Serious Scientific Answers to Absurd Hypothetical questions

[sevenperforce]

Paradoxically, a physically smaller bomb requires a greater amount of fissile material.

[Jovus]

On 11/27/2017 at 4:22 PM, Jovus said:

Just for funsies:(

Gravitational force at the surface of the Earth (in 'disk' configuration):

GMm/r^2

M = 5.972 × 10²⁴ kg

Now then, the average density of the Earth remains unchanged at 5.51 g/cm³ or 5.51 x 10^-9 kg/m³ (because I said so)

Assuming the disk is a smooth disk with h = 10m, that makes V_disk = 2 πhR_disk = 3.29 m³, so R_disk = 5.237 x 10¹⁴ m²

G = 6.67408 × 10^-11 m³ kg^-1 s^-2

so F (on a mass at the 'rim' of the Earth) = m*(1.45m/s²)

 

Or 1.45m/s²) is our new 'g'

To be continued when I get home/feel like it

Now continued.

 

First, found a couple stupids. We assume the density of the Earth, and therefore the volume of the Earth, is unchanged. Modeling the Earth as a perfect sphere of uniform density (close enough for our purposes), we find:

V_Earth = (4/3)*π*R_earth³= (4/3)*π*(6.371 x 10⁶ m)³ = 1.083 x 10²¹ m³

Now we can say V_disk = V_Earth = 1.083 x 10²¹ m³ = 2πhR_disk² , or R_disk = sqrt((1.083 x 10²¹ m³)/2π(10m)) = 4.152 x 10⁹ m which is a lot more reasonable.

so the force on a mass m at the surface of the disk (on the outer edge, not on one of the faces) is F_grav = GMm/R_disk² = m*(2.3 x 10^-5 m/s²)

So our new 'g' (acceleration by gravity at or near the surface) is 2.3 x 10^-5 m/s²

Putting that into perspective, g_Pluto = 0.62 m/s² and g_Gilly = 0.049 m/s², though I find those numbers suspect. (The Pluto-Charon system has a barycentre outside itself, so g_Pluto necessarily depends heavily on where you are on the planet, and Gilly is lumpy and small, so the same can be said of it.)

Since the Earth didn't lose any mass in the transition, angular momentum should be conserved. Since the disk is spinning through the same axis, and it's also a prinicpal axis of the disk, the calculations for moment of inertia are simple:

L_Earth = I_Earth * ω_Earth

L_disk = I_disk * ω_disk

L_Earth = L_disk so

I_Earth * ω_Earth = I_disk * ω_disk

Moment of inertia of a sphere is 0.4*mr², so I_Earth = (0.4)*(5.972 × 10²⁴ kg)*(6.371 x 10⁶ m)² = 9.696 x 10³⁷ kg*m²

Moment of inertia of a disk is 0.5*mr², so I_disk = (0.5)*(5.972 × 10²⁴ kg)*((6.371 x 10⁶ m)² = 5.148 x 10⁴⁸ kg*m²

T_Earth =  1 day = 86400 sec, so ω_Earth = 2π/86400sec = 7.272 x 10^-5 rev/sec

I_Earth * ω_Earth = I_disk * ω_disk so ω_disk  = (I_Earth/I_disk) * ω_Earth = ((9.696 x 10³⁷ kg*m²)/(5.148 x 10⁴⁸ kg*m²)) * (7.272 x 10^-5 rev/sec) = 1.370 x 10^-10 rev/sec

Which means that a day on the disk is roughly twice as long. (hah! That's very wrong. But 'twice' is revealing to the way astrophysicists think, so I'm leaving that in.)

T_disk = 2π/ω_disk = 2π/(1.370 x 10^-10 rev/sec) = 4.587 x 10¹⁰ sec, or 5.31 x 10⁵ Earth days.

Now then, the force from gravity at the rim of the disk as shown above is m*(2.3 x 10^-5 m/s²) (where m is the mass of a particle on the rim), but

centrifugal force on a spinning object is -m*ω x (ω x r). The cross-products, on a disk spinning through the centre around the z axis, resolve to radial out, and

F_cent = -m*ω_disk²*R_disk = m*(1.370 x 10^-10 rev/sec)²*(4.152 x 10⁹ m) = m*7.790 x 10^-11

Compared to F_grav = m*(2.3 x 10^-5 m/s²), we can say that this object wouldn't spin itself apart.

So much for simple mechanics. I'll leave worrying about interactions with other bodies, weather, effect on life, thermodynamic effect of injecting enough energy to accomplish this feat, etc. to other minds.

[TheSaint]

On 11/23/2017 at 7:17 AM, Zeiss Ikon said:

because what space libertarian would go anywhere without a gun?

This sentence made my day.

On 11/29/2017 at 4:18 AM, ARS said:

Could we possibly make a nuclear grenade? Aka a nuke that can fit inside your hand, can be thrown like a grenade, and having the "mushroom cloud" explosion the size of grenade blast, and still leaves radiation in a relatively small area compared to tac nukes

Probably not, at least with our current technology. The W54 was the smallest (declassified) warhead ever produced. (There have been rumors of smaller warheads being produced, both in the US and the USSR, but I've never seen any concrete evidence of that.) It weighed 51 pounds, which puts it in the "man-portable" category (Which lead to the creation of the Special Atomic Demolition Munition, essentially a nuclear satchel charge). But I really couldn't see you throwing that, at least by hand.

The problem is that you need your fissile material to reach critical mass. There are a lot of ways to do that in a controlled manner that doesn't endanger your bomb thrower, but they all tend to make the weapon a minimum size. There's a fairly decent discussion of those problems here. Then, if you are planning on being healthy enough to celebrate your victory, you probably want to shield the device as well, which is going to add size and weight. It just doesn't seem likely that you'd be able to make a fission device that small.

Now, if you're going to start talking about exotic technologies, then things may start moving forward. Some kind of miniaturized micro-fusion could conceivably put a 1-ton yield micronuke in your hand. But at that point you're kind of just making things up as you go. Just make sure you have a good pitching arm.

I would think that if you were going to go to all the trouble of creating such a device, putting it in a grenade would be a waste. (And potentially lethal to the thrower.) It would be a much better option to put it in a tactical missile, something like a Javelin. It would stand a better chance of getting what is probably a very expensive warhead onto its intended target, and getting the warhead far away from friendly troops before it detonates.

[DAL59]

How much antimatter would you need for an antimatter bullet (a special, very expensive round that would be a normal bullet with antimatter magnetically contained inside) that could fit inside a handgun to destroy a building?  

[ARS]

23 minutes ago, DAL59 said:

How much antimatter would you need for an antimatter bullet (a special, very expensive round that would be a normal bullet with antimatter magnetically contained inside) that could fit inside a handgun to destroy a building?  

Not much actually

[Mayer]

Uhm, 1 gram of antimatter is the equivalent of ~43kt of TNT or about 3-times the Hiroshima bomb,  slightly overkill..

In controlled demolitions, structures are weakened with only small charges of ammonium nitrate which has 1/2 the power of TNT.

1nanogram antimatter (1/10 global production) should be like 5t TNT which is enough to kill skycrapers, 1 microgram is therefore 5kg TNT which should destroy everything within 25m.

Edited December 1, 2017 by Mayer

[Diche Bach]

Here is one I came up with on the C++ fun times Discord I hang out on. Since the first electronic computer were fired up in the 1940s, how many instructions have been processed? Is it even possible to come up with such an estimate? What is the rate of acceleration in the increase in number of instructions processed by time? At what point will more instructions be processed during one discrete and relatively short time span (say one year or maybe even one month or less) than during the entire preceding history of computing? At what point will the number of instructions exceed the estimated number of fermions in the universe?

(I suppose one could consider the fire-control types of machines that existed since the 1910s, or even the mechanical arithmetic machines that go back to the late 18th century to be "computers" but I'm not sure if they fit the question, so lets just limit it to the ones that used general purpose encoding schemes and electrical states, i.e., the 1930s or 1940s and onwards).

When I initially posed this question on the board, I had just been dazzled with some of the inner workings of 3d graphics, the sheer numbers of putPixel calls involved in a mere few seconds of a typical 3d graphical animation, and I opined something like "I bet the total number of instructions that have occurred since the first computers were fired up in the 1940s already exceeds the number of particles in the universe. The master programmer of the board said "not even close" and came up with an estimate that he thinks was probably even still an overestimate of the total number of instructions so far, and that was still many orders of magnitude less than the 10^80 number of fermions he quoted.

Edited December 1, 2017 by Diche Bach
elaboration

[kerbiloid]

1 hour ago, Diche Bach said:

Since the first electronic computer were fired up in the 1940s, how many instructions have been processed?

Probably you can take just last ten years. All before this was just a statistical error, an equipment warming.

1 hour ago, Diche Bach said:

What is the rate of acceleration in the increase in number of instructions processed by time?

Doesn't matter anymore. In several decades it will probably uncontrollably grow as an avalanche, so any estimation could be only momentary and predict nothing.

1 hour ago, Diche Bach said:

At what point will more instructions be processed during one discrete and relatively short time span (say one year or maybe even one month or less) than during the entire preceding history of computing?

Once the solid and/or mechanical storages begin getting widely replaced with molecular and (quantum? field?) storages.

1 hour ago, Diche Bach said:

At what point will the number of instructions exceed the estimated number of fermions in the universe?

Long before this happens, post-humans will realize that the Universe is a self-emulating quantum supercomputer, and fermions are just standard data packages of its low-level protocol.
Also that its CPU runs in arbitrary number of "times" at once, so there is no "when".

[Green Baron]

14 hours ago, Jovus said:

*snip*

I admire your eagerness :-) I should add that irl a disk of that mass cannot exist, it'll cramp into a sphere and/or the outer parts fly away if it rotates. Or the laws of gravity must be changed for some of the assumptions to work, they partly stay intact for others.

I have the following examples: Density of the sphere cannot be the same as that of a disk as the pressure is missing. Iron in the earth's core is compressed to >13g/cm³. Also angular momentum of a disk is different than that of a sphere as more mass is out and away from the center. The mass of a sphere can, for ease of calculation, be regarded as being concentrated in a point, for a disk that doesn't work.

Yeah, geo-discodynamics wouldn't work AT ALL, as NONE of the forces that drive them on and inside of earth exist on a disc. Beginning with convection inside, going over coriolis force and not stopping at insolation. None of the observations would be explainable with a disc-concept if we keep the laws of physics like they are.

 

Edit: moment of inertia of a disk is m*r²/4 ? Or did i get something wrong ?  I didn't: you took the moment in z-axis, but it should be in x/y-axis as you want to calculate the forces on the rim of the thing. Also, i am not sure if assuming the disk as being represented by a point mass is correct in this context. It's really not ... ?

Edited December 1, 2017 by Green Baron

[wumpus]

12 hours ago, kerbiloid said:

Probably you can take just last ten years. All before this was just a statistical error, an equipment warming.

In 2006 Intel introduced the "core" microarchitecture.  Current Intel chips strongly resemble it and it probably ran about half as fast (flat out) as current chips doing the same.  2000-2007 will probably include a fairly significant error to "just the last 10 years", but before that Moore's Law* had to be obeyed.  On the other hand, if you count individual "word" actions used for 3d graphics calculations in GPUs, the last 10 years should be fine (and nearly all of them done in GPUs).  Part of that is that the GPUs are that much more powerful, but also because the CPUs are simply waiting for a human to ask them to do something (typically redraw another keystroke), and games simply run the GPU (and somewhat the CPU) flat out.

* Moore originally claimed in his law that the amount of transistors per ship would exponentially increase.  He had to change the constants on the exponents twice, but the law outlived him as flash memory is still happily scaling away, admittedly by stacking transistors on top of each other instead of making them smaller.  Don't expect computers to become more powerful merely by making silicon transistors smaller though.

[kerbiloid]

45 minutes ago, wumpus said:

2000-2007 will probably include a fairly significant error to "just the last 10 years", but before that Moore's Law* had to be obeyed. 

Just add "25..50%" to the result of the "last 10 years."
I believe, this would cover everything since abacus.

[Diche Bach]

So what you guys are saying . . . is that we big-brained apes, have already, unwittingly, created a multi-verse in which we are likely lost

 

[Jovus]

On 12/1/2017 at 4:46 AM, Green Baron said:

I admire your eagerness :-) I should add that irl a disk of that mass cannot exist, it'll cramp into a sphere and/or the outer parts fly away if it rotates. Or the laws of gravity must be changed for some of the assumptions to work, they partly stay intact for others.

I have the following examples: Density of the sphere cannot be the same as that of a disk as the pressure is missing. Iron in the earth's core is compressed to >13g/cm³. Also angular momentum of a disk is different than that of a sphere as more mass is out and away from the center. The mass of a sphere can, for ease of calculation, be regarded as being concentrated in a point, for a disk that doesn't work.

Edit: moment of inertia of a disk is m*r²/4 ? Or did i get something wrong ?  I didn't: you took the moment in z-axis, but it should be in x/y-axis as you want to calculate the forces on the rim of the thing. Also, i am not sure if assuming the disk as being represented by a point mass is correct in this context. It's really not ... ?

Well, yeah, the hypothetical is absurd. It would definitely condense into a sphere; that's in fact what actually happened according to our current models. (Though it may take a good long while on a human time-frame.) However, I should point out that the whole point about centrifugal force being lower than gravitational force was to see whether this construction would immediately fly apart at the rim, which it wouldn't due purely to mechanical forces of the object itself. (In 'reality', of course, perturbations from the other celestial bodies would take care of it pretty quickly.)

As to the moment of inertia: no, you want the moment of inertia around the z-axis. Moment of inertia is relative to the axis of rotation, and the axis of rotation for this disk is through the z-axis (that is, I took it as rotating like a top or the disk in a gyroscope, instead of like a coin spun on a table).

As for the rest, I was explicitly not treating of anything other than the barest, simplest-order mechanics. You can if you like. I admit there would be other effects; I just don't care enough about the question to work them out.

[DAL59]

At what point in history would the world's combined computers be able to run lowest setting ksp?  

[bigcalm]

On 30/11/2017 at 7:31 PM, Jovus said:

Now then, the average density of the Earth remains unchanged at 5.51 g/cm³ or 5.51 x 10^-9 kg/m³ (because I said so)

Just a minor nitpick - 5.51 g/cm^3 == 5.51 x 10^3 kg/m^3

Might be why you're out further on, though I couldn't see you using density anywhere.  Also radians/s is the unit of omega (rotational angular velocity) not rev / s

[Jovus]

Oi. Yes, it is.

Density was used to find the radius of the 'disk' (cylinder). So the radius is off by a factor of 10³, which means acceleration due to gravity on the surface of the sphere is off by a factor of 10⁶, and centrifugal force is off by a factor of 10³; but still, the disk won't fly apart under the force of its own rotation.

[Mayer]

On 6.12.2017 at 9:03 PM, DAL59 said:

At what point in history would the world's combined computers be able to run lowest setting ksp?  

I reckon quite early. The paper "The World’s Technological Capacity to Store, Communicate, and Compute Information." by Hilbert, M., & López, P. published in the Science magazine, puts the number of general-purpose compution of the year 1986 to be 0.06 MIPS(Million Instructions per second) per capita. There were around 5 billion people in the world, which means 300,000,000 MIPS. The average person also had access to 539MB of storage space, so also 2,695,000,000 GB in total..

In contrast KSP minimum requirements specify a Core 2 Duro processor (which can do about 27,000 MIPS), 3GB RAM, 512 MB VRAM  and 1 GB of HD storage space.

[Zeiss Ikon]

On 12/6/2017 at 3:03 PM, DAL59 said:

At what point in history would the world's combined computers be able to run lowest setting ksp?  

All the world's computers combined?  Well, let's see -- equivalent to a fast Core2Duo, say 2+ GHz, or (as noted previously) around 27,000 mips.  Honestly, there was probably more mainframe capacity than that in just the United States before 1980.  Banks had gone heavily into computers starting as early as the 1970s.  Of course, there were the mainframes at NASA, NOAA, and heaven only knows what the CIA and NSA had then -- but I'd bet it was at least equal to NASA.  Every university of consequence had a mainframe or two.  When I took Fortran in 1979-1980, we had an IBM 360 and there was an Amdahl 470 at the neighboring school eight miles (13 km) away -- and I was at University of Idaho.  So, even if we ignore the mini-computers and early micros (there were never that many of the latter anyway), US installed mainframes alone likely exceeded the required mips count and RAM size for minimum KSP by 1980.

[Green Baron]

Floating point power may be thee limiting figure. That had to be emulated then, became widely used as late as the 80s. Today graphics are based on it.

[kerbiloid]

Probably we would have a look at KSP 0.7 system requirements. The system should definitely support that version of Unity.

 

Was the music sent on Voyagers protected with DRM?
How would we stop the ET from copying it?

[Zeiss Ikon]

7 hours ago, kerbiloid said:

Probably we would have a look at KSP 0.7 system requirements. The system should definitely support that version of Unity.

 

Was the music sent on Voyagers protected with DRM?
How would we stop the ET from copying it?

As I recall, the actual music on the Voyager disks was analog, and video data encoded in analog form.  The package even included a phonograph cartridge, to ensure the aliens would have a suitable reproduction device (turning the record at the correct angular rate, encoded on the package, and electronic processing of the signal output from the cartridge was left as an exercise).

I wonder if there was any thought  given to correct pre-amp characteristics?  Some years ago, I hooked up my magnetic-cartridge turntable directly to the microphone jack on my sound card (because, with no pre-amp in the turntable, the signal wasn't strong enough for line-in), and the resulting music was extremely tinny.  Turns out vinyl disks (at least in the post-1970 stereo era) were equalized to compensate for pre-amp response curves that heavily favored bass.  I was later able to apply a digital filter to weaken the high frequencies, in emulation of a suitable pre-amp, but that music was painful to listen to before then.

Edited December 9, 2017 by Zeiss Ikon

[Diche Bach]

Watched a video recently on project Azorian on Amazon video and it was quite good. Definitely the type that you engineers and scientists might enjoy. Anyway, I ALSO watched "Ghost Armies" recently, and that one was about an American deception unit in WWII European theatre. It made mention of a technology I had never even heard of: Wire recording.

And so, my absurd question: what would be the practical implications if all the worlds recordings were transferred to wire recordings?

If that one doesn't have any traction then how about one from Azorian: what if every metal hulled vessel which sank and remains capsized in the last 150 years were recovered from the ocean floor? How much would it cost? How much scrap/science/benefit would it garner?

[Zeiss Ikon]

Wire recording was, of necessity, monaural.  There was never a way to keep a wire aligned so as to put two audio tracks on it, and a dual-wire system would be impractical (twists and desynchronization as the wires stretched or broke and were spliced).  As a result, it was also possible to accidentally play a wire backward, as there was no simple way to tell the beginning from the end (there was a generic takeup spool on the machines, but if you have a wire on the same kind of spool, you have no way to know if it just didn't get rewound).

Some of these issues would be solved by emulating the earliest video recording technology, which used a metal ribbon (very hazardous to be around -- the fast-moving ribbon would slice like a razor blade if it came off the spools or slots, and moved fast enough that unspooling was a real issue).

Aside from loss of stereo separation (most like you'd mix everything down to mono during re-recording), you'd wind up with an LP or CD replaced by a wire spool about the size of a large roll of electrical tape.  That was the one hour size, though some older LPs would fit on a half hour spool, and a few maximum length CDs wouldn't quite fit on the one hour size.  A "single," needing only about five minutes of wire, would fit on a spool hardly bigger than an old style silver dollar (roughly twice the diameter and half again as thick as a modern golden dollar), if a little thicker.  The wire was somewhat fragile (because very thin and made of soft steel, so it would accept magnetization easily); it could break during either recording or playback.  Splicing was tricky, and produced a very audible "bump" in the recording, as the twisted splice ran over the playback head and the magnetized wire got closer and further from the head as the splice passed.  Both record and playback heads were subject to significant wear as the steel wire slid over the head surface -- I don't know what their working life was, but it can't have been comparable to that of a magnetic tape recorder/player, where there's a very thin, very smooth overcoat protecting the heads from the metal oxide (or, in a few ultra-premium tapes, sputtered metal) recording material.

An afternoon in the sun, on a car dashboard or similar, wouldn't harm a wire recording at all (unlike vinyl records, plastic cassette shells, or CD cases and disks; wire spools were usually stored in card stock boxes), but one would have to take care to store them in dry conditions, so the wire wouldn't rust.  Rust on the wire wouldn't have much immediate effect on the recorded sound (the rust is magnetic, too, and should pick up the field from the parent metal as it forms), but iron oxide is abrasive, so would further accelerate wear on the recording and playback heads.

[ARS]

150 years is a very long time. Things got claimed by marine life VERY FAST. Salvaging the sunken ships for metals is not worth the price and effort. The metal would be barely recognizable and poor quality because of rust and getting rid the corals that grew around it is simply too much hassle when all you want to get is just metal. The most that you can get is science, but that can be done without lifting the whole ship to the surface. Sunken ship becomes an underwater ecosystem for a variety of marine life. Lifting that ship isn't gonna be a good idea, since you potentially destroying that ecosystem. Aside from that there's an unwritten rule for sailors to simply left the sunken ship underwater and not disturbing it to honor those who died aboard the ship and sunk with her

[Spaced Out]

If you sat in an office chair and fired an AK-47 that was placed so the thrust from recoil would be directed through the Center of Mass, how far would you move per shot?