One precise burn - possible?

[laukejas]

Hi,

I've been thinking, if interplanetary travel in Orbiter is done with corrections (usually), these corrections correct previously inprecise burns. What if very precise burn would be made when departing, for example, Earth - Mars?

Just imagine: launch, orbit ascent, and direct departure in one single burn, but with such precision, that ship would arrive at mars with correct inclination and distance, precise enough for direct reentry to base.

Of course, it would require very, VERY precise burn. But is it possible? What do you think?

 

[Arrowstar]

The problem is more in that we have an imperfect knowledge of how gravity affects spacecraft. The "point mass" approximation is just that, an approximation. In reality, you get additional, higher order terms in the gravity force equation that reflect the 3D nature of the body in question. Those terms are not necessarily easy to compute, and hence we have an inaccurate (though very good!) model of solar system dynamics. Thus, our mid-course correction maneuvers are necessary to keep the arrival point on target at wherever it is we're going.

(That being said: I am aware of the probabilistic nature of thruster/engine firings, and that certainly comes into play here, too. I simply mean to say that gravity models aren't perfect, either. )

 

[laukejas]

That's what I'm saying! If gravity predictions could be made more accurate, that is, including almost all bodies in solar system (since such precision can be affected by even gravity of far stars), could it be possible?

 

[Arrowstar]

It's not a matter of including all the bodies in the solar system, though that might help. The problem is two-fold.

First, we can't compute all the higher order gravity force terms (and there are an infinite number of them). It's just not feasible.

Second, thrust firings are not precise. You have mechanical valves, fluid flow, all that. The amount of matter you expel through those engines/thrusters is not exactly what you want: you can only get close. That's why I said it was a probabilistic problem. You may have the highest probability of burning some amount of propellant, but you can't target that to an infinite accuracy.

In short, you're always going to be off. It's not a matter of perfection, it's a matter of understanding how wrong you can be and still be "right".

 

[laukejas]

Well, of course infinite accuracy is not possible - but I'm talking about enough precision, for my example - Earth to Mars trip, with correct inclination ant altitude at arrival, so deorbiting is possible without any burns.

Talking of engines, this problem can be solved by implementing micro-RCS engines, designed for micro translation burns. Smaller, less powerful engines would provide much better accuracy.

 

[IgnoreThisBarrel]

This is a question I've asked myself all the time. If the gravitational models would some day be accurate enough to safely accomplish everything in one burn (maybe several more for orbit insertion and landing on non-atmospheric bodies) it would be truly amazing.

 

[Keatah]

This is a question I've asked myself all the time. If the gravitational models would some day be accurate enough to safely accomplish everything in one burn (maybe several more for orbit insertion and landing on non-atmospheric bodies) it would be truly amazing.

As far as orbit insertion, you need to be going at x speed. And x speed is usually much much slower than y speed. y speed is your interplanetary cruise velocity. Unless you're gonna travel real slow to your destination, you'll need a burn on arrival. Or you will need an atmosphere to decelerate, aerocapture!!

 

[Quick_Nick]

Think about this: TransX or IMFD could plan perfect burns if they used the same methods as Orbiter. (But then where's the fun in that?) But to do this, planning would take far far far longer than the realtime flight.
You can get closer and closer estimates but it's going to take that much more sacrifice of time. Besides, course corrections are usually a few m/s dV.

 

[Artlav]

In Orbiter - possible, all you need is to calculate the burn relative to the model Orbiter uses and the time accelerations that would be set.
Also, fixed timesteps are necessary - that makes the simulation totally deterministic and predictions possible.
Then, you'll just burn till your vectors converge with the prediction, and off you go to the coin on Mars.

I wonder what numerical precision would there be at the end of that trip, we'll obviously be able to get through only quite grainy targets due to it.

In real, such trips are impossible with current understanding. You'll need to find an exact solution in a chaotic problem with unknown unknowns.