Satellite bombs?

[Rocket In My Pocket]

So I've been toying with this idea lately, as I've heard it mentioned by others. If you aren't familiar with the concept you take a core vessel and basically cover it in small, one way satellites; then when it get's to it's destination it "explodes" and spreads a cloud of mini-sats everywhere. Neat concept, difficult in execution lol.

Does this approach have any merit? Has anyone used to it success? Is it just silly? I'm interested in hearing any and all opinions on the matter. The main advantages I see to this are mostly time saving, getting a ton of relay's in orbit in a relatively small amount of player time.

The hurdles/roadblocks I've been encountering so far are as follows:

1. No real way to "explode" it. Alternatives I've explored are SRB's on each mini-sat that send it flying away, or spinning the core vessel at high speed so it "throws" them. The first approach is a lot better but leads to issue #2.
2. Physics bubble. During my testing I noticed something interesting, if I "rode" along with the mini-sat it went farther! If I fired it off and stayed with the core craft, it didn't go nearly as far. I'm assuming the SRB's stop producing thrust as soon as it gets out of physics range. So any thrust needs to be produced as fast as possible.
3. They just don't go far enough. Space is a big place, a bunch of mini-sats in a tight cloud doesn't help much.
4. Expense. Although in theory this method should be cheap, I'm not finding it to be.
5. Direction. Invariably some of the mini-sats are thrown in a bad direction like retro, which generally doesn't help anyone. I suppose losing a few is part of the bargain though.

I've begun to wonder if an "orbital satellite gun" style approach would be better? Aiming and firing them off one by one from a "clip" to desired locations? Would take a little more time, and loses some of the charm of an actual "satellite bomb" lol. Anyways, let me know what your thoughts are!

(I know everyone likes pics in a thread, so not that it's really needed; but here's a shot of a test rig I was using.)

Obviously this is just one mini-sat launcher assembly (With a "dummy core" vessel), the real deal would have 6-12 of these at least I'm thinking.

Edited November 2, 2018 by Rocket In My Pocket

[Xd the great]

1. Someone has done this before.

2. Try throwing them out on a place between 2 moons. Gravitational slingshots may help. May.

[Geonovast]

I've seen this mentioned once before, and I think it was done at the very edge of Jool's SOI.  I think you'd have to extended your physics range big enough to account for the engines on the sats, try to go with as high thrust as possible so they burn out quick.... then, go see a movie, I guess.  I know your computer could handle it.  Might need to wait for a couple decades before it's efficient, though.

 

What's your endgame here?  If you're just hanging out in Kerbin's SOI, just do a mothership with a resonant orbit and release them one by one.

[Zhetaan]

I've not myself tried it; the closest I've come is a stack with separators or a set of three mounted on an inverted tri-coupler where the sats have enough of their own fuel to make course corrections but rely on a common booster to place them in their respective orbits.

However, I think that this sort of thing can work for the right application, and I think that your best applications are in an extremely elliptical orbit or in an interplanetary transfer.

If you release the cloud at the apoapsis of an elliptical orbit, then you have the benefit of a vast range of periapsides--of course, you have to deal with losing potentially half of your cloud to the planet--but such a cloud will disperse much more quickly.

On the other hand, if you send out a sort of shotgun blast of cubesats while still some ways out from an interplanetary target, then so long as the central core is a dummy that you don't mind sending straight into the planet, the cloud will pass around it in a sort of rough ring for the same reason that interplanetary correction manoeuvres are so inexpensive:  when you have enough distance, a pulse of even centimetres per second can make a difference of millions of kilometres at the destination.  What use this could be to you is the subject of a different discussion, but for my part, I've always wanted to try fly-by scanning with SCANSat, and this may be a way to get more than one track of coverage.  It probably will be hideously expensive, though.

[Ultimate Steve]

What I've done for Kerbin once, as a "scientific" test of migration, was taking 800 or so probes to a Mun encounter, spinning up the spacecraft, and releasing them at different times. This was before CommNet but I imagine that it should work just as well. Eventually the probes will crash into the Mun or Kerbin or orbit the sun, but unless you do a few interplanetary missions it should stay serviceable. I'd imagine doing it at Jool would produce more scatter, but constellation lifetime would be greatly reduced because of the three large moons.

[ZL647]

I'm not sure I see the point of this. What is your goal, and how does a cluster of mini sats help accomplish it? 

If we know that, we might better understand your question.

[Rocket In My Pocket]

2 hours ago, Xd the great said:

1. Someone has done this before.

Yeah, I know lol. I said that. That's why I'm trying it, it sounded interesting.

1 hour ago, Geonovast said:

What's your endgame here?  If you're just hanging out in Kerbin's SOI, just do a mothership with a resonant orbit and release them one by one.

Yeah, that's what I usually do. To be honest I was just bored and wanted to try something different even if it ended up being worse.

24 minutes ago, ZL647 said:

I'm not sure I see the point of this. What is your goal, and how does a cluster of mini sats help accomplish it? 

If we know that, we might better understand your question.

The goal is to make a satellite bomb, because. That's it, just because. I don't have a goal, I already have effective relay coverage via traditional satellites, I'm just exploring the concept because it's a potential new avenue of engineering challenges for me.

I'm not asking a question, so much as sparking a discussion on the topic.

Edited November 2, 2018 by Rocket In My Pocket

[Ultimate Steve]

This thread may be of interest:

 

[ZL647]

37 minutes ago, Rocket In My Pocket said:

 The goal is to make a satellite bomb, because

My suggestion is to just put a couple separatrons on them then. Those will complete their burns within physics range and impart a decent amount of velocity so long as the probe is light weight enough. Then let time and random encounters do the rest.

[Rocket In My Pocket]

1 hour ago, ZL647 said:

My suggestion is to just put a couple separatrons on them then. Those will complete their burns within physics range and impart a decent amount of velocity so long as the probe is light weight enough. Then let time and random encounters do the rest.

As you can see from my picture, that's my current setup. However I find squeezing enough on to get the Dv I want is harder than I initially imagined. It seems the hit box's on them are a bit temperamental, cluster them too close and kaboom! I'd like to keep each mini sat assembly as small as possible, and given that you can't stage once launched away, it grows out wider with every added separatron.

Edited November 2, 2018 by Rocket In My Pocket

[ZL647]

18 minutes ago, Rocket In My Pocket said:

As you can see from my picture, that's my current setup

From your picture, in my opinion your sats are too large. You'll give more mileage out of it if they are much smaller.

You could drop most of the structural things, drop the SAS, drop the antenna down to a smaller one. That would help a bit I think.

Keep in mind that the upper bound for performance you can achieve is whatever the performance of a single separatron by itself can achieve. The more separatrons you add, it only makes it approach that theoretical value. So, the best thing you can do is actually try to eliminate as much extra as you can.

Edited November 2, 2018 by ZL647

[Rocket In My Pocket]

14 minutes ago, ZL647 said:

From your picture, in my opinion your sats are too large. You'll give more mileage out of it if they are much smaller.

You could drop most of the structural things, drop the SAS, drop the antenna down to a smaller one. That would help a bit I think.

Keep in mind that the upper bound for performance you can achieve is whatever the performance of a single separatron by itself can achieve. The more separatrons you add, it only makes it approach that theoretical value. So, the best thing you can do is actually try to eliminate as much extra as you can.

The actual payload is just the bit at the top. Box probe, relay, solar panel; that's it.

The rest is the launcher assembly which needs another probe core/reaction wheel to keep it going straight during the initial thrust. At that point the mini-sat itself is already decoupled from the launcher assembly and is just being "pushed" along in front of it by the plate.

Appreciate the suggestions though.

Edited November 2, 2018 by Rocket In My Pocket

[Snark]

6 hours ago, Rocket In My Pocket said:

So I've been toying with this idea lately, as I've heard it mentioned by others. If you aren't familiar with the concept you take a core vessel and basically cover it in small, one way satellites; then when it get's to it's destination it "explodes" and spreads a cloud of mini-sats everywhere. Neat concept, difficult in execution lol.

Does this approach have any merit? Has anyone used to it success? Is it just silly? I'm interested in hearing any and all opinions on the matter. The main advantages I see to this are mostly time saving, getting a ton of relay's in orbit in a relatively small amount of player time.

It seems perfectly viable to me.  Bear in mind that you really don't need very much dV at all to the separate components in the cloud, if you're orbiting in a gravitationally "tight" place such as low orbit over a planet.

That's because they will drift apart and "diffuse" over time.  If you disperse your dandelion-seed cloud initially with very low relative velocity, then sure, they'll be a small tight cluster to start with-- but every orbit they'll diverge farther, and after a while they'll be smeared all over the place.

If you're dealing with a region of space that's "flatter" (e.g. high Jool orbit, or solar orbit), then yeah, you'd probably want to impart a bit of dV.  But then you odn't have to worry much about "wrong orbits" causing you to lose satellites.

My suggestion for the low-orbit scenario would be to just load up the little satellites attached with radial symmetry around a central core.  When the time comes to disperse, orient your central core so it's pointing , and then start spinning rapidly about the "roll" axis, and then stage all the decouplers at once.  Your various satellites will move apart with a velocity of a few meters per second.  Give that a few orbits, and pretty soon they'll be smeared all over the place.

[Snark]

@Rocket In My Pocket, I tried a sample ship on the "just spin and release" method, and it worked great for Kerbin orbit.

Here's a "dandelion seeds" relay distributor that I slapped together:

Puts 12 micro relay sats with one HG-5 each into Kerbin orbit, plus the central relay sat with an RA-2.  Total cost of the vessel is about 35K .

Here it is in 1000 km circular orbit, ready to deploy the microsats:

What you're looking at there:  12 microsats, arranged in three rings of 4 microsats each.  Each microsat has a battery, a couple of solar panels, an HG-5 antenna, a probe core, and a sepratron.  They're all mounted using TD-06 stack decouplers.  The staging is set up so that all the sepratrons fire together in one stage, then all the stack decouplers release in a subsequent stage.

To deploy, I just point the main craft in the direction, hit the space bar to activate all the sepratrons (which starts it spinning madly), wait until they're mostly spent, and then hit the space bar a second time to scatter all the satellites.  (This is actually overkill, I didn't really need one sepratron per microsat.  Probably could have done with only putting one ring of them with the sepratrons.)

Result:  They end up in orbits like this:

^ The above picture also includes a bit of clutter from some extraneous satellites that are unrelated to this test.  But you see that "family" of a dozen or so orbits, that all intersect at one spot?  That's where all the microsats ended up.  You can see that none of them were "lost" (i.e. none of them will crash into Kerbin, or get snagged by the Mun, or something).  And they're in sufficiently divergent orbits that after only one or two orbits, they're smeared around Kerbin basically at random and there you've got your cloud.

So yes, your idea works just fine. 

[Rocket In My Pocket]

Nice @Snark!

I'm surprised the spin and throw method works better than actual thrust, but then again; the physics bubble limitation means instant velocity is way better than something that takes time.

I just hope those HG-5's aren't folded! (Hard to tell in the pic) gonna be a big pita to unfold them one by one. Lol.

[Geonovast]

8 minutes ago, Rocket In My Pocket said:

Nice @Snark!

I'm surprised the spin and throw method works better than actual thrust, but then again; the physics bubble limitation means instant velocity is way better than something that takes time.

I just hope those HG-5's aren't folded! (Hard to tell in the pic) gonna be a big pita to unfold them one by one. Lol.

You could use some SmartParts to trigger the unfolding on staging.

[Snark]

20 minutes ago, Rocket In My Pocket said:

I just hope those HG-5's aren't folded! (Hard to tell in the pic) gonna be a big pita to unfold them one by one. Lol.

They're folded, though I suppose with some creative rearrangement you could have them unfolded before dispersal.

In my case, I didn't worry too much about it because I've got my DefaultActionGroups mod set up so that every deployable antenna automatically goes into action group 3 for toggle-deployment, by default at ship construction time.  So all I have to do after dispersal is to alternately hit ] followed by 3.  Do that rapidly a dozen times, and everything's all unfolded. 

[Gargamel]

7 hours ago, Snark said:

They're folded, though I suppose with some creative rearrangement you could have them unfolded before dispersal.

I'd guess the dispersal would be quick enough that if you placed the antennas and separators in the same action group, the sats would be far enough apart before the antenna could hit anything else.