The ARM megathread

[Kasuha]

Oh, and by the way, assuming the claw connection between the ship and the asteroid is rigid, then pulling has the same effect as pushing. Thinking otherwise means falling for well-known pendulum fallacy.

[Awaras]

He's talking about game mechanics dude..not reality.

Well, excuse me for thinking that a mission pack made in cooperation with NASA should have some basis in reality...

[AlamoVampire]

I still hope they give us a robotic arm

[KvickFlygarn87]

So, who's gonna watch tomorrow's stream? Sadly for me, Laserdome. Can someone take notes of new intel for everyone who will miss it?

[LethalDose]

Physics.

Even if you're not thrusting against the asteroid's CoM, you're giving it acceleration in the direction of your thrust as long as you avoid the ship turning around common CoM (and thrusting in different direction).

So the answer is torque.

The video provided is at best only tangentially relevant.

In the video shows an inelastic collision: all the energy of the bullet is transferred to the block in a single instant*, and then the block + bullet body fly off in a given energy state, and only experiences acceleration at the intitial impact in the direction of the bullet's original travel. The bullet does not continue to exert force on the block after the collision in a direction perpendicular to the blocks surface.

*not instanteous, but the acceleration takes place over so short a time to be negligible.

With rockets, the force is applied over time, and when the asteroid spins (as your video indicates that it would), the asteroid will be thrusted in a direction different than it was at the beginning. In the relevant case I posted about, the asteroid will eventually just spin in a circle, because the direction of thrust will also spin in a circle around the CoM. Despite all that thrust, you won't effectively change the asteroids velocity.

If you want to see this in game, build a Delta IV clone, then shut off the center booster and side booster, fire up the other side booster and see how far you get.

You may also want to check your facts RE: Pendulum rocket fallacy ("PRF"). It has nothing to do with the efficacy of the rocket pushing or pulling a payload. Instead, the PRF has to do with the stability of the rocket firing against gravity. In this case, it states that the stability rockets pushing & pulling actually are different, which seems to be in exact opposition to your statement, that there is no difference between pushing and pulling.

So I'm not sure if you were trying to say I was correct or incorrect in worrying about aligning CoT and CoM in the rocket-asteroid system, but the points you raised barely have anything to do with the discussion I'm trying to have, which is:

"Have the devs mentioned what the solution is? If so, what is it?"

Asteroids are frequently just loose piles of rubble...

And even if they aren't, I highly doubt the material they are made of would have enough tensile strength to allow you to pull the entire asteroid. You would likely end up pulling away from the asteroid with a claw full of gravel.

I see your point, but like Seshins pointed out, IRL is only relevant as far as the game simulates IRL. From everything that's been shown, it's safe to assume that asteroids will be rigid bodies without gravity, not aggregates held together by gravity.

Edited March 23, 2014 by LethalDose

[Kasuha]

The video provided is at best only tangentially relevant.

...

I would appreciate if you took everything I wrote into account before you start commenting on it.

The instaneity of the impulse transfer is only relevant because it ensures whole impulse is transferred in the same direction. But you don't need to transfer your impulse instantly to transfer it all in one direction. There are other ways how to achieve that.

If you push your rocket engine exhaust gases in one direction, the complex you+asteroid is accelerated in the opposite direction. That's basic physics. And it does not depend on how are you arranged in space.

Edited March 23, 2014 by Kasuha

[Linear]

...If you push your rocket engine exhaust gases in one direction, the complex you+asteroid is accelerated in the opposite direction. That's basic physics. And it does not depend on how are you arranged in space.

Haven't read the above debate..But i can't believe someone is managing to argue with you?

[Daze]

How C7 could swing their Kerbal if the crawl is a solid connection without wire?

[Twreed87]

If you push your rocket engine exhaust gases in one direction, the complex you+asteroid is accelerated in the opposite direction. That's basic physics. And it does not depend on how are you arranged in space.

No one is arguing that. The problem is that when the ship turns, the direction of thrust changes. The solution you've proposed is not very efficient. You've got a stack of SAS modules nearly as tall as the fuel tank.

The answer should probably be engines with higher gimbal range. Seems like something that would be nice to have anyways, for building shuttles and stuff.

[LethalDose]

Okay, at least now it's clear that you're disagreeing with me instead of just saying words.

I would appreciate if you took everything I wrote into account before you start commenting on it.

The instaneity of the impulse transfer is only relevant because it ensures whole impulse is transferred in the same direction. But you don't need to transfer your impulse instantly to transfer it all in one direction. There are other ways how to achieve that.

http://i.imgur.com/pWuEXN4.jpg

Yes, I agree that there are other ways to achieve that. I play KSP, so obviously I am aware of that.

I did read everything you posted before commenting, and you did state that you have to avoid the ship turning around the axis. When you said "torque" you actually meant "Lots of SAS". I assumed you were talking about the torque created by applying a force to a body oblique to it's CoM, causing rotation, because that's what the linked video demonstrated.

I appreciate the picture provided to elucidate your point, but theres a big problem with scale. You show a 36 t tank, a 6 t engine and a 1.6 t stack of SAS (43.6 t) pushing an offset 4 t payload. The "pusher" section is > 10x the mass of the offset payload. So even in a system in which a 4 t offset payload is ~ 9% of the mass of the system, you needed 8 reaction wheels to offset the torque produced by not having CoT through CoM. SQUAD has stated the asteroids will be 10-30 t, which is 250% to 800% greater than the Mk1-2's mass. It would take way more reaction wheels to offset the rotational

And maybe that will your solution to the problem when the ARM comes, and it's very "kerbal" solution: MOAR SAS!!!! I think the engines in the image above behind the grapple may actually be to counter act torque.

I would prefer to just line it up as best I can with the CoM and be done with it. Ounce of prevention vs a pound of fuel, so to speak.

If you push your rocket engine exhaust gases in one direction, the complex you+asteroid is accelerated in the opposite direction. That's basic physics. And it does not depend on how are you arranged in space.

The arrangement of forces and bodies in space does depend the orientation of forces and bodies in space. That is also basic physics. If youdon't think the arrangement of the parts in space is important, then why did you put 8 large advanced SAS units on that rocket? Would those 8 reaction wheels be needed if that craft were built with CoM inline with CoT?

Anyway, I've stated my case, you've stated yours. We should let the thread go back to talking about the ARM.

I swear I started a new thread to discuss this, but maybe I was just tired.

[LethalDose]

The answer should probably be engines with higher gimbal range. Seems like something that would be nice to have anyways, for building shuttles and stuff.

That's an excellent point! I had overlooked thrust vectoring as a control method. Gimbaled engines would allow the thrust vector to be directed through the CoM, even if the axis of the pusher wasn't in line with the asteroid's CoM. Those radial engines facing the asteroid in the image look a lot like Mk 55's.

See, this is what I'm lookin' for!

[Nemrav]

I guess since we are talking about how to move these asteroids and center of thrust etc.

although it might involve blowing up your ship if not done properly, you might consider aiming for the center of the asteroid and simply ramming it with enough force... We all know how strong a tower of struts is. And that in Newtonian physics, given there is no friction (no real impressive friction in the vacuum) when object A going X speed hits object B, then object A halts and object B gets speed X (transfered from A). granted this is based off both object A and B being the same mass.... But this means if you simply ram an asteroid at enough speed in relation to it, it will move, also if the mass difference is a problem, then physics also says we can make the difference up in speed (speed = mass*accelation). in other words a ballistic suicide missile (if the asteroid can't be shattered) would change the direction of the rock and save Kerbin from being blown up...

I will probably get lectured by somebody about how bad I explained things, all the exceptions I missed, the proper terms etc.

[RocketPilot573]

Can someone remember to make a transcript of the ARM reveal that will be streamed later today/tomorrow please. It's in the middle of the night or me (and Squad), and since Twich doesn't allow quality changes, I wont be able to rewatch the stream (bad internet, I will get 5 seconds a minute at such a high quality).

[SkyRender]

I figured the solution to moving an asteroid is obvious. All you have to do is ask yourself how Danny2462 would handle it. And how would he handle it?

With a Mass Relay! (Of course!)

[RocketPilot573]

I figured the solution to moving an asteroid is obvious. All you have to do is ask yourself how Danny2462 would handle it. And how would he handle it?

With a Mass Relay! (Of course!)

Mainsails aren't powerful enough, but maybe the new engine can do it?

[LethalDose]

I guess since we are talking about how to move these asteroids and center of thrust etc.

although it might involve blowing up your ship if not done properly, you might consider aiming for the center of the asteroid and simply ramming it with enough force... We all know how strong a tower of struts is. And that in Newtonian physics, given there is no friction (no real impressive friction in the vacuum) when object A going X speed hits object B, then object A halts and object B gets speed X (transfered from A). granted this is based off both object A and B being the same mass.... But this means if you simply ram an asteroid at enough speed in relation to it, it will move, also if the mass difference is a problem, then physics also says we can make the difference up in speed (speed = mass*accelation). in other words a ballistic suicide missile (if the asteroid can't be shattered) would change the direction of the rock and save Kerbin from being blown up...

I will probably get lectured by somebody about how bad I explained things, all the exceptions I missed, the proper terms etc.

No, what you're talking about is actually legit. You're basically describing a "kinetic kill" vehicle, and according to

(Solution #2, while the ARM is really #1), it's actually one of the proposed methods for dealing with asteroids IRL. It's also similar to the the video linked by Kashua above (bullet into a block of wood). So, yeah, you could redirect the KSP asteroids by slamming rockets into them, and I guess that makes it a valid topic for this ARM megathread.

The only problem... well, A problem, with doing it this way is that you'll have very little control over how much force is applied to the asteroid, and since you're trying to capture the asteroid in a Kerbin orbit, kinetic impacts would be a suboptimal strategy, IMO.

[cantab]

Yeah, the grapple is obviously not a flexible connection, its basically just a docking port. But putting the engines in a tractor configuration instead of a push could be interesting.

Indeed. We can at least already make flexible "chains" using stacks of parts. You'd want to be lined up as well as possible, but I reckon having a little flexibility would help keep you on course. Whether it's better than simply using regular controls like reaction wheels is another matter.

[Nemrav]

No, what you're talking about is actually legit. You're basically describing a "kinetic kill" vehicle, and according to

(Solution #2, while the ARM is really #1), it's actually one of the proposed methods for dealing with asteroids IRL. It's also similar to the the video linked by Kashua above (bullet into a block of wood). So, yeah, you could redirect the KSP asteroids by slamming rockets into them, and I guess that makes it a valid topic for this ARM megathread.

The only problem... well, A problem, with doing it this way is that you'll have very little control over how much force is applied to the asteroid, and since you're trying to capture the asteroid in a Kerbin orbit, kinetic impacts would be a suboptimal strategy, IMO.

the structural pylon has a crash tolerance of 999m/s , though visibly trusses have more... it looks like asteroids are a single object which doesn't break up so...

for control over how much you push the asteroid, you don't need to know the center of mass... all you need to know is the mass... this can be done mathematically by taking "speed = mass*acceleration" and reversing it to get the mass or "speed / acceleration = mass" though I'm not sure how you determine the acceleration, it would be a method. and after determining the mass, you accelerate to the required speed in relation to the asteroid...

oh well, realistically when ARM comes out I might try it out once, but otherwise it will be me trying to dock with the claw and pulling the asteroid out...

edit : I won't discuss this method any further, feel free to hold me responsible if I do push it further...

edit 2 : its possible to get the acceleration with this : http://wiki.kerbalspaceprogram.com/wiki/Double-C_Seismic_Accelerometer

(read "usage")

Edited March 23, 2014 by Nemrav

[Insomn]

Why is Asteroid spelt "Astroid" in that?

Why is Moon Mun?

[ddavis425]

Why is Moon Mun?

Not the same thing. Astroid was a typo that Yargnit later corrected.

[Seshins]

Well, excuse me for thinking that a mission pack made in cooperation with NASA should have some basis in reality...

It does...the mission itself. Can you imagine the cpu load if the game was to create 100k small peebles per asteroid?

[Kasuha]

although it might involve blowing up your ship if not done properly, you might consider aiming for the center of the asteroid and simply ramming it with enough force...

Something is telling me that you could actually blow that asteroid up by that. Asteroid is technically a part and it will have certain impact speed limit. It probably won't be very high, otherwise you could pile up asteroids on Kerbin just by following them in freefall and deploying your own chute less than 2.5 km above terrain. But who knows, maybe that will be possible.

If that worked, though, then you don't have to aim at CoM. You can hit the asteroid anywhere with the same effect on its trajectory (assuming inelastic collision and realistic impulse transfer, otherwise you need to aim at surface perpendicular to your trajectory).

Edited March 23, 2014 by Kasuha

[Red Iron Crown]

If that worked, though, then you don't have to aim at CoM. You can hit the asteroid anywhere with the same effect on its trajectory (assuming inelastic scattering and realistic impulse transfer, otherwise you need to aim at surface perpendicular to your trajectory).

Is that how it would work? Intuitively, I would expect it to behave like billiard balls, where the relationship between impact point and CoM has a large effect on the resultant vector.

[Kasuha]

Is that how it would work? Intuitively, I would expect it to behave like billiard balls, where the relationship between impact point and CoM has a large effect on the resultant vector.

Billiard balls are a good example of elastic collisions. Balls bounce off each other. And notice that the sideways momentum of the target ball (hit off center) is compensated by equal opposite sideways momentum of the cue ball. So your rocket would need to hit that asteroid e.g. at 1 km/s and bounce off it and continue to the side at 1 km/s to give it impulse at 45 degrees instead of parallel to its original trajectory.

In real world, most impacts are inellastic. Real rocket would not bounce off the asteroid. Its parts would get deflected, but at negligible speeds compared to the impact speed and that means giving it negligible sideways impulse.

Of course, KSP is not real world and these cases may not be implemented 100% correctly in the PhysX engine.

Edited March 23, 2014 by Kasuha

[cantab]

Something is telling me that you could actually blow that asteroid up by that. Asteroid is technically a part and it will have certain impact speed limit. It probably won't be very high, otherwise you could pile up asteroids on Kerbin just by following them in freefall and deploying your own chute less than 2.5 km above terrain. But who knows, maybe that will be possible.

It could be kludged so that when the asteroid hits a part, a very high impact tolerance is used, but when it hits a celestial a lower one is used.