Dead Horses - Discussion

Loren Pechtel wrote:
The thing is, you need not only your location but velocity. If you fire at a target a light-hour away you need to know your velocity relative to the target down to several meters per hour.

Why? A photon moves at C whether the emitter is at at rest or moving at .95C.

kzt wrote:

Loren Pechtel wrote:
The thing is, you need not only your location but velocity. If you fire at a target a light-hour away you need to know your velocity relative to the target down to several meters per hour.

Why? A photon moves at C whether the emitter is at at rest or moving at .95C.

Well, you do need to know your lateral velocity.

Though that's why you do this remote sensing. As I've argued, in any system that has something worth shooting at, determining your exact position and velocity relative to known sources inside that system is not difficult at all.

kzt wrote:Orbital data for most interesting objects will be easily available in navigational databases to milimeter accuracy. And you are not going to fire one shot and leave. You are goong to saturate the area to cover your error predictions. I have no real idea how fast rapid fire on a graser is, but lets assume two shots per second. You have say 20 weapons shooting at that rate for say 5 minutes, then you hyper out and await the tragic news to arrive. The odds are that you’ll get multiple hits, possibly a lot of hits, on a multi-km sized object and probably ruin a lot of people’s weekends.

Also to consider: just getting a low power graser hit is really bad if you are a mammal. Intense gamma rays are bad, and it doesn’t have to be nearly powerful enough to do any physical damage to a space raft to kill everyone inside.

Though the low power you'd need in the Honorverse is presumably much higher than the low power gamma ray hit it'd take to, say kill everyone on the ISS tomorrow. OTOH they have ludicrously more powerful gamma emitters so that extra power, even if short of a level destructive to gross physical objects, shouldn't be a problem.

The Honorverse does have nearly magical rad shields on their ships and I've no reason to think they wouldn't mount powerful rad shields on their stations as well -- provide protection against solar flares or semi-nearby reactor accidents. But against a graser - almost irrelevant.

Loren Pechtel wrote:Tens or hundreds of meters? We can survey locations far more accurately than that!

Yes, but my point was that a VLB doesn't need the exact position of each telescope since that is solved when you number-crunch the measurements from each telescope.

kzt wrote:Orbital data for most interesting objects will be easily available in navigational databases to milimeter accuracy.

Uhm, no. Calculating orbits will never be accurate, because any planetary orbit is never stable enough to be predictable enough to get millimeter accuracy. In Honorverse, someone lighting off their impellers will throw all your calculations off.

Doing orbital mechanics with multiple bodies is impossible to solve accurately, just go and look up the n-body problem.

Joat42 wrote:Doing orbital mechanics with multiple bodies is impossible to solve accurately, just go and look up the n-body problem.

Very timely: PBS Space Time had da bonus episode this weekend that was about Solving the Three Body Problem.

TL;DR: you can't solve it analytically, but you can do sufficient numerical integration to calculate the positions of the bodies in a star system in a million years' time.

Joat42 wrote:Doing orbital mechanics with multiple bodies is impossible to solve accurately, just go and look up the n-body problem.

ThinksMarkedly wrote:Very timely: PBS Space Time had da bonus episode this weekend that was about Solving the Three Body Problem.

TL;DR: you can't solve it analytically, but you can do sufficient numerical integration to calculate the positions of the bodies in a star system in a million years' time.

The video states that the three-body problem can be accurately approximated for reasonable times by numerical iteration (and integration) and a phase space analysis can shed light on what the resulting two-body system might be when one of the bodies is ejected. However a solar system is NOT a three-body system (ours can be considered that, if you ONLY include the Sun, Jupiter and Saturn), so even the position of the Earth and its moon cannot be accurately positioned in anything like the time scale you suggest.

Note that the smaller the time slice you use the more accurate you are at figuring position after a given period of time. I cannot image what size time slice and how much computing time you would have to use to get the correct result for our solar system after just a few years time. The reason we know our orbits so well is that we have had more than a century of observations, so we can use the observed results rather than recomputing them each time.

tlb wrote:

Joat42 wrote:Doing orbital mechanics with multiple bodies is impossible to solve accurately, just go and look up the n-body problem.

ThinksMarkedly wrote:Very timely: PBS Space Time had da bonus episode this weekend that was about Solving the Three Body Problem.

TL;DR: you can't solve it analytically, but you can do sufficient numerical integration to calculate the positions of the bodies in a star system in a million years' time.

The video states that the three-body problem can be accurately approximated for reasonable times by numerical iteration (and integration) and a phase space analysis can shed light on what the resulting two-body system might be when one of the bodies is ejected. However a solar system is NOT a three-body system (ours can be considered that, if you ONLY include the Sun, Jupiter and Saturn), so even the position of the Earth and its moon cannot be accurately positioned in anything like the time scale you suggest.

Fair enough. But if you've got an energy weapon with a light-hour range you only need to accurately predict the target's orbit an hour in advance - so inability to calculate for millions or years, or even a single year, doesn't really matter.

Plus as others pointed out you're not restricted to firing this thing once. Calculate the orbit and possitional-error cloud around that target (caused by uncertainties in your measurement or probably perturbation over that time), then do the same for anything you DON'T want to hit (like the planet) and then fire a pattern of shots that pepper the possitional-error cloud of the target wherever it doesn't overlap the possitional-error cloud of anything you're avoiding hitting.

And even if you're firing at a DD in orbit that's a target that's at least a couple dozen meters across on it's smallest axis. So you hardly need mm level accuracy to begin with.

Jonathan_S wrote:Fair enough. But if you've got an energy weapon with a light-hour range you only need to accurately predict the target's orbit an hour in advance - so inability to calculate for millions or years, or even a single year, doesn't really matter.

Plus as others pointed out you're not restricted to firing this thing once. Calculate the orbit and possitional-error cloud around that target (caused by uncertainties in your measurement or probably perturbation over that time), then do the same for anything you DON'T want to hit (like the planet) and then fire a pattern of shots that pepper the possitional-error cloud of the target wherever it doesn't overlap the possitional-error cloud of anything you're avoiding hitting.

And even if you're firing at a DD in orbit that's a target that's at least a couple dozen meters across on it's smallest axis. So you hardly need mm level accuracy to begin with.

I agree that a range of a light hour is much more doable, but the post to which I replied was talking about distances of a light week (to a light month!). Note that you are using one hour old data to compute the target location in one hour time, but the measurement equipment and the computing power probably makes this practical. The advantage of using something like Mistletoe instead, is that you are not going to blow up a passenger liner, instead of your target, because it just arrived in the system within your two hour period of ignorance.

Jonathan_S wrote:

tlb wrote:The video states that the three-body problem can be accurately approximated for reasonable times by numerical iteration (and integration) and a phase space analysis can shed light on what the resulting two-body system might be when one of the bodies is ejected. However a solar system is NOT a three-body system (ours can be considered that, if you ONLY include the Sun, Jupiter and Saturn), so even the position of the Earth and its moon cannot be accurately positioned in anything like the time scale you suggest.

With enough computing power, you can integrate for more than three bodies. But even if you couldn't, you don't need to resolve for the entire star system, only for the body you're interested and other massive objects that will affect the position of your target within the margin of error.

Jonathan_S wrote:Fair enough. But if you've got an energy weapon with a light-hour range you only need to accurately predict the target's orbit an hour in advance - so inability to calculate for millions or years, or even a single year, doesn't really matter.

I was really thinking of a light-week to a light-month out. A single light hour is within the range of most systems' hyper detection sensor net. At least, any system with strong enough defences you'd try this trick at anyway. A translation a light-hour out will be noticed within 1 minute of the ship arriving; one light-week will give 2 hours and 35 minutes until detection.

And even if you're firing at a DD in orbit that's a target that's at least a couple dozen meters across on it's smallest axis. So you hardly need mm level accuracy to begin with.

You're not going to fire at a DD or any ship or anything that is moving. You could fire at mothballed ships that are in stable orbits, but an active ship, even one idling with impellers shut down could activate them and get underway in less than an hour.