lowest deltaV Mun transfer for landing?

[herbal space program]

Howdy,

I've been trying for some time now to build a SSTMAB (single stage to Mun and back) space plane, and after endless optimization of the design and flight to orbit, I am *almost* there with this honking Mk II-based ship:

Unfortunately, the best I have managed to do so far is get back to low Munar orbit with this puppy before running out of go-juice. I estimate I need maybe about another 300m/s or so of deltaV from where I run out of gas to get all the way back home. I feel like there must be someplace I can squeeze that out if I optimize everything.

So my question is: What is the most efficient trans-munar insertion trajectory for both getting captured by the Mun, and once captured landing with the least deltaV? My guess is that it would be to come in to the Mun's SOI low and from behind, so that the gravity well accelerates the ship both outwards and prograde to match velocities. Has anybody worked this all out for KSP physics? If so, is there a tutorial or general procedure for setting up a maneuver node to do this? TIA!

Edited October 27, 2014 by herbal space program

[mhoram]

This would be really tricky given the Mun's bumpy surface, but I've been wondering if it would be possible to do on the perfectly flat Minmus flats. In theory you could put your periapsis right at zero on the flats, touch down on wheels, and just use wheel braking (and a little downforce) to kill your forward velocity. That would be the ultimate no-burn suicide burn.

Does anyone know if this has ever been done?

Not exactly what you asked for, but close enough: Pinhole Marksman Challenge

[Laie]

[Horizontal landing on the Minmus Flats] Does anyone know if this has ever been done?

Just touching the flats isn't very difficult, however, compared to Minmus' diameter, the hills surrounding the flats are huge. You'd need to come in from a very eccentric orbit. Staying on the ground would require a lot of down-force -> dedicated engines that serve no other purpose and the fuel to drive them. It makes for a nice challenge, but I don't think it saves weight.

[herbal space program]

Yes they are.

So are all the various ladder parts, losing those won't give any advantage.

Do they still have drag? Because I think that's where all those little extra bits would deal me the biggest performance hit. If I could just get 200-300 more m/s out of my jets when I top out, I'm pretty sure I'd make orbit with enough deltaV left to go to Mun and back using the engines I've got.

[Pecan]

...To my gratification, I got a great deal of very helpful advice and ideas from a bunch of friendly people here (thank you!). From you, not so much. Why?

Because, in answer to my question "WHY do you care?" you only answered "Because it's a challenge" - which left me with no more idea of what it was you wanted advice with. Specifically on the subject of Mun transfer my only point was that the possible savings are minute compared to a change in vehicle design or (with a single stage) flight through atmosphere.

I'm sure I could go look them up if I really wanted to, but the whole point of the game is to figure it out yourself as much as possible, isn't it?

So would you have thanked me if I'd just posted "This craft can single-stage to Mun and back ... fly it like this ..." or would it have ruined the challenge? See why I asked WHY you cared?

[Rhomphaia]

Do they still have drag? Because I think that's where all those little extra bits would deal me the biggest performance hit. If I could just get 200-300 more m/s out of my jets when I top out, I'm pretty sure I'd make orbit with enough deltaV left to go to Mun and back using the engines I've got.

Nope, no drag either.

[numerobis]

If you're up for modifying the craft, there's lots to do, but we need to know your design requirements first.

Using a lander can instead of the cockpit saves you a lot of mass. Then you can save on fuel. Two lander cans if you need two Kerbals, runs to 1.2t rather than 2.1t. But it doesn't look like a plane anymore. Even crazier, a seat (or a pair of seats).

Intake spamming: you cool with that? If you spam intakes you can get you almost orbital off just the jets, which could convert a few hundred m/s from rocket to jet, and more importantly, lets you convert your heavy engines to tiny ones. A single LV-909 or a pair of 48-7S will land your plane on the Mun.

With enough wings, you can get a 40t plane into the air with just two jets. That's 1.2t saved.

The small control surface produces far more lift than wings do for the same mass; put a few of them on, tilted 60 degrees, and tweaked to not respond to control inputs (turn off yaw/roll/pitch). Then when you pitch up 30 degrees they are at 90 degrees to the airflow, which gives their maximum lift. This is getting heavily into exploiting Kerbal physics. The main disadvantage flight-wise is that landing will be tough if you put too many on: too much lift, can't land.

With all that engine and structure mass savings, you could reduce fuel a bunch and still get to the Mun and back. But your craft will be a contraption, not a plane.

[mdg583]

My guess is that it would be to come in to the Mun's SOI low and from behind, so that the gravity well accelerates the ship both outwards and prograde to match velocities.

I've thought about this - the thing is that if you kerbin-relative apogee is just barely at the mun, then when you reach the mun, your vessel velocity (relative to kerbin) will be much lower than the mun, so it will come in behind you, and slow you down even more (relative to kerbin) as it pulls you back. I'm been guessing that it may be better if your kerbin-relative apogee would be higher than the mun orbit (larger delta-v at kerbin) so that you come in to the mun more from the side, allowing your mun-relative velocity to be less when you reach it. But I could easily be wrong here. Or it may not be much different.

[KerikBalm]

If I could just get 200-300 more m/s out of my jets when I top out

How fast are you getting on your jets now? make sure to specify if it is surface velocity/orbital velocity

How high are you getting?

[herbal space program]

With all that engine and structure mass savings, you could reduce fuel a bunch and still get to the Mun and back. But your craft will be a contraption, not a plane.

For my part, I want it to look like a plane, but these things you listed are obviously legitimate design choices. If I post a challenge around this, I think I would have different categories for freestyle, Mk1, Mk2, and Mk3 cockpit-based designs, because there ought to be a reward for actually making it look like a plane. I also think I would have all-stock and no extreme part clipping as ground rules.

[herbal space program]

How fast are you getting on your jets now? make sure to specify if it is surface velocity/orbital velocity

How high are you getting?

Right now, I'm starting to gain speed only very slowly at around 1800 m/s (orbital) at 30km. I usually get it up to around 1900 by the time all my extra jet fuel runs out, at which time I fire the rockets and pitch up. Flameout usually happens around 33km. I assume less drag would allow me to go faster at this stage before the fuel runs out.

[herbal space program]

Nope, no drag either.

So do the science experiments, i.e. barometer, gravioli, etc. have drag? I could pitch 4 of those over the side.

[herbal space program]

Even crazier, a seat (or a pair of seats).

I tried this before with a probe core as the root part and I couldn't get a Kerbal pre-loaded into the seat. Is there some trick to this or did I maybe just not have the part placed correctly?

[cantab]

The wiki should say what parts are physicsless. The small experiments are.

You can't put a ship on the pad with an occupied chair. You need to move the Kerbal over manually. Either put a pod on a decoupler and discard said pod before takeoff, or have a kerbal waiting on the ground away from the runway or launchpad.

But anyway a Mun spaceplane is perfectly possible without needing external seats, or nuclear or ion engines. Once you have an efficient flightplan and as many intakes per engine as you're OK with, you simply need to add more fuel.

[wasmic]

Have you tried doing a bi-elliptic transfer instead of a Hohmann transfer? When you're in a 70 km circular orbit, you simply thrust prograde until your apoapsis is right at the edge of Kerbin's SOI. Then once you're at your apoapsis, you burn prograde until your periapsis is as high as the Mun and wait for an encounter. It might seem counterintuitive, but AFAIK it's actually more efficient for the Mun. Of course, it'll take more time and much more skill to pull off such a maneuver instead of a normal Hohmann transfer.

[numerobis]

There was a challenge for min deltaV to the Mun from LKO a while back. What came out of it was:

1. Bielliptic transfer with gravity assist. You can get several gravity assists off the Mun, but there's a cost to lining them up (either in sanity or in m/s), so there's diminishing returns after the easy stuff.

2.

landing video shows the fuel-optimal landing. You set periapsis to be very low, from as far away as possible. When you get close, you burn mostly retrograde but also a bit up, to maintain constant vertical speed as your horizontal speed falls and you drop out of orbit.

3. You hit the surface as hard as possible without blowing up, letting the lunar rocks slow you down rather than using fuel.

[numerobis]

At 1900 m/s orbit, you're only getting 1725 m/s surface speed from the jets. Tack on more intakes (preferably the ram air intakes) and you can get over 2 km/s from jets. IIRC (I did the math once) it's worth having 1 ram air intake per two tonnes of vessel mass for the Laythe and back trip, which is about as much rocket dV as Mun and back.

[tsotha]

Rumor is landing wheels are no longer massless as of 0.25. I have not checked the settings myself, though.

[PLAD]

I tried to get the old Kerbal-X to the Mun and back again with the 6854m/s of dV it has, and failed, but I got from Kerbal's surface to Mun's surface and back again for the lowest delta-V I've seen. Here's the post

I think this method is a little better than the bi-elliptic method because of the trick of only starting with 845m/s to get to the inner edge of Mun's SOI, then using multiple Mun flybys to get flung way out, then doing a small second burn to arrive at Mun with a very small relative velocity. But check it out, I tried to show everything so others could try it. I would be fascinated to see it be beaten. Summary of dV used:

Kerbal surface to a 70x70 orbit(I used normal rockets): 4352m/s

TMI: 845m/s

Mid-course corrections and deep space burn: 2+13=15m/s

Braking into a 6.3x8.8 Munar orbit: 208m/s

Munar orbit to surface (5908 meters altitude): 575m/s

So far it's 1644m/s from LKO to Munar surface, 5996m/s surface-to-surface.

Munar surface to a 9.3x9.3 Munar orbit:584m/s

Munar escape: 205m/s

Course adjustments and deep space burn: 3+21=24m/s

Total from Mun surface to Kerbin surface:813m/s

MISSION TOTAL:6809 m/s.

Of course, it took 26 days, including waiting for the emergency refueling mission to be sent from Kerbin.

[Yakky]

2.

landing video shows the fuel-optimal landing. You set periapsis to be very low, from as far away as possible. When you get close, you burn mostly retrograde but also a bit up, to maintain constant vertical speed as your horizontal speed falls and you drop out of orbit.

Kosmo-not's landing is some incredible piloting, and is OK from an efficiency standpoint, but it's not close to fuel-optimal. Why? His thrust to weight ratio is low and it takes him a long time to slow down. As he slows down, he has to devote an increasingly large fraction of his thrust to fighting off gravity (i.e. pointed off-retrograde). You can see this in how much down-angle he has to apply to keep from falling to the surface. In an ideally efficient landing, the slowdown would be nearly instantaneous and there would be very little wastage on holding the spacecraft aloft.

To be specific, he wastes about 70 m/s delta-v on holding himself aloft while slowing down. You can see this by noting his starting speed is about 580 m/s but he burns 650 m/s of delta-v during his landing.

Rather than use a trio of LV-1 ant engines, he could improve efficiency by using a single 48-7S. It weighs about the same but has 2.5x the thrust of the combined ants, and it has a higher Isp. That means you stop 2.5x as fast and you waste a lot less fuel holding yourself off the ground. Plus it's a more efficient engine. I would suspect he could salvage about 50 of the 70 wasted delta-v by swapping engines (and as a bonus, his fuel tank would hold more total delta-v since the engine is more efficient).

I don't mean to diminish the impressiveness of his amazing landing. I'm just pointing out that it's not fuel-optimal.

Edited October 27, 2014 by Yakky

[numerobis]

For a given spacecraft, his approach is the best anyone could come up with a year ago, and I believe was proved to be essentially the theoretical optimal flight plan (except for the bit of avoiding mountains).

His spacecraft isn't optimal, sure, but that's a different question.

[herbal space program]

Thank you everybody for all the good advice. After replacing my Vesta engines with a couple of nukes and dropping about 4 tons of fuel, my plane can now make it easily!

[NikkyD]

Similar question bothers me lately and quite honestly the answers here are not sufficient.

Given any craft with good TWR and the requirement of a really soft landing on the mun,

how do i get the most out of the mun gravity ?

Let's ignore the part how i start from kerbin. If i go vertically and my AP is exactly in muns orbit, if timed correctly, i would enter muns SOI "standing still" seens from kerbin, but seens from mun i would enter its SOI with 542 m/s straight towards the surface. I would accelerate some more and had to fight all that, doesn't seem smart.

Now lets say i do the same but enter the SOI about 45° towards kerbin. There is a chance that the speed is so high that the mun pulls me towards it but flies by too fast. Same if i enter the SOI from 45° "higher".

I could time it so that i enter before or after reaching AP and still have some velocity up/down (from kerbin). But being in front of the mun is not helping.

If in theory i could be 1 km above muns surface with the above velocities, from muns system it would be like i got ejected with 542 m/s from mun and gravity would slow me down and i would fall from a certain altitude and only had to brake that down.

I am playing with windows paint trying to figure out if there is a way that i enter the muns SOI in such a way, that it wouldnt require me to do something wasteful before and so that i can use its gravity for maximum deacceleration, but i just cant figure out whats more beneficial.

In addition, if going bi-elliptic, i could enter the muns SOI from low and behind. My orbital speed would mean i dont need to brake down the full mun orbital speed and if the AP (from kerbin) is close to the muns surface i wouldn't have to fight much gravity... but the fact that no one pulled this off by now somehow makes me think its impossible and i am missing something!

Help me out here

[Yakky]

The answer is that you want to enter the Mun's SOI with as little speed as possible relative to the Mun. It's easiest to think about it in energy terms. Any object entering the Mun's SOI and going to the surface is going to pick up the same, unavoidable, amount of potential energy as it falls through the Mun's gravity well. If you just free fall, all of that potential energy gets converted into kinetic energy (i.e. speed), which needs to be dissipated if you don't want to crash.

Any speed with which you enter the Mun's SOI is just additive to your total energy that will ultimately need to be dissipated. If you enter with non-zero velocity, now you have the same amount of potential energy as before, but you also have extra kinetic energy. All of the potential energy, plus all the kinetic energy, is going to be converted to kinetic energy (speed) by the time you reach the surface, which means you will just be going faster once you get there. There is no way for your initial SOI entry speed to give you negative kinetic energy-- it will always be positive-- so you can't design a lower-energy Munar approach than one that enters the SOI as close to zero-speed (relative to the Mun) as possible. No clever entry trajectories can get around this fact, although flybys that subsequently exit the Mun's SOI can sometimes help set up a better (slower) final encounter later.

In practice the means you want your pre-entry orbit around Kerbin to be very similar to the Mun's orbit, so that you just sort of barely drift into the Mun's SOI.

Hope this helps.

Edited January 15, 2015 by Yakky

[NikkyD]

There was a challenge like "least deltaV from LKO to mun surface", can't find the thread, can anyone else find it ?

Edit:

thx @ Yakky, u are right

in addition:

if maccollo is right with http://forum.kerbalspaceprogram.com/threads/73696-Landing-from-Low-vs-High-Orbit?p=1045038&viewfull=1#post1045038

a perpendicular drop from 2.400.000m with initial velocity 0 would result in an impact velocity of 546 m/s

same with initial 500 would result in 650 m/s

My thoughts were that if you, in fact, race towards the planet, you spend less time in its gravity well.

If above is correct, any attempt to waste fuel on entering muns SOI with lower speeds is counterproductive!

So the cheapest way is in fact to reach muns SOI with the least deltaV possible and have high TWR for a good short suicide burn.

comments welcome !!!

Edit #2:

somehow the math of maccollo seems off, the interesting part remains.

Just performed an almost complete stop at the corner of mun SOI and would have impacted with 770 m/s (roughly didnt watch)

the same without any correction (vertical speed, mechjeb, -350 m/s) results in 850 m/s impact!

Edited January 15, 2015 by NikkyD