He who radiates is lost

4136 views
Skip to first unread message

Brett Evill

unread,
Sep 4, 1998, 3:00:00 AM9/4/98
to
So, what are plausible technologies and tactics for space combat?

Missiles/torpedoes will be vulnerable to ECM, I guess. Will it be possible
to have guided-missile frigates in space with buckets of ECCM, guiding
their torpedoes by remote control?

Will it be worth attempting stealth technologies to avoid observation, or
is IR always going to give you away? How conspicuous will spaceships'
drives be at range?

Will spaceships be painted black to prevent telescopic detection, or
mirrored to bounce laser-beams? Perhaps ships with heat-superconducting
armour will not be susceptible to significant damage by laser beams,
because they will have large radiating surfaces. Will that protect them
from particle beams?

How effective are nuclear warheads going to be in space? Sure they won't
do blast damage, but neutrons and gamma-rays might still be effective.
What about nuclear claymore mines? Nuke-pumped grasers?

Are KE weapons going to be effective, or is guidance and delivery going to
cause problems? A ship might kill itself running into a crowbar, but how
easy is it to get a crowbar in front of a ship?

What ranges are plausible for engagement? Are these long enough that
random evasion will be effective against beam weapons?

--
Brett Evill

To reply by e-mail, remove 'spamblocker.' from <b.e...@spamblocker.tyndale.apana.org.au>

Erik Max Francis

unread,
Sep 4, 1998, 3:00:00 AM9/4/98
to
Brett Evill wrote:

> Missiles/torpedoes will be vulnerable to ECM, I guess. Will it be
> possible
> to have guided-missile frigates in space with buckets of ECCM, guiding
> their torpedoes by remote control?

Not necessarily by remote control, but your torpedoes will have to be
clever to make it to the target.

> Will it be worth attempting stealth technologies to avoid observation,
> or
> is IR always going to give you away? How conspicuous will spaceships'
> drives be at range?

If you're talking about something like a fusion feed drive, then
basically whenever you fire your engines you'll be visible across the
whole system. Lay-in-wait stealth will probably be very important, but
once you make your move you will be a sitting duck. Which means you'll
only move or open fire when you're assured of a kill.

> Will spaceships be painted black to prevent telescopic detection, or
> mirrored to bounce laser-beams?

It's doubtful that the coating of the ship is really going to be the
problem, since what's _really_ going to give you away is lighting your
drive or firing ordnance.

I don't know if it would be _mirrored_, though, since that's almost
doing your best to be seen. Perhaps some kind of corner-mirrored
surface which is designed to ablate easily, and still stay somewhat
shiny. Underneath this reflective armor would be normal armor to try to
deflect detonations or kinetic damage.

> How effective are nuclear warheads going to be in space? Sure they
> won't
> do blast damage, but neutrons and gamma-rays might still be effective.

Not terribly, but they're still good for something. A near miss will
send the ship reeling, and a contact hit will certainly take care of the
ship.

In space combat it pretty much looks like the first to get a hit wins.
This isn't really surprising; it's true of most combat these days (air
combat, submarine combat, etc.). The weapons will be devastating enough
that if you land a hit it does so much damage that the ship, even if it
has survived, is out of the fight.

> What about nuclear claymore mines? Nuke-pumped grasers?

Mines are not terribly likely, unless you're talking about some sort of
lay-in-wait device that launches itself at intruders. Nuke-pumped gamma
ray lasers are theoretical.

> Are KE weapons going to be effective, or is guidance and delivery
> going to
> cause problems? A ship might kill itself running into a crowbar, but
> how
> easy is it to get a crowbar in front of a ship?

Kinetic weapons have the disadvantage of that they're unguided, but have
an advantage in that their unguided, as well; they can't be
countermeasured. Since they don't have their own exhaust trail, it also
means you can't easily see them coming quite as easily as a drive-guided
missile or torpedo.

Further you can have a number of different kinds of kinetic weapons,
including the equivalent of shot (many small pebbles), or even
projectiles intended to deform into large, mostly-connected ring shapes,
for instance.

In normal ballistics, the goal is to deposit all the bullet's energy
into the victim, because that way maximum damage is done. That probably
won't be the case with space combat with kinetic weapons; you want the
projectiles to move fast so that they can't be dodged, and plus a clean
puncture through gives you two breathing-air holes for the price of one.

> What ranges are plausible for engagement? Are these long enough that
> random evasion will be effective against beam weapons?

Here's one way to look at it -- ships that engage in combat within easy
beam weapons range won't be in combat for very long.

My thinking is that it will be very much like submarine warfare; ships
slink around, looking for an enemy. When they find one, it becomes a
game of cat-and-mouse until one slips up. Then kinetic or missile
ordnance is fired, and when a hit is landed, the winner closes in for a
kill with beam weapons.

That's the way I see it, anyway.

--
Erik Max Francis / email m...@alcyone.com / whois mf303 / icq 16063900
Alcyone Systems / irc maxxon (efnet) / finger m...@sade.alcyone.com
San Jose, CA / languages En, Eo / web http://www.alcyone.com/max/
USA / icbm 37 20 07 N 121 53 38 W / &tSftDotIotE
\
/ Sanity is a cozy lie.
/ Susan Sontag

Joe

unread,
Sep 4, 1998, 3:00:00 AM9/4/98
to
On 4 Sep 1998 14:12:39 GMT, b.e...@spamblocker.tyndale.apana.org.au
(Brett Evill) wrote:

>So, what are plausible technologies and tactics for space combat?
>

snip

All of this was thrashed out pretty thoroughly about 5 or six months
ago. Here Check Deja News for 'realistic space battles' thread.

I reckon the conclsuion was that stealth was worthless except under
very special circumstances under any plausible near future technology.

As for what weapons well that depends on your technology assumptions

regards
Joe

--
***********************************************************************
Since my earlier anti spam measure did not work]
Hitting the reply button will not work. My email address is jdineen at
iol dot ie

Tommy the Terrorist

unread,
Sep 5, 1998, 3:00:00 AM9/5/98
to
In article <b.evill-0509...@tynslip1.apana.org.au> Brett Evill,

b.e...@spamblocker.tyndale.apana.org.au writes:
>Will spaceships be painted black to prevent telescopic detection, or
>mirrored to bounce laser-beams? Perhaps ships with heat-superconducting
>armour will not be susceptible to significant damage by laser beams,
>because they will have large radiating surfaces. Will that protect them
>from particle beams?

Definitely mirrored.

1) A black ship is a "blackbody", which is to say, unless your crew likes
life at 3 Kelvin, it's going to glow a cherry red to anybody with IR
accessories.

2) "heat superconduction" has been claimed here to be a myth of Niven's
unrelated to reality, and it seems believable. Heat by its nature is
distributed between all sorts of random vibrational modes - how can it
move QUICKLY, without pausing or getting sidetracked?

3) As another thread pointed out, a black ship is a COLD ship, since it
loses all that energy it's busy radiating, and gets back only weak
distant sunlight. Unless it's near the sun - then it bakes.

4) A mirrored ship sounds like a bright object. But what's it going to
reflect? The sky in space is none too bright. So long as it has flat
surfaces (like a Stealth bomber) it will send sunlight or bright
starlight in a narrow and controllable range of directions.

5) For those who say meaningless: the ship will spend most of its time in
ballistic flight. During that time it will be almost invisible, a mere
floating meteor in boundless space. In theory, a skilled adversary would
have calculated its course precisely when its engines were running, or
would be able to detect it by powerful radars or other resource-intensive
techniques. In practice, the opponent of a ship might well turn out to
be a lone pirate with a weak infrared scanner, travelling a little
quickly or slowly in a major "shipping lane" (cheap transfer orbit).
Even if the adversary is a major government with many resources, it is
more likely that they will miss a lone smuggler, fugitive, or rebel "crop
duster" if that ship is particularly inobtrusive. They might track the
launch and guess the trajectory, but if there is a steady flow of
smugglers and refugees and illegal immigrants and so forth, they may not
have the manpower to catch a major fraction of the "offenders". And with
the right virus, you only need one ship...

Ian

unread,
Sep 5, 1998, 3:00:00 AM9/5/98
to
b.e...@spamblocker.tyndale.apana.org.au (Brett Evill) wrote:

>So, what are plausible technologies and tactics for space combat?
>

>Missiles/torpedoes will be vulnerable to ECM, I guess. Will it be possible
>to have guided-missile frigates in space with buckets of ECCM, guiding
>their torpedoes by remote control?

Lightspeed communications delay makes guiding anything by remote control
pretty much unworkable. Everything must be self-guided over the sorts of
distances likely to be involved.

ECM is of limited usefulness in space battles, where ships and weapons
will be seperated by vast empty distances. It's harder to confuse the
enemy about where you are in such an environment. Of course since ECM
isn't tremendously hard, and the enemy is going to see you even if you
don't use it, it will probably be used. Just not to huge effect.

>Will it be worth attempting stealth technologies to avoid observation, or
>is IR always going to give you away? How conspicuous will spaceships'
>drives be at range?

Stealth is only useful under very limited circumstances. You have to
build a huge radiator structure to dissipate heat in one direction, you
can't accelerate at any point in which you are in the enemy's view, and
you are screwed if he has sensors in the direction you are radiating
heat from. In other words, it could be useful to sneak your ships into
close range from deep space for an intricately planned surprise attack
(involving months drifting under special radiator shields, and then
ditching the shields when you reach the point where you will be detected
anyway). It is not a useful strategy in battle. Sensors small enough to
be carried on board (large) ships could detect pretty much any engine in
operation anywhere in the solar system.

>Will spaceships be painted black to prevent telescopic detection, or
>mirrored to bounce laser-beams?

A mirrored surface probably wouldn't be very useful. It would only
affect lasers, and even then would only reduce the power of the laser
attack for a fraction of a second before the mirror at that spot boiled
away.

>Perhaps ships with heat-superconducting
>armour will not be susceptible to significant damage by laser beams,

"Heat-superconducting armor"? To my knowledge, no such thing exists.
Normal superconductors superconduct electricity, but _not_ heat.

>How effective are nuclear warheads going to be in space?

They will destroy a ship if they explode very close to it. If they don't
explode very close, then they are little more than an annoyance that
blinds exposed sensors. Even the largest warheads won't seriously damage
a shielded space warship if they explode over a kilometer away, and in
space a kilometer is pretty short range.

>Sure they won't
>do blast damage, but neutrons and gamma-rays might still be effective.

>What about nuclear claymore mines? Nuke-pumped grasers?

Nuclear-pumped lasers would do impressive damage, if they hit. If you're
in a situation where it is easier to evade beam weapons than to hit with
them, and each side fires a hundred laser pulses for every one that
hits, then nuke-pumped lasers aren't very useful because they only fire
one shot, and a short one at that. Tremendous power only matters if you
can be reasonably sure of a hit.

>Are KE weapons going to be effective, or is guidance and delivery going to
>cause problems? A ship might kill itself running into a crowbar, but how
>easy is it to get a crowbar in front of a ship?

Crowbar-sized objects would have to be guided, and subject to
countermeasures. Filling the area around a ship with pebble-sized or
smaller debris, however, will cause serious damage unless it can
outright fly around the area.

>What ranges are plausible for engagement? Are these long enough that
>random evasion will be effective against beam weapons?

Range is entirely dependant on the sorts of weapons used. Spacecraft may
launch missiles at _huge_ distances, depending on the technology, and
most of the battle may consist of trying to keep the other guys' weapons
out of range. Beam weapon engagements between opponents reasonably
capable of manouvering, will happen at a few hundred thousand kilometers
at most, otherwhise virtually every shot will miss.

At a guess, though, space battles will not involve a lot of strategic
manouvering. Both sides know where the other guy is, where the other guy
wants to go, and roughly what course he has to take to get there. If one
side wants to completely avoid the other and has any capability of doing
so, they do, and no battle happens. Otherwhise the "battle" is a pretty
straighforward approach of the two forces, with both conducting small
manouvers to make sure the enemy doesn't hit them from extreme range. At
some point before firing starts, both sides launch their long-range
missiles, and possibly a screen of interceptors. As the missiles and
interceptors pass each other, they duke it out with energy weapons,
kinetic-kill, and a few nukes and then whatever is left from each side's
launch goes past toward the opposing fleet. Each fleet uses
point-defense countermeasures against the enemy, tries last-minute
invasions, and then takes whatever damage the missiles deal. Repeat this
for multiple volleys of missiles, until the fleets are within a few
hundred thousand kilometers when they start pelting each other with
energy weapons. They can fly past each other shooting, in which case
likely one side or the other will be utterly destroyed, or one side can
decide it is defeated and try to break off from battle as quickly as
possible, preserving what it has left. This may not be feasible
depending on how remaining fuel reserves compare to their velocity.

The dominant rules of space battle are:

1. You can't hide. The enemy probably knows where you are the moment you
launch.

2. You can run or fight, but you will typically have to pick one of the
two long before battle is joined, and then you're stuck with it.

3. There isn't much room for tactics and strategy. Pretty much
everything is automated, and the what your computers calculate as the
best possible attack strategy probably really is the best possible
attack strategy. If battle is joined, and both sides have a good idea of
what the capabilities of the other side's ships are, the result will be
a lot more predictable than we are used to.

4. If the engagement moves into short range (ie beam weapon range), it
will probably be decided in a single pass. This is because at those
kinds of ranges, it is very easy to hit the enemy and very hard to avoid
being hit. Both sides will keep hitting each other with deadly accuracy
until one of them is no longer capable of shooting. The exception is if
both sides pass each other at _extremely_ high speed, but this isn't
likely to happen if the primary objective of one side is to do as much
damage to the other side as possible (in which case it will decelerate,
slowing relative velocity).

5. You may know where the other guy is from extreme range, but due to
lightspeed delay, none of your weapons actually have a chance of hitting
him until they get fairly close. The range of the weapons themselves
isn't the determing factor at all, range of engagement is determined by
the size and manouverability of the enemy. Thus, your weapons are
effective against large, stationary, or slow-moving targets at far, far
longer ranges than they are against the average warship.

6. Anything which can't manouver had better be very heavily protected,
or had better not let you get anywhere near it. This tends to elimenate
the "happy medium" of space stations, they die if an enemy fleet gets in
close. Anything that can't move, and move fast, had better have huge
amounts of shielding and lots and lots of countermeasures,
point-defense, and whatnot. Only fortified targets on planets, moons, or
within large asteroids have a chance of surviving a close attack by an
enemy space fleet. These bases will compensate for their lack of
mobility with means of defense that spacecraft cannot have due to weight
restrictions. They will be extremely heavily armored (probably with most
of the important parts deep underground), and have large numbers of
defensive batteries that can destroy spacecraft at long range simply by
putting up so much fire that something is bound to be hit.

7. The most effective way to destroy planetary installations is not by
using warships, but by using the warships to clear out enemy space
forces so you can bombard the planet with asteroids or mass nuclear
assault. Conventional invasion is effectively impossible, since the
defenses will destroy your invasion force if they are stil functional,
and the only reliable way to take out the defenses is with mass
bombardment. A successful planetary assault will not allow you to
capture any installations intact except those that are deep underground,
and to capture those you have to send in ground forces that were very
expensive to transport across space. This means capture of installations
isn't really a viable alternative, unless you get the enemy to surrender
upon the threat of destruction.


Ian

unread,
Sep 5, 1998, 3:00:00 AM9/5/98
to
Erik Max Francis <m...@alcyone.com> wrote:

>If you're talking about something like a fusion feed drive, then
>basically whenever you fire your engines you'll be visible across the
>whole system. Lay-in-wait stealth will probably be very important, but
>once you make your move you will be a sitting duck. Which means you'll
>only move or open fire when you're assured of a kill.

Lay-in-wait stealth isn't very useful, since maintaining basic minimal
power and running ships' computers, let along actually having a crew,
radiates enough heat to be detectable at extremely long range. Modern
telescopic IR sensors could detect a space shuttle _manouvering
thruster_ firing beyond the asteroid belt.

>> Will spaceships be painted black to prevent telescopic detection, or
>> mirrored to bounce laser-beams?
>

>It's doubtful that the coating of the ship is really going to be the
>problem, since what's _really_ going to give you away is lighting your
>drive or firing ordnance.

Actually, being anything except a totally inert piece of junk is going
to produce enough IR to be detectable at long range.

>> Are KE weapons going to be effective, or is guidance and delivery
>> going to
>> cause problems? A ship might kill itself running into a crowbar, but
>> how
>> easy is it to get a crowbar in front of a ship?
>

>Kinetic weapons have the disadvantage of that they're unguided, but have
>an advantage in that their unguided, as well; they can't be
>countermeasured. Since they don't have their own exhaust trail, it also
>means you can't easily see them coming quite as easily as a drive-guided
>missile or torpedo.

Actually you can see them when they're launched and compute their exact
trajectory with ease. If they use engines to change course, you can see
that too. About the only unknown is that once they get very close, they
could explode fragments all over the place and you might not have time
to manouver. Still, the only real way to beat them is to outmanouver
them entirely, if you fly through a cloud of pebbles you have to accept
a chance of being hit.

>> What ranges are plausible for engagement? Are these long enough that
>> random evasion will be effective against beam weapons?
>

>Here's one way to look at it -- ships that engage in combat within easy
>beam weapons range won't be in combat for very long.
>
>My thinking is that it will be very much like submarine warfare; ships
>slink around, looking for an enemy. When they find one, it becomes a
>game of cat-and-mouse until one slips up. Then kinetic or missile
>ordnance is fired, and when a hit is landed, the winner closes in for a
>kill with beam weapons.
>
>That's the way I see it, anyway.

Cat and mouse doesn't happen. It's been discussed on this NG before, but
the simple fact is, it's a LOT easier to detect infrared emissions in
space than you probably think.

Ian

unread,
Sep 5, 1998, 3:00:00 AM9/5/98
to
Tommy the Terrorist <may...@newsguy.com> wrote:

>5) For those who say meaningless: the ship will spend most of its time in
>ballistic flight. During that time it will be almost invisible, a mere
>floating meteor in boundless space. In theory, a skilled adversary would
>have calculated its course precisely when its engines were running, or
>would be able to detect it by powerful radars or other resource-intensive
>techniques. In practice, the opponent of a ship might well turn out to
>be a lone pirate with a weak infrared scanner, travelling a little
>quickly or slowly in a major "shipping lane" (cheap transfer orbit).
>Even if the adversary is a major government with many resources, it is
>more likely that they will miss a lone smuggler, fugitive, or rebel "crop
>duster" if that ship is particularly inobtrusive. They might track the
>launch and guess the trajectory, but if there is a steady flow of
>smugglers and refugees and illegal immigrants and so forth, they may not
>have the manpower to catch a major fraction of the "offenders". And with
>the right virus, you only need one ship...

This is only relevant for engagements against "lone wolf" types who are
trying to sneak through some sort of blockade - in which case it's
probably just as easy to sneak in something on a "legitimate" transport
anyway. Not really relevant to the concept of space warfare, only
interplanetary terrorism. And with interplanetary travel technology,
viruses aren't likely to be any sort of global-scale threat, technology
to effectively contain outbreaks probably exists. Worst a terrorist
could do is nuke a few major cities. Lots of damage, sure, but it won't
likely win a war.

Erik Max Francis

unread,
Sep 5, 1998, 3:00:00 AM9/5/98
to
Ian wrote:

> Lay-in-wait stealth isn't very useful, since maintaining basic minimal
> power and running ships' computers, let along actually having a crew,
> radiates enough heat to be detectable at extremely long range. Modern
> telescopic IR sensors could detect a space shuttle _manouvering
> thruster_ firing beyond the asteroid belt.

If they already know where to look. It's a big solar system.

> Actually you can see them when they're launched and compute their
> exact
> trajectory with ease. If they use engines to change course, you can
> see
> that too.

And the missile can compute your course changes and adjust accordingly
as well.

--
Erik Max Francis / email m...@alcyone.com / whois mf303 / icq 16063900
Alcyone Systems / irc maxxon (efnet) / finger m...@sade.alcyone.com
San Jose, CA / languages En, Eo / web http://www.alcyone.com/max/
USA / icbm 37 20 07 N 121 53 38 W / &tSftDotIotE
\

/ The more violent the love, the more violent the anger.
/ _Burmese Proverbs_ (tr. Hla Pe)

J. Clarke

unread,
Sep 5, 1998, 3:00:00 AM9/5/98
to
They could detect it _if_ they were pointed at it.

--

--John

Reply to jclarke at eye bee em dot net.


Ian wrote in message <35fca23b...@news.ktchnr1.on.wave.home.com>...


>Erik Max Francis <m...@alcyone.com> wrote:
>
>>If you're talking about something like a fusion feed drive, then
>>basically whenever you fire your engines you'll be visible across the
>>whole system. Lay-in-wait stealth will probably be very important, but
>>once you make your move you will be a sitting duck. Which means you'll
>>only move or open fire when you're assured of a kill.
>

>Lay-in-wait stealth isn't very useful, since maintaining basic minimal
>power and running ships' computers, let along actually having a crew,
>radiates enough heat to be detectable at extremely long range. Modern
>telescopic IR sensors could detect a space shuttle _manouvering
>thruster_ firing beyond the asteroid belt.
>

>>> Will spaceships be painted black to prevent telescopic detection, or
>>> mirrored to bounce laser-beams?
>>
>>It's doubtful that the coating of the ship is really going to be the
>>problem, since what's _really_ going to give you away is lighting your
>>drive or firing ordnance.
>
>Actually, being anything except a totally inert piece of junk is going
>to produce enough IR to be detectable at long range.
>
>>> Are KE weapons going to be effective, or is guidance and delivery
>>> going to
>>> cause problems? A ship might kill itself running into a crowbar, but
>>> how
>>> easy is it to get a crowbar in front of a ship?
>>
>>Kinetic weapons have the disadvantage of that they're unguided, but have
>>an advantage in that their unguided, as well; they can't be
>>countermeasured. Since they don't have their own exhaust trail, it also
>>means you can't easily see them coming quite as easily as a drive-guided
>>missile or torpedo.
>

>Actually you can see them when they're launched and compute their exact
>trajectory with ease. If they use engines to change course, you can see

Ian

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
Erik Max Francis <m...@alcyone.com> wrote:

>Ian wrote:
>
>> Lay-in-wait stealth isn't very useful, since maintaining basic minimal
>> power and running ships' computers, let along actually having a crew,
>> radiates enough heat to be detectable at extremely long range. Modern
>> telescopic IR sensors could detect a space shuttle _manouvering
>> thruster_ firing beyond the asteroid belt.
>

>If they already know where to look. It's a big solar system.

A modern telescope would need to know where to look, yes. The
warship-mounted "sensor of the future" (TM) can likely scan the entire
solar system in ten minutes or less.

>> Actually you can see them when they're launched and compute their
>> exact
>> trajectory with ease. If they use engines to change course, you can
>> see
>> that too.
>

>And the missile can compute your course changes and adjust accordingly
>as well.

Which is my entire point. Everyone knows exactly where everything is (or
at least where it was, adjusting for lightspeed delay). Problem is that
in a contest between a complex kinetic missile and a spacecraft, this
gives the spacecraft and it's point defense and/or antimissiles the
advantage. The best strategy for kinetic weapons is the "shotgun"
effect, fragmenting into a cloud of particles at whatever the optimal
range to ensure the enemy can't get out of the way is.


Ian

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
"J. Clarke" <nos...@nospam.nospam> wrote:

>They could detect it _if_ they were pointed at it.

Today's telescopes could detect it if they were pointed at it.
Tomorrow's scanning IR sensor could easily trace enough sky to scan the
solar system in hours, or even minutes, and it _will_ pick up anything
whose temperature is above background.


Tom Breton

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
Tommy the Terrorist <may...@newsguy.com> writes:

> >Will spaceships be painted black to prevent telescopic detection, or

> >mirrored to bounce laser-beams? Perhaps ships with heat-superconducting


> >armour will not be susceptible to significant damage by laser beams,

> >because they will have large radiating surfaces. Will that protect them
> >from particle beams?
>
> Definitely mirrored.
>
> 1) A black ship is a "blackbody", which is to say, unless your crew likes
> life at 3 Kelvin, it's going to glow a cherry red to anybody with IR
> accessories.

I'd expect a ship to glow in IR, yes, unless extraordinary measures
are taken.

But not because of the crew's need for warmth. The surface of the
spaceship could be considerably colder than room temperature and still
be at thermal equilibrium, between 3K outside and room temperature
inside.

Rather, all sources of waste heat "count", and the engines will be a
big deal for any spaceship you can think about combat with.

Stealth measures mite include refrigerating most of the spaceship's
surface and moving the heat into (say) a pit on the surface so that
from some angles the ship is easily visible, while from other angles
it's not showing IR at all.

Another stealth approach, if far enuff away from the enemy, mite be to
spread out a really large surface and trickle just a little heat into
it, so that even tho its radiation is lower frequency, it still
rejects a lot of heat.

That has exponential problems, tho. Lowering the temperature linearly
reduces the heat energy rejected exponentially. If (EG) it was at a
temperature of 90 K, it would reject 1/100th as much heat per unit
area as if it were at room temperature (c.233 K). And 90 K is still
solidly in the far infrared.

I am mostly retreading last discussion.

> 2) "heat superconduction" has been claimed here to be a myth of Niven's
> unrelated to reality, and it seems believable. Heat by its nature is
> distributed between all sorts of random vibrational modes - how can it
> move QUICKLY, without pausing or getting sidetracked?


Indeed, "heat superconduction" is meaningless. How would heat be
"lost to friction/resistance/etc" anyways? "lost heat" is just heat
again!

> 3) As another thread pointed out, a black ship is a COLD ship, since it
> loses all that energy it's busy radiating, and gets back only weak
> distant sunlight. Unless it's near the sun - then it bakes.

Yes, altho you could say that underneath that black skin it's really
well insulated or something.

> 4) A mirrored ship sounds like a bright object. But what's it going to
> reflect? The sky in space is none too bright. So long as it has flat
> surfaces (like a Stealth bomber) it will send sunlight or bright

> starlight in a narrow and controllable range of directions.

Exactly.

Tom

J. Clarke

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
Ian wrote in message <35f7e532...@news.ktchnr1.on.wave.home.com>...


Given that all the astronomers and all the telescopes in the world haven't
yet succeeded in locating all the asteroids, I think that finding a single
spacecraft by pointing an infrared sensor at it is exceedingly unlikely
unless it is very close.

Chris Lawson

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
iadm...@undergrad.math.uwaterloo.ca (Ian) wrote:

>Today's telescopes could detect it if they were pointed at it.
>Tomorrow's scanning IR sensor could easily trace enough sky to scan the
>solar system in hours, or even minutes, and it _will_ pick up anything
>whose temperature is above background.

OK, folks, let's assume for the sake of argument that anything
that radiates IR will be easily detected. This leaves several
possible tactics.

1. Cold, ballistic weapon. This would need to be launched out of
range (which could mean months between launch and strike).

2. Directional IR radiation. As far as I can tell, a rocket that
only radiates in one direction would be a pretty impressive bit
of engineering.

3. Using shields. A rocket could accelerate all it likes provided
it was hidden by a planet or star. The problems are (i) it's
pretty hard to hide behind a planet unless you're already
in-system; (ii) it requires that the target occupt a limited
volume of space -- any two widely-separated, communicating ships
would be immune to this form of stealth.

4. Overkill. Make the missile so massive and fast that it doesn't
matter if the enemy detects it. This would only work against
unmaneuvrable targets such as planets, moons, very large orbital
stations, etc. And it's not really stealth :-)

5. Shotgun missile. Once the missile gets within range it
separates into (pick a number) warheads. It would be extremely
difficult for the target to defend against this sort of attack.
Even if the attacker couldn't afford a thousand warheads per
missile, he can still use say 10 warheads and 990 decoys to
exhaust the target defences.

Suffice to say, in a space battle, the crucial technology may be
scanner range rather than weapon power.

Any other possible tactics?

regards,
Chris Lawson


Ian

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
"J. Clarke" <nos...@nospam.nospam> wrote:

>Ian wrote in message <35f7e532...@news.ktchnr1.on.wave.home.com>...
>>"J. Clarke" <nos...@nospam.nospam> wrote:
>>

>>>They could detect it _if_ they were pointed at it.


>>
>>Today's telescopes could detect it if they were pointed at it.
>>Tomorrow's scanning IR sensor could easily trace enough sky to scan the
>>solar system in hours, or even minutes, and it _will_ pick up anything
>>whose temperature is above background.
>

>Given that all the astronomers and all the telescopes in the world haven't
>yet succeeded in locating all the asteroids, I think that finding a single
>spacecraft by pointing an infrared sensor at it is exceedingly unlikely
>unless it is very close.

Last I checked, asteroids were not detected by infrared. They are, after
all, exactly the same temperature as the space around them, what with
not having electronics, crews, and power systems inside them generating
heat.

Infrared is by far the best detection method in space, providing you
want to detect warm objects. Radar and visual-wavelength searches are a
hell of a lot harder.


Ian

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
cl...@ozemail.com.au (Chris Lawson) wrote:

>iadm...@undergrad.math.uwaterloo.ca (Ian) wrote:
>
>>Today's telescopes could detect it if they were pointed at it.
>>Tomorrow's scanning IR sensor could easily trace enough sky to scan the
>>solar system in hours, or even minutes, and it _will_ pick up anything
>>whose temperature is above background.
>

>OK, folks, let's assume for the sake of argument that anything
>that radiates IR will be easily detected. This leaves several
>possible tactics.
>
>1. Cold, ballistic weapon. This would need to be launched out of
>range (which could mean months between launch and strike).

"Out of range" means "outside the solar system". We are talking
_billions_ of kilometers. A completely cold projectile also can't have
electronics active to watch out for interception - the enemy still has a
chance to detect it at closer range using non-IR means, and intercept
it.

>2. Directional IR radiation. As far as I can tell, a rocket that
>only radiates in one direction would be a pretty impressive bit
>of engineering.

Directional radiation is a viable method of IR stealth. The problem is
that the limits to its effectiveness are such that the whole ship still
has to be relatively cold. It doesn't work when your engines are on,
especially since the exjaust plume itself probably radiates a lot of IR.
And its "directionality" is limited, the most you could compress it
(even with a _large_ directional radiator assembly) would be to a 60
degree or so cone. Better make sure there are no enemy forces in a whole
60 degrees of sky...

>3. Using shields. A rocket could accelerate all it likes provided
>it was hidden by a planet or star. The problems are (i) it's
>pretty hard to hide behind a planet unless you're already
>in-system; (ii) it requires that the target occupt a limited
>volume of space -- any two widely-separated, communicating ships
>would be immune to this form of stealth.

It also requires that the side trying to do the detecting doesn't have
other ships or recon drones in a position to observe the "hidden"
acceleration.

>5. Shotgun missile. Once the missile gets within range it
>separates into (pick a number) warheads. It would be extremely
>difficult for the target to defend against this sort of attack.
>Even if the attacker couldn't afford a thousand warheads per
>missile, he can still use say 10 warheads and 990 decoys to
>exhaust the target defences.

Depends if they are ballistic or homing warheads. If kinetic, the enemy
could simply avoid them. If homing (ie each of them is actually a
fair-sized missile itself), you'r just using a lot of ordnance to try
and flood target defenses.

Decoys are of limited usefulness depending on scanner technology. It's
plausible that decoys light enough that you could deploy a truly
substantial number of them, could be identified as such very quickly by
a sufficiently advanced sensor/computer combination. Among other things,
a projectile is identified as a real missile the moment it uses its
engine. So "ballistic decoys" _might_ be workable, but they could be
avoided much more easily than guided projectiles, and if it's a
ballistic kinetic-kill weapon, then decoys are a bit pointless because a
"decoy" is also a kinetic-kill projectile. Putting guided missiles in
with decoys is fairly unworkable too - the defender can just put his
priority on shooting projectiles which have fired an engine, displaying
that they are real missiles.

>Suffice to say, in a space battle, the crucial technology may be
>scanner range rather than weapon power.

I'm fairly sure that this won't be true. In a plausible battle between
space warships, scanner tech for both sides will likely be much more
than enough so that the real range limit for scanners is caused by
lightspeed delay. At "combat ranges" of a few million kilometers or less
(when energy and projectile weapons have a chance of hitting), both
sides are likely to have scanners that see everything they really need
to see.


Paul Dietz

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
> Given that all the astronomers and all the telescopes in the world haven't
> yet succeeded in locating all the asteroids, I think that finding a single
> spacecraft by pointing an infrared sensor at it is exceedingly unlikely
> unless it is very close.

Most astronomers aren't looking for asteroids, and do not
observe in the IR (especially from above the atmosphere).
Most telescopes also have very narrow fields of view.
The "fast" telescopes with wide fields of view used
in NEO searches are few in number, funded miserably,
and work in visible light.

Paul

J. Clarke

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
Ian wrote in message <35fbc3f2...@news.ktchnr1.on.wave.home.com>...

>"J. Clarke" <nos...@nospam.nospam> wrote:
>
>>Ian wrote in message <35f7e532...@news.ktchnr1.on.wave.home.com>...
>>>"J. Clarke" <nos...@nospam.nospam> wrote:
>>>
>>>>They could detect it _if_ they were pointed at it.

>>>
>>>Today's telescopes could detect it if they were pointed at it.
>>>Tomorrow's scanning IR sensor could easily trace enough sky to scan the
>>>solar system in hours, or even minutes, and it _will_ pick up anything
>>>whose temperature is above background.
>>
>>Given that all the astronomers and all the telescopes in the world haven't
>>yet succeeded in locating all the asteroids, I think that finding a single
>>spacecraft by pointing an infrared sensor at it is exceedingly unlikely
>>unless it is very close.
>
>Last I checked, asteroids were not detected by infrared. They are, after
>all, exactly the same temperature as the space around them, what with
>not having electronics, crews, and power systems inside them generating
>heat.

All infrared is is a little bit different wavelength. Doesn't alter the
nature of the problem. And asteroids are not "at exactly the same
temperature as the space around them", the are at an equilibrium temperature
considerably above that of the cosmic background radiation. In point of
fact in the visible and the infrared an asteroid is a bright spot in the
darkness, and you problem is (a) detecting that bright spot and (b)
distinguishing it from all the other bright spots.

>Infrared is by far the best detection method in space, providing you
>want to detect warm objects. Radar and visual-wavelength searches are a
>hell of a lot harder.

The trouble is that (a) a dim warm object takes a pretty sensitive detector
with a small enough field of view that other bright warm objects don't mask
the target and (b) space is absolutely _full_ of "warm objects".

J. Clarke

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
So how is detecting an object which glows in infrared against a dark
background a different problem from detecting an object which glows in
visible against a dark background? And you'd be surprised at how many
astronomers are looking for asteroids. And given that astronomers have been
looking for asteroids ever since the first one was found, it seems that if
there was some quick way to find them it would have been implemented by now.
Only way you do it is to point a detector at a piece of sky, image, and then
analyze the image. Some rapidly whirling infrared detector may spot an
incoming ship, but it's also going to spot the Sun, 9 planets, dozens of
moons, thousands or millions of asteroids, and billions and billions of
stars. So how do you tell which of those is a ship?

If an incoming spaceship was the only object out there radiating infrared it
would be one thing, but it is not, and it is likely to be a fairly dim
object in that regard as well.

--

--John

Reply to jclarke at eye bee em dot net.


Paul Dietz wrote in message <35F2CA9D...@interaccess.com>...


>> Given that all the astronomers and all the telescopes in the world
haven't
>> yet succeeded in locating all the asteroids, I think that finding a
single
>> spacecraft by pointing an infrared sensor at it is exceedingly unlikely
>> unless it is very close.
>

John Schilling

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
"J. Clarke" <nos...@nospam.nospam> writes:

>Ian wrote in message <35f7e532...@news.ktchnr1.on.wave.home.com>...
>>"J. Clarke" <nos...@nospam.nospam> wrote:
>>
>>>They could detect it _if_ they were pointed at it.
>>
>>Today's telescopes could detect it if they were pointed at it.
>>Tomorrow's scanning IR sensor could easily trace enough sky to scan the
>>solar system in hours, or even minutes, and it _will_ pick up anything
>>whose temperature is above background.

>Given that all the astronomers and all the telescopes in the world haven't
>yet succeeded in locating all the asteroids, I think that finding a single
>spacecraft by pointing an infrared sensor at it is exceedingly unlikely
>unless it is very close.


Given that most astronomers are engaged in research so intently focused on
tiny, remote corners of the universe that they would probably never notice
Jupiter in their entire professional careers, I don't know that this really
means much.

The fraction of the astronomical community's resources devoted to the search
for near-Earth, or even anywhere-in-the-solar-system, objects, is miniscule.
And the astronomical community as a whole has only a tiny fraction of the
resources of the world's military forces. The ability or inability of the
world's astronomers to comprehensively map the asteroids says nothing about
the ability of future spacefaring military forces to accomplish similar feats.


--
*John Schilling * "You can have Peace, *
*Member:AIAA,NRA,ACLU,SAS,LP * or you can have Freedom. *
*University of Southern California * Don't ever count on having both *
*Aerospace Engineering Department * at the same time." *
*schi...@spock.usc.edu * - Robert A. Heinlein *
*(213)-740-5311 or 747-2527 * Finger for PGP public key *

Erik Max Francis

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
Ian wrote:

> "Out of range" means "outside the solar system". We are talking
> _billions_ of kilometers. A completely cold projectile also can't have
> electronics active to watch out for interception - the enemy still has
> a
> chance to detect it at closer range using non-IR means, and intercept
> it.

You keep saying this, but can we have some figures on detection
capabilities?

For instance, let's say that our hypothetical infrared sensor can
resolve to magnitude 20 in the J band. In the ISO system, the J band
(wavelength 1.24 um, frequency 242 THz) has its zeroth magnitude at 1587
Jy, which is 1.587 x 10^-23 W/(m^2 Hz). At that frequency, that
corresponds to a flux of 3.84 x 10^-9 W/m^2.

Magnitude 20 is [100^(1/5)]^20 times dimmer, or exactly 10^8 times
dimmer, which puts the infrared flux for magnitude 20 at 3.84 x 10^-17
W/m^2.

As an example of how good this is, at a distance of 1 au = 150 Gm, you
detection threshhold for infrared emissions would be 10.9 MW. Anything
dimmer (in infrared) and you wouldn't be able to see it.

Do you have any figures backing up your claims on the effectiveness of
infrared for basically making _everything_ in the solar system instantly
trackable?

--
Erik Max Francis / email m...@alcyone.com / whois mf303 / icq 16063900
Alcyone Systems / irc maxxon (efnet) / finger m...@sade.alcyone.com
San Jose, CA / languages En, Eo / web http://www.alcyone.com/max/
USA / icbm 37 20 07 N 121 53 38 W / &tSftDotIotE
\

/ It is only to the individual that a soul is given.
/ Albert Einstein

J. Clarke

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
Well, given that the astronomical community has been about it for a rather
long time and not completed the job, I seriously doubt that anyone is going
to come up with some magic sensor that can do it in thirty seconds. And
while the paid astronomers aren't for the most part hunting for asteroids,
there are thousands of amateurs doing so, with considerable success. The
nature of the problem doesn't change because the military is working on it.

--

--John

Reply to jclarke at eye bee em dot net.


John Schilling wrote in message <6suqmn$99h$1...@spock.usc.edu>...

Paul Dietz

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
> 2. Directional IR radiation. As far as I can tell, a rocket that
> only radiates in one direction would be a pretty impressive bit
> of engineering.

The problem is radiation from the rocket plume.
This was discussed in the previous incarnation of
this thread some month ago. You can get at least
some performance (several km/s) with very low IR
signature by using helium as reaction mass in a
nuclear thermal engine. Helium, being monatomic,
has no rotational or vibrational modes to store
energy for emission in the IR, and its electronic
excited states are at ~20 eV and above, which is
very high by thermal standards. The only concern
would be detection of the plume by interaction with
solar UV.

Direct radiation from the engine itself could
be restricted to a small solid angle with shields/
mirrors.

Paul

Paul Dietz

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
> So how is detecting an object which glows in infrared against a dark
> background a different problem from detecting an object which glows in
> visible against a dark background?

It's different because (1) the atmosphere is not transparent to IR
in some of the wavelengths of interest, and (2) the telescope and
atmosphere *radiate* IR in the wavelengths of interest. You can
overcome this by putting cryogenic telescopes in space, but that's
currently expensive. In a future with space battleships it would
not be.


> And you'd be surprised at how many
> astronomers are looking for asteroids.

The number is a small fraction of the astronomical community,
both in number of observers and fraction of funding. And,
they're using visible light.


> And given that astronomers have been
> looking for asteroids ever since the first one was found, it seems that if
> there was some quick way to find them it would have been implemented by now.
> Only way you do it is to point a detector at a piece of sky, image, and then
> analyze the image.

Asteroids have been ignored, or reviled, by most astronomers for much of
this
century. They interfere with observation of deep sky objects.

Truly productive scans for NEOs use a time integration technique
in which the image is moved (by rotation of the earth) over a CCD
(or CCDs), with the charge being clocked to track the image. This
scans a strip across the sky. Repeat twice and look for moving
dots and/or streaks. This had to wait for sufficiently good CCDs,
cheap enough computers, software, and the funding to put it all
together. Only recently has this started to happen.

> If an incoming spaceship was the only object out there radiating infrared it
> would be one thing, but it is not, and it is likely to be a fairly dim
> object in that regard as well.

Imagining that current astronomical technology will still
be used in a future where interplanetary battleships are built
is an interestingly selective failure of the imagination.

Paul

Erik Max Francis

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
Ian wrote:

> Well, there are a lot of technological variables to consider, so as a
> baseline let's use today. What would happen if we mounted a single, 2
> meter aperture modern telescope to detect visual and near-IR radiation
> in space (We could use it to sweep across the entire sky within a few
> hours)? Turns out we can detect a space shuttle manouvering thruster
> firing from at least 15 million kilometers. Space shuttle main engines
> would be detectable from 20 billion kilometers (waaay outside the
> orbit
> of Pluto). A fairly small (not terribly larger than space shuttle)
> size
> ship using ion thrusters to accelerate at 1/1000 of a g would be
> detectable from 100 million kilometers or so. These calculations are
> from a post by John Schilling on 1/2/98.

I dug up his post, it's at

http://www.dejanews.com/getdoc.xp?AN=312184581

Quite frankly, I think there's something wrong with his numbers. The
detection threshhold that he gives is 2.5 x 10^-17 W/m^2, which is not
too different from the estimate I got based on an assumption of a
limiting magnitude of 20 (3.84 x 10^-17 W/m^2).

With a limiting flux of 2.5 x 10^-17 W/m^2, for a plume to be visible at
20 Tm, the source power must be 130 GW, all in the bands in which the
sensor is operating (he notes that this is a visual/near-infrared
system). Are the Space Shuttle main engines capable of 130 GW? I don't
have any figures in front of me.

And at a distance of 15 Gm, the limiting power is about 71 kW, which
sounds more reasonable.

John Schilling

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
"J. Clarke" <nos...@nospam.nospam> writes:

>Well, given that the astronomical community has been about it for a rather
>long time and not completed the job, I seriously doubt that anyone is going
>to come up with some magic sensor that can do it in thirty seconds. And
>while the paid astronomers aren't for the most part hunting for asteroids,
>there are thousands of amateurs doing so, with considerable success. The
>nature of the problem doesn't change because the military is working on it.


The nature of the problem doesn't change, but the ammount of resources
being devoted to the problem does. By about three or four orders of
magnitude. And that makes for qualitatively different results. You
may say that the astronomical community has been about "it" for a very
long time, but all the manpower and equipment devoted to asteroid-hunting
in all of recorded history probably wouldn't add up to a month's worth
of the global military budget for sensors, sensor operators, sensor
platforms, and assorted logistical support.

You might as well judge the feasibility of antisubmarine warfare by
the frequency with which fisherman, oceanographers, and recreational
divers stumble across submarines.

Leonard Erickson

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
t...@world.std.com ("Tom Breton") writes:

>> 2) "heat superconduction" has been claimed here to be a myth of Niven's
>> unrelated to reality, and it seems believable. Heat by its nature is
>> distributed between all sorts of random vibrational modes - how can it
>> move QUICKLY, without pausing or getting sidetracked?
>
>
> Indeed, "heat superconduction" is meaningless. How would heat be
> "lost to friction/resistance/etc" anyways? "lost heat" is just heat
> again!

As Niven envisioned it, "heat superconductivity" is essentially
infinite thermal conductivity. That is, the temp of the object is
uniform, even if one portion is receiving substantially more heat input
than the rest.

So if you had a long rod, and placed one end in the fire, all of it
would be too hot to handle.

>> 3) As another thread pointed out, a black ship is a COLD ship, since it
>> loses all that energy it's busy radiating, and gets back only weak
>> distant sunlight. Unless it's near the sun - then it bakes.
>
> Yes, altho you could say that underneath that black skin it's really
> well insulated or something.

In which case the crew or equipment roasts in a *very* short period of
time. Heat not radiated winds up as an increase in internal temp.

--
Leonard Erickson (aka Shadow)
sha...@krypton.rain.com <--preferred
leo...@qiclab.scn.rain.com <--last resort

John Schilling

unread,
Sep 6, 1998, 3:00:00 AM9/6/98
to
Erik Max Francis <m...@alcyone.com> writes:

>Ian wrote:

>> Well, there are a lot of technological variables to consider, so as a
>> baseline let's use today. What would happen if we mounted a single, 2
>> meter aperture modern telescope to detect visual and near-IR radiation
>> in space (We could use it to sweep across the entire sky within a few
>> hours)? Turns out we can detect a space shuttle manouvering thruster
>> firing from at least 15 million kilometers. Space shuttle main engines
>> would be detectable from 20 billion kilometers (waaay outside the
>> orbit
>> of Pluto). A fairly small (not terribly larger than space shuttle)
>> size
>> ship using ion thrusters to accelerate at 1/1000 of a g would be
>> detectable from 100 million kilometers or so. These calculations are
>> from a post by John Schilling on 1/2/98.

>I dug up his post, it's at

> http://www.dejanews.com/getdoc.xp?AN=312184581

>Quite frankly, I think there's something wrong with his numbers. The
>detection threshhold that he gives is 2.5 x 10^-17 W/m^2, which is not
>too different from the estimate I got based on an assumption of a
>limiting magnitude of 20 (3.84 x 10^-17 W/m^2).


There was a factor of 4*pi error in my initial post; I don't recall
exactly where offhand, but one of the readers caught it fairly quickly
and I was able correct for it. IIRC, the range of the original sensor
as described should be reduced by sqrt(4*pi), or seventy percent, giving
space shuttle detection at "only" six billion kilometers. A mark II
sensor restoring the original performance required a somewhat more
expensive detector array, but still (barely) within reach of current
technology.


>With a limiting flux of 2.5 x 10^-17 W/m^2, for a plume to be visible at
>20 Tm, the source power must be 130 GW, all in the bands in which the
>sensor is operating (he notes that this is a visual/near-infrared
>system). Are the Space Shuttle main engines capable of 130 GW? I don't
>have any figures in front of me.

The SSMEs produce a shade under 5 GW each, and the solid boosters 33 GW.
So figure 80 GW at launch. Not all of that is radiated, of course. Not
even most of it. But the radiated fraction times the missing 4*pi is
probably fairly close to 130 GW.


>And at a distance of 15 Gm, the limiting power is about 71 kW, which
>sounds more reasonable.

The smallest attitude control thrusters on the space shuttle run at
140 kw, so again the numbers tend to match up.


Which is the missing half of the equation in most analyses. Space
travel requires, well, astronomical energies, and it is remarkably
difficult to hide astronomical energy releases against effectively
zero background. One can handwave about hyperefficient drives or
clever directional radiators or whatnot, but even a tiny fraction
of a percent of the energy leaking out as isotropic radiation will
kill you. Until someone repeals the second law of thermodynamics,
I am quite skeptical about proposals to do *anything* with an
efficiency that requires more nines to express than I can count on
one hand, which is pretty much what practical stealth in space would
require.

Ian

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
Erik Max Francis <m...@alcyone.com> wrote:

>> "Out of range" means "outside the solar system". We are talking
>> _billions_ of kilometers. A completely cold projectile also can't have
>> electronics active to watch out for interception - the enemy still has
>> a
>> chance to detect it at closer range using non-IR means, and intercept
>> it.
>
>You keep saying this, but can we have some figures on detection
>capabilities?

Well, there are a lot of technological variables to consider, so as a


baseline let's use today. What would happen if we mounted a single, 2
meter aperture modern telescope to detect visual and near-IR radiation
in space (We could use it to sweep across the entire sky within a few
hours)? Turns out we can detect a space shuttle manouvering thruster
firing from at least 15 million kilometers. Space shuttle main engines
would be detectable from 20 billion kilometers (waaay outside the orbit
of Pluto). A fairly small (not terribly larger than space shuttle) size
ship using ion thrusters to accelerate at 1/1000 of a g would be
detectable from 100 million kilometers or so. These calculations are
from a post by John Schilling on 1/2/98.

So even with _today's_ technology, any major engine burn can be detected
to Pluto's orbit and beyond, use of manouvering thrusters or low-thrust,
highly efficient engines out to many millions of kilometers. Detection
of objects which are merely radiating, not thrusting, is a bit harder to
calculate since their surface area and exact temperature must be taken
into account among other things.

By the time interplanetary warships are feasible, it's likely that IR
detection technology will be at least 10 times as sensitive as it is
today, perhaps much more - and dedicated observation stations, keeping
in comminication with warships for long-range threat warning, could use
telescopes with apertures _far_ larger than 10 meters.

So it's plausible to have warships that can detect extraordinarily small
engine firings, or large ships that are not firing engines at all, over
100 million kilometers away, and that can detect main engines firing
well beyond Pluto's orbit. Dedicated observation stations, using arrays
of very large telescopes, can likely detect any operational warship out
to billions of kilometers whether its engines are firing or not,
communicate this information to the warships, who can then use pinpoint
tracking telescopes to get a lock on these far-off targets and keep tabs
on them.


Ian

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
"J. Clarke" <nos...@nospam.nospam> wrote:

>So how is detecting an object which glows in infrared against a dark
>background a different problem from detecting an object which glows in
>visible against a dark background?

Asteroids don't radiate in visible, they only reflect what visible light
from the stars and the sun hits them. In contrast, any object with a
temperature above that of the background actually _radiates_ infrared.
Warm objects in deep space are simply much brighter in infrared than
they are in the visible spectrum.

>analyze the image. Some rapidly whirling infrared detector may spot an
>incoming ship, but it's also going to spot the Sun, 9 planets, dozens of
>moons, thousands or millions of asteroids, and billions and billions of
>stars.

No, it's NOT going to spot most of those.

It's not going to spot asteroids, they're not warmer than background.
And it's not going to spot most of the stars, they _are_ the background.
Sure it'll spot the sun and the planets and particularly bright stars,
but it will instantly disregard these because their exact positions are
known.


Ian

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
"J. Clarke" <nos...@nospam.nospam> wrote:

>Well, given that the astronomical community has been about it for a rather
>long time and not completed the job, I seriously doubt that anyone is going
>to come up with some magic sensor that can do it in thirty seconds. And
>while the paid astronomers aren't for the most part hunting for asteroids,
>there are thousands of amateurs doing so, with considerable success. The
>nature of the problem doesn't change because the military is working on it.

No, the scope of resources dedicated is what changes. By a LOT. The
telescopic array of a single spaceborn, dedicated IR observation station
would easily exceed the combined capabilities of every telescope on
Earth ever used for searching for near-Earth objects. Sort of like the
difference between a pre-WWII research radar, and the modern US air
defense net.


Ian

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
Erik Max Francis <m...@alcyone.com> wrote:

>With a limiting flux of 2.5 x 10^-17 W/m^2, for a plume to be visible at
>20 Tm, the source power must be 130 GW, all in the bands in which the
>sensor is operating (he notes that this is a visual/near-infrared
>system). Are the Space Shuttle main engines capable of 130 GW? I don't
>have any figures in front of me.

130 GW is a lot of power, but a liquid fueled rocket capable of lifting
a space shuttle into orbit is a pretty big, bright engine.


John Park

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
Guesstimate: The heat of formation of water from hydrogen and oxygen is
about 16 MJ/kg. Guessing that the combustion of the shuttle's fuel produces
about 5000 t of water, and neglecting the kinetic energy of the shuttle
itself, the energy released is about 80 TJ. If the burn time is about
1000 sec this gives a power of ca. 80 GW. In the right ballpark at least.

--John Park


Ian

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
sha...@krypton.rain.com (Leonard Erickson) wrote:

>As Niven envisioned it, "heat superconductivity" is essentially
>infinite thermal conductivity. That is, the temp of the object is
>uniform, even if one portion is receiving substantially more heat input
>than the rest.
>
>So if you had a long rod, and placed one end in the fire, all of it
>would be too hot to handle.

Problem is there are no known substances which superconduct heat, nor
any known mechanism by which this could occur. Niven seemed to simply
assume that "superconductor" meant superconductor in all senses, when it
really just means "of electricity".

Superconducting electricity just means that the electricity travels
through the object as fast as it normally does, but resistance doesn't
occur. There is no "resistance" to heat - it's just a bunch of
vibrational kinetic energy transmitted between atoms. You'd have to
increase the speed of transmission of this energy to lightspeed for heat
"superconductivity", which means that the atoms in question would have
to be vibrating at lightspeed... oops, that's impossible.


Ian

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
schi...@spock.usc.edu (John Schilling) wrote:

>Which is the missing half of the equation in most analyses. Space
>travel requires, well, astronomical energies, and it is remarkably
>difficult to hide astronomical energy releases against effectively
>zero background. One can handwave about hyperefficient drives or
>clever directional radiators or whatnot, but even a tiny fraction
>of a percent of the energy leaking out as isotropic radiation will
>kill you. Until someone repeals the second law of thermodynamics,
>I am quite skeptical about proposals to do *anything* with an
>efficiency that requires more nines to express than I can count on
>one hand, which is pretty much what practical stealth in space would
>require.

Stealthed engines, yeah, basically impossible. It's just barely possible
to coast at stealth levels that would reduce detection ranges from most
angles tremendously, _if_ your ship is capable of running very cold (ie
no engines, no main power, any crew in suspended animation and
nonessential computers deactivated). Sort of makes stealth useless for
anything but getting up to speed billions of kilometers away, and
coasting for months to get within a few million kilometers of the enemy
before detection.


Brett Evill

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
In article <35f29...@news1.ibm.net>, "J. Clarke" <nos...@nospam.nospam> wrote:


>Given that all the astronomers and all the telescopes in the world haven't
>yet succeeded in locating all the asteroids, I think that finding a single
>spacecraft by pointing an infrared sensor at it is exceedingly unlikely
>unless it is very close.

alt.non.sequitur

Asteroids are in thermal equilibrium. They won't show up on IR. Spaceships
cannot be in thermal equilibrium, they must radiate in IR.

The question that seems of more concern to me is whether these sensors
will be able readily to distinguish asteroids from *stars*.

--
Brett Evill

To reply by e-mail, remove 'spamblocker.' from <b.e...@spamblocker.tyndale.apana.org.au>

Erik Max Francis

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
Ian wrote:

> Problem is there are no known substances which superconduct heat, nor
> any known mechanism by which this could occur. Niven seemed to simply
> assume that "superconductor" meant superconductor in all senses, when
> it
> really just means "of electricity".

To be fair to Niven, I think it's clear that he knew this. He
specifically talked about "heat superconductors," clearly distinguishing
them from normal superconductors. While superconduction of heat is
physically baseless, he _was_ specifically distinguishing between normal
superconduction and the science-fictional premise which he was using in
his stories.

Paul Dietz

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
> Asteroids are in thermal equilibrium. They won't show up on IR. Spaceships
> cannot be in thermal equilibrium, they must radiate in IR.

No.

(Hint: asteroids are typically not very reflective. Where
does the energy absorbed from sunlight go?)

Paul

Brett Evill

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
In article <35F3CDAC...@interaccess.com>, Paul Dietz
<di...@interaccess.com> wrote:

It warms them up until they radiate heat as fast as they absorb it. This
condition is called 'thermal equilibrium'. Ooh! Look! That's the condition
that I said that they would be in.

An ideal black body, that absorbs every milliwatt of radiation that falls
on it can still be in thermal equilibrium with its surroundings. And when
you do the physics, it turns out that you expect such a black body to be
in thermal equilibrium when is is the same temperature as its surroundings
(averaged by subtended angle).

But a working spaceship, that has to radiate all of its waste heat as well
as reflecting or reradiating the sunlight that falls on it. So a working
spaceship will always be warmer than an asteroid at the same distance from
the Sun, and will have a higher-frequency and brighterr IR signature.

J. Clarke

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to

--

--John

Reply to jclarke at eye bee em dot net.


Paul Dietz wrote in message <35F324B7...@interaccess.com>...


>> So how is detecting an object which glows in infrared against a dark
>> background a different problem from detecting an object which glows in
>> visible against a dark background?
>

>It's different because (1) the atmosphere is not transparent to IR
>in some of the wavelengths of interest, and (2) the telescope and
>atmosphere *radiate* IR in the wavelengths of interest. You can
>overcome this by putting cryogenic telescopes in space, but that's
>currently expensive. In a future with space battleships it would
>not be.

You haven't demonstrated any difference that would reduce the level of
effort required, merely.

>> And you'd be surprised at how many
>> astronomers are looking for asteroids.
>
>The number is a small fraction of the astronomical community,
>both in number of observers and fraction of funding. And,
>they're using visible light.

Which is irrelevant, unless perhaps you can demonstrate that is easier to
build an IR sensor that can scan wide areas and distinguish a spaceship from
all the other emitters than it is to build a visible spectrum sensor to
perform that task.

>> And given that astronomers have been
>> looking for asteroids ever since the first one was found, it seems that
if
>> there was some quick way to find them it would have been implemented by
now.
>> Only way you do it is to point a detector at a piece of sky, image, and
then
>> analyze the image.
>
>Asteroids have been ignored, or reviled, by most astronomers for much of
>this
>century. They interfere with observation of deep sky objects.

Yes, they do, and they are being cataloged among other things so that the
astronomers know when an asteroid will be interfering in that manner.

>Truly productive scans for NEOs use a time integration technique
>in which the image is moved (by rotation of the earth) over a CCD
>(or CCDs), with the charge being clocked to track the image. This
>scans a strip across the sky. Repeat twice and look for moving
>dots and/or streaks. This had to wait for sufficiently good CCDs,
>cheap enough computers, software, and the funding to put it all
>together. Only recently has this started to happen.

And how long does it take to scan the entire sky with such a sensor?

>> If an incoming spaceship was the only object out there radiating infrared
it
>> would be one thing, but it is not, and it is likely to be a fairly dim
>> object in that regard as well.
>
>Imagining that current astronomical technology will still
>be used in a future where interplanetary battleships are built
>is an interestingly selective failure of the imagination.

I'm sorry, but optics is optics. If you want to postulate some baloneytron
that detects spaceships by magic, that's fine, but if you are going to claim
that they can be detected by infrared then you're going to have to make a
case for it, which you have not.
>
> Paul

J. Clarke

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
Ian wrote in message <36012655...@news.ktchnr1.on.wave.home.com>...

>"J. Clarke" <nos...@nospam.nospam> wrote:
>
>>So how is detecting an object which glows in infrared against a dark
>>background a different problem from detecting an object which glows in
>>visible against a dark background?
>
>Asteroids don't radiate in visible, they only reflect what visible light
>from the stars and the sun hits them. In contrast, any object with a
>temperature above that of the background actually _radiates_ infrared.
>Warm objects in deep space are simply much brighter in infrared than
>they are in the visible spectrum.

Are they? Do you have numbers to support this contention? In any case,
they are still objects against a background containing many other objects,
so how do you distinguish?

>>analyze the image. Some rapidly whirling infrared detector may spot an
>>incoming ship, but it's also going to spot the Sun, 9 planets, dozens of
>>moons, thousands or millions of asteroids, and billions and billions of
>>stars.
>
>No, it's NOT going to spot most of those.
>
>It's not going to spot asteroids, they're not warmer than background.

Asteroids are not warmer than "background"? What exactly do you believe the
temperature of an asteroid to be and what do you believe the temperature of
the background to be?

>And it's not going to spot most of the stars, they _are_ the background.

They are? Then what is all that black area up there? You really should
read an elementary astronomy book before you pontificate you know.

>Sure it'll spot the sun and the planets and particularly bright stars,
>but it will instantly disregard these because their exact positions are
>known.

I see. So all that an enemy has to do is align himself between your sensor
the sun or a bright star and it ignores him. Great design. In any case
until the entire galaxy is mapped this will work only in developed systems,
not in the hinterlands where a ship is operating alone.

J. Clarke

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
Ian wrote in message <3602278f...@news.ktchnr1.on.wave.home.com>...

>"J. Clarke" <nos...@nospam.nospam> wrote:
>
>>Well, given that the astronomical community has been about it for a rather
>>long time and not completed the job, I seriously doubt that anyone is
going
>>to come up with some magic sensor that can do it in thirty seconds. And
>>while the paid astronomers aren't for the most part hunting for asteroids,
>>there are thousands of amateurs doing so, with considerable success. The
>>nature of the problem doesn't change because the military is working on
it.
>
>No, the scope of resources dedicated is what changes. By a LOT. The
>telescopic array of a single spaceborn, dedicated IR observation station
>would easily exceed the combined capabilities of every telescope on
>Earth ever used for searching for near-Earth objects.

Why do you keep harping on "near earth objects"? They are not the only
asteroids out there.

>Sort of like the
>difference between a pre-WWII research radar, and the modern US air
>defense net.

This is fine if you can afford to dedicate a station to this task. But how
about if your ship is out in the boonies? Is is going to tow your station
along everywhere it goes?

In my first post on this topic, which you obviously neglected to read, I
stated clearly that the infrared method _might_ work in a star system with
an established industrial base, but not for a ship operating alone. So
kindly in the future restrain your remarks to that context and not the one
which I have already stipulated.

J. Clarke

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
So how many sensors will a spacegoing destroyer be able to carry?

--

--John

Reply to jclarke at eye bee em dot net.


John Schilling wrote in message <6svorr$aj7$1...@spock.usc.edu>...


>"J. Clarke" <nos...@nospam.nospam> writes:
>
>>Well, given that the astronomical community has been about it for a rather
>>long time and not completed the job, I seriously doubt that anyone is
going
>>to come up with some magic sensor that can do it in thirty seconds. And
>>while the paid astronomers aren't for the most part hunting for asteroids,
>>there are thousands of amateurs doing so, with considerable success. The
>>nature of the problem doesn't change because the military is working on
it.
>
>

>The nature of the problem doesn't change, but the ammount of resources
>being devoted to the problem does. By about three or four orders of
>magnitude. And that makes for qualitatively different results. You
>may say that the astronomical community has been about "it" for a very
>long time, but all the manpower and equipment devoted to asteroid-hunting
>in all of recorded history probably wouldn't add up to a month's worth
>of the global military budget for sensors, sensor operators, sensor
>platforms, and assorted logistical support.
>
>You might as well judge the feasibility of antisubmarine warfare by
>the frequency with which fisherman, oceanographers, and recreational
>divers stumble across submarines.
>
>

J. Clarke

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
Well, now, Brett, you've just stated that they "radiate heat". Last I hear,
"radiating heat" was another way to say "emitting infrared". So what does
that do to your argument?

Yes, the object is the same temperature as its surroundings, averaged by
subtended angle. The trouble with this is that those surroundings are mostly
very cold sink with one very hot source, so the equilibrium temperature is
somewhere between.

What do you believe the radiation temperature of the background to be, and
the radiation temperature of an asteroid?

--

--John

Reply to jclarke at eye bee em dot net.


Brett Evill wrote in message ...

J. Clarke

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
First item, how long an exposure is required to achieve these results.
Second item, what is the probability that a spacecraft operated by someone
who understands the capabilities of your sensors is going to be doing one of
these things at the exact time that you have the sensor pointed at it?

--

--John

Reply to jclarke at eye bee em dot net.


Ian wrote in message <36002309...@news.ktchnr1.on.wave.home.com>...


>Erik Max Francis <m...@alcyone.com> wrote:
>

Brett Evill

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
In article <35f3f...@news1.ibm.net>, "J. Clarke" <nos...@nospam.nospam> wrote:

>Asteroids are not warmer than "background"? What exactly do you believe the

>temperature of an asteroid to be and what do you believe the temperature of
>the background to be?

The IR signature of the cosmic background is equivalent to a black body
radiating at an effective temperature of 3K.

Thermal equilibrium at the orbit of the Earth is about 260-270K. It varies
inversely with the square root of distance from the Sun.

Black-body radiation varies with the fourth power of the temperature, and
the frequency of the maximum on the power curve is proportional to the
temperature.

It would seem that the obvious way to distinguish stars, inert lumps, and
working spacecraft is to measure their brightness (CCD cameras will do
this) at several different wavelengths in the IR and visible. This will
allow you to work out roughly at what wavelength the maximum of power is,
and that will let you sort them into temperature categories.

Of course, this is hard to do with ground-based instruments because the
atmosphere is largely opaque to IR, especially in the longer wavelengths,
and because all of your instruments are sitting in a warm bath of air. To
detect deep IR photons, you really need to use instruments that are too
cool to emit much thermal radiation at the relevant frequencies. But
cyrogenic cooling is conparatively easy in vaccuum.

It certainly seems to me that detection of working spacecraft can be made
rapid and mechanical. The only problem I can see it that if the sweeping
telescopes have a narrow field of view a complete 4.pi steradian sweep
might take a while. Can anyone think of a reason why the field of view
need be small or the sweeping slow.

Paul Dietz

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
>> You can
> >overcome this by putting cryogenic telescopes in space, but that's
> >currently expensive. In a future with space battleships it would
> >not be.
>
> You haven't demonstrated any difference that would reduce the level of
> effort required, merely.

(with annoyance) I had assumed some intelligence on the part
of the reader. Think -- the whole spacewar scenario assumes
a level space ability far beyond what we have today, with
much lower cost to reach space and operate in space. That
by itself makes space telescopes (cryogenic or otherwise)
less difficult.

> >The number is a small fraction of the astronomical community,
> >both in number of observers and fraction of funding. And,
> >they're using visible light.
>
> Which is irrelevant, unless perhaps you can demonstrate that is easier to
> build an IR sensor that can scan wide areas and distinguish a spaceship from
> all the other emitters than it is to build a visible spectrum sensor to
> perform that task.

No, it is not irrelevant. The small effort being expended
on asteroid searches shows that you can't use current results
as a limit of what a larger scale program would do.


> >Imagining that current astronomical technology will still
> >be used in a future where interplanetary battleships are built
> >is an interestingly selective failure of the imagination.

> I'm sorry, but optics is optics. If you want to postulate some baloneytron
> that detects spaceships by magic, that's fine, but if you are going to claim
> that they can be detected by infrared then you're going to have to make a
> case for it, which you have not.

Bigger apertures, better sensors, better computers. All
of this is easier than building space battleships.

Paul

Paul Dietz

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
> >> Asteroids are in thermal equilibrium. They won't show up on IR. Spaceships
> >> cannot be in thermal equilibrium, they must radiate in IR.
> >
> >No.
> >
> >(Hint: asteroids are typically not very reflective. Where
> >does the energy absorbed from sunlight go?)
>
> It warms them up until they radiate heat as fast as they absorb it. This
> condition is called 'thermal equilibrium'. Ooh! Look! That's the condition
> that I said that they would be in.


And your contention that they would not show up on IR
is completely wrong. Yes, they warm up until they are in
a kind of equilibrium (not thermodynamic equilibrium in
its precise sense). BUT... they are then much warmer
than 2.73K, and would show up quite nicely in the IR.

Look, IR observations are *already* used to estimate
albedos of asteroids (if not to search for them).


> An ideal black body, that absorbs every milliwatt of radiation that falls
> on it can still be in thermal equilibrium with its surroundings. And when
> you do the physics, it turns out that you expect such a black body to be
> in thermal equilibrium when is is the same temperature as its surroundings
> (averaged by subtended angle).

There's your confusion: you are confusing a state
of dynamic equilibrium with the much more strictly
defined state of thermodynamic equilibrium. This
asteroid cannot in thermodynamic equilibrium with
its surroundings, since the surroundings itself
(sun, microwave background) are not in thermodynamic
equilibrium.

Paul

Ian

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
"J. Clarke" <nos...@nospam.nospam> wrote:

>Are they? Do you have numbers to support this contention? In any case,
>they are still objects against a background containing many other objects,
>so how do you distinguish?

There is a nice device which, given the sensory return from an object
and a catalog of all known objects with their positions, is capable of
determining whether the sensory return is of a known or unknown object,
and giving ideas as to such characteristics as its location, velocity,
and temperature.

This miracle of science is called a "computer".

Your assertion is sort of like "an airplane flying at mach 2 is just an
object against a background of other objects, with other similar objects
around it, so how does a radar distinguish enemy aircraft"?

If one is fighting space battles in the solar system with advanced IR
observation technology, the following factors are likely to be true:

1. The positions of space objects, bright stars, friendly ships, and
other IR emitters will be recorded and known.

2. Detecting any object which is accelerating, emitting anything other
than IR, or doing anything un-rocklike, and then keeping track of which
object it is, is trivial.

3. Relative motion to multiple observers can easily be used to calculate
distance to an object, which allows strength of emission to be used to
calculate the temperature of the object. Since live ships must be hotter
than floating rocks, any object within the solar system that is hotter
than a floating rock is man-made and should be investigated further.

4. The scanning IR sensor is not the only sensor the ship possesses - if
anything is suspicious or hard to identify, more powerful and/or
different sensors may be pointed specifically at it to find out what it
is.

>>Sure it'll spot the sun and the planets and particularly bright stars,
>>but it will instantly disregard these because their exact positions are
>>known.
>
>I see. So all that an enemy has to do is align himself between your sensor
>the sun or a bright star and it ignores him.

Actually he will show up as an unexpected cold or warm spot unless his
distance from you and his level of radiation mean he is just as bright
as what he's between you and.

Also, he'd have to _stay_ between you and the object, which is going to
be damn hard if you're both moving. He'd need to accelerate either to
get in position, or stay in position. And if he accelerates, he is _far_
more detectable, and probably far brighter than any natural radiating
body (excepting the sun, at least).

>Great design. In any case
>until the entire galaxy is mapped this will work only in developed systems,
>not in the hinterlands where a ship is operating alone.

"Developed systems"? Developed SYSTEMS? This is NOT a discussion about a
world where FTL exists, since we cannot have any sort of plausible
discussion about that at all. Maybe ships can radiate their waste heat
into hyperspace, or maybe warp power cores can be detectable clear
across the galaxy! The context of this discussion has been
interplanetary war within the solar system.


Ian

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
"J. Clarke" <nos...@nospam.nospam> wrote:

>>Sort of like the
>>difference between a pre-WWII research radar, and the modern US air
>>defense net.
>
>This is fine if you can afford to dedicate a station to this task. But how
>about if your ship is out in the boonies? Is is going to tow your station
>along everywhere it goes?

What, exactly, are the "boonies" when a pair of observation stations in
Earth orbit can scan the entire solar system?

>In my first post on this topic, which you obviously neglected to read, I
>stated clearly that the infrared method _might_ work in a star system with
>an established industrial base, but not for a ship operating alone.

You are apparently the only person in this entire discussion which is
assuming inter_stellar_ travel, which is an entirely different ball game
from the interplanetary war the rest of us are talking about.


Ian

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
"J. Clarke" <nos...@nospam.nospam> wrote:

>First item, how long an exposure is required to achieve these results.
>Second item, what is the probability that a spacecraft operated by someone
>who understands the capabilities of your sensors is going to be doing one of
>these things at the exact time that you have the sensor pointed at it?

They don't know when our sensors are pointed at it (which is likely once
an hour or more frequently - with multiple ships and multiple
observation stations, quite plausibly, once a minute). Meaningful
acceleration requires _sustained_ burn. Not to mention that the exhaust
plume of any system that leaves one, doesn't cool instantly. And that
powerful enough sensors will detect any spaceship with running power.


Steve Hix

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to

> > >> Asteroids are in thermal equilibrium. They won't show up on IR.
> > > >> Spaceships cannot be in thermal equilibrium, they must radiate in IR.
> > >
> > >No.
> > >
> > >(Hint: asteroids are typically not very reflective. Where
> > >does the energy absorbed from sunlight go?)
> >
> > It warms them up until they radiate heat as fast as they absorb it. This
> > condition is called 'thermal equilibrium'. Ooh! Look! That's the condition
> > that I said that they would be in.

You don't understand equilibrium, apparently.

It does *not* mean "no emission", nor does it mean "emitting at the
3 degree Kelvin level" (in this case).

All it means is that the body is at equilibrium, it isn't changing
it's temperature. It's a steady state condition.

The moon can be looked at as a (really big) asteroid...does it
show up in IR?

Brett Evill

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
> 2) "heat superconduction" has been claimed here to be a myth of Niven's
> unrelated to reality, and it seems believable. Heat by its nature is
> distributed between all sorts of random vibrational modes - how can it
> move QUICKLY, without pausing or getting sidetracked?

I find that when I stir a cup of tea with a sterling silver spoon my
finger get hot a lot more quickly than when I stir a cup of coffee with a
stainless-steel spoon. Perhaps it's something to do with the thermal
differences between coffee and tea.

Sarcasm aside, a spaceship threatened by a laser weapon will probably spin
to spread the heat out, and it will probably have a shiny surface to
reflect the beam away (will this be a few atoms thickness of lithium of
half an inch of silver?).

Another obvious strategy is to use a hull material with high thermal
conductivity so as to spread the heat out through a large thermal mass and
over a large radiating surface. The people who make strongboxes use a
layer of copper under the hardened stell shell to accomplish this. Is
there something that would be better than copper for filling this role on
spacecraft?

Tom Breton

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to J. Clarke
"J. Clarke" <nos...@nospam.nospam> writes:

>
> Ian wrote in message <3602278f...@news.ktchnr1.on.wave.home.com>...
> >"J. Clarke" <nos...@nospam.nospam> wrote:
> >

> >>Well, given that the astronomical community has been about it for a rather
> >>long time and not completed the job, I seriously doubt that anyone is
> going
> >>to come up with some magic sensor that can do it in thirty seconds. And
> >>while the paid astronomers aren't for the most part hunting for asteroids,
> >>there are thousands of amateurs doing so, with considerable success. The
> >>nature of the problem doesn't change because the military is working on
> it.
> >

> >No, the scope of resources dedicated is what changes. By a LOT. The
> >telescopic array of a single spaceborn, dedicated IR observation station
> >would easily exceed the combined capabilities of every telescope on
> >Earth ever used for searching for near-Earth objects.
>
> Why do you keep harping on "near earth objects"? They are not the only
> asteroids out there.

Because earth-crossing asteroids are the ones that are most heavily
being searched for. They are the most important to find, because they
mite smack into Earth, an event whose significance should not be
underestimated.

Tom

Tom Breton

unread,
Sep 7, 1998, 3:00:00 AM9/7/98
to
b.e...@spamblocker.tyndale.apana.org.au (Brett Evill) writes:

>
> In article <35F3CDAC...@interaccess.com>, Paul Dietz
> <di...@interaccess.com> wrote:
>

> >> Asteroids are in thermal equilibrium. They won't show up on IR. Spaceships
> >> cannot be in thermal equilibrium, they must radiate in IR.
> >
> >No.
> >
> >(Hint: asteroids are typically not very reflective. Where
> >does the energy absorbed from sunlight go?)
>
> It warms them up until they radiate heat as fast as they absorb it. This
> condition is called 'thermal equilibrium'. Ooh! Look! That's the condition
> that I said that they would be in.

Yes, they are in thermal equilibrium wrt the radiative panorama
around them. Well, close to it: consider eccentric orbits and
rotation that rearranges its albedo wrt radiative sources & sinks. EG,
an asteroid that was half white and half black would have distinctly
different equilibrium temperatures depending on which part was facing
the sun. Anyways...

However, their thermal equilibrium is not the same as the 3K
background radiation, which is the backdrop it is to be detected
against.

Tom

J. Clarke

unread,
Sep 8, 1998, 3:00:00 AM9/8/98
to
Paul Dietz wrote in message <35F42679...@interaccess.com>...

>>> You can
>> >overcome this by putting cryogenic telescopes in space, but that's
>> >currently expensive. In a future with space battleships it would
>> >not be.
>>
>> You haven't demonstrated any difference that would reduce the level of
>> effort required, merely.
>
>(with annoyance) I had assumed some intelligence on the part
>of the reader. Think -- the whole spacewar scenario assumes
>a level space ability far beyond what we have today, with
>much lower cost to reach space and operate in space. That
>by itself makes space telescopes (cryogenic or otherwise)
>less difficult.

If you actually read the messages to which you were responding and thought
about the issues being raised for a while then you would be less annoyed.
The level of space ability is irrelevant. It takes x number of observations
with y sensitivity and z field of view in order to search the entire sky for
an object. It doesn't matter one iota whether what frequency the object is
radiating. The frequency affects the details of the device, but not the
logistics of the search.

Furthermore, if you had actually read the messages you would see that I have
stipulated, several times now, that an incoming spacecraft could be detected
by the sort of sensors that would be present in a _developed_ star system.
The sort of sensors that a single ship could carry would be quite different,
unless you are suggesting that all the military resources of a civilization
would be present in a single ship.

>> >The number is a small fraction of the astronomical community,
>> >both in number of observers and fraction of funding. And,
>> >they're using visible light.
>>
>> Which is irrelevant, unless perhaps you can demonstrate that is easier to
>> build an IR sensor that can scan wide areas and distinguish a spaceship
from
>> all the other emitters than it is to build a visible spectrum sensor to
>> perform that task.
>
>No, it is not irrelevant. The small effort being expended
>on asteroid searches shows that you can't use current results
>as a limit of what a larger scale program would do.

A red herring sir. Tell me why the _logistics_ are different for IR.
Remember that to reliably detect incoming spaceships this "small effort" has
to be repeated every few minutes. That makes it a _much_ larger effort.

>> >Imagining that current astronomical technology will still
>> >be used in a future where interplanetary battleships are built
>> >is an interestingly selective failure of the imagination.
>
>> I'm sorry, but optics is optics. If you want to postulate some
baloneytron
>> that detects spaceships by magic, that's fine, but if you are going to
claim
>> that they can be detected by infrared then you're going to have to make a
>> case for it, which you have not.
>
>Bigger apertures, better sensors, better computers. All
>of this is easier than building space battleships.

How do bigger apertures help? One can see most asteroids with a back yard
telescope. How do better sensors help? The ones that your typical backyard
astronomer can buy are sufficient to the task. How do better computers
help? Computers can't collect data, just interpret it.

>
> Paul

J. Clarke

unread,
Sep 8, 1998, 3:00:00 AM9/8/98
to
Brett Evill wrote in message ...
>In article <35f3f...@news1.ibm.net>, "J. Clarke" <nos...@nospam.nospam>
wrote:
>
>>Asteroids are not warmer than "background"? What exactly do you believe
the
>>temperature of an asteroid to be and what do you believe the temperature
of
>>the background to be?
>
>The IR signature of the cosmic background is equivalent to a black body
>radiating at an effective temperature of 3K.
>
>Thermal equilibrium at the orbit of the Earth is about 260-270K. It varies
>inversely with the square root of distance from the Sun.
>
>Black-body radiation varies with the fourth power of the temperature, and
>the frequency of the maximum on the power curve is proportional to the
>temperature.
>
>It would seem that the obvious way to distinguish stars, inert lumps, and
>working spacecraft is to measure their brightness (CCD cameras will do
>this) at several different wavelengths in the IR and visible. This will
>allow you to work out roughly at what wavelength the maximum of power is,
>and that will let you sort them into temperature categories.

So let's see, basically you're going to take spectra with several different
instruments? How long is it going to take you do this for the entire sky?

>Of course, this is hard to do with ground-based instruments because the
>atmosphere is largely opaque to IR, especially in the longer wavelengths,
>and because all of your instruments are sitting in a warm bath of air. To
>detect deep IR photons, you really need to use instruments that are too
>cool to emit much thermal radiation at the relevant frequencies. But
>cyrogenic cooling is conparatively easy in vaccuum.
>
>It certainly seems to me that detection of working spacecraft can be made
>rapid and mechanical. The only problem I can see it that if the sweeping
>telescopes have a narrow field of view a complete 4.pi steradian sweep
>might take a while. Can anyone think of a reason why the field of view
>need be small or the sweeping slow.

The larger the field the more objects are in it. The rate of sweep is going
to be controlled by the amount of time necessary to secure enough data for
your analysis. For your spectrum you'll need multiple images with different
filters as a minimum, or else you'll need to do spectrographic analysis of
every object in the field. How long does that take?


>
>--
>Brett Evill
>
>To reply by e-mail, remove 'spamblocker.' from
<b.e...@spamblocker.tyndale.apana.org.au>

--

J. Clarke

unread,
Sep 8, 1998, 3:00:00 AM9/8/98