Flying Duna

[Geschosskopf]

I've spent the last week trying and finally succeeding in making an airplane that flies reasonably well on Duna. I made this because airplanes, where they're possible, have a lot of advantages for exploring large areas compared to rovers and multiple lander hops. Doing this was by far the greatest design challenge I've come across so far in over 300 hours of playing, and it resulted in something not only quite useful but also fun. So, if you want to cover long distances on Duna, or create an air service to ferry Kerbals or whatever between your scattered bases there, you might want to try making a Duna airplane yourself.

Design Requirements:

1. The vehicle must create its lift by the motion of aerodynamic surfaces through air. So, airplanes and helicopters are OK, things held up by rockets and balloons are not.

2. The vehicle must be capable of being delivered from the ground on Kerbin to the ground on Duna. However you get it there, apart from HyperEdit, is fine.

3. On Duna, the vehicle must be able to take off under its own power from a location with an altitude of at least 2500m and be able to climb to and maintain an altitude of at least 5000m. Higher is better

4. The vehicle must have sufficient range to fly at least 1/4 of the way around Duna and return to base, or at least 1/2way around Duna to another base. Longer range is better.

5. The vehicle must be capable of indefinite use without further support from Kerbin. IOW, if it burns fuel, you have to be able to refuel it on Duna, and it must be able to make repeated flights, so no jettisoning parts to enable single flights. If the vehicle has parachutes, you must provide access to them so Kerbals can repack them between flights.

6. The vehicle must be able to carry at least 2 Kerbals. More is better, as is enclosed seating.

7. The vehicle needs to be able to land in short, fairly rough places to enable ground examination in any area of potential interest. VTOL beats STOL.

Other Rules:

1. Mods: You have to use mods to do this. In general, do what you gotta do but try not to use things that are way overpowered, like the electric engines from SM. I used B9, Firespitter, Kethane, Modular Multiwheels, KAS for refueling, the GPS unit from ISA MapSat to help get to anomalies, and MJ to keep an eye on the weight.

2. CFG Tweaks: To make sure we're all doing the same thing, don't tweak CFG files.

Scoring:

You get 1 point for creating a vehicle within the above rules that meets all the design requirements.

Bonus points as follows:

• +1 for every extra 1000m in service ceiling above 5000m obtained
  
• +1 for being able to circumnavigate Duna without refueling
  
• +1 for carrying more than 2 Kerbals
  
• +1 for carrying something besides Kerbals, like a rover or some sort of cargo
  
• +1 for having internal Kerbal seating
  
• +1 for VTOL
  
• +1 for using FAR
  

Tie-breaking will come down to measuring details, like the highest altitude if both are in the same 1000m band, or the number of Kerbals carried, etc.

My Version

Here's what I came up with. I call it the D'OH, the most recent in a family that included the DORK, the DOPE, and the DRUNK. It meets the minimum design requirements with 1 bonus point for being able to circumnavigate Duna.

Propulsion is a Kethane-electric hybrid system combining solar panels with the Kethane-burning generator. They're hard to see but there are 6 of the radial Kethane tanks on top of the lower wing, carefully positioned so as not to affect CoM as they drain. The generator has the virtue of only kicking on as needed, like the gas-burning engine in my Prius. The solar panels do most of the work but I have the Kethane for extra power if needed, for when the sun's at a bad angle, and for night operations. The generator is in the rear fuselage just in front of the tail fins. All this feeds 2 of the Firespitter electric motors.

The Mk 1 fuselage tubes, the wings, and the 16 air brakes are from B9. The main landing gear and the small wheels under the wingtips are from TT's Multiwheels, and the tailwheel (this is a taildragger to save weight) is also from Firespitter. There are also some of the B9 omnilights scattered around. The fuselage ends in a flat surface where the stage separator for the delivery rocket used to be, and the KAS docking ports are back there. It seats 2 Kerbals in lawn chairs on the nose.

Edited August 19, 2013 by Geschosskopf

[Geschosskopf]

Whoops, I spoke too soon here. D'OH is right :/

Yup, it flies on Duna just fine, but I can't get the damn thing off the ground at Kerbin yet. I thought I'd already done that, but it must have been an earlier version. Anyway, it's got so much lift that when put atop a rocket, it either pulls the whole rocket over or breaks off. So, I'm pursuing 2 alternative methods. 1 is to build a big winged jet/rocket stage for a horizontal take-off, a sort of quasi-SSTO. As usual with my SSTOs, this ALMOST works but not enough rocket delta-V left to circularize at the top. I'm thinking I can make this work eventually but it's such a pain, just like SSTOs. The other method is to add some detachable downforce plates under the D'OH for a vertical launch. This is showing promise, in that I've got it narrowed down to a range of plate size and angle. Too little and the same problem happens as before. Too much and the thing goes the other way. Maybe some linear RCS up there, too, although I want to avoid that because it would be a permanent addition.

But anyway, until I can meet the requirements myself, this is on hold. However, if you figure out both ends of the problem (and both ends appear about equally challenging), feel free to make me look like and even bigger idiot

[Bothersome]

Yeah, I've done this quite a while ago... Worked on several designs to fly around on Duna for exploration.

All these planes were brought to Duna by normal methods. No cheats were done and all craft are 100% stock. More pictures where those came from but I figure that's enough for now.

[Zeroignite]

Ok, you just gave me a reason to set up permanent infrastructure on Duna

[Geschosskopf]

Yeah, I've done this quite a while ago... Worked on several designs to fly around on Duna for exploration.

(Pics snipped)

All these planes were brought to Duna by normal methods. No cheats were done and all craft are 100% stock. More pictures where those came from but I figure that's enough for now.

Those are very elegant, realistic looking planes there. Bravo . Nice low part counts, too. The D'OH has like 150 parts (about 50 of which are solar panels--wish they had procedural ones). It's fine when it's off by itself but things get a bit laggy when it's near my Duna base. But what powered them? RCS? Or am I missing something? Anyway, I suppose if they can SSTO like that they can circumnavigate Duna, but you'll have to add some KAS ports to refuel them on the ground and strap a lawnchair on them so they can carry the minimum of 2 Kerbals.

Anyway, you've given me hope. I was up till 0300 this morning struggling with various ways of launching the D'OH. I might have to try launching the D'OH sideways or backwards like that, although as this whole this is about the aerodynamics, I can't help but feel such non-aerodynamic launches go against the spirit of it all. But OTOH, the real point is to overcome the aerodynamic issues of Duna's thin, noncombustible air; the realization that Kerbin's thick air could cause a problem on the other end was an unexpected surprise, in that I thought I'd overcome it long ago. And I did say it didn't matter how you got it to Duna, provided it started on the ground and flew there.

I didn't, however, say that the plane had to leave the ground in 1 piece, and I did say to "do what you gotta do" with mods. So technically, if you used Orbital Construction to replace the aerodynamic problem of Kerbin with the tedium of building the orbital shipyard and sending up all the needed materials, that would be OK, too. The Kerbin issue wasn't originally part of the idea anyway.

So, what do you think? Add some bonus points for launching the thing in 1 piece and nose-first from Kerbin, and subtract points for using Orbital Construction?

[zarakon]

Very nice, Bothersome

Do those planes have any propulsion, or are they all just gliders?

How did you handle launching them from Kerbin, since they don't appear to be friendly to sticking on the top of a rocket?

Here are some pics of an LV-N powered all-stock plane I sent to Duna a while ago. Doesn't meet your requirements for the challenge though. It's also big and a bit ugly, but it worked well!

Edited August 20, 2013 by zarakon

[Bothersome]

Thanks, they are all gliders. Getting them to orbit was a simple matter of strapping on some rockets in just the right alignments to keep the wings from flipping it on the way up.

If you want to see one I'll set one on the runway so you can have peek.

OK I loaded up Glider F into the SPH because it was dark outside, figured you could see more inside with more lighting.

It was hard to find a camera angle that showed the important parts. Basically all the planes were launched with similar methods.

Edited August 20, 2013 by Bothersome
Added some info

[Zeroignite]

This graph I made a while back might be useful:

5km on Duna is the same atmospheric pressure as about 17km on Kerbin. Datum altitude is equivalent to ~6.5km on Kerbin.

[Bothersome]

That's an interesting graph but there is another component missing. Gravity.

What we need to know is how high a plane will fly on Duna, if it will only fly X high on Kerbin. For example, if my plane will only fly to 15000 on Kerbin, how high will the same plan fly on Duna? Gravity is less so there is less pulling it down. And yeah, I'm asking cause I have NO idea how to figure that out.

[Geschosskopf]

Here are some pics of an LV-N powered all-stock plane I sent to Duna a while ago. Doesn't meet your requirements for the challenge though. It's also big and a bit ugly, but it worked well!

That's an amazing plane there! Think you could tweak it a bit to fit it in or would you have to start over from scratch?

I'm totally impressed that you were able to launch it, too. That long, skinny neck between it and the lifter looks like a great place to RUD.

Thanks, they are all gliders. Getting them to orbit was a simple matter of strapping on some rockets in just the right alignments to keep the wings from flipping it on the way up.

Impressive engineering! Getting something with high lift off the ground is a real pain .

That plane appears to have RCS tanks and the order of your previous batch of pics appeared to show it taking off and docking with a ship in orbit like an RCS-powered SSTO. Or was the order of the pics reversed? Anyway, that inspired me to try an RCS-powered plane for Duna but it didn't work. Not enough thrust to do the job on a tank of gas.

What we need to know is how high a plane will fly on Duna, if it will only fly X high on Kerbin. For example, if my plane will only fly to 15000 on Kerbin, how high will the same plan fly on Duna? Gravity is less so there is less pulling it down. And yeah, I'm asking cause I have NO idea how to figure that out.

I have no idea, either, but in my experience so far, it appears that a rule of thumb that works OK is to get as high as you can on Kerbin, add 1-2km, and then use the chart to convert the total into Duna altitude.

Still, the easiest way is to actually test it on Duna. Build your plane and stick a small de-orbit booster on it, then use HyperEdit to put it into your desired orbit at Duna (say about 45-50km). Then use the booster to start you down the desire path, ditch the booster, and fly it from there. That way, you can learn how well it enters the atmosphere and how to slow it down for a safe landing. And if you get past that, then you can see if you can take off again and how well you can fly. Think of this as the guys back at KSC running a simulator. You do this in a separate game from your real stuff, then copy over the craft files when you're happy with them.

Just be careful doing this. If you clutter up Duna's orbit with a bunch of different test planes that you HyperEdited there, you might forget which ones you launch-tested and which ones you just HyperEdited straight from KSC, so you'll end up like me with something that flies great on Duna but is extremely difficult, if not impossible, to launch from Kerbin . Although I did manage to get it off the ground after several launches using the Orbital Construction mod. But seeing your fancy tinkering above, that makes me feel like I weaseled out of part of the challenge, even if the Kerbin side wasn't the original intent.

In any case, one of the major issues with flying on Duna is getting slow enough to land safely, especially if you want to land within walking distance of a chosen spot (your base, an anomaly, etc.). Because the air is so thin, any given wing has to be moving relatively fast to generate enough lift to fly. But there doesn't seem to be anything resembling level ground on Duna these days. There are very occasional spots that are only slightly sloped but they tend to be only about 100m wide, all surrounded by rather big, rolling sand dunes of variable height and the valleys between them which make up the bulk of the terrain. This puts a premium on managing to fly slow in thin air, which apart from using VTOL-type thrusters means you have to have a really big wing on as light a plane as you can build, which makes the whole thing big and difficult to launch.

Another thing about this thin air is that even if your plane can fly slow in it, the air won't slow it down very quickly, and the low gravity reduces the effectiveness of brakes. So, even if your plane can land at about 30m/s with the brakes already locked when it touches, you'll still roll and bounce about 500m, and it goes up rapidly for higher landing speeds. This will take you up and down at least a couple of dunes. If you're plane's only about 30m/s and built tough, it can usually survive this treatment OK, but higher speeds can easily make you faceplant in the side of the dune ahead, plus you end up miles from your landing target. So, there seems to be a need to either be able to fly very slow (for Duna) and/or have retro rockets, drag chutes, etc.

Here's the D'OH in one of its test flights on Duna. I built it as part of this Kethane-based long-term mission I'm doing so, so I copied that saved game over to another one so I could build the thing to fit what I've already got on the ground, in this case a Kethane refueling station at 2500m. In this case, the D'OH arrived via HyperEdit as described above.

If you can't read the fine print, I'm about 17km from my base about to start my descent. I've found I have to chop throttle and pop the 16 air brakes at about 5km out to have much hope of stopping anywhere close. But the plane can taxi up and down dunes OK so I'm happy ending up within about 1-2km .

And here's one of my failed attempts to get it off the ground. Ironic, the very big, slow, lumpy thing moving this fast .

This was a 3STO configuration. On the back is a stage with the huge rear wings and the jet engines. The middle stage is rockets and a docking port so the thing could link up with an interplanetary tug once in orbit. But the rear wings are necessary all the way through the atmosphere, so what happens when the jets finally die is I drag the rockets down into their stage so the rockets can fire while the wings are still attached. Then ditch the wings and jets once above 70km to expose the docking port and give the rockets less mass to push.

As with most of my spaceplanes, however, there just never seems to be quite enough rocket fuel. I could get it out of the air but it was always sub-orbital after many attempts. GEEZ, I hate spaceplanes. Not only are they very finicky like this, but each attempt takes 15-30 minutes of real time, so it's a very long, slow process getting them to work. I figure I'm close enough, however, that I could probably do it in another couple of days, but I'd really rather not. Orbital Construction. Yay!

[Bothersome]

Geschosskopf, it's nice to see some one take a strong interest in planes on Duna. I think most of the engineering that is needed to get one there is overlooked by most people. It takes a bit of study to get all the pieces in the right places because, in my case, I've never hyper-edited anything anywhere. I have never even opened up a debug interface or edited a single file. So that meant that I needed to think ahead on what was going to be needed. Of course I didn't get it just right most times. That's why you see 4 different versions of about 8 that I made on Duna. Glider F was the most successful because it had the largest wing surface area for the weight of the craft. The RCS tanks were two purpose intended. One was of course to allow me some maneuvering to dock onto the transport ship in Kerbin orbit. The other was as a ballast tank. You notice two tanks on Glider F, what I would do is empty one of the tanks so I could move the monoprop fuel fore and aft to obtain a certain balance.

I had done some extensive research on what makes an efficient infini-glide plane, and balance had a play in that. To fly fast, balance needed to be forward, and to allow slow flying for landing (which is the hardest part for these gliders) I needed to move the weight aft. That was the main purpose for the RCS tanks.

You are absolutely right about the thin air on Duna making it quite a challenge to slow down. That's the purpose of the parachutes. Get over your target (no higher than 600 meters above it) and pop the parachute, keep the nose up and she comes down just right.

The model with 4 rover wheels is actually one of my better gliders (I forget the model off the top of my head) because it acts more of a rover that flies. It has enough wing surface to just get over the mountains on Duna and still fly slow enough to take-off from 4000 meter altitude on Duna. Have to stay under 60 to keep the tires from popping. Yet it could rover around and climb like a rover if needed. It's hard to see in the pictures but Glider F has two "wing seats" for extra passengers. One of the other models has an extra outside seat too.

A tip you might could use on getting one to space is to put the rocket engines about the center of the plane's COG. Let the wings act as rocket fins to help stabalize the whole structure as it goes up. Once it's out of the atmosphere it no longer matters on the wings but CG of the rocket/plane system needs to be pretty close. In my case, the wings already had a 5 degree incline to assist in level flight as a plane. So at lift-off I would put the nose at about 5 to 10 degrees toward east because the wings were trying to assist in lifting the craft the other direction just a bit. I couldn't go over too far until the air got thin because if you try to turn over too soon, the wings would "hold up the ass" and the nose would go over too soon. So basically the flight path up was not going to be optimal for fuel use. Basically it goes 10 degrees all the way to 20k then prograde to 40k. Something like that, it's been a while since I've launched one.

I hadn't offered the craft file, but if you'd like a stab at it, I'll put on my company thread if you request it. The reason I hadn't offered it before is because none of those planes in the photos I've posted out perform the 6SKK from my company thread. But, the 6SKK doesn't do Duna well from my own estimation. I haven't tried it there yet but from flight characteristics of its performance on Kerbin and Laythe, I would think it doesn't have enough wing surface for Duna. But a modified version could be easily made for Duna though. And if I were to want another plane on Duna, that's what I would do, modify the 6SKK for a Duna version.

Like I said earlier, it's nice to see some one taking an interest in Duna planes. Keep up the good work.

[Geschosskopf]

Geschosskopf, it's nice to see some one take a strong interest in planes on Duna.

Well, if there's a way to fly around on a world instead of having to drive everywhere, I'm all for it. I find rovers boring, slow, and likely to wreck at any moment unless you pay very strict, constant attention to them. Besides, one of my favorite books as a kid was Clarke's Rendezvous with Rama, in which this guy made the statement that if you could fly on Mars, you could fly anywhere. And that's true in KSP, too. Can't use air-breathing engines and there's barely enough air to fly.

I think most of the engineering that is needed to get one there is overlooked by most people. It takes a bit of study to get all the pieces in the right places because, in my case, I've never hyper-edited anything anywhere. I have never even opened up a debug interface or edited a single file. So that meant that I needed to think ahead on what was going to be needed. Of course I didn't get it just right most times. That's why you see 4 different versions of about 8 that I made on Duna.

I don't have the patience for traditional KSP trial-and-error (unless I'm just going to Mun or Minmus, where I don't have to wait months for launch windows). Besides, I can't imagine even a Kerbal space program being run without all sorts of calculation and simulations to develop a reasonable expectation of success before spending the money to do something. Pretty much everybody uses F5/F9 to take mulligans on execution so I see HyperEdit as a way to take mulligans on design, a pre-launch simulation to work the bugs out before you really try it.

As a side note, I really wonder how career mode will deal with this. Pricetags don't mix well with the spectacular failures of KSP trial-and-error engineering....

Glider F was the most successful because it had the largest wing surface area for the weight of the craft. The RCS tanks were two purpose intended. One was of course to allow me some maneuvering to dock onto the transport ship in Kerbin orbit. The other was as a ballast tank. You notice two tanks on Glider F, what I would do is empty one of the tanks so I could move the monoprop fuel fore and aft to obtain a certain balance.

I had done some extensive research on what makes an efficient infini-glide plane, and balance had a play in that. To fly fast, balance needed to be forward, and to allow slow flying for landing (which is the hardest part for these gliders) I needed to move the weight aft. That was the main purpose for the RCS tanks.

(Light suddenly dawns) Fascinating. I never think about infini-gliders because in some future update they'll suddenly stop working, so it hadn't even occurred to me that's what you were doing here. Quite clever design, the 1st I've heard of that can actually land. That does kinda side-step a lot of the intended challenge, though, in that you don't have to solve the problem of lugging around a propulsion system. Which is another reason the D'OH is so big.

You are absolutely right about the thin air on Duna making it quite a challenge to slow down. That's the purpose of the parachutes. Get over your target (no higher than 600 meters above it) and pop the parachute, keep the nose up and she comes down just right.

The D'OH weighs 15 tons without Kethane so can't carry enough parachutes to land that way without becoming too heavy to fly. I tried a drag chute to shorten the landing roll but that just tended to make it land nose-first and then disappeared upon touchdown without providing any real braking. So, I made the plane strong enough to survive rough landings, arranged the landing gear so it won't nose over, and made sure it could taxi up steep sand dunes to get back to the target area. Brute force and ignorance .

A tip you might could use on getting one to space is to put the rocket engines about the center of the plane's COG. Let the wings act as rocket fins to help stabalize the whole structure as it goes up.

Brilliant! I'll give something like that a try. It also occurs to me that putting 2 D'OHs facing opposite directions on the same rocket might balance each other out, plus give me some redundancy. Too bad doing that with KSP's editors is such a pain.

[Geschosskopf]

So, I wanted to try Bothersome's suggestions on how to vertically launch the D'OH. I was going to make a sort of "conjoined twin" of the D'OH on the other side of the rocket, so their wings would hopefully balance out, and to do this I thought the symmetry of the SPH would work better than the VAB, so I went there. And then I decided, WTF, I'll work on the spaceplane version instead.

I totally scrapped the entire ascent portion I'd been using before and started over from scratch. This time I decided to give the SABRE engines (for them as don't know, you can switch them from jet to rocket) from B9 a try in hopes of saving weight. Then I was going to build the usual "conjoined D'OH twin" around that, as shown in the fail pic above, but then I saw the B9 "heavy wing" sitting there. I'd never used this before, mostly because the 0.20 SPH was too small to see them, but with the larger facility now, that's no problem.

Eventually, after some major tinkering, I came up with this:

All 32.1 tons of it, 15.2 being the payload D'OH. Not a bad ratio if I do say so myself . If you look close, you can see the big wings are attached to radial separators mounted on the small I-beams. The whole engine/intake/fuel tank assembly between the big wings is attached to the tail of the D'OH with an inline separator, and there's a Clamp-O-Tron Sr. centered on the rear for the interplanetary tug to attach to. Despite all the separators, this is an SSTO. The plan: to get safely in orbit then ditch the big wings, then send up one of my standard interplanetary tugs to dock on and push it to Duna. At Duna, the tug puts the D'OH in a 50km orbit then detaches and goes on to other things, leaving the SABREs in the remains of the SSTO assembly with just enough delta-V to de-orbit. So do that burn, dump the SABREs and fuel tanks, and the plain old D'OH is left to enter Duna's atmosphere, land, and do its thing, which I already know it can do.

Much to my continuing disbelief and amazement, this all worked. I have struggled NO END with spaceplanes and SSTOs but this one not only worked the 1st attempt, it was EXTREMELY well-behaved all the way up, mostly flying itself with zero input from me, whereas all my others have always wanted to wander off course and have actively fought my corrective measures. Hell, this one even flew itself off the runway and settled into a nice climb all by itself (one of the benefits of being a tail-dragger, I suppose). Of course, being so big and heavy, it was in no hurry to pick up speed, although it climbed like a champ at very low pitch angles due to all that wing. As you can see, though, by 15.5 minutes in, I was finally at 24km doing 900m/s

Also note I'm not climbing very fast, trying to maximize speed on jets before running out of intake air. I eventually got to about 1200m/s at 26km, at which point the right-hand SABRE flamed out the thing went into a pirouette, just yawing around (too much wing, and too much moment of inertia in roll, to do anything else, I guess). Never having used SABREs, part of the purpose of this flight was to determine where they flamed out, so I could work around that in subsequent attempts. Still, I pressed on, hitting action groups to switch the SABREs to rocket mode, closing the intakes, and stomping the rudder pedals to stop the rotation to the right. After about 2 revolutions I had it back under control, both climbing and accelerating under rocket power. Nothing broke off in all this but I'm not surprised; I build my ships tough because the pilot's usually been drinking .

Anyway, switching quickly to the map view, keeping 1 eye on the resources tab and the other on my apoapsis, I cut power when I saw I'd reach 77km and still had what looked like enough bars left to circularize. So I set up an Ap node while still at about 50km which, by the time I got to it, was off-target a bit. My 77km Ap ended up being my 72km Pe and the burn gave me an 80km Ap on the other side of Kerbin. GOOD E-(explicative deleted)-NUFF! So I ditched the big wings after this burn:

And I had just enough dV to circularize at 80km with 113m/s to spare. Yay me! Next time, I'll switch to rocket mode at 25.5km so I won't waste fuel stopping the off-center flameout gyrations .

Then it was just launching my standard interplanetary tug with my standard 30-ton lifter and have it dock with the D'OH which, not having any RCS, made the tug do all the work. But it's built for that and more. So I ended up with this:

4600m/s dV is enough to go almost anywhere so I figure I've got the problem solved, given the D'OH can handle everything on the Duna end.

Getting this cow into orbit was at least as hard as making it fly on Duna. All in all, as Sherlock Holmes used to say, it was a "two-pipe problem". Because I don't smoke opium or shoot up cocaine like he did, it was for me about 10 strong ales and half a gallon of bourbon .

[Bothersome]

Congratulations on a job well done. Now, when is Duna gonna be in position?

[Geschosskopf]

Congratulations on a job well done. Now, when is Duna gonna be in position?

(Pushes aside empty bottles from celebratory feast). Um, er, what? Oh, let me check. Let's see, in the parallel universe of the Kethane Travelling Circus, it's Y1D135. According to Alexmoon's web site, the 2nd Wave won't be able to leave until Y1D290. That day will see the depature of the D'OH, its support bowser, and some Kerbals to pump gas and fix flats. All this will attempt to land within at least a few Kms of the kethane drill/habitat module already down there. Now that I've got all the components working, most of this should happen very quickly once I have time to play again, which won't be until this coming weekend.