KSP orbital mechanics divergence from real life

[NERVAfan]

How does the way KSP calculates orbits differ from reality? I know there are no Lagrangian points...

Is it just that KSP uses pure Keplerian orbits (for on-rails objects) IE no orbital perturbations from other bodies (so Mun gravity won't influence stuff in Kerbin orbits), exosphere drag/orbital decay, radiation pressure etc.?

[lammatt]

KSP only simulates a two-body interaction.

while people say it'll be asking too much for a domestic computer to sim multibody interactions, i would say it's hard to write good efficient codes to do the simulation.

Edited March 20, 2014 by lammatt

[MalfunctionM1Ke]

KSP is about fun-based, not realism-based. Having your Ships flip out of their Orbits when you dont look at them, might be realistic but not funny at all. But I would love n-body-physics just for testing.

[MockKnizzle]

I think the problem with N-body orbital simulations is when timewarp is involved. Realtime simulation isn't really an issue for normal computers, but when you're warping at 10,000x normal speed on the same machine you lose a ton of precision in the integration.

[SpaceSphereOfDeath]

I think the problem with N-body orbital simulations is when timewarp is involved. Realtime simulation isn't really an issue for normal computers, but when you're warping at 10,000x normal speed on the same machine you lose a ton of precision in the integration.

You can see this problem is you have a copy of universe sandbox, as simulation speed increases, bodies closer to the parent start to become less and less circular, even down to triangular orbits, and then they shoot off into infinity.

[yaur]

While the perforamce issue I think is the main reason... you should go play http://deltav.corpsmoderne.net/ and think about how not fun that would be with multiple missions going on. Anyway to answer the OP's question ksp:

uses patched conics instead instead of true n-body

ignores all forms of acceleration other than the power you provide and in atmosphere drag

has floating point errors

doesn't properly conserve momentum when shifting your center off mass in orbit

i'm sure there are other little details that I have missed.

[cantab]

By the standards of creative works in general, the orbital mechanics in KSP is very good. Though that says more about creative works in general.

The planets and moons are "on rails" with no physics simulation at all. Their orbits are in accord with Kepler's laws, but effects like the Sun's orbit around the solar system barycentre, and the precession of moons' orbits, aren't implemented. On the plus side, this ensures the system is indefinitely stable.

Spacecraft use the Patched conic approximation. It's pretty accurate in the short term for most trajectories, but it doesn't factor in Lagrange points, perturbations from other bodies, or the effects of the primary being non-spherical and non-uniform. (Non-uniformity of the real Moon created challenges for the Apollo missions.) On the plus side, this means a closed orbit above the atmosphere and clear of any other SOIs will stay put indefinitely; drifting orbits would be a PAIN.

[Tidus Klein]

If you really want to know what the difference is or just looking to experiment i know that there is a mod in progress for N body physics

[jfull]

Technically, KSP doesn't need n-body physics.

In Kerbal orbit, it only needs to account for the gravity of Kerbin, the Mun, and Minmus.

In Jool orbit its a bit more complicated, but the principle is the same. You don't need n-body equations, you can do it with a few finite-number-body-equations.

[Nemrav]

while it does sound neat lagrange points, and other possible fuel saving things with n-body physics. It sounds pretty infuriating depending on how far out of a body's gravity well you are....

[cantab]

If we did have proper n-body physics, then the Interplanetary Transport Network would be a complete Disk One Nuke in Career Mode. Imagine your first mission, or at least your first mission after getting probes and solar cells, being a Moho, Eve, Duna, Dres, Jool, and Eeloo flyby - all in one.

[Kasuha]

If we did have proper n-body physics, then the Interplanetary Transport Network would be a complete Disk One Nuke in Career Mode. Imagine your first mission, or at least your first mission after getting probes and solar cells, being a Moho, Eve, Duna, Dres, Jool, and Eeloo flyby - all in one.

You can do grand tour using gravity slingshots using current game mechanics with no problem. It's actually easier than to do that with N-body gravity.