Trajectories right after launch - why is that ?

[thyriel]

Well, this is something that confuses me since a while. In stock game (prior 0.90) it wasn't much of a problem at all, as there was no need to make any ascents other then to the east. It drove me mad playing Real solar system making the ascents i liked to from different starting locations, and now with 0.90 it's driving me again mad in stock game with the contracts that force me to fly somewhere for a crew report:

Ascents, other then to the east, are always on completely different trajectories then the prograde is facing to:

Starting right to the west displays a trajectory like this:

http://cloud-4.steampowered.com/ugc/541886693881987755/26D5BEAE52D42E95852A254F8501E5DAFF7CD538/

As you see, the prograde is facing perfectly west, but the trajectory looks like i would fly to the east. And in fact, the capsule did not fly the trajectory shown, it landed exactly a couple hundred meters south of the KSC.

Ascending right to the South / North leads to trajectories like this, while the prograde is facing perfectly north / south:

http://cloud-4.steampowered.com/ugc/541886693882078717/23D5E8BB88B99760A81975EC3A33519E3EE2D916/

I'm aware that east ascents are "supported" by Kerbins rotation, and that i have to counter that rotation if i want to go somewhere else, but i just don't get it how exact paths are done then. And i don't think that the rotation makes so much difference on a trajectory that has only half a minute flytime. Plus, i don't think kerbins rotation should affect things still in the atmosphere. On earth if i throw a rock straight in the air, it still has its horizontal velocity from earths rotation, so i don't know why its affecting trajectories inside the atmosphere at all.

And even worse then that, if i start with a nearby contract i need to fly to and just aim at it on the navball, the prograde seems to perfectly show in the direction of the site marker, but it's just missing it because of that strange behavior, so the only way i found so far is just reverting launch until i got a feeling for that specific location on how far my prograde needs to be sideways to the contract marker.

So, how do you do a perfect north facing launch for example ? (that would end in a perfect 90° inclination on orbit)

Edited December 16, 2014 by thyriel

[Taki117]

As soon as you lift off the surface you are on a sub-orbital trajectory with most of your orbital velocity in the 090 direction (Due east) Now, if you change where your gravity turn is going (say due north) you are effectively performing a plane change maneuver. Your surface velocity is pointing due north, but your orbital velocity is off center due to the eastward rotation of the planet. Switch your Navball to Orbit mode (by clicking the speed indication) and you will see what I am talking about.

[thyriel]

As soon as you lift off the surface you are on a sub-orbital trajectory with most of your orbital velocity in the 090 direction (Due east) Now, if you change where your gravity turn is going (say due north) you are effectively performing a plane change maneuver. Your surface velocity is pointing due north, but your orbital velocity is off center due to the eastward rotation of the planet. Switch your Navball to Orbit mode (by clicking the speed indication) and you will see what I am talking about.

ah ok.. so suborbital trajectories are shown as if they where orbital but they are not ?

edit: and how do i determine then where i'm really flying to on low paths (especially those contracts where i need to fly below 18km) ?

Edited December 16, 2014 by thyriel

[rkman]

Trajectories have an orbital velocity, regardless of whether or not the trajectory is an orbit. The velocity is a different thing than the trajectory.

Your ship 'inherits' the orbital velocity of Kerbin's surface; 174m/s due east. That velocity vector adds to the current velocity vector that the ship has relative to the surface.

[Taki117]

ah ok.. so suborbital trajectories are shown as if they where orbital but they are not ?

edit: and how do i determine then where i'm really flying to on low paths (especially those contracts where i need to fly below 18km) ?

Technically all trajectories are orbital, the only difference between a sub-orbital and an orbital trajectory is that a sub-orbital trajectory has a periapsis below the surface of the parent body (be it a planet, moon, or asteroid) whereas an orbital trajectory the periapsis is above the surface of the parent body.

Now, to answer your question if you click in map view and select the marker (assuming you have a contract that says fly over area) it will show the pink target markers on your Navball, then just align your surface prograde marker to the target prograde marker and enjoy your flight, don't worry about what your trajectory is showing on the map. These contracts are best completed with aircraft.

[thyriel]

Now, to answer your question if you click in map view and select the marker (assuming you have a contract that says fly over area) it will show the pink target markers on your Navball, then just align your surface prograde marker to the target prograde marker and enjoy your flight, don't worry about what your trajectory is showing on the map. These contracts are best completed with aircraft.

Thanks i will try it. Maybe it just confused me that in map view the trajectories are shown to be far off my contract target while the prograde on the navball is perfectly targeting it. (surface mode)

[thyriel]

Oh what i forgot now, how do i determine then the perfect direction for an ascent with a specific inclination i want upon orbit (an explanation for different launching sites then KSC would be perfect) ?

Let's say my launchsite is in europe (so it's not equatorial) and i want to launch in the correct direction to end with a 315° inclination in orbit (Northwest). Switching navball to orbit mode before launch ?

[Cleric2145]

To clarify a little bit:

Just sitting on the Kerbin surface, you are moving 174m/s Eastwards. If you launched perfectly upwards, that lateral velocity would still be there, and your trajectory would look similar to your first picture. Sitting on the surface, you could say you are on a perfectly circular orbit, with the trajectory making a ring across the surface. Now imagine you are in an 'actual' orbit at 100km, perfectly circular and equatorial, and you burn directly west or north. Your trajectory doesn't immediately about-face (or left/right-face for North/South), but it slowly tilts in the direction as you overcome your orbital velocity. This is same thing that you do at launch when you take off headed anywhere other than East.

Since you are moving East at 174m/s at the same speed as the ground, your surface velocity (the difference between your speed and the speed of the planet) is zero. When you start boosting West, you are actually still moving East at 174m/s but you are essentially decelerating to a 'stop' before moving West. Once you 'stop' your orbital velocity, you will appear to be moving Westwards, but it's actually the planet spinning under you as you 'stand still.' When you keep accelerating, you will gain a retrograde velocity and see your trajectory start shifting.

EDIT: And to answer your question about launching into highly-inclined orbits, you'll need to aim closer to West than you want to end up. Again, because you start with a high Eastward velocity, once your navball ticks into Orbital velocity, your vector marker will jump towards the East. I suppose starting with your navball in orbital might actually help, but you'll be losing a lot of energy to fighting your surface velocity until you leave the denser atmosphere, and if you're using FAR, you'll probably lose control because you'll be aiming pretty far off your velocity and you'll get flipped by the airflow.

Edited December 16, 2014 by Cleric2145

[thyriel]

Since you are moving East at 174m/s at the same speed as the ground, your surface velocity (the difference between your speed and the speed of the planet) is zero. When you start boosting West, you are actually still moving East at 174m/s but you are essentially decelerating to a 'stop' before moving West. Once you 'stop' your orbital velocity, you will appear to be moving Westwards, but it's actually the planet spinning under you as you 'stand still.' When you keep accelerating, you will gain a retrograde velocity and see your trajectory start shifting.

But shouldn't that only apply to things outside the atmosphere ? If i throw a stone to the west, it doesn't need to counter any east velocity from rotation first. Even more, if i throw a stone with 100 Newtons to the east it will fly exactly the same length as if i throw it with 100 Newtons to the west.

Edited December 16, 2014 by thyriel

[RainDreamer]

Somewhat related, but not really question: How much more dV do you need for changing the trajectory direction from east to west? 174m+/s?

[thyriel]

Somewhat related, but not really question: How much more dV do you need for changing the trajectory direction from east to west? 174m+/s?

I think it depends on the altitude you are countering the rotation. It would be 174m/s at zero height, but the higher you are, the more delta-v you need. The background is pretty simple: Think of the planet as a centrifuge, the further you are away from its center, the more speed you will have that you would need to counter to stop rotation.

Edited December 16, 2014 by thyriel

[Cleric2145]

But shouldn't that only apply to things outside the atmosphere ? If i throw a stone to the west, it doesn't need to counter any east velocity from rotation first. Even more, if i throw a stone with 100 Newtons to the east it will fly exactly the same length as if i throw it with 100 Newtons to the west.

If you threw the rock at a speed of 2 m/s to the East, it would briefly travel 2 m/s faster in the "ground orbit" before hitting the ground. If it was thrown West at 2 m/s, it would be travelling East 2 m/s slower than everything around it. It's a very strange frame of mind to think about, but by throwing the rock to the West, you are decelerating it. If you stood on Earth's equator and fired a bullet West at ~465 m/s, it would essentially be standing still in the reference of Earth's rotation, getting battered by the atmosphere soup that spins along with the earth. If you fired it to the East, it would be moving twice as fast as the Earth is spinning, and have almost the same effect on the bullet. Once you start firing North and South you start getting all sorts of fun effects.

In the frame of reference that we live in, it's almost impossible to see or perceive the effects that the orbital velocity of the Earth has on movement. Leaving the atmosphere doesn't effect anything differently other than allowing our rocks to not immediately vaporize when we throw them fast enough to see the effects. If you are in orbit going 1000 m/s Eastwards, and you send one rocket to the West and one to the East, one will fall straight down to the planet, the other will nearly fly off to Duna. Being in the atmosphere wouldn't change the idea other than having to deal with the all the bothersome junk like air.

EDIT:

I think it depends on the altitude you are countering the rotation. It would be 174m/s at zero height, but the higher you are, the more delta-v you need. The background is pretty simple: Think of the planet as a centrifuge, the further you are away from its center, the more speed you will have that you would need to counter to stop rotation.

And this the great thing about KSP: set up a scenario like this with hyper edit and test it and find the answer to your question. Then blow up a Kerbal for fun. It's like being a scientist, but with more screaming and explosions!

Edited December 16, 2014 by Cleric2145

[thyriel]

I think slowly i got it now At least switching the navball to orbit mode before launch makes it possible now to orbit directly into a specific inclination. Guess it will still take me some time getting a feeling for it to reduce course corrections to a minimum, but it's far better now. (And at least im not confused anymore why my orbit doesn't end like i thought it would)

And this the great thing about KSP: set up a scenario like this with hyper edit and test it and find the answer to your question. Then blow up a Kerbal for fun. It's like being a scientist, but with more screaming and explosions!

Definitly It's just a geek's dream of gaming fusing with science... And pretty regardless of how much you think you now know about rocket science and maneuvers, there's always something more to learn for perfectionists like me

[Cleric2145]

I think slowly i got it now At least switching the navball to orbit mode before launch makes it possible now to orbit directly into a specific inclination. Guess it will still take me some time getting a feeling for it to reduce course corrections to a minimum, but it's far better now. (And at least im not confused anymore why my orbit doesn't end like i thought it would)

Definitly It's just a geek's dream of gaming fusing with science... And pretty regardless of how much you think you now know about rocket science and maneuvers, there's always something more to learn for perfectionists like me

We have an incredible tool, having KSP. Being able to have all of the orbital mechanics automatically calculated and visualized is a huge deal. Back in the Gemini days of space travel, when astronauts where getting shot up into space by teams of the smartest living human beings, the astronauts and mission controllers could not, for several hours, understand why the hell the pilot could thrust directly at a target and miss wide to the right until they pulled out their slide-rules, and sketched some circles and realized they weren't looking at orbital mechanics properly. I believe Buzz Aldrin wrote a thesis before he went into space about that type of orbital movement, but I could be wrong, I'm very sleep deprived at the moment...

[thyriel]

We have an incredible tool, having KSP. Being able to have all of the orbital mechanics automatically calculated and visualized is a huge deal. Back in the Gemini days of space travel, when astronauts where getting shot up into space by teams of the smartest living human beings, the astronauts and mission controllers could not, for several hours, understand why the hell the pilot could thrust directly at a target and miss wide to the right until they pulled out their slide-rules, and sketched some circles and realized they weren't looking at orbital mechanics properly. I believe Buzz Aldrin wrote a thesis before he went into space about that type of orbital movement, but I could be wrong, I'm very sleep deprived at the moment...

Just a couple of days ago i have thought about how the first astronauts must have felt sitting in that small capsule completely relaying on someones calculations to know where he will end (and not crash into the moon or get lost into space). Must have been an awkward feeling somehow...

[Claw]

I think it depends on the altitude you are countering the rotation. It would be 174m/s at zero height, but the higher you are, the more delta-v you need. The background is pretty simple: Think of the planet as a centrifuge, the further you are away from its center, the more speed you will have that you would need to counter to stop rotation.

It looks like your other question is answered, but I think there is some misunderstanding with this one.

It would take 174 m/s at zero height (and gets higher as you go up) because you aren't in orbit yet. If you were actually orbiting at zero height (assuming flat surface and no atmosphere...), you would be going a lot faster than 174 m/s.

But the real key point I'm heading toward is that as you get further away you need less dV to stop (or reverse) an orbit, not more.

For example, LKO orbital velocity is around 2,300 m/s. At the Mun's orbital distance, orbital velocity around Kerbin is roughly 800 m/s. So reversing an orbit in LKO takes 4,600 m/s, while reversing an orbit out at the Mun's distance takes 1,600 m/s.

This is the reason inclination changes are "cheaper" to do when you execute the burn at AP. There is more information and math that goes into this, but I wanted to help clarify the orbital speed vs. altitude without completely changing the thread topic.

Cheers,

-Claw

[FleshJeb]

Drawn for a specific case, but you get the gist of it:

EDIT: Equation: LAUNCHANG = arctan (ORBVEL sine(INCLIN) / (ORBVEL cosine(INCLIN) - SIDEREALV * cosine(LATITUDE)))

Edited December 16, 2014 by FleshJeb

[thyriel]

It looks like your other question is answered, but I think there is some misunderstanding with this one.

It would take 174 m/s at zero height (and gets higher as you go up) because you aren't in orbit yet. If you were actually orbiting at zero height (assuming flat surface and no atmosphere...), you would be going a lot faster than 174 m/s.

Thanks, i should have mentioned that i ment atmospheric for that case

But the real key point I'm heading toward is that as you get further away you need less dV to stop (or reverse) an orbit, not more.

For example, LKO orbital velocity is around 2,300 m/s. At the Mun's orbital distance, orbital velocity around Kerbin is roughly 800 m/s. So reversing an orbit in LKO takes 4,600 m/s, while reversing an orbit out at the Mun's distance takes 1,600 m/s.

This is the reason inclination changes are "cheaper" to do when you execute the burn at AP. There is more information and math that goes into this, but I wanted to help clarify the orbital speed vs. altitude without completely changing the thread topic.

Cheers,

-Claw

That's pretty interesting now Somehow i mostly tend to make course corrections (even big ones like changing inclination) better soon then late. I always thought it's generally the lesser fuel usage.

Would be pretty interesting now to test if a north or a west high orbit is cheaper via direct launch, or eastern orbit with low Pa, high Ap and then turning at Ap... think i gotta install mechjeb and take a look

[Cleric2145]

Thanks, i should have mentioned that i ment atmospheric for that case

That's pretty interesting now Somehow i mostly tend to make course corrections (even big ones like changing inclination) better soon then late. I always thought it's generally the lesser fuel usage.

Would be pretty interesting now to test if a north or a west high orbit is cheaper via direct launch, or eastern orbit with low Pa, high Ap and then turning at Ap... think i gotta install mechjeb and take a look

Launching directly into the inclination should still be most efficient. Imagine your rocket's delta-v as a full bucket, and your orbital velocity in each direction is an empty bucket. Instead of pouring half your bucket into the East bucket, then pouring that heavy East bucket into the North bucket, just start filling the North bucket to begin with. c: I love metaphors.

[Claw]

That's pretty interesting now Somehow i mostly tend to make course corrections (even big ones like changing inclination) better soon then late. I always thought it's generally the lesser fuel usage.

Oh yeah, sorry. You are indeed correct to generally say "making corrections sooner is better than later." If you switch to a westerly direction during the initial stages of launch, you spend much less dV than trying to change your orbit direction later. In that case, you're spending 174 m/s * 2. Whereas if you did it in orbit, you need a lot more (as illustrated in my earlier post). But that's not to say that being closer to the planet is the reason. It's because your velocity in relation to orbit is already very low.

Course corrections are another case too (such as on the way to Mun). Making them early rather saves you dV because small angles far away make big changes later on. Whereas once you get closer, you need to add more dV to get the same distance offset (or inclination change).

I'm trying to be short, so hopefully that's still clear.

Cheers,

~Claw

[Mechie_PhD]

Well, this is something that confuses me since a while. In stock game (prior 0.90) it wasn't much of a problem at all, as there was no need to make any ascents other then to the east. It drove me mad playing Real solar system making the ascents i liked to from different starting locations, and now with 0.90 it's driving me again mad in stock game with the contracts that force me to fly somewhere for a crew report:

Ascents, other then to the east, are always on completely different trajectories then the prograde is facing to:

Starting right to the west displays a trajectory like this:

http://cloud-4.steampowered.com/ugc/541886693881987755/26D5BEAE52D42E95852A254F8501E5DAFF7CD538/

As you see, the prograde is facing perfectly west, but the trajectory looks like i would fly to the east. And in fact, the capsule did not fly the trajectory shown, it landed exactly a couple hundred meters south of the KSC.

Ascending right to the South / North leads to trajectories like this, while the prograde is facing perfectly north / south:

http://cloud-4.steampowered.com/ugc/541886693882078717/23D5E8BB88B99760A81975EC3A33519E3EE2D916/

I'm aware that east ascents are "supported" by Kerbins rotation, and that i have to counter that rotation if i want to go somewhere else, but i just don't get it how exact paths are done then. And i don't think that the rotation makes so much difference on a trajectory that has only half a minute flytime. Plus, i don't think kerbins rotation should affect things still in the atmosphere. On earth if i throw a rock straight in the air, it still has its horizontal velocity from earths rotation, so i don't know why its affecting trajectories inside the atmosphere at all.

And even worse then that, if i start with a nearby contract i need to fly to and just aim at it on the navball, the prograde seems to perfectly show in the direction of the site marker, but it's just missing it because of that strange behavior, so the only way i found so far is just reverting launch until i got a feeling for that specific location on how far my prograde needs to be sideways to the contract marker.

So, how do you do a perfect north facing launch for example ? (that would end in a perfect 90° inclination on orbit)

You might want to google Coriolis effect and Eotvos effect... Since you are traveling from one point to another in a rotating body, you're in a non-newtonian reference frame. This applies even when in atmosphere (That is how hurricanes(cyclones) get their shape).

[thyriel]

Was anyway interesting to see the dV differences with mechjeb

If anyone's interested:

All flights where done with mechjeb (ascent: 100km orbit, 1.2km turn start, 1.5TWR, 100km turn end, 15° final angle, 90% shape), NEAR and deadly reentry until 90° inclined circular geosynchronous orbit (2878.75km)

[TABLE=width: 500]

[TR]

[TD][/TD]

[TD]A[/TD]

[TD]B[/TD]

[TD]C[/TD]

[/TR]

[TR]

[TD]dV left[/TD]

[TD]3294[/TD]

[TD]2555[/TD]

[TD]1951[/TD]

[/TR]

[TR]

[TD]difference[/TD]

[TD]739[/TD]

[TD]604[/TD]

[TD]1343[/TD]

[/TR]

[/TABLE]

A) Ascent directly to 90° (N), raise Ap, circularize

Ascent to 0° (E), raiseAP, 90° inclinitation @Ap, circularize

C) Ascent to 0° (E), raiseAP, circularize, 90° inclination

So as expected, most efficient way is launching directly into the target inclination. Being already in orbit the most efficient way is to have a high Ap and low Pe. (Will try what happens to dV usage when Pe is lowered to zero right before changing inclination). dV difference between changing inclination at circular 100km orbit or at geosync. orbit is nearly 3k m/s.

EDIT:

Somewhat related, but not really question: How much more dV do you need for changing the trajectory direction from east to west? 174m+/s?

I tested this now too:

A) same ascent path as above to 100km orbit - East: 4454 dV left - West: 4278 dV left = 176 m/s difference

same as above but start turn at 15km - East: 4321 dV left - West: 4048 = 273 m/s difference

Somehow i expected a lot more dV difference between low and high altitude turn.

Edited December 17, 2014 by thyriel

[Janos1986]

For a polar orbit you may want to point at vector 315 until the orbit prograde marker is almost facing North, then start pointing north with the craft until both nose and marker collide on vector 360. From then on you just get to orbit and chill. K.E.R. can help you with fine tuning of your inclination until it is precisely 90 degrees.

For a 180 orbit, well you just point west and keep pushing. It requires a lot more deltaV to get to a stable orbit, obviously.

Note: The ORBIT prograde, mind you, not the surface one.

Edited December 17, 2014 by Janos1986

[DrD]

We have an incredible tool, having KSP. Being able to have all of the orbital mechanics automatically calculated and visualized is a huge deal. Back in the Gemini days of space travel, when astronauts where getting shot up into space by teams of the smartest living human beings, the astronauts and mission controllers could not, for several hours, understand why the hell the pilot could thrust directly at a target and miss wide to the right until they pulled out their slide-rules, and sketched some circles and realized they weren't looking at orbital mechanics properly. I believe Buzz Aldrin wrote a thesis before he went into space about that type of orbital movement, but I could be wrong, I'm very sleep deprived at the moment...

Yes, MIT 1963.

It can be found here: http://dspace.mit.edu/handle/1721.1/12652

[thyriel]

It requires a lot more deltaV to get to a stable orbit, obviously.

Not really, look at the numbers above It's only ~180m/s difference between east and west launch.