Why is creating a Kerbin orbit so unpredictable?

[JackBush]

I asked an earlier question about the awkwardness of camera movement in map mode but there's apparently nothing to be done about it. This is a similar question because when I pull the Prograde thingy out I never know what kind of orbit I'm going to get. This is a good example of what I just got. But other times it will be close to a circle. There seems to be no reason to it. Or is there? If so, what am I missing or doing wrong. I would prefer to get something close to a circle every time.

http://imgur.com/X1vgKqv

Looking forward to your answers -- eagerly!

[Alshain]

Its because your node is slightly to one side of the Ap marker. You need to be right on it for a circular orbit.

[Just Jim]

You can move the node itself around the orbit (or sub-orbit in your pic). Left-click and hold the center circle of the node to move it where you want. This should help you get a circlular orbit

[Alshain]

One other note, The more narrow your arc, the more impact imperfect placement will cause. Your Pe is still very far inside the planet. You need to practice launcing more horizontal to the planet, it will minimize this effect and it is more efficient and will save fuel.

[Snark]

To get a circular orbit:

1. place node right at Ap, as Alshain has pointed out (you can drag the node itself to adjust position, as Jim notes)

2. drag until the Ap node post-maneuver starts to slide away from its pre-maneuver position

3. when the new Ap and Pe are about 90 degrees before/after your original Ap, you should be pretty close to a circle

4. you can check how close you are by mousing over Pe and Ap to see if they're close to equal

5. if they're not quite right, you can slide the node forward or backward a bit to adjust. Alternatively, you can drag slightly radial-out (if the new Ap is behind you) or radial-in (if it's ahead of you) to fine-tune.

[paul23]

I typically make sure I burn most/all my orbit just *before* the apoapsis. As you probably have seen, it is the last few delta V that make the biggest change. And by burning before the apoapsis, the apoapsis will move "away" from where it is now (so you can stay stable at, say, 10 seconds before apoapsis. The amount you have to be before depends on your TWR as well as how circular your orbit is already.

Not only is this more "stable" and easy, but as you probably has seen the last few delta-v make the biggest changes on the orbit (visually at least). This allows me to burn the last 100 or so delta-v exactly on the apoapsis.

Giving me a neat stable near circular orbit. Sure I don't have a LKO of 80 km. But I'm around 90-100 km above surface - less efficient. But unlike others I can turn very little very late, allowing me to leave atmosphere faster which is easier and more efficient. (I go straight up till around 20km).

Anyways if this is advice on reaching a specific orbit, I'd recommend untill your proficient and understand it completely to do the 3 steps separate of each other. Yes it costs maybe a few 100 delta-v more, but it's way more easy & very very much less frustrating.

The 3 steps:

1) get in orbit above 70km (as low as possible)

2) circularize the orbit by doing propagate burn at apoapsis.

3) Do orbital changes.

(sub steps for orbital changes):

3a) Fix the inclination at ascending/descending node

3b) Burn at the wanted orbit's periapsis propagade (or reverse apoapsis and retrograde burn if wanted orbit is smaller): align camara so the required orbit is a simple "line" -looking at the periapsis before the orbitted body-, with your current orbit being also on the same line, then click at the place of the periapsis on "node now".

3c) When your orbit touches the other orbit (apoapsis should align now) stop with the burn.

3d) When at apoapsis now burn untill the periapsis matches also the wanted orbit.

If you did step 2 correctly the orbit should match nearly perfectly. If it isn't (or you skipped step 2) nudge the orbit a bit, this is most efficiently done by burning pro/retrograde at points halfway the orbit.

More advanced flight patterns include launching already in such a way that the inclination matches the wanted orbit (huge delta-v can be gained here more than 9000 delta-V in worst case). Skipping step 2 (though for certain wanted orbits this is nearly the most efficient approach anyways), and combining the nudging & steps 3b, c & d.

Edited October 25, 2015 by paul23

[Alshain]

I typically make sure I burn most/all my orbit just *before* the apoapsis. As you probably have seen, it is the last few delta V that make the biggest change. And by burning before the apoapsis, the apoapsis will move "away" from where it is now (so you can stay stable at, say, 10 seconds before apoapsis. The amount you have to be before depends on your TWR as well as how circular your orbit is already.

Not only is this more "stable" and easy, but as you probably has seen the last few delta-v make the biggest changes on the orbit (visually at least). This allows me to burn the last 100 or so delta-v exactly on the apoapsis.

Giving me a neat stable near circular orbit. Sure I don't have a LKO of 80 km. But I'm around 90-100 km above surface - less efficient. But unlike others I can turn very little very late, allowing me to leave atmosphere faster which is easier and more efficient. (I go straight up till around 20km).

The most accurate way is to split the difference. If your burn time is 1 min, burn 30 seconds before the node and 30 seconds after. This means any innacuracies caused by burning early are negated by innacuracies of burning late. Unfortunately stock burn time estimation sucks. Kerbal Engineer does it better.

[Clipperride]

And also remember that when in a near circular orbit, it doesn't take much delta V to tweak the Ap or Pe. If you have RCS available its often easier to use those thrusters for small changes as you can be much more precise.

[JackBush]

Its because your node is slightly to one side of the Ap marker. You need to be right on it for a circular orbit.

You're now my very best friend, Alshain! Thank you.

- - - Updated - - -

Yes, this I don't understand because the more horizontal I am, the longer it takes to get up to 70K+. So aren't I wasting fuel? And two posts below is Paul 23 saying "allowing me to leave atmosphere faster which is easier and more efficient. (I go straight up till around 20km)" Is this not contradictory? I'm always running out of fuel when it's time to come back to Kerbin so I need help here.

[paul23]

The most accurate way is to split the difference. If your burn time is 1 min, burn 30 seconds before the node and 30 seconds after. This means any innacuracies caused by burning early are negated by innacuracies of burning late. Unfortunately stock burn time estimation sucks. Kerbal Engineer does it better.

This is true in the case you wish to keep your peri/apoapsis at the same point: but not necessary when trying to make a circular orbit. Especially an orbit quickly at the "end of burn" (and thus you wish to have the end of your burn at the apoapsis).

Oh just a slight remark: the inaccuracies are not "negated" (first obvious is the fact that you need more dv to get to the right orbit if not burning at the best point). But more fundamentally you produce more delta-V per second (more acceleration) as your rocket gets lighter. (Simple newton shows F = m*a shows this). So more dV is done at the end/second half of the burning. Ideally you would wish to throttle the engine in such a way not the force but the acceleration is kept constant through the burn, in real life this is what typically happens in autonomous satellites.

Yes, this I don't understand because the more horizontal I am, the longer it takes to get up to 70K+. So aren't I wasting fuel? And two posts below is Paul 23 saying "allowing me to leave atmosphere faster which is easier and more efficient. (I go straight up till around 20km)" Is this not contradictory? I'm always running out of fuel when it's time to come back to Kerbin so I need help here.

This depends: faster leaving atmosphere means less drag. And lower requirement on attitude/pitch/roll control (so rockets can be lighter). However it also typically means a higher orbit when trying to circularize (you have to "start much more before highest altitude" as the burn time to reach an orbit is way longer). Ideally you would start turning already in a circular curve so that you are exactly horizontal when in orbit - when disregarding drag.

Remember also that the the main problem in spaceflight is not getting up there:

It's only 100km (or in case of KSP: 70) away. What's hard is staying up there, you have to create a major horizontal speed increase, and the best way to do that is to go horizontally. 7.8 km/s for earth 2.5 km/s for Kerbin.

Btw if you have problems launching your first rocket: remember that a return is virtually free: once your periapsis is above 70km you are officially in orbit. So you only have to apply very very small thrust (retrograde at apoapsi) to get the periapsis back at 65km, from where drag eventually captures you. (takes a dozen or so pass-throughs...).

Edited October 25, 2015 by paul23

[Red Iron Crown]

Yes, this I don't understand because the more horizontal I am, the longer it takes to get up to 70K+. So aren't I wasting fuel? And two posts below is Paul 23 saying "allowing me to leave atmosphere faster which is easier and more efficient. (I go straight up till around 20km)" Is this not contradictory? I'm always running out of fuel when it's time to come back to Kerbin so I need help here.

Alshain is correct. It may take longer to get your Ap up over 70km, but you are accumulating horizontal velocity that will stay with you and reduce the magnitude of your circularization burn (which helps make it easier to be precise in circularizing). It's also more efficient; while your craft will spend more time in atmosphere and lose more to aero drag it will also experience lower gravity losses, and gravity losses are usually more significant than drag losses (assuming your craft is reasonably streamlined).

[Snark]

Yes, this I don't understand because the more horizontal I am, the longer it takes to get up to 70K+. So aren't I wasting fuel? And two posts below is Paul 23 saying "allowing me to leave atmosphere faster which is easier and more efficient. (I go straight up till around 20km)" Is this not contradictory? I'm always running out of fuel when it's time to come back to Kerbin so I need help here.

So there are two issues here that are kind of battling one another:

• On the one hand: You get the best rocket efficiency when going from A to B if you thrust in a straight line from A towards B. It would be less efficient to thrust way off to one side, then make a right-angle turn towards B. By launching more horizontally, you're killing two birds with one stone: you're not only lifting your ship out of atmosphere, but you're also contributing to the huge horizontal speed you need to be in orbit. You're also losing much less to gravity losses because your engines aren't battling gravity as much.
  
• On the other hand: for rockets, flying through atmosphere is a Bad Thing and you want to be out of the atmosphere quickly so you're not fighting drag.
  

It's the contradiction between these two that's confusing you.

Here's the answer, and why Alshain and Paul23 are both right: The atmosphere gets thin pretty quick with elevation.

By the time you get up to just 10 km of altitude on Kerbin, air pressure is already down to only 16% of sea level. When you get up to 25 km, it's under 1%!

This means that effectively, once you're above 25 km the air resistance doesn't matter much. Even at 15 km you're dealing with much less atmosphere than sea level.So a good launch profile for a vertical-takeoff rocket typically looks something like this:

1. Launch straight up.
   
2. Almost immediately, tip eastwards just a tiny bit.
   
3. Do a gravity turn that gets you to a roughly 45 degree angle at 10-12 km of altitude (throttling down if necessary).
   
4. Punch the throttle and blast out of atmosphere as fast as possible, following prograde.
   
5. By the time you leave atmosphere, you should be mostly horizontal and only gently climbing. If you have a really high TWR rocket such that it's still climbing at close to 45 degrees when it leaves atmosphere, you may need to nudge its nose downward a bit after you get over 20 km.
   

[ForScience6686]

I would like to see buttons near the burn dv bar that allow finer control. However my thoughts are the node tool is just to help plan missions. I often close the node before I've even reached my target velocity as it doesn't seem very precise on long burn times.

[Alshain]

Yes, this I don't understand because the more horizontal I am, the longer it takes to get up to 70K+. So aren't I wasting fuel? And two posts below is Paul 23 saying "allowing me to leave atmosphere faster which is easier and more efficient. (I go straight up till around 20km)" Is this not contradictory? I'm always running out of fuel when it's time to come back to Kerbin so I need help here.

Vertical velocity is consistently eaten by gravity's pull and aerodynamic drag. Horizontal velocity is eaten only by aerodynamic drag, which gets thinner rather quickly and since you initially launch straight up, by time you get your gravity turn completed (usually around 10 degrees above the artificial horizon) you will be at about 30km give or take, where the atmosphere is getting thinner. In fact it is practically non existent by 45km. This is true for the new 1.0 atmosphere. In pre 1.0 we had something lovingly referred to as the soup-o-sphere. The atmosphere remained almost constant thickness all the way up to about 45km where it just kinda ended very suddenly. In that case, powering above the atmosphere quickly was in fact the better choice. It just isn't anymore. Gravity has more effect than Aerodynamic drag in the new world order.

You can't get to orbit going straight up, and in fact the less you go straight up, the better. You still need to go up a little, but your objective is speed, not height.

- - - Updated - - -

This is true in the case you wish to keep your peri/apoapsis at the same point: but not necessary when trying to make a circular orbit. Especially an orbit quickly at the "end of burn" (and thus you wish to have the end of your burn at the apoapsis).

I'm not sure you are making a direct comparison. If you want a circular orbit on a launch, you get your Ap at the desired altitude, then burn half before and half after it. Trying to incorporate some kind of radial or normal burn to raise the Ap while circularizing is going to be a waste of fuel. If you are getting a near circular orbit, it really won't matter in the end where the Ap and Pe are because they are so close it won't realistically effect anything. You could start your next maneuver from anywhere in the orbit and any cost penalty would be negligible.

Oh just a slight remark: the inaccuracies are not "negated" (first obvious is the fact that you need more dv to get to the right orbit if not burning at the best point). But more fundamentally you produce more delta-V per second (more acceleration) as your rocket gets lighter. (Simple newton shows F = m*a shows this). So more dV is done at the end/second half of the burning. Ideally you would wish to throttle the engine in such a way not the force but the acceleration is kept constant through the burn, in real life this is what typically happens in autonomous satellites.

This effect is so minuscule you won't be able to register it. By burning half before and half after, you don't end up moving the Ap or Pe at all.

Edited October 25, 2015 by Alshain

[Red Iron Crown]

A note: KER's node burn time calculator splits the delta-V of the burn, not the time, so you end up expending the same dV on either side of the node.

[paul23]

I'm not sure you are making a direct comparison. If you want a circular orbit on a launch, you get your Ap at the desired altitude, then burn half before and half after it. Trying to incorporate some kind of radial or normal burn to raise the Ap while circularizing is going to be a waste of fuel.

I'm not stating that: I'm stating that to *get* into an orbit from suborbital it is oftentimes better to start burning before, so that the apoapsis moves along, giving you "more time" before the critical point comes at which you should have to make the orbit circular. (I often have 2-3 minute long burns to reach orbit, so if I would do half before, half after as you suggest I'd be near 70km altitude, hopelessly doing a huge burn just to get the periapsis up.)

[Alshain]

I'm not stating that: I'm stating that to *get* into an orbit from suborbital it is oftentimes better to start burning before, so that the apoapsis moves along, giving you "more time" before the critical point comes at which you should have to make the orbit circular. (I often have 2-3 minute long burns to reach orbit, so if I would do half before, half after as you suggest I'd be near 70km altitude, hopelessly doing a huge burn just to get the periapsis up.)

Ok, fair enough, but if you are launching horizontally as we have been discussing, by the time your Ap reaches the desired altitude, you will probably be 5 to 10 minutes out from the Ap. (or you need to build a better upper stage)

When I launch if my Ap hits the desired target and my Pe is positive, I consider it a best possible launch. If it is negative but higher than -100km, I consider it a mildly successful launch. In either of those two scenarios it is going to be very difficult to have a Time to Ap under a minute unless you keep shutting off the engines.

- - - Updated - - -

A note: KER's node burn time calculator splits the delta-V of the burn, not the time, so you end up expending the same dV on either side of the node.

Hmm, I didn't know that. Usually when I compare the burn time and the time to burn, it's half the time as well. That could just be a coincidence though.

Edited October 25, 2015 by Alshain

[Waxing_Kibbous]

You may want to look into the Kerbal Engineer Redux mod if you haven't already, it will show you how much dV you are packing, which helps a lot. I'm still pretty new and I try to get my first stage (aka the main launch sequence that can include solid rockets and maybe a simple asparagus setup) comfortably above 3500 dV, the second stage can wrap the orbit up and get me where I need to go after orbit. Since my launches are still random/sloppy I generally like to go a bit overkill- it's a lot better than losing a flight. Once you get into the 2.5m rockets getting into orbit becomes much easier, and expensive, but it's almost fun to do.

This is a good series:

https://youtu.be/6ffdsuZxAfc?t=682

[paul23]

When I launch if my Ap hits the desired target and my Pe is positive, I consider it a best possible launch. If it is negative but higher than -100km, I consider it a mildly successful launch. In either of those two scenarios it is going to be very difficult to have a Time to Ap under a minute unless you keep shutting off the engines.

Maybe as I progress I can make smoother turns: but when you start out turning late = less aerodynamic forces. And mainly the body lift force keeps forcing my aircraft to "tumble" & the gravitational forces rip my aircraft apart. - But that might be having to stack almost a dozen FL-T400s due to lack of science in 100% science settings. (Now trying to reach Duna with only electrics and space exploration in the 90-science branch, maybe I should unlock aviation instead of storming heavy rocketry/fuel systems).

Anyways my point is: a high apoapsis is not necessary bad, and turning late makes the flight actually easier & predictable.

[Alshain]

Maybe as I progress I can make smoother turns: but when you start out turning late = less aerodynamic forces. And mainly the body lift force keeps forcing my aircraft to "tumble" & the gravitational forces rip my aircraft apart. - But that might be having to stack almost a dozen FL-T400s due to lack of science in 100% science settings. (Now trying to reach Duna with only electrics and space exploration in the 90-science branch, maybe I should unlock aviation instead of storming heavy rocketry/fuel systems).

Anyways my point is: a high apoapsis is not necessary bad, and turning late makes the flight actually easier & predictable.

Ok, so in your case I would suggest if you are using a dozen FL-T400s, you need to work on better rocket building skills (that's not intended to be an insult, but constructive criticism). If you are trying to SSTO, it's going to cost more fuel, but if you combine a swivel first stage with a terrier second stage, the efficiency of the terrier in the upper atmosphere will make your rocket a lot smaller, cheaper, and easier to fly. This is a link to my lifter Subassemblies. ALL of my tier 4 lifters have half a dozen FLT400's. While I strongly suggest you learn to do this on your own, feel free to use them as examples. I'm not saying the are the best but they are pretty good. Try giving one or two of them a flight and see if you can get them into orbit. It may take practice.

[paul23]

I'm not having any difficulty, I'm not the original poster. I'm trying to aim (as said) for Duna. So right now working on a 9k dV vehicle (monstrosity).

Btw I find that rockets with two boosters tend to start rolling when I try to make a turn.

[Alshain]

I'm not having any difficulty, I'm not the original poster. I'm trying to aim (as said) for Duna. So right now working on a 9k dV vehicle (monstrosity).

http://i.imgur.com/k1IPG76.png

Btw I find that rockets with two boosters tend to start rolling when I try to make a turn.

Lol, Duna at Tier 4? That is quite the challenge.

If your rockets that roll are like that one, it's most likely becuase you are using SRB's only. They don't have gimbal (which is sad because real ones do) and controlling a rocket with reaction wheels alone can be difficult. That is why I use the Swivel as my center engine, the gimbal gives me steering control without the strange behavior. I start turning by 1km and I never lose control.

[Wolf Baginski]

You may want to look into the Kerbal Engineer Redux mod if you haven't already, it will show you how much dV you are packing, which helps a lot.

I agree.

The HUD displays maybe don't give you much extra info, but periapsis and apoapsis heights are continuously shown, rather than having to click on nodes in map mode. That makes things much easier.

The current version also gives Mach Number on the RH HUD. It can make for a better launch to throttle back so as to stay below Mach 1 for the first few thousand metres. I aim to stay subsonic below about 6000m. The drag can be very high in the trans-sonic speed range, and many craft either struggle to fly faster at lower altitudes, or burn a lot of fuel for not much gain.

Mach number is more important for air-breathing engines. Depending on the engine/intake combination, thrust can increase a huge amount above about Mach 1.5

[JackBush]

Ok, so in your case I would suggest if you are using a dozen FL-T400s, you need to work on better rocket building skills (that's not intended to be an insult, but constructive criticism). If you are trying to SSTO, it's going to cost more fuel, but if you combine a swivel first stage with a terrier second stage, the efficiency of the terrier in the upper atmosphere will make your rocket a lot smaller, cheaper, and easier to fly. This is a link to my lifter Subassemblies. ALL of my tier 4 lifters have half a dozen FLT400's. While I strongly suggest you learn to do this on your own, feel free to use them as examples. I'm not saying the are the best but they are pretty good. Try giving one or two of them a flight and see if you can get them into orbit. It may take practice.

http://i.imgur.com/7l5TZ4U.png

Alshain: your "lifter Subassemblies" link isn't working; not for me, at least. (This is the best discussion I've come across in weeks.)

- - - Updated - - -

Okay, I'm going to play with going much more horizontal than I have been. Would clicking the little prograde button (and letting it guide the ship automatically) achieve the ideal? (I'm playing version 1.04 if that makes a difference)

[Red Iron Crown]

Alshain: your "lifter Subassemblies" link isn't working; not for me, at least. (This is the best discussion I've come across in weeks.)

Working for me, at least.

Okay, I'm going to play with going much more horizontal than I have been. Would clicking the little prograde button (and letting it guide the ship automatically) achieve the ideal? (I'm playing version 1.04 if that makes a difference)

This is pretty much the standard procedure for launching now. Launch, tip a little bit, then follow prograde all the way up.