efficient landing - from circular or elliptic orbit

[seyss]

Hello,

I'm trying to land on Tylo (again) and I'd like to ask if it's better to go into a low-altitude circular orbit around Tylo before landing or, from the orbit below, burn Prograde at Apoapsis and land more vertically?

Thanks.

edit: fixed the image

[TheXRuler]

Raise your periapsis to a few kilometers and Land using technique shown in the attached video, it is by far the easiest (won't smash you into the ground) and the most efficient also.

By landing with this profile you will be taking maximum advantage of the Oberth Effect. Also you will reduce gravity losses to a minimum by not freefalling from 7km

Edit: No need to have a circular orbit beforehand, except that it makes things a little easier.

[Starwhip]

Whoah, yeah, circularize at periapsis beforehand. That orbit is way too elliptical.

Not that you couldn't land with that orbit, but circularizing will be much simpler.

[Warzouz]

When you follow an ellipitc orbit you go very slow on apoaps, but very fast on periaps. That's why we retrograde at periaps to reduce apoaps to circularize.

But with a high TWR, maybe you would be able to do both. Circularise an orbit has a cost too.

You should circularize first. Another advatage of circular orbits : you can easily select your landing area. With elipitic orbits you have to start landing procedure at periaps. That could leads to land by night.

[Red Iron Crown]

Most efficient would be to lower your periapsis from apoapsis to as low as you can without hitting any terrain, then at Pe perform your deorbit and suicide burn. During that deorbit burn you will pass a point where the orbit has been circularized.

[Scarecrow88]

The problem with Tylo is that there is no atmosphere to help slow you down, so all deceleration has to be done with engines. I have found that the best way to land there is to have my interstellar drive put me in to a low circular orbit, and to use a separate dedicated lander to get to the surface and back.

As others have said, if you are trying to land from an eccentric orbit, your speed at periapsis will be higher than it would be from a circular orbit, and you have got to lose that speed at some point in order to land.

[Starhawk]

The posts above pretty much sum it up.

Most efficient is described by Red Iron Crown.

Safest/easiest is, as several people point out, circularize and then land.

Happy landings!

[seyss]

Raise your periapsis to a few kilometers and Land using technique shown in the attached video, it is by far the easiest (won't smash you into the ground) and the most efficient also.

Edit: No need to have a circular orbit beforehand, except that it makes things a little easier.

I'm trying this method however Tylo's terrain is not flat as the Moon's.. I'm doing fine at ~5km altitude then boom a mountain kills me....

[Warzouz]

What is you TWR now ? Is it your very low TWR lander ?

Starting at 5 km seems very bold. I think there are some mountains at 4km high. That leave you only 1000m. So you MUST keep your vertical velocity at 0 while killing your orbital. With Low TWR, this is nearly impossible on the long term.

Landing on highlands is a good idea (less fuel down and up) but you should start higher.

[herbal space program]

It's a treadeoff of thrust vs. engine efficiency.The most efficient thing in terms of deltaV is as R.I.C. said, to lower periapsis to just above the surface and then burn very hard at periapsis to stop yourself. From that high apoapsis, this will require a very high TWR or you will drill a hole. If you circularize at a low altitude first, you will need less thrust to pull off the landing, but will have to burn more fuel overall. In terms of total fuel consumption, I'm not quite sure. Engines that can deliver the kind of thrust required for the suicide burn landing tend to be heavy and inefficient. It may actually be better to use more deltaV to land from a lower orbit, because you can do that with a weaker and more efficient engine. Thoughts?

[Red Iron Crown]

Engines that can deliver the kind of thrust required for the suicide burn landing tend to be heavy and inefficient. It may actually be better to use more deltaV to land from a lower orbit, because you can do that with a weaker and more efficient engine. Thoughts?

Cluster the weaker, more efficient engines and have the best of both worlds.

Seriously though, Tylo's dV requirements are high enough (~6km/s for a narrow margin landing/ascent) that the high efficiency engines will win out. If it was the Mun or somewhere else with lower requirements then a case might be made for using 48-7Ss or some other high TWR, lower Isp engines.

[OhioBob]

During that deorbit burn you will pass a point where the orbit has been circularized.

Important point. Performing a suicide burn at periapsis of an elliptical orbit is really no different in terms of ÃŽâ€v than first circularizing and then performing a suicide burn. The only difference is that in the first case you burn right through the point where the orbit is circularized, while in the second case there is a pause in the action. The advantage or circularizing first, as Warzouz said, is that you can then select any site along your orbital path to land. In the other case you're pretty much committed to landing at the site of the initial periapsis.

Most efficient would be to lower your periapsis from apoapsis to as low as you can without hitting any terrain

I think that's true of everyplace but Tylo. Orbital velocity around Tylo is much higher than any other airless body, which means that the duration of the suicide burn is much longer than other places. Unless you have a very high TWR, this can result in a considerable loss of altitude during the burn. The initial altitude must be high enough to account for this. For my one and only Tylo landing, the stage I used for the suicide burn had an initial TWR of only about 1. I started the burn at an altitude of about 47 km and was down to about 8 km when the burn ended. If I had to do it again, I'd consider increasing the TWR.

[Warzouz]

Well we had a discussion about that on another topic. The suicide burn on retrograde is far from being the easier solution for landing. You have to have a decent TWR. With lower TWR, this will always end crashing.

With lower TWR, you have to kill horizontal while keeping vertical under control. Then switch to retrograde and land when horizontal is nearly killed.

This is less efficient than suicide burn, but this is the only option with low TWR.

With this technique, I was able to land and reorbit a 14T SSTO rocket with used dv of 5900m/s (out of 6050). starting at 30km

[OhioBob]

Well we had a discussion about that on another topic. The suicide burn on retrograde is far from being the easier solution for landing. You have to have a decent TWR. With lower TWR, this will always end crashing.

With lower TWR, you have to kill horizontal while keeping vertical under control. Then switch to retrograde and land when horizontal is nearly killed.

This is less efficient than suicide burn, but this is the only option with low TWR.

With this technique, I was able to land and reorbit a 14T SSTO rocket with used dv of 5900m/s (out of 6050). starting at 30km

I just kept burning retrograde until I killed nearly all my velocity (both horizontal and vertical). I then free fell from about 7-8 km AGL until I started a final breaking burn at about 2.5-3 km AGL. The stage I used for final breaking had a TWR of about 3.5. This higher TWR stage was also used at the very end of the suicide burn, killing the last 250-300 m/s. Using this method my total landing ÃŽâ€v was about 3400 m/s. My ascent/reorbit ÃŽâ€v was right around 2640 m/s, for a total of 6040 m/s. I'm sure that if I used a higher TWR for the majority of the suicide burn, and started from a lower altitude, I could get down to your 5900 m/s or maybe even a little less.

Edited April 14, 2015 by OhioBob

[P.Lumumba]

Can someone please explain why a suicide burn, or anything else, is more efficient than killing all velocity at Ap and guiding it straight down from there ?

[Wolf Baginski]

If you have KER, go to the VAB and build a test vehicle, same fuel and same empty mass as the mass of the landed vehicle. (Yes, you will need fuel to take off again) and try a few different engines, watching the TWR and the deltaV. The gotcha is the destination-gravity. Your TWR needs to be greater than local gravity. I have built a few landers with those small radial rockets. 1000Kg landed on Mun needs more thrust to balance the gravity than 1000Kg landed on Minmus.

The basic Rocket Equation shows that the faster you burn the fuel in these conditions, the more deltaV you get. That's what a suicide burn does. If you need to burn X units of fuel per second to balance gravity and your system can burn X + Y in total, only the Y gives you deltaV. The precise calculation is a bit more tricky, so I shall keep it quick and dirty. if you need a 10-second burn, the total fuel is 10 ( X + Y ) and 10 Y is what stops you. Double the excess thrust, halve the burn time (This is where the rocket equation comes in: it isn't really this simple) and the total is 5 X + 10 Y So you have saved 5 X fuel.

Now, what does this mean for system weight?

Check the http://wiki.kerbalspaceprogram.com/wiki/Parts#Liquid_Fuel_Engines for the stock parts.

The first thing to check is the TWR for the engine. Both the OV-10 Monopropellant Engine and the Rockomax 24-77 have the same TWR of 22.7, which is pretty good. The Rockomax has a slight edge on isp, which means it will use a little less fuel, and if you think you will need it in atmospere, the margin is bigger. Engine mass is 90 Kg.

Let's assume 20000Kg on Minmus, which is rather large for a lander. Both these engines have 20 kN of thrust, which mean 1 Newton per kilogram, which is 1 m/s² total acceleration. Not enough.

Two of these engines gives us 2 m/s² and Minmus gravity is 1.635 m/s² so we have the excess to stop. But the useful acceleration is only 0.365 m/s² which is rather pitiful.

Add a third engine, which feels more balanced as well although two will work, and your useful thrust is is nearly quadrupled to 1.365 m/s²

A fourth engine, for a total increase of 180 Kg over the minimum, gives us 2.365 m/s² That reduces the burn time a lot, by a factor of more than 6. What does this mean for the fuel consumption? That table shows the Rockomax burning 9.3 Kg/s to balance Minmus gravity. So for every 6 seconds of the minimal system's burn you can save 5 seconds worth of fuel, which is conveniently half the mass of one of those engines. So if you expect more than 24 seconds, using four of those engines instead of two will pay off in mass.

24 seconds with the minimal 2-engine rig is less than 9 m/s

Radial Engines?

I do have a preference for radial engines for landers. There is also the Rockomax 48-7S which has a thrust of 30 kN and a TWR of 30.7 to work with. It's a bit more of a fiddle to place them, and for a practical suicide burn the weight difference seems marginal. You don't need automatically need the weight of landing legs for radial engines, though they give you more margin for error.

The Tsiolkovsky rocket equation.

This is really old, and at the core of rocket design. It's described on the Kerbal Wiki in this tutorial: http://wiki.kerbalspaceprogram.com/wiki/Tutorial:Advanced_Rocket_Design#Delta-V and you will notice that time doesn't really come into it. Well, I hope I have showed why, when working against gravity, a faster fuel burn pays off.

Other Factors.

This is where there might be a weakness in the simulation. The structure of the craft has limits in Kerbal Space Program. Connections between parts will break under sufficient load. But these limits are generally quite high. You could build a real lander with less inherent strength and reduce weight, because of the low gravity it operates in. KSP sets these limits quite high. Some real-world rockets have needed them insanely high. The Sprint missile, developed in the Apollo era, accelerated at 100g http://en.wikipedia.org/wiki/Sprint_%28missile%29 . The warhead massed about 70 kg which meant the support structure has to be able to support 7 tonnes in flight.

On the other hand, while the Apollo LM was lightly constructed to land on the Moon, it had to survive launch too. The design started with three landing legs, the most stable configuration was five, but the engineers compromised on four to save weight.

Also, a lot of the landing process is the deceleration from orbit. The gravity effect only bites in killing the vertical speed. Which is also why just blasting straight up on launch wastes fuel. You can do a vertical flight to Mun, but it's a rather crude and unsubtle way of doing the job.

[Captain H@dock]

Can someone please explain why a suicide burn, or anything else, is more efficient than killing all velocity at Ap and guiding it straight down from there ?

Edit: Not so sure my answer is the correct answer to your question because it looks as if you meant Pe. I left it because of the confusion over what a suicide burn is. But what you want to look at is vectorial sum (here we're talking about vector components, so it's really just Pythagorean theorem).

Your method requires burning dV_orbital at first, then dV_vertical on the surface (suicide burn for a vertical free fall), in two separate burns.

Real suicide burn (with a small deorbit burn to be on a balistic trajectory) do these two burns at the same time, burning retrograde. Since these two burn are component on the vertical and "orbital" axis, their values are the same (or very close) to the ones in your methods, but by burning them together, they require SQRT(dV_orbital² + dV_vertical²) dV, which is less.

Because gravity is an acceleration, which means it will increase your vertical speed for every unit of time you spend on a ballistic trajectory.

Since you want to nullify vertical speed for landing (else it's called a crash), all gained vertical speed will need to be shed burning vertically (consuming deltaV).

Yet reducing your vertical speed means you'll reach the surface later, thus increasing the amount of time you spend on a ballistic trajectory, thus increase your vertical speed more, thus requiring more deltaV to nullify it.

Therefore you want to reduce your vertical speed as late as possible, yet early enough that it is about 0 when you reach the surface. That's the definition of a perfect suicide burn.

This assumed atmosphere-less body, obviously.

Edited April 15, 2015 by Captain H@dock

[Red Iron Crown]

Can someone please explain why a suicide burn, or anything else, is more efficient than killing all velocity at Ap and guiding it straight down from there ?

Because of the Oberth effect. Landing is effectively subtracting energy from your trajectory until your Ap is at surface level and your speed is the same as surface rotational speed, and it is most efficient to change a trajectory's energy when moving fastest.

[celem]

When going into Tylo Im normally in a 10km low orbit, then start scrubbing horizontal over a considerable period of time. Doing this without losing so much vertical that you hit a mountain at hypersonic velocity means burning a good chunk above the retrograde horizon, which in turn wastes dV. The lower your TWR then more you lose.

While people are correct in saying these suicide burns are more efficient that is most true for decent TWR or higher. A weak engine will have to pitch so much to maintain vertical that it begins to waste again. (steerage losses) A nuke lander i'll normally bomb in from high orbit and make a very lengthy (and very risky) suicide burn which is mainly vertical since it can then just hug retrograde. With clusters of 48-7S or something i'll do the horizontal suicide from a 10-12km low orbit. This problem is only noticably powerful at Tylo, everything else with enough G to be challenging has atmosphere which does 99% of the work for you.

In an ideal world your suicide burn ends at like 50cm. In practice you shoot for a bit higher, maybe 1km, enter a freefall and then make a second vertical suicide burn at maybe 100m. Done this way you can convert your first burn into something that kills horizontal but not all of the vertical (horizon behind retrograde) Maintaning a moderate descent rate across the last 1km/500m might be safe, but it is costly, very costly.

Edited April 15, 2015 by celem

[Streetwind]

Can someone please explain why a suicide burn, or anything else, is more efficient than killing all velocity at Ap and guiding it straight down from there ?

A different way to look at it:

Let's say that falling one kilometer in distance towards a planet requires you to run your engines for 1 second to null out the velocity that that falling has given you.

Even though that's a crass simplification, you can clearly see that you want to minimize the time spent falling straight down, because it means you must run your engines less. Killing your velocity at Ap, however, ensures that you are falling the longest possible distance. It's a very bad idea.

Edited April 15, 2015 by Streetwind
Misread question at first

[P.Lumumba]

Thanks for trying everyone.

I have a basic understanding of physics, and it got a bit better by playing KSP the last few years, but this is beyond me.

let's say I want to land on one of the poles, so the bodies rotation doesn't come into play all too much.

I based myself on the law of conservation of energy. Velocity and altitude combined give you the starting energy level, when you land it should be close enough to zero. (low speed, zero altitude)

You use the engine the make it go down. (I know this is also energy at the start, but you're using it in the opposite direction)

Am I right this far ?

And if yes, does it mean that you get more energy out of 1 ton of fuel if you burn it up quickly (suicide thing) rather then slowly (slow controlled descend)

Edited April 15, 2015 by P.Lumumba

[Starhawk]

Can someone please explain why a suicide burn, or anything else, is more efficient than killing all velocity at Ap and guiding it straight down from there ?

Yet another way to answer this question is to consider Pythagorus/vector addition.

If you expend 1m/s burning East and 1m/s burning South, you end up with a net change of 1.41m/s SE. But if you just burn 1.41m/s pointing SE you get the same result, saving you .59m/s.

And if yes, does it mean that you get more energy out of 1 ton of fuel if you burn it up quickly (suicide thing) rather then slowly (slow controlled descend)

You don't get more energy out by burning more quickly. You do however spend less time fighting gravity, and therefore less energy fighting gravity.

Happy landings!

Edited April 15, 2015 by Starhawk

[Red Iron Crown]

Thanks for trying everyone.

I have a basic understanding of physics, and it got a bit better by playing KSP the last few years, but this is beyond me.

let's say I want to land on one of the poles, so the bodies rotation doesn't come into play all too much.

I based myself on the law of conservation of energy. Velocity and altitude combined give you the starting energy level, when you land it should be close enough to zero. (low speed, zero altitude)

You use the engine the make it go down. (I know this is also energy at the start, but you're using it in the opposite direction)

Am I right this far ?

For the most part, yes. You still have some potential energy in the orbital frame of reference when landed, because orbital radius is measured from the center of the body, but that is largely an academic distinction.

And if yes, does it mean that you get more energy out of 1 ton of fuel if you burn it up quickly (suicide thing) rather then slowly (slow controlled descend)

Not so much that you burn it quickly, but that you burn it when travelling at a higher speed. Burning one ton of fuel gives more energy change the faster the vessel is moving, so it is better delta-V-wise to do as much of the landing burn as possible as close to the surface as possible, all other things being equal. In practise, it's sometimes better to leave yourself some margin, both to compensate for lower TWR (gotta make sure you don't hit the ground when scrubbing off that horizontal speed) and to allow for some of us having less than perfect piloting skills.