So... Spaceplanes for Duna

[juanml82]

So, now that I'm finishing my Duna exploration during career by using some rather frustrating spaceplanes, I wonder what's the forum's conventional wisdom about designing spaceplanes for use in Duna.

Here's the design I've been using

And these are the problems I've found using HOTOL vehicles in Duna:

Little control authority through aerodynamic surfaces

The atmosphere barely brakes the spacecraft

Which tends to result in high speeds during touchdown

On the plus side, little drag means stuff can be mounted radially without a serious performance hit. Or so I think

The terrain is rather rolling and uneven, which can lead to the spaceship tipping when it hits the ground

No oxygen (duh!)

Exploring the planet can eat up some serious amount of fuel, so 3 km/s of dV seems to be the minimum

So:

Engines: That design has aerospikes, which have good thrust, are short and perform well in Duna's thin atmosphere. However, they lack gimbal - and we already have little control, so a thrust vectoring engine can be a good idea. But is there an engine that can achieve good fuel efficience (the dV target for such a ship should be above 3 km/s IMHO)?

I also tried nukes, but their increase in ISP is negated by their weight, their large size (which makes them even more likely to crash into the ground during landing), and their pitiful thrust.

Cutting horizontal speed: It looks like retrorockets are a must. I've used two radially mounted Twitch, but they have too little thrust. The one with nukes I've tried (not pictured, it uses 3 nervs) has four but, again, they take to long to cut horizontal velocity. I was thinking about adding terriers instead, or even a forward facing aerospike.

Control: RCS, lots of it, as well as reaction wheels. That spaceship I've posted completely lacks vernors, and that's a mistake. Rover wheels for use in the ground also seem like a good idea, since I usually overshoot my landing destination. But the landing gears and firing up the engines work well anyway.

Landing: Well, I think this is the harder part of HOTOL vehicles in Duna. Aerobreaks help, but they don't slow down a lot. There is also the matter of tipping when hitting the ground, so I think the medium (or large) landing gears are better than the smaller ones, despite the extra weight. I usually design the landing gears with a tricycle arrangement, but I often find ships tipping forward and, when the forward gear hits, the ship tips violently towards some side, which leads to breaking the wings. Would four gears instead of three help improve things?

Parachutes are needed to bring the ship to a halt once it hits the ground (again, I've missed those in this ship) but they can't be used to slow down before touch down (well, they can, but that tends to lead to an unplanned disassembly). Also, that means an engineer must be among the crew in order to repack them later.

The other option is to spam vernors and make a vertical landing (or use any other engine arrangement for that). So, kill H/S with a retrorocket, land vertically. But then the question is: if it's landing vertically, if you want to get the wheels into the air ASAP when taking off, and you don't get too much out of aerodynamic control surfaces anyway, why bother to attach wings to the thing?

So, ideas, thoughts, opinions, hints?

[Kelderek]

You should take a look at this recent tutorial, it may have some additional tips for you:

Flying on Duna in 1.0.4 Some General Guidelines

[Geschosskopf]

Well, Kelderek linked my post (not really a tutorial, just guidelines) so I don't have that much more to say, just emphasize a few points.

Duna and Kerbin are totally different environments so something that flies well on Duna might not fly at all on Kerbin, and something that can fly on Kerbin might be able to reenter and fly around on Duna but probably won't be able to land safely (except in a few, isolated places). So you kinda have a choice in making something that flies well in 1 place or the other. You might be able to do both but on Duna the thing will be severely limited to where it can land.

The guidelines linked above are mostly for airplanes that will never leave Duna. However, Duna-only spaceplanes aren't much different given the happen circumstances currently prevailing there. That plane of mine I illustrated could easily be a Duna-only SSTO with just a tiny bit of rocket power, given that it can throw itself out of the atmosphere with just its prop.

The hardest part of flying on Duna is landing safely. Both wheel brakes and airbrakes are largely ineffecrtive at slowing you down and slowing down is vital because long, reasonably flat stretches of terrain are limited to certain areas of the big basins and canyons. Therefore, you pretty much require retro thrust to land before some lump in the terrain kills you, whehter you fly it in all the way or use chutes for the final stage of the landing. Because of all this, it's very nice to have scads of wing so you can keep flying in the thin air despite being slow enough to land safely. And it's this huge wing that makes the plane such a dog on Kerbin.

So really, the best thing is to design a plane or spaceplane solely for Duna and have it dock at a station for refueling and crew transfer to a ship that cycles back and forth between Duna and Kerbin. At Kerbin, this ship docks with another station serviced by a Kerbin-based spaceplane. Yes, you can make a single ship that can make the whole trip, but it's performance and usefulness will be compromised at both ends.

[jwenting]

And of course, because of the very much thinner atmosphere, lift is greatly reduced therefore you need either more speed and/or far larger wing area to stay in the air.

Think a U-2 going as fast as an SR-71 just to have the amount of lift generated by a small Cessna on Kerbin.

[juanml82]

Well, Kelderek linked my post (not really a tutorial, just guidelines) so I don't have that much more to say, just emphasize a few points.

Duna and Kerbin are totally different environments so something that flies well on Duna might not fly at all on Kerbin, and something that can fly on Kerbin might be able to reenter and fly around on Duna but probably won't be able to land safely (except in a few, isolated places). So you kinda have a choice in making something that flies well in 1 place or the other. You might be able to do both but on Duna the thing will be severely limited to where it can land.

The guidelines linked above are mostly for airplanes that will never leave Duna. However, Duna-only spaceplanes aren't much different given the happen circumstances currently prevailing there. That plane of mine I illustrated could easily be a Duna-only SSTO with just a tiny bit of rocket power, given that it can throw itself out of the atmosphere with just its prop.

The hardest part of flying on Duna is landing safely. Both wheel brakes and airbrakes are largely ineffecrtive at slowing you down and slowing down is vital because long, reasonably flat stretches of terrain are limited to certain areas of the big basins and canyons. Therefore, you pretty much require retro thrust to land before some lump in the terrain kills you, whehter you fly it in all the way or use chutes for the final stage of the landing. Because of all this, it's very nice to have scads of wing so you can keep flying in the thin air despite being slow enough to land safely. And it's this huge wing that makes the plane such a dog on Kerbin.

So really, the best thing is to design a plane or spaceplane solely for Duna and have it dock at a station for refueling and crew transfer to a ship that cycles back and forth between Duna and Kerbin. At Kerbin, this ship docks with another station serviced by a Kerbin-based spaceplane. Yes, you can make a single ship that can make the whole trip, but it's performance and usefulness will be compromised at both ends.

Yeah, I wasn't thinking in making a single-stage-to-Duna. Even if possible, it's not worth the effort.

And yes, landing on Duna is by far the most complicated part. What I get from your post is that I really need a lot more wing, and adding small landing gears in the wing tips is a great tip.

I think I'll settle for 2/3 aerospikes for propulsion and 1 as a retrorocket, so I can cut H/S quickly.

That, and conventional rockets with rover wheels instead of planes

[Geschosskopf]

And yes, landing on Duna is by far the most complicated part. What I get from your post is that I really need a lot more wing, and adding small landing gears in the wing tips is a great tip.

I think I'll settle for 2/3 aerospikes for propulsion and 1 as a retrorocket, so I can cut H/S quickly.

That, and conventional rockets with rover wheels instead of planes

[juanml82]

Yes, you do need gobs of wing. It's the ratio of lift value from the wing parts to the mass of the ship that's important. I don't recommend going lower than about 1.4 lift/ton or your landing speed will be dangerously high (for STOL) or your landing accuracy will suffer (for parachute STOVL).

This ratio of lift to mass reinforces a practical upper limit on plane size that is imposed originally by the terrain. The bigger the plane, the more likely it is to high-center its belly on a bump, or for a wingtip to catch a bump and flip the rest of the plane or just break off. Thus, Duna favors relatively small plane dimensions, which means that if your plane is heavy, you'll want to go with layers of wings to keep the wingspan down while still having sufficient lift. The problem there, however, is that short wingspans mean low leverage for the control surfaces, while the higher mass of the plane requires more force to change its direction, so you start running into real problems with control authority at low speeds. You might end up needing RCS to suppliment your aerodynamic control surfaces, which adds more mass, which means more wing, etc.

I say this because aerospikes and their fuel are heavy so having several of them will present problems. But planes don't need much at all in the way of thrust to achieve hypersonic speeds. I'd bet that you could make a perfectly good Duna SSTO with just 1 aerospike on the back and some of the small orange radial engines on the nose for braking when landing. Experiment

Right, but one of the issues I've found when trying to land is having to abort a landing because I'm falling too fast and onto an undulation in the terrain - and that's when I need thrust to get back into the air and "jump" over the terrain.

But I still need to design something with a lot of wing that doesn't mean I have to try a landing more than ten times because I keep breaking the wings.

[Geschosskopf]

Right, but one of the issues I've found when trying to land is having to abort a landing because I'm falling too fast and onto an undulation in the terrain - and that's when I need thrust to get back into the air and "jump" over the terrain.

High sink rates are the result of insufficient lift. MOAR wing . Adding thrust just prolongs the problem because it speeds you back up so now you have to slow down again.

[Mesklin]

High sink rates are the result of insufficient lift. MOAR wing

Agreed, MOAR wings, wide placed gears and small thrust for maintain horizontal speed about 50-70 m/s and vertical descent speed 1...5 m/s works very good for Dune. Also 2 wide placed drogue parachutes opened as soon as possibly after touchdown can prevent override craft and slowly and safely decrease speed.

Javascript is disabled. View full album

Edited August 23, 2015 by Mesklin

[Jason Patterson]

Agreed, MOAR wings, wide placed gears and small thrust for maintain horizontal speed about 50-70 m/s and vertical descent speed 1...5 m/s works very good for Dune. Also 2 wide placed drogue parachutes opened as soon as possibly after touchdown can prevent override craft and slowly and safely decrease speed.

+1 to this. I really like drogue chutes on Duna.

[Mister Dilsby]

I've had good success with a comparatively heavy (~70t fueled) miner/refiner powered by two aerospikes. Wing area is not big enough for an unpowered landing, especially at higher altitudes, but it does have a Vernor thruster based VTOL system that enabled it to land in all biomes of Duna and Ike. Really as long as you have enough thrust/dV and some kind of landing system that doesn't rely solely on the wings (VTOL or 'chutes) you should be able to get away with almost anything on Duna.

- - - Updated - - -

Cutting horizontal speed: It looks like retrorockets are a must. I've used two radially mounted Twitch, but they have too little thrust. The one with nukes I've tried (not pictured, it uses 3 nervs) has four but, again, they take to long to cut horizontal velocity. I was thinking about adding terriers instead, or even a forward facing aerospike.

Right--with this design I was able to cut horizontal speed quite well using the VTOL engines and a cobra maneuver (pitch nose up to 70-80 degrees) and thrust UP K with the Vernors to slow down. No additional engines required, but you will need lots of aero/RCS/gyro control authority to pitch up that quickly.

[panzer1b]

Interesting thread, ive honestly never thought about this as 90% of everything that went to duna was either touch down and get back out (hence dV needs arent much), or a permanenet base/dropped vehicle that i just parachute spammed my way to the ground.

As for planes, all i can say is the more lift and control surface area the better. Duna at ground level is more or less like flying at ~15-20km on kerbin, its enough to do something, but unless you have excessive wing area, expect alot of lithobraking, which in itself is quite tough to actually do without breaking apart (Duna is notorious for uneven ground in 90% of teh entire planet, and even flat looking area have alot of what id call death jumps, areas where the terrain will send u into the air slightly, and if ur speed isnt high enough to maintain lift, but also isnt low enough to stop u from going airborne, itll whack u into the ground behind hill at rather high velocity. Then again, ive landed on duna a few times in a laythe/kerbin designed SSTO, ofc my landing speed was ~200m/s, and at that speed it took a few quicksaves before i managed to survive the ordeal. That SSTO was ~20 tons heavy, and has ~6 or so total lift area (not counting the vertical tail surfaces). If you come across a death jump, blast ur engine for a split second to help cut or vertical velocity down so u dont slam into the ground at >50m/s vertical speed. Aside from that, if you can manage a flat area, braking is usually workeable, if slowish. Once below 50m/s you are safe unless u fly off a mountain since the death jumps seem most dangerous around 100m/s where you dont have any more lift whatsoever, but are going fast enough to get slammed into the ground rather badly.

Anyways, ive found lift area and control surfaces to be better then airbrakes in every situation (and aside from a few tests when they were released i dont even use airbrakes for anything except drop pod style escape pods, as in no parachute retrorocket pod, that lithobrakes, but the airbrakes are needed to not go into ground at 500m/s).

Ohh, and if you can make one that works well, and is balanced, VTOL makes every landing a joke in terms of difficyulkty, only issue is that it needs to be very balanced CoM wise, and you need to have more fuel to drag those extra VTOL drives on a plane.

Also, one question, what do you gusy reccomend for engines? Aerospikes have the bets ISP for a LFO engine (and unless u need higher thrust are pretty good mass wise too, 1.5t is nothing now when jets are 2 tons now), but nukes seem to have so much more ISP that even if you spend forever climbing, they seem to work well. Its getting to the pojnt i feel nukes are the only viable engine is KSP for general purpose non atmo work (tylo aside ofc).

[Mister Dilsby]

Also, one question, what do you gusy reccomend for engines? Aerospikes have the bets ISP for a LFO engine (and unless u need higher thrust are pretty good mass wise too, 1.5t is nothing now when jets are 2 tons now), but nukes seem to have so much more ISP that even if you spend forever climbing, they seem to work well. Its getting to the pojnt i feel nukes are the only viable engine is KSP for general purpose non atmo work (tylo aside ofc).

I used to fly LV-Ns on Duna, but in 1.0.4 aerospikes are only 1t, and with nukes are 3t--that plus the slight nerfing of thrust in Duna atmosphere (<55kN below 5km) has changed my mind

[panzer1b]

I used to fly LV-Ns on Duna, but in 1.0.4 aerospikes are only 1t, and with nukes are 3t--that plus the slight nerfing of thrust in Duna atmosphere (<55kN below 5km) has changed my mind

true, but does that thruist nerf actually mean anything when we are talking about such a massive change to ISP (its still ~ 2X that of a aerospike at ground level on duna). Ofc the longer you burn, the more fuel u are using to do teh maneuver, but arent nukes just the only viable engine to use for this situation due to sheer difference in ISP? (id die for a shorter nuke that doesnt look bad on SSTOs). That and ions ofc, but whos going to have 50 ion engines on one craft? Well maybee im not that patient, but i cant stand and less then 2KN thrust per ton of craft, that means 30t is the absolute max ill even touch with a single nuke, unless its a super short burn in the 1st place.