sure why. Glass fibres don't seem to get stuck in the lungs like asbestos fibres do; this could be purely because of the size difference, or because they don't have the same rough, almost barbed sides...

Asbestos fibres

...as many asbestos fibres (that's another USGS picture).

Glass is also slightly soluble in water, and - it is theorised - fibres in the lungs can thus be slowly eliminated via blood or sputum. For macroscopic objects the water-solubility of glass is essentially zero; you can run water through a glass tube in a laboratory for years with no visible change, and you don't need to worry about rain wearing through your windows. But the thinner the fibres, the greater will be the surface area of those fibres relative to their volume. So even extremely slight solubility can, the theory goes, get rid of the fibres usefully quickly.

It's important to realise that we're not just talking about cancer, here. Most people with asbestos-related lung disease don't have mesothelioma; they've got "asbestosis", a non-cancerous inflammatory disease which can, nonetheless, very effectively destroy your quality of life and in extreme cases kill you. Again, it's the super-fine fibres of asbestos that make it particularly nasty here, but you can get similar syndromes by inhaling various other particulate substances that get stuck in your lungs, like coal dust, and also little bits of fibreglass.

Realistically, even someone who stuffs fibreglass into speaker boxes for a living, without so much as a face mask, isn't at a huge risk of lung disease - cancer, or "just" an asbestosis-like condition. Usually it's people like surfboard manufacturers or insulation installers who get sick, and then only if they don't use a respirator while they're sanding boards or stuffing insulation batts into unventilated roof or floor cavities.

So if you've got ported speakers with fibreglass in them, don't worry about it. Even if you open the speakers up to replace a blown crossover or something, you're in no real danger. (And if you've got un-ported, sealed "infinite baffle" speakers, there is of course even less risk.)

The loudspeaker industry switched to using cellulose acetate or polyester fluff...

Speaker kit parts

...as the normal low-cost anti-resonance speaker-lagging material...

Inside of small speaker enclosures

...some time ago, but I think the change was mainly because these fibres are easier to cut and place, and not a prickly skin irritant, rather than for health reasons.

Incidentally, white asbestos and talc, as in talcum powder and...

Sticks of French chalk and holder

..."French chalk", are chemically the same, both magnesium silicate. Asbestos can metamorphose into talc, though I'm not sure if talc can go the other way. In any case, industrial-grade talc can be expected to contain some asbestos-like filaments.

This fact caused a certain amount of panic among people who've applied talcum powder liberally to their baby, or discovered that standard children's wax crayons contained a small amount of talc, which in turn did or did not contain - depending on who you asked - a tiny amount of actual identifiable asbestos.

(Some people went so far as to allege that because they're chemically the same, talcum powder is asbestos. By this logic, it should be easy to drive nails with a Brillo pad. But have no fear, highly independent thinkers stand ready to help you remove asbestos from your spine with magnets!)

In response to this, more than a decade ago the big-brand crayon formulations were changed to contain no talc. How crayon-talc was supposed to get into kids' lungs in the first place, I'm not sure. Embedding asbestos in wax strikes me as an excellent way of rendering the stuff harmless, even when kids stick crayons up their noses. Perhaps some asbestos could lodge in the digestive tract if they ate it, but I think the normal regeneration of the gut lining would carry it away. Unless you actually lit a crayon fire, I doubt any significant exposure was even theoretically possible.

The asbestos-in-talcum-powder scare was more rational, because people unquestionably do inhale some talcum powder when they use the product in normal everyday ways. There's a difference between bulk industrial talc and the super-fine stuff used for talcum powder, though. Major talcum-powder companies hotly protested that there was no asbestos in their talc at all.

It's rational to take at least some care to prevent you or your baby from inhaling talcum powder, because, as discussed above, inhaling fine insoluble powders in general is a bad idea. But there doesn't seem to be any serious reason to boycott the product entirely.

(There's also a popular belief that laser-printer or photocopier toner is deadly poisonous if inhaled. Actually, toner is just yet another insoluble fine powder. So, once again, you should avoid inhaling it if you can, and wear an appropriate mask if you have chronic exposure to it. But there's no need to panic if you snort a little of the stuff by accident.)

Psycho Science is a regular feature here. Ask me your science questions, and I'll answer them. Probably.

And then commenters will, I hope, correct at least the most obvious flaws in my answer.