Kerbal Space Program - Orbiting Duna

[collinsmark]

... Continued from Part 4.

The Sands of Duna

Part 5: On your mark, get set, ...

Now that we're in a low, roughly circular, Kerbin orbit, we need to find the position within that orbit that we will choose for are upcoming periapsis. We need to find the place in that orbit that we will choose for all our prograde thrust maneuvers.

Recall the discussion a few posts ago around Figure 13. It's going to be right around there.

So let's set up a maneuver node there, and drag out the prograde symbol until it extends all the way out to Duna's orbit. We won't be following through on this maneuver, but we use it to choose the future, central location of our maneuvers. Figure 30 shows the map view.

[Figure 30: Maneuver node to help pinpoint location of new periapsis around Kerbin's orbit.]

Now zoom in and make sure that the projected path is roughly symmetric around Kerbin's solar orbit. The goal is to adjust the maneuver node around the low Kerbin orbit such that the projected orbit always stays outside, on other side, of Kerbin's orbit around the sun.

[Figure 31: Make sure projected orbit (purple, dotted line) is symmetric around Kerbin's solar orbit. The projected orbit should never cross Kerbin's orbit, on either side.]

Keeping the projected orbit as such means that maneuver's location around Kerbin's low orbit is a fine location to ensure a good Hohmann transfer orbit, as that location becomes the new periapsis.

It's okay now to reduce the prograde adjustment of the maneuver node, now that we know the maneuver node's location. We won't be able to perform the entire Duna injection/Kerbin escape burn in a single burn anyway.

[Figure 32: Location of manuever node with respect to Kerbin's day/night sides (1/2)]

[Figure 33: Location of manuever node with respect to Kerbin's day/night sides (2/2)]

Figures 32 and 33 show the maneuver node's location with respect to Kerbin's day and night transition. For a Hohmann transfer between planets, the maneuver node's location will always be roughly around there if the target planet has a higher, solar orbit than the starting planet. You might need to adjust it slightly using methods outline above, depending on the starting and target planets, initial orbital elevation, etc., but it will always like somewhere around that general location.

Adjust the prograde symbol of the maneuver node such that the burn time acceptably short. Starting at a low Kerbin orbit, I'd suggest that the burns be kept down to less than a minute. Thirty second burns will be even more efficient.

If you haven't figured it out yet, we are going to gradually increase the apoapsis over the course of several orbits. Each orbit we will burn at periapsis only.

This is the same, general strategy used by the Mars Orbiter Mission (MOM) of the Indian Space Research Organization (ISRO).

[Orbit trajectory diagram (not to scale) of MOM.]

For more information on the MOM, see http://www.isro.org/pslv-c25/mission.aspx

The idea here is that we don't have enough thrust (particularly when we use our main engine cluster) to perform the entire Duna injection/Kerbin escape burn in a single burn. But we can break it up over several obits, keeping are periapsis low, and thus fully exploiting the Oberth effect for each burn.

[Figure 34: ... Go! And we're off.]

To be continued...

 

[collinsmark]

If you change the ship's heading with yaw everyone is going to fall over. To avoid this you need to bank so that the heading can be changed with pitch alone.

I still don't buy it. Your claim would only be correct if there were some sort of gravitational acceleration that the ship is fighting, making a definite "down" direction, *AND* only if there was a force directed away from the ship's prograde velocity vector. However, there is no such "down" direction and there are no external forces slowing the craft.

The ships I describe when "banking" were in relatively open space. The occupants would not experience any forces merely by drifting at a constant velocity. The heading change, before applying thrust, could be done along any, arbitrary axis. The occupants would feel no forces besides that of a gentle rotation. [Edit: And once thrust is applied, they would pulled to the back of the seats, as they are accustomed to.]

Now, if there's some sort of "artificial gravity" inside the small ship (which doesn't exist, so I don't feel the need to comment on that much), that might even negate the rotational perception. But in any case, this "banking" rationale makes no sense.

These conventions are learned at Starfleet Command School.

Okay, I'll take your word for it.

 

[pbuk]

I still don't buy it. Your claim would only be correct if there were some sort of gravitational acceleration that the ship is fighting, making a definite "down" direction, *AND* only if there was a force directed away from the ship's prograde velocity vector. However, there is no such "down" direction and there are no external forces slowing the craft.

The ships I describe when "banking" were in relatively open space. The occupants would not experience any forces merely by drifting at a constant velocity. The heading change, before applying thrust, could be done along any, arbitrary axis. The occupants would feel no forces besides that of a gentle rotation.

A gentle rotation? Let's say the ship is 800 m long and so someone at the front of the ship is 400 m from the centre of rotation. The ship performs a yaw of 90° in 10 s. That's an average centripetal acceleration of 1 g, let alone the tangential acceleration starting and finishing the yaw - they are going to fall over.

Now, if there's some sort of "artificial gravity" inside the small ship (which doesn't exist, so I don't feel the need to comment on that much), that might even negate the rotational perception. But in any case, this "banking" rationale makes no sense.

There clearly is artificial gravity in both star trek and battlestar technology, but in order to keep people on the floor it only needs to act in one axis and provide a uniform field throughout the ship. In order to keep people from being crushed by the main thrusters (which presumably accelerate a lot faster than 1 g otherwise it would take forever to get anywhere) it also needs to act on a second axis and again provide a uniform field. In order to compensate for yaw you would need a non-linear artificial gravity field - even if this technology exists there would be no need for the additional weight and expense when you can avoid it by simply following a banked turn manoeuvre adapted from the one you learn in aerodynamic flight school.

 

[collinsmark]

A gentle rotation? Let's say the ship is 800 m long and so someone at the front of the ship is 400 m from the centre of rotation. The ship performs a yaw of 90° in 10 s. That's an average centripetal acceleration of 1 g, let alone the tangential acceleration starting and finishing the yaw - they are going to fall over.

But even if they bank, those at the front (fore) of the ship will still experience approximately 1 g, and those at the back (aft) of the ship will experience approximately -1 g. By that logic, somebody will be "falling over" regardless.

There clearly is artificial gravity in both star trek and battlestar technology, but in order to keep people on the floor it only needs to act in one axis and provide a uniform field throughout the ship. In order to keep people from being crushed by the main thrusters (which presumably accelerate a lot faster than 1 g otherwise it would take forever to get anywhere) it also needs to act on a second axis and again provide a uniform field. In order to compensate for yaw you would need a non-linear artificial gravity field - even if this technology exists there would be no need for the additional weight and expense when you can avoid it by simply following a banked turn manoeuvre adapted from the one you learn in aerodynamic flight school.

If there is "artificial gravity" within the vessel, then there is no need to perform banking to keep people from falling over. The artificial gravity would presumably compensate. (If there was such a thing. It doesn't exist in reality.)

By the way, I was referring to smaller ships "banking."

Airplanes can and do bank, because the atmospheric drag and aerodynamic lift forces can and do work together to produce a "normal"-like, external resultant force. This resultant force can be used to change the airplane's direction of motion. But in open space, there are no drag forces; there are no aerodynamic lift forces. In the old Battlestar Galactica, small spaceships flew as if they were airplanes. But they are not airplanes. Empty space lacks aerodynamics. In space, banking maneuvers are superfluous.

Here is an example of this ridiculous banking. Notice the only thrusters on the Cylon ships are in the rear of the craft -- there are no thrusters on the craft's "bottom":

https://www.youtube.com/watch?v=YPL7QPCfVZc

[Edit: Ha! watching that video brings back memories. I used to love that stuff so. Looking back on it, it's so silly. Rofl! :rofl:]

 

[pbuk]

But even if they bank, those at the front (fore) of the ship will still experience approximately 1 g, and those at the back (aft) of the ship will experience approximately -1 g. By that logic, somebody will be "falling over" regardless.

The artifical gravity can easily compensate for these forces which are in the same direction as it needs to act to keep people on the floor.

If there is "artificial gravity" within the vessel, then there is no need to perform banking to keep people from falling over. The artificial gravity would presumably compensate. (If there was such a thing. It doesn't exist in reality.)

I have already explained why completely different artifical gravity technology would be required to compensate for the (ficticious) forces in a yaw which are non-uniform in direction and orthogonal to the direction required to keep people on the floor in freefall.

Here is an example of this ridiculous banking. Notice the only thrusters on the Cylon ships are in the rear of the craft -- there are no thrusters on the craft's "bottom"

And that's (another reason) why ships turn with bank and pitch rather than yaw - all you need is a pair of main thrusters which can be independently vectored in the vertical axis. Point one slightly up and one slightly down and the ship banks while accelerating forwards. In order to yaw with main thrusters you would need to be able to vector them sideways which again increases complexity and therefore adds cost, inertial mass and moment, decreases MTBF etc.

 

[collinsmark]

The artifical gravity can easily compensate for these forces which are in the same direction as it needs to act to keep people on the floor.

[...]

I have already explained why completely different artifical gravity technology would be required to compensate for the (ficticious) forces in a yaw which are non-uniform in direction and orthogonal to the direction required to keep people on the floor in freefall.

Okay, fine, I'll take you word for that. But I'd rather not discuss hypotheticals that do not exist in reality. Doing so might quickly derail this thread. "Artificial gravity" does not exist in KSP nor does it exist in real life. So let's leave that distraction behind.

[Edit: Yes, I admit that I did open the door by mentioning (and rambling about) such things in a previous post. My point here is that I think we've discussed the "artificial gravity" topic enough. Any more and I fear that it might derail this thread.]

And that's (another reason) why ships turn with bank and pitch rather than yaw - all you need is a pair of main thrusters which can be independently vectored in the vertical axis. Point one slightly up and one slightly down and the ship banks while accelerating forwards. In order to yaw with main thrusters you would need to be able to vector them sideways which again increases complexity and therefore adds cost, inertial mass and moment, decreases MTBF etc.

That may be true for airplanes experiencing aerodynamic forces, but it does not apply to ships in space. Optimally, thrust is applied in the direction of desired $\Delta \vec v$ only*. Applying thrust in various directions as you turn is a complete waste of fuel. And applying thrust with multiple thrusters, aligned in different directions for any purpose other than changing (or stabilizing) orientation, is again a waste of fuel.

*[Edit: in relatively open space -- in a region of space considered local.]

 

[pbuk]

Okay, fine, I'll take you word for that. But I'd rather not discuss hypotheticals that do not exist in reality. Doing so might quickly derail this thread. "Artificial gravity" does not exist in KSP nor does it exist in real life. So let's leave that distraction behind.

But it is you that are making statements about the "absurdity" of the manoeuvres depicted based on your hypothetical assumptions. In contrast I am stating that based only on observation of the technology depicted in the Star Trek and Battlestar environments (uni-directional artificial gravity), banked turns would be preferable to yaw turns.

That may be true for airplanes experiencing aerodynamic forces, but it does not apply to ships in space.

No, vectorable impulse motors (rockets) work just as well in space.

Optimally, thrust is applied in the direction of desired $\Delta \vec v$ only*. Applying thrust in various directions as you turn is a complete waste of fuel. And applying thrust with multiple thrusters, aligned in different directions for any purpose other than changing (or stabilizing) orientation, is again a waste of fuel.

But changing orientation is exactly what we are talking about! And you have pointed out that you can observe the Battlestar ships turning using only their main thrusters so your assumption that this is due to some inability of the engineers and pilots to understand the technology available to them (whereas you assume knowledge of both the limits and unobserved capabilities of this technology) is what is absurd.

As for the "waste of fuel" of applying thrust in various directions, again you are assuming that there is a more efficient way of turning using the available technology. I can hypothesise (not assume) that the cost (in an aggregate sense i.e. money, maintenance, mass/moment penalty etc.) of having and using (including start-up and shut-down inefficiencies) 3-axis orientation motors is greater than the inefficiencies of using the already burning main thrusters. Observation of the ships in flight supports this hypothesis.

 

[collinsmark]

But it is you that are making statements about the "absurdity" of the manoeuvres depicted based on your hypothetical assumptions. In contrast I am stating that based only on observation of the technology depicted in the Star Trek and Battlestar environments (uni-directional artificial gravity), banked turns would be preferable to yaw turns.

Let's just keep this discussion relevant to reality, or at least to KSP, shall we?

[Edit: Allow me to repeat: Yes, I admit that I did open the door by mentioning (and rambling about) such things in a previous post. My point here is that I think we've discussed the "artificial gravity" topic enough. Any more and I fear that it might derail this thread.]

But changing orientation is exactly what we are talking about! And you have pointed out that you can observe the Battlestar ships turning using only their main thrusters so your assumption that this is due to some inability of the engineers and pilots to understand the technology available to them (whereas you assume knowledge of both the limits and unobserved capabilities of this technology) is what is absurd.

As for the "waste of fuel" of applying thrust in various directions, again you are assuming that there is a more efficient way of turning using the available technology. I can hypothesise (not assume) that the cost (in an aggregate sense i.e. money, maintenance, mass/moment penalty etc.) of having and using (including start-up and shut-down inefficiencies) 3-axis orientation motors is greater than the inefficiencies of using the already burning main thrusters. Observation of the ships in flight supports this hypothesis.

Consider a ship in empty space that is moving in some direction. Let's just call that direction East, for lack of a better term. Assume that the ship is presently pointing East (the ship's heading is East). The ship's motors are not applying any thrust.

Now assume that the ship adjusts its heading slightly to the North (let's say that "North" is 90^o from East). And if you wish, you may assume that the ship adjusts its roll such that the adjustment toward North can be done by using a pitch adjustment alone. (Although it doesn't really matter.)

Do you think that just because the ship's heading is pointing more toward North that it's direction of motion will also change? I can assure you that it will not.

Now consider that the ship continues this adjustment and now it's heading points due North. Has it's velocity changed? It has not. The ship remains moving due East

Now suppose the ship adjusts its heading to point 180^o from its original direction (let's call this direction West). Is the ship's velocity in the direction of West? No it is not. The ship is still traveling due East at exactly the same speed it started with. It matters not that the ship is pointing West. It is still moving East.

So what is the optimal method for the ship to change its velocity such that it travels due West? Assuming no gravitational bodies are nearby, it needs to change its heading such that it points West and then apply thrust. It's speed will first drop to zero and then pick up in the Westerly direction.

Had the ship instead started applying thrust when pointing in any other direction (such as North, or any direction that has a component anything else than West), it would just be fuel wasted. If it had thrusted with a Northerly component, it would later have to compensate by thrusting with a component to the South (again, a waste).

Instead, what if the ship desired to change its velocity direction to the North (with the same Easterly speed it is presently traveling at)? The optimal method is to change its heading to be 45^o West of North, and then thrust. The Easterly speed would eventually drop to zero and the Northerly speed would pick up to the original speed. Note that no "banking" is necessary.

 

[pbuk]

Let's just keep this discussion relevant to reality, or at least to KSP, shall we?

Sure: if you have found yourself in a hole I can understand why you want to stop digging.

So what is the optimal method for the ship to change its velocity such that it travels due West? Assuming no gravitational bodies are nearby, it needs to change its heading such that it points West and then apply thrust. It's speed will first drop to zero and then pick up in the Westerly direction.

But how does it "change its heading such that it points West"? Whichever way it does this introduces a cost (inefficiency) and you are assuming that this cost is less than the cost of fuel "wasted" by thrust vectoring. How is this relevant to KSP? Well if I can get into orbit just using the thrust vectoring of an LV-T45 I can save the weight of RCS thrusters and fuel allowing a greater payload.

 

[pbuk]

Instead, what if the ship desired to change its velocity direction to the North (with the same Easterly speed it is presently traveling at)? The optimal method is to change its heading to be 45^o West of North, and then thrust.

In what way is that optimal?

Clue: is "change velocity direction to the North (with the same Easterly speed you are presently traveling at)" a sufficiently constrained navigational requirement, or does it require further assumptions?

 

[collinsmark]

In what way is that optimal?

Clue: is "change velocity direction to the North (with the same Easterly speed you are presently traveling at)" a sufficiently constrained navigational requirement, or does it require further assumptions?

Changing a ship's heading requires very little resources. It's almost free.

Changing a ship's velocity -- be that speed or direction of motion -- is expensive. These maneuvers must be done wisely. That means only thrusting in the direction of desired $\Delta \vec v$, within a local region of space. It means not wasting it by thrusting in other directions to perform useless "banking" maneuvers.

After rotating the craft to the proper heading for a burn, the only force any occupant of the craft would feel is a force toward the rear of the ship when the burn occurs.

Nothing about this force would cause anybody to "fall over."

Prior to the burn, it matters little if the attitude change occurs via pitch or yaw (ignoring possible differences of moments of inertia along different axes. And that would depend on the ship). People near the bow might feel a small force forward (toward the bow), and people near the stern might feel a small force toward aft. [That and small, initial and final nudges.] That's true regardless of whether the attitude adjustment was done using a pitch or a yaw maneuver. Nobody falls over.

Edit: And btw, the only assumption that needs to be made is that this deals with reality, or at least a fairly realistic model of real-world physics. The physics in KSP is realistic enough.

 

[pbuk]

Why are you still talking about banking manoeuvres? I have never suggested that banking is either necessary or desireable for manoeuvring a ballistic craft in either 21 century Earth or Kerbal technology worlds. You brought the whole thing of banking up in the context of Battlestar technology where you stated it was "absurd". I pointed out that not only did you have insufficient knowledge of Battlestar technology to make that statement, but from what can be observed of that technology it is reasonable to assume that banking is in fact the preferred method of manoeuvring a Battlestar ship.

Can we stop talking about banking, Star Trek and Battlestar now?

Back to your proposed manouevre to "change velocity direction to the North (with the same Easterly speed you are presently traveling at)" - my apologies, I misread this. The most efficient way to achieve this (in a ship capable of 3-axis control) is almost certainly as you say to initiate and maintain thrust at 45° West of North.

However in a ship that is not required to perform such radical changes of direction, such as a rocket launching a payload into (Earth or Kerbal) orbit, it may be more efficient overall to accept the inefficiences of thrust vectoring in order to make an initial minor change of heading (and therafter use a gravity turn to complete the acceleration to orbital velocity) and thereby do completely without the requirement for motors and fuel for 3-axis control.

 

[collinsmark]

However in a ship that is not required to perform such radical changes of direction, such as a rocket launching a payload into (Earth or Kerbal) orbit, it may be more efficient overall to accept the inefficiences of thrust vectoring in order to make an initial minor change of heading (and therafter use a gravity turn to complete the acceleration to orbital velocity) and thereby do completely without the requirement for motors and fuel for 3-axis control.

Of course thrust vectoring is a valid method for attitude control (along with other possible methods such as RCS thrusters). It's perhaps the preferred method while launching and initially reaching orbit. Thrust vectoring can still be used for attitude control in space by performing a short, minimal burst to start a slow rotation then after waiting until the desired attitude is achieved, another short, minimal burst can be done to stop the rotation. In the process the ship's heading deviates significantly than its velocity vector. It might involve additional corrective bursts, but is certainly an option.

This is still a fundamentally different concept than banking. Maintaining even moderate thrust for an extended period to keep the velocity roughly aligned the ship's heading, to model some sort of banking, would be prohibitively expensive. It would be a waste of fuel.

In my rambling of Post #13, I only brought any of this up for the purpose of praising KSP for its realistic physics (at least in concept), and to contrast it against so many other movies, TV series, and other video games that at best are unrealistically wasteful or at worse violate the fundamental laws of physics.

And that's (another reason) why ships turn with bank and pitch rather than yaw - all you need is a pair of main thrusters which can be independently vectored in the vertical axis. Point one slightly up and one slightly down and the ship banks while accelerating forwards. In order to yaw with main thrusters you would need to be able to vector them sideways which again increases complexity and therefore adds cost, inertial mass and moment, decreases MTBF etc.

(Boldface mine)

This does not explain why pitch is preferable to yaw for a ship in space. By the same logic you could cut cost, inertial mass and moment, etc., by only implementing thrust vectoring in the yaw direction. There's nothing inherently preferential about pitch. [Edit: ignoring any possible differences in moments of inertia around different axes, which depends on the ship.]

Can we stop talking about banking, Star Trek and Battlestar now?

Thank goodness, yes!

Why are you still talking about banking manoeuvres? I have never suggested that banking is either necessary or desireable for manoeuvring a ballistic craft in either 21 century Earth or Kerbal technology worlds.

Oh, in that case I'm sorry. I must have misunderstood. Allow me to correct my misinterpretations,

I remember the old Battlestar Galactica series (I'm speaking of the original series, back in the '70s. I haven't watched the new ones), spaceships bank when they turn. Why are they banking? Airplanes bank because they get a normal force from the atmosphere, but there is no atmosphere in space! Why bank, I ask. Why?

So the crew don't fall over.

If you change the ship's heading with yaw everyone is going to fall over. To avoid this you need to bank so that the heading can be changed with pitch alone.

And that's (another reason) why ships turn with bank and pitch rather than yaw

Allow me to extend an apology then. When you kept repeatedly mentioning that ships in space bank, for some reason I thought that you were talking about banking. Silly me.

[Edit: are you confusing banking with roll? Ships in space can roll, but they do not bank. One needs aerodynamic forces, or at least some sort of external force to effectively bank (that is, without unduly wasting fuel).]

I'm glad we have finally cleared up the confusion. Now that we are in agreement that banking maneuvers are unrealistic in space, let's move on to other things.

And concerning the business of people falling over and the unrelenting preference for pitch vs. yaw, I still do not follow your reasoning. But that's okay to me, we can let that go too. It's a mystery I'm fully prepared to live with.