Better to push or pull payload?

[JaySmoka]

Like the title says, I would like to know if it's better to push or pull thing when launching our rocket? I mean, put the payload at thee top or at the bottom of our rockets.

Please give me your opinion and explain why one way is better than other?

Thanks!

Edited July 5, 2013 by JaySmoka

[unwisemonk]

If in the case of a rocket already in orbit with multiple sections it is more stable to pull. On launch it is best to push the rocket.

[Sirrobert]

Ships that you assamble in orbit with docking ports, definitly pull.

When launching from kerbin, push them. Use struts to secure anything that might flex under the force

[HoY]

Also if you pull the payload you can use the tug to land the payload as well, without the payload having its own engines, saving part count and mass

[Vanamonde]

Pulling has some drawbacks. The experts who know about real rocket science insist that it's much less stable and much harder to steer, and my own experience has born that out. Also, the way KSP rocket engines are programmed to gimbal works backwards on engines that are placed ahead of the center of mass. Pulling can be done, but seems inadvisable.

[Astrobug]

Pulling has some drawbacks. The experts who know about real rocket science insist that it's much less stable and much harder to steer, and my own experience has born that out. -snip-

Beyond things that KSP models, I have to wonder what the effects might be of putting your payload "downwind" of your propulsion. This sounds like it would have less-than-ideal results, especially in the case of nuclear thrusters.

[Scotius]

My interplanetary tugs can do both - at once if it's needed I found i get best results when engines are on roughly the same level as CoM of the ship - it's more stable then, and easier to maneuver around. Nonetheless you will still need a lot of RCS blocks and fuel to turn around.

[numerobis]

Currently, gimballing engines must be below the center of mass, or else they gimbal backwards. That's just a bug, but there's no good workaround yet. You can mount them above if you lock the gimbals, of course.

Aerospikes are slightly better off being above the payload when flying in the atmosphere: they have a lower drag coefficient, so if you mount them on the bottom, your spacecraft is aerodynamically unstable (with the drag being ahead of the center of mass). You can overwhelm this problem with a few control surfaces.

Tractor designs are more stable when you have an articulated object, like an understrutted KSP spacecraft (particularly one held together by docking ports).

Push or pull doesn't matter when you have a perfectly rigid design.

Push designs are far easier to build when staging, since engines have an unexplainable tendency to fall downwards when you drop them.

PS: proverb from Kerbal would be more like "Mieux vaut exploser que prévenir"

Edited July 4, 2013 by numerobis

[Guest]

Beyond things that KSP models, I have to wonder what the effects might be of putting your payload "downwind" of your propulsion. This sounds like it would have less-than-ideal results, especially in the case of nuclear thrusters.

Off topic, there are some technical hurdles but the benefits outweigh them, at least from what I understand. The big one is mass; pulling is more efficient on the engine-to-payload connection. See the ship from Avatar, it's actually modeled after an antimatter powered vessel from some novel I can't remember.

Anyway, in KSP you generally want to push because of gimballing. You can probably get around that by disabling the gimballing on your engines and bringing along RCS or "torque" from command pods.

[Sirrobert]

Pulling has some drawbacks. The experts who know about real rocket science insist that it's much less stable and much harder to steer, and my own experience has born that out. Also, the way KSP rocket engines are programmed to gimbal works backwards on engines that are placed ahead of the center of mass. Pulling can be done, but seems inadvisable.

That's not really a problem in vaccum though. And nuclear engines (best for in space) don't gimbal anyway

[Vanamonde]

That's not really a problem in vaccum though. And nuclear engines (best for in space) don't gimbal anyway.

Hmm? The gimballing of the engines is the single most powerful stabilization force KSP ships have, and it actively maintains the heading of your ships in vacuum or atmosphere. Also, the LV-Ns do indeed gimbal. Watch them the next time you fly one of your ships that use them.

[Guest]

Hmm? The gimballing of the engines is the single most powerful stabilization force KSP ships have, and it actively maintains the heading of your ships in vacuum or atmosphere. Also, the LV-Ns do indeed gimbal. Watch them the next time you fly one of your ships that use them.

Confirming this. I built a puller style ship with nuclear engines thinking the gimballing alone would be enough to manuever. Let me tell you, was I ever wrong.

[Crass]

Push!: http://en.wikipedia.org/wiki/Pendulum_rocket_fallacy

[bsalis]

Push!: http://en.wikipedia.org/wiki/Pendulum_rocket_fallacy

Say hello to docking ports --- http://en.wikipedia.org/wiki/Pushing_on_a_string

For strutted rockets launching from Kerbin, pushing is much more practical and fine. However when docking crafts in orbit for transit, pulling is more stable. Just turn the gimbals off - you should not need them anyhow.

For me i started to make things modular with docking ports on top and bottom. This tug works brilliantly...

As you can see, the land tanker also has port top and bottom, so I can stack crafts into a train.

If need I can stack two of those tugs to make a double if i needed the oompf.

[Sirrobert]

Hmm? The gimballing of the engines is the single most powerful stabilization force KSP ships have, and it actively maintains the heading of your ships in vacuum or atmosphere. Also, the LV-Ns do indeed gimbal. Watch them the next time you fly one of your ships that use them.

My bad, I always asumed they didn't gimbal.

Never noticed any backward gimballing on my mun train though

[JaySmoka]

PS: proverb from Kerbal would be more like "Mieux vaut exploser que prévenir"

Like that one.

OK then, better to push when going in space, pull when in space. What about this type of design (sorry for ascii, but I cant test my concept right now)?

Dots represents nothing just to keep the shape. Imagine a rocket strapped to the left and the payload to the right. CoM represented by the + sign

_

. |

. |.... +

. |

. |_

Is that pulling or pushing? I think it's pushing since the force that make the thing going up is applied athe the bottom of the payload. What do you think (or know) about this?

[Geschosskopf]

PS: proverb from Kerbal would be more like "Mieux vaut exploser que prévenir"

Ça c'est bon, ça!

The main problem I have with tractors is that they necessarily can't have engines on the centerline. This add the complexity of needing to balance fuel in outboard tanks and/or using fuel lines, not to mention the part count increase to mount and strut the outboard engines. Thus, the rocket has more mass and, while in the atmosphere, more drag. All can be overcome, it's just more hassle.

[HoY]

You can actually have engines out on pylons away from the core without using fuel lines. Tho yes, it still uses some extra parts and struts, but as I said last time, you can use the tractor to skycrane land the payload without the payload having its own engines. That saves fuel, parts, and complexity in itself.

[numerobis]

A niche use for a tractor design: engines outboard, fuel tanks on the centerline. Then you can use half as many decouplers, or decouple twice as often, slightly increasing the deltaV you can achieve. And dramatically increasing the complexity.

[leopardenthusiast]

It's been said before, but I'm saying it again: While a pulling rocket doesn't tend to point in any given direction, it DOES tend to reduce the vessel's flexing. Now, there are downsides, and doing this on a launch vehicle is just silly, but when you're building an interplanetary spacecraft in orbit, having the engines pull is nice. There's less chance of your entire spacecraft simply snapping in half.

[JaySmoka]

Ok you answered my question. Thanks!

Push to get in orbit, pull in space.

[Vanamonde]

I don't doubt the pendulum problem itself, but the Wikipedia explanation of it doesn't sound right to me. The idea that imperfectly balanced thrust is the root of the problem wouldn't seem to explain why it's any worse at the front than the back, and if you have active steering with the thrust, it seems to me that you could average out any imbalance in the static thrust anyway. Am I missing something?

[numerobis]

The argument isn't that it's worse at the front; it's that Goddard believed (wrongly) that it's better at the front. In a rigid rocket, all that matters is the torque.

[DerekL1963]

The argument isn't that it's worse at the front; it's that Goddard believed (wrongly) that it's better at the front. In a rigid rocket, all that matters is the torque.

And that's the heart of the problem - in KSP, rockets aren't as rigid as IRL. (Plus nobody puts large engines in a 'pull' configuration IRL because protecting against their exhaust is... problematic.)

[Cope]

Is that true? I'm having some issues lifting heavy payloads because right around the time when I start the gravity turn my payload at the top weighs much more than the emptying tanks (and jettisoned stages, obviously) at the bottom. This tutorial on the wiki references pushing upwards on a broom balanced on your hand: it's much easier when the heavy part is at the top rather than the bottom.

Assuming KSP physics and parts rigidity, assuming you could avoid any exhaust damage issues, and acknowledging the backwards-gimbal issue, couldn't a case be made for pulling huge payloads into space?

I'm picturing a massive payload with fuel tanks mounted below it, with engines way up on outriggers connected to the fuel by fuel lines.

I can't test it until later today, but it seems like that should work fairly well, no?