Where do the engines generate thrust from? Problems with towing an asteroid...

[GreatPlainsDrifter]

How does the game calculate thrust? Is it based on the end of the nozzle or end of the flame? If an asteroid is attached using a grabbing unit can that block it? If so, is it because the thrust is "bouncing" off the rock or because the game is trying to calculate the acceleration inside of the asteroid?

I have a ship with 16 LV-Ns grabbed onto a B-size asteroid; running at full power there was zero velocity change. The engine nozzles themselves were clear of the rock (not touching\clipping into the rock, but the asteroid diameter is greater than the diameter of the ship), however the ends of the cones was clipping inside the asteroid. If i set to 3x-4x physical time compression I start to see thrust, also am generating dV with the maneuvering thrusters.

Unfortunately at 3x-4x physical compression my ship starts to rip itself apart .

Edited October 8, 2014 by GreatPlainsDrifter
Satisfactory solution

[UmbralRaptor]

If the thrust is pointed at another part of the "same" (stuff grabbed with the claw counts), craft, there is 0 net thrust. Check the in-game log (alt+F2) for any mention of damage from engine exhaust?

[KerikBalm]

"however the ends of the cones was clipping inside the asteroid. "

Huh? I think a screenshot would help here

[Kryxal]

It sounds like a puller design without enough clearance for the thrust to me.

[MarvinKitFox]

How does the game calculate thrust? Is it based on the end of the nozzle or end of the flame? If an asteroid is attached using a grabbing unit can that block it? If so, is it because the thrust is "bouncing" off the rock or because the game is trying to calculate the acceleration inside of the asteroid? ...

The thrust point is at the bottom of the engine's nozzle, HOWEVER if the game physics engine realizes that there is an obstruction in the way, the thrust is also counter-applied to the obstruction. If this obstruction is attached to your ship, then all your rocket is doing is trying to rip your own ship apart.

But... Kerbal phizzics can be fooled.

If you put a separation distance of about 1.5 orange tank lengths between the engine ant the obstruction, the phizzics is too lazy to detect the collision.

The reason you got thrust under X4 warp was because to phizzics operating under those conditions is like asking you to do a math quizz on a rollercoaster. Too much happening at the same time!

So, if you want to tow, not push, an asteroid...

Have a straight-line gap of open space behind your engines, at least 1.5 orange tanks length.

It helps to have the engines and the back bit separated by a decoupler, so that they are effectively in different "stages" of the rocket. Dunno why.

[EdFred]

This is the classic problem of trying to lift yourself off the floor by grabbing your ankles and lifting.

Edited October 5, 2014 by EdFred

[Xavven]

This is the classic problem of trying to lift yourself off the floor by grabbing your ankles and lifting.

LOL! Love the visual!

To the OP -- another thing you can do if you feel that MarvinKitFox's solution is cheating a little (no offense intended!) is to angle your engines outward so the thrust clears the asteroid. You lose net thrust to cosine losses, but it is a way to tow an asteroid without fooling the fizzicks engine. You can reduce the angle of your engines if you have a long "tow cable" made out of girders.

Ship
 ^
|||
/ \   engines angled outwards
 |
 |    tow "cable"
 |
###   asteroid
###
###

[eataTREE]

Something I have been wondering:

Suppose I build a "puller" type potato-fetcher like Xavven's design above, but using an actual flexible cable and grappling hook from KAS.

The cable is long enough that you can still thrust, so that's fine. But what happens when I want to change direction?

I fire my engines, they apply a force vector to the ship, ship starts moving at an angle to the original vector. But the 'roid is still travelling at the same vector. Suddenly, I have an enormous swinging pendulum attached to the bottom of my ship. That doesn't sound like a happy situation; it sounds like the whole system would start rotating with no very good way to correct this rotation using only the ship's RCS. Does this problem make a literal "tow cable" unworkable, or is there a solution I'm not thinking of?

[Sean Mirrsen]

Something I have been wondering:

Suppose I build a "puller" type potato-fetcher like Xavven's design above, but using an actual flexible cable and grappling hook from KAS.

The cable is long enough that you can still thrust, so that's fine. But what happens when I want to change direction?

I fire my engines, they apply a force vector to the ship, ship starts moving at an angle to the original vector. But the 'roid is still travelling at the same vector. Suddenly, I have an enormous swinging pendulum attached to the bottom of my ship. That doesn't sound like a happy situation; it sounds like the whole system would start rotating with no very good way to correct this rotation using only the ship's RCS. Does this problem make a literal "tow cable" unworkable, or is there a solution I'm not thinking of?

I tried it, and the biggest problem with using a tow line is the line's elasticity. You also need to line up perfectly to the asteroid's center of mass, and use at least three tow lines spaced out relatively equally both on your ship and on the asteroid, so that you have a multi-point attachment. But after you've got the setup right, you have to move slow. Slowly, very slowly, you have to pull the wires to get them taut. Then gently increase acceleration until you're going forward. If you start bouncing, your best bet is starting over - killing velocity, stopping rotation of the asteroid, and trying again. To turn, you need to remember that your ship and the asteroid are now one ship, in principle. The thrust of your engines - constant thrust, otherwise you start bouncing - is the "gravity" that keeps the connection stable. Your ship is actually the "nose" of the ship-asteroid contraption. So to turn the whole thing around, you need to strafe, with RCS, and then counter-strafe when you're at the needed angle. Slowly, again, because oscillations on KAS cables are near-impossible to stop.

But ultimately, it's doable.

[GreatPlainsDrifter]

Thanks for the help, I'll have to make my designs a little less compact.