What I learned from ARM

[Monger]

I just caught my first class D asteroid which was on direct collision course to Kerbin. Things I learned from that mission:

• "Deep Impact" was pretty stupid. You don't need nuclear bombs to move asteroids, you just have to catch them early enough and give them a little nudge.
  
• "Armageddon" was insanely stupid. For several obvious reasons, but especially that if the Earth vista is already filling the Asteroid's horizon, you are WAY too late.
  
• kind of shocking: even if you discover an Asteroid really early on (say, a year before impact), it might be practically impossible to find a Rendez-Vous point. My class D asteroid had a really benign orbit (same orientation as Kerbin's orbit, only slightly crossed), and still I just caught it like 40 days earlier (of 170 days discovered before impact)
  
• Generally speaking: finding a proper interception path is not simple. I am so used to more or less planar orbits, and that I can choose where to intercept my target, I really had to rethink how and where to intercept.
  

[MKI]

Well i have heard NASA said a last ditch effort would be firing mutliple nukes into the sky regardless of the implications. As the massive energy released by firing multiple nukes COULD work in destroying an asteroid. But then again firing multiple nukes anywhere isn't a smart idea and might just make things worse, but again its the last possible choice.

Do not forget, both movies had asteroids MUCH larger than what is currently in KSP. E class asteroids still are small and probably wont do much if one hit earth or kerbin. If i remember correctly, Armageddon's asteroid was the size of texas.

PS My issue in Armageddon is wouldn't it be easier to teach astronauts how to deep drill, rather than teach deep drillers how to be astronauts? (answer: Micheal Bay film)

[Jouni]

I once did some calculations on what could actually be done in an Deep Impact-type scenario for an RPG campaign. With some assumptions (nuclear pulse propulsion actually works; politics, economics, and logistics are efficient; etc.), the asteroid could actually be deflected, if detected 5-10 years in advance.

[The Destroyer]

Well i have heard NASA said a last ditch effort would be firing mutliple nukes into the sky regardless of the implications. As the massive energy released by firing multiple nukes COULD work in destroying an asteroid. But then again firing multiple nukes anywhere isn't a smart idea and might just make things worse, but again its the last possible choice.

Do not forget, both movies had asteroids MUCH larger than what is currently in KSP. E class asteroids still are small and probably wont do much if one hit earth or kerbin. If i remember correctly, Armageddon's asteroid was the size of texas.

PS My issue in Armageddon is wouldn't it be easier to teach astronauts how to deep drill, rather than teach deep drillers how to be astronauts? (answer: Micheal Bay film)

Size of Texas. AKA the size of Kerbin

[Monger]

Do not forget, both movies had asteroids MUCH larger than what is currently in KSP. E class asteroids still are small and probably wont do much if one hit earth or kerbin. If i remember correctly, Armageddon's asteroid was the size of texas.

True. Still: if you are able to reach the asteroid, you should have enough time on the Asteroid's surface. Not only a few hours, but rather weeks. Attaching a few nuclear drives that continuously nudge it out of harm's way should be much more efficient than a single undirected nuclear blow. It is no fair comparison, but I needed to move my asteroid about 10 m/s to move it away from a center hit to a near miss (i.e. Periapsis at 400km). Of course, this will get increasingly more difficult with bigger objects, but you don't have to knock it around like a bowling pin. Getting there is probably the hard part.

[Scotius]

Size of Texas. AKA the size of Kerbin

Which makes the movie even dumber. Smallest asteroid found so far was about 10 meters in diameter. Texas is roughly 1200 km wide. Body this big would be a dwarf planet, and visible decades prior to impact.

[Jouni]

There is a nice estimate in the Wikipedia: "It has been estimated that a velocity change of just 3.5/t × 10^{−2} ms^{−1} (where t is the number of years until potential impact) is needed to successfully deflect a body on a direct collision trajectory."

If we had a tiny ship with a million tonnes of fuel, we could deflect a nice 10 km asteroid a few years in advance, if the Isp of its engines was at least 1500 s.

[Streetwind]

Indeed, the earlier you get to an asteroid the easier it is to deflect it. If you get there early enough you only need the most minute of nudges... you may in fact not even have to touch the asteroid.

That's actually one proposed IRL solution: something called a gravity tractor. In contrast to KSP, a real asteroid will in all likelyhood be spinning around at least one axis, so attaching a tug spacecraft will be nearly impossible. However, if you have enough time, you don't have to. Simply intercepting the asteroid might be enough if you can make your spacecraft big enough. Then, all you need to do is keep station a dozen meters or so beside the asteroid for a year or two, and the attraction of gravity between the asteroid and the fairly massive spacecraft will apply the small few cm/s dV you need to turn a hit into a near miss.

Of course, this implies that we're able to put a spacecraft into orbit that has enough dV to intercept a rogue asteroid as far away from earth as possible and then, when it has succeeded in doing so, still weighs 50-100 tons. That is a very tall order even with rockets like the Falcon Heavy and the SLS coming up. It would probably be something that would require multiple launches and assembly in orbit, along with a non-chemical propulsion solution. However, I'd wager that if an approaching "planet killer" was detected, we'd be able to pool the resources of our various spacefaring nations to pull it off. With the Ad Astra VF-200 we have a production-ready (if little tested) plasma thruster available too. The question then will not be one of cost or launch capability, but much rather one of "can we build it quickly enough and get it there early enough".

[Kasuha]

• kind of shocking: even if you discover an Asteroid really early on (say, a year before impact), it might be practically impossible to find a Rendez-Vous point. My class D asteroid had a really benign orbit (same orientation as Kerbin's orbit, only slightly crossed), and still I just caught it like 40 days earlier (of 170 days discovered before impact)
  
• Generally speaking: finding a proper interception path is not simple. I am so used to more or less planar orbits, and that I can choose where to intercept my target, I really had to rethink how and where to intercept.
  

I found asteroid rendezvous rather simple if you're not stuck to equatorial orbits.

- put your ship to launchpad

- select the asteroid as target

- time warp until the target is on horizon of the navball

- launch and perform gravity turn towards the target (or towards target retrograde if it's the more eastern one). This will end up in an inclined orbit approximately matching asteroid's inclination relative to Kerbin. It will probably not match its trajectory within Kerbin SOI but you don't need to care about that.

- put a maneuver node on the orbit on the side away from the target and add just enough dv to exit Kerbin SOI. The exit direction is generally towards the asteroid. Then fine-tune it for decent rendezvous (10,000 km nearest approach is ok).

- exit Kerbin SOI and perform classic orbital rendezvous using navball.

In fact, I did not really learn anything substantial with asteroids as they work more or less exactly how I expected.

In contrast to KSP, a real asteroid will in all likelyhood be spinning around at least one axis

I don't think real object can be spinning around more than one axis, physics laws don't allow that IMO - unless the spin is forced. Yes, real asteroids are spinning and the axis may be random. But there's just one axis for each.

Edited April 8, 2014 by Kasuha

[Twreed87]

"Armageddon" was insanely stupid. For several obvious reasons, but especially that if the Earth vista is already filling the Asteroid's horizon, you are WAY too late.

Did you know NASA actually train their employees using this movie, by have them point out every scientific mistake that's made? There's a lot of them.

An object as big as the one described would be hundreds of times as bright as any star in the sky way before it got as close as it is in the movie. Plus, a nuke dug 800 feet into a 1200km wide rock is going to split it in half? C'mon!

[Comwarrior]

PS My issue in Armageddon is wouldn't it be easier to teach astronauts how to deep drill, rather than teach deep drillers how to be astronauts? (answer: Micheal Bay film)

You can train a normal physically fit person to be an astronaut in less than 2 weeks... provided their not required to use any space craft systems...

But... you'd never get travel insurance...

[Streetwind]

I don't think real object can be spinning around more than one axis, physics laws don't allow that IMO - unless the spin is forced. Yes, real asteroids are spinning and the axis may be random. But there's just one axis for each.

That's why I said "at least one". It allows for all scenarios

[Rakaydos]

Nukes arnt supposed to crack apart an asteroid. It's more an impromptu Orion drive, without the need for pusher plates. You shoot a nuke at the asteroid , and the blast applies a few M/S of delta V to the target. repeat until nuclear arsonals are depleted.

[Rune]

I just caught my first class D asteroid which was on direct collision course to Kerbin. Things I learned from that mission:

• "Deep Impact" was pretty stupid. You don't need nuclear bombs to move asteroids, you just have to catch them early enough and give them a little nudge.
  
• "Armageddon" was insanely stupid. For several obvious reasons, but especially that if the Earth vista is already filling the Asteroid's horizon, you are WAY too late.
  
• kind of shocking: even if you discover an Asteroid really early on (say, a year before impact), it might be practically impossible to find a Rendez-Vous point. My class D asteroid had a really benign orbit (same orientation as Kerbin's orbit, only slightly crossed), and still I just caught it like 40 days earlier (of 170 days discovered before impact)
  
• Generally speaking: finding a
  proper interception path is not simple. I am so used to more or less planar orbits, and that I can choose where to intercept my target, I really had to rethink how and where to intercept.
  

Hey! I pretty much already knew those things, but getting it done was very instructing too. Not the same having read the theory as playing around with a system and experimenting on what variations do.

Plus, I also got a stark reminder about why inclination is important: I launched my first tug into an equatorial orbit, and I had to change the objective of the mission because the delta-v hit to correct inclination on the kerbin escape burn ate more than a km/s out of my 2km/s kerbin escape stage. So as Kashua says it's crucial to launch into a more-or less well inclined orbit, cosine losses get really steep once you get far from ideal, even if they are very forgiving of small inaccuracies.

Other than that, I didn't have many problems to hit <50km rendezvouses from low kerbin orbit into solar orbit, with weeks to change their trajectories before they come into Kerbin's SOI. I'm getting good at this! My trick foe judging inclinations was to look at the projected kerbin pass of the asteroid, though. If it shows as coming from above, I would aim to launch into a polar orbit that allowed me to escape outwards from the sun while going up to meet the asteroid. I'm probably picking up lots of relative vertical velocity... but I also have days to adjust speed with the rocks before I grapple them!

Rune. It's 20% theory, 80% practice until you "get" the concept intuitively.

[Guest]

Well, if you did shove a nuke into a hole (either dug or simply made by smacking it real hard into the rock), you'd would probably achieve good results as long as the asteroid was solid. Not really cracking it, but firing off a high-energy jet of vaporized material and particles. That should pretty much give the asteroid all the dV it needs to miss Earth. Not only that, you could probably get it into an usable orbit with some well-planned nuking. In fact, a single bomb optimized for propulsion would deflect it, 2 or 3 would probably be enough to orbit it. Another one could be used to correct the inclination for easy launches.

For a pile-of-gravel type asteroid, it'd be somewhat more difficult, though if it was big enough, perhaps it'd hold together long enough to be deflected.

[KerbMav]

wouldn't it be easier to teach astronauts how to deep drill, rather than teach deep drillers how to be astronauts? (answer: Micheal Bay film)

Train someone to work in space or to work a space ship?

Astronaut is the same as (or even less than) a sailor, he will have basic knowledge, but not necessarily be able to navigate the ship.

You shoot a nuke at the asteroid , and the blast applies a few M/S of delta V to the target.

Well, if you did shove a nuke into a hole (either dug or simply made by smacking it real hard into the rock), you'd would probably achieve good results as long as the asteroid was solid. Not really cracking it, but firing off a high-energy jet of vaporized material and particles.

This is also theorized to be possible without using nukes: Mirrors bundling sun light and heating the surface of an asteroid - most likely more viable to use on comets though.

[Guest]

Mirrors would only be good on comets, heating a solid rock like that would take time. Probably quite some time indeed. And if it's close to Earth for an orbital insertion, it won't do (though you could probably deflect it that way, if you got to it early enough). The only problem with nukes is that they're heavy. Current ICBMs can't put their usual load into orbit, to do that, it'd be necessary to use a HLV (probably hijacking a commercial launch, because normal scheduling takes a few years).

Also, note that braking by an asteroid is a nontrivial problem. It's closing in with the Earth, so you first have to take off Earth, accelerate to escape, the brake from that speed and then match the asteroid's velocity. In KSP, it's easy, but IRL, dV are multiplied by a lot. No such problem with a nuke, you can impact the asteroid at full speed, either from the front or from the side. Nuclear bombs have been investigated as a means of propulsion, and while I don't think it'd work for a manned craft, asteroid probably isn't going to mind.

[Kasuha]

Comet's nucleus is already absorbing sunlight with half of its surface (producing rather uneven and random thrust) so a mirror would only help if its area is comparable with the nucleus size. I don't see that very realistic, comet nuclei are usually rather big.

[RW-1]

Indeed, the earlier you get to an asteroid the easier it is to deflect it. If you get there early enough you only need the most minute of nudges... you may in fact not even have to touch the asteroid.

That's actually one proposed IRL solution: something called a gravity tractor. In contrast to KSP, a real asteroid will in all likelyhood be spinning around at least one axis, so attaching a tug spacecraft will be nearly impossible. However, if you have enough time, you don't have to. Simply intercepting the asteroid might be enough if you can make your spacecraft big enough. Then, all you need to do is keep station a dozen meters or so beside the asteroid for a year or two, and the attraction of gravity between the asteroid and the fairly massive spacecraft will apply the small few cm/s dV you need to turn a hit into a near miss.

Of course, this implies that we're able to put a spacecraft into orbit that has enough dV to intercept a rogue asteroid as far away from earth as possible and then, when it has succeeded in doing so, still weighs 50-100 tons. That is a very tall order even with rockets like the Falcon Heavy and the SLS coming up. It would probably be something that would require multiple launches and assembly in orbit, along with a non-chemical propulsion solution. However, I'd wager that if an approaching "planet killer" was detected, we'd be able to pool the resources of our various spacefaring nations to pull it off. With the Ad Astra VF-200 we have a production-ready (if little tested) plasma thruster available too. The question then will not be one of cost or launch capability, but much rather one of "can we build it quickly enough and get it there early enough".

I apologize if brought up before I posted, replying without looking at posts after yours ...

There is another school of thought as well for nudging an asteroid farther out, you attach a mylar solar reflector to it, and the actual change in the amount of light and solar wind can also provide the additional nudge over time between applied and eventual planet intercept ...

[Monger]

Plus, I also got a stark reminder about why inclination is important: I launched my first tug into an equatorial orbit, and I had to change the objective of the mission because the delta-v hit to correct inclination on the kerbin escape burn ate more than a km/s out of my 2km/s kerbin escape stage. So as Kashua says it's crucial to launch into a more-or less well inclined orbit, cosine losses get really steep once you get far from ideal, even if they are very forgiving of small inaccuracies.

I made the same mistake: launching into space on a default inclination instead of looking where and how the asteroid actually hits Kerbin. However, since I had to hit escape velocity anyhow, I just changed my inclination on a very high orbit around Kerbin, way behind Minmus. Luckily, inclination changes on high orbits are very cheap.

[RW-1]

You can train a normal physically fit person to be an astronaut in less than 2 weeks... provided their not required to use any space craft systems...

But... you'd never get travel insurance...

It's said that dunrod completed training in 1.7 ....

[NecroBones]

I found asteroid rendezvous rather simple if you're not stuck to equatorial orbits.

- put your ship to launchpad

- select the asteroid as target

- time warp until the target is on horizon of the navball

- launch and perform gravity turn towards the target (or towards target retrograde if it's the more eastern one). This will end up in an inclined orbit approximately matching asteroid's inclination relative to Kerbin. It will probably not match its trajectory within Kerbin SOI but you don't need to care about that.

- put a maneuver node on the orbit on the side away from the target and add just enough dv to exit Kerbin SOI. The exit direction is generally towards the asteroid. Then fine-tune it for decent rendezvous (10,000 km nearest approach is ok).

- exit Kerbin SOI and perform classic orbital rendezvous using navball.

In fact, I did not really learn anything substantial with asteroids as they work more or less exactly how I expected.

This is quite good advice.

What I've been doing is probably a slightly easier intercept maneuver, but requires more thrust I think.

* Focus on an existing ship/satellite that's in a nearly perfect equatorial orbit

* Target the asteroid

* Look at an ascending/descending node, and take note of the relative inclination

* Put ship on pad

* Warp until pad is in the plane of the asteroid's Kerbion-SOI trajectory

* Launch and angle the gravity turn's heading according to the inclination from above (easy with MechJeb, but you can just do the addition/subtraction yourself)

* Get into a stable orbit until you're ready to escape for rendezvous

* Wait/warp for asteroid to barely enter Kerbin's SOI

* Plot a maneuver that intercepts when the asteroid has gone about 1/3 of its approach toward periapsis, which takes into account that the asteroid is moving more slowly further out, so it's a good "half-way" point.

This requires a good amount of thrust, because you're doing a hard-turn to match velocity with the asteroid inside the SOI. And then altering the trajectory of course is happening within Kerbin's SOI rather than much earlier. But if your goal is to put it in a parking orbit rather than just make it miss the planet, you're going to be doing hard burns inside the SOI in any case.

D-class asteroid captured this way (now in a polar orbit around Kerbin):

Edited April 8, 2014 by NecroBones

[rkman]

Generally speaking: finding a proper interception path is not simple.

And even if you do, the maneuver node system might not indicate an intercept until separation is only a few km.

[Kasuha]

Oh, just one more thing:

Mun is priceless for asteroid redirection. Particularly using it for gravity slingshots. You can fix your inclination and send the asteroid towards your target for minimum dv. The less the larger is the distance where you prepare it.

[zarakon]

I apologize if brought up before I posted, replying without looking at posts after yours ...

There is another school of thought as well for nudging an asteroid farther out, you attach a mylar solar reflector to it, and the actual change in the amount of light and solar wind can also provide the additional nudge over time between applied and eventual planet intercept ...

I've seen a similar idea, to basically cover an asteroid with reflective paint to change how much the sun's light pushes on it. You would have to do it a long time in advance, but it would be one of the cheapest, simplest options