Flying on Duna in 1.0.4 -- Some General Guidelines

[Geschosskopf]

INTRO: Why Fly on Duna?

Because airplanes are way faster and (mostly) safer than rovers. Thus, if you can get an airplane to Duna, you can see a lot more of it in a lot less time than if you drive. More biomes, more science, visiting all the anomalies, etc. Plus, if you have multiple bases on the ground, airplanes are the best way to move stuff between them.

The aero model in 1.0.4 makes flying on Duna very easy. In fact, everything involving Duna's atmosphere (aercapture, reentry, landing a rocket on a specific spot, using parachutes, taking off again, etc.) is way easier than it was pre-1.0. As such, there is no good reason NOT to fly on Duna.

But flying on Duna is rather different that flying on Kerbin, so I've put this guide together to provide a starting point for folks who've never designed a Duna plane before. Use at your own risk and be advised that 1.1 will almost certainly change things.

I. General Advice

Compared to Kerbin, Duna has considerably less gravity and thinner, oxygen-free air. Strangely, these factors come together in 1.0.4 to make a nearly ideal flight environment, provided you can find a propulsion system than doesn't need free-range oxygen. However, a plane that will fly superbly on Duna will likely be a complete dog on Kerbin. Therefore, I strongly recommend setting up a separate sandbox game with HyperEdit installed so you can do your testfights on Duna---you won't get much useful info flying the plane on Kerbin. So, you need:

* A propulsion system that doesn't need atmospheric oxygen (more about this later)

* HyperEdit

* A sandbox testing lab

II. Basic Flight Parameters

LIFT: Because Duna's air is much thinner than Kerbin's, you need lots more lift on Duna to fly the same mass. Somewhere in the region of 1.4 lift per ton of ship works very well as it gives a stall speed below 50m/s even at relatively high terrain elevations. This is EXTREMELY important for getting the plane stopped on landing before some lump in the terrain kills it (more on this below). You can go with less wing and rely on rockets or chutes for landing, but then you also need lots of horsepower to take off again before encountering a fatal bump in the terrain.

DRAG: Most of the drag of a plane comes from the lift of its wings, but this is very much a function of air density. The way things work at present, a plane with 1.4 lift/ton will have a very low top speed on Kerbin due to gobs of drag, but on Duna there's hardly any drag at all so the plane can reach hypersonicc speeds. Also, because there's hardly any drag on Duna, airbrakes aren't very effective at all. You need way more airbrakes than you do on Kerbin.

THRUST: Because there's so little drag on Duna, and because the gravity is so much lower, you need way less thrust on Duna to achieve insane performance. A ratio of just 4Kn/ton is sufficient on Duna to climb steeply while still accelerating, reach any altitude desired, and exceed orbital velocity in atmospheric flight. But on Kerbin, the plane will hardly fly at all with this little thrust.

CONTROL AUTHORITY: This is largely a function of both control surface lift values and air density. Thus, for crisp, responsive controls, a Duna plane needs WAY more control surface area, with as much leverage as possible, than a Kerbin plane. But even so, the controls will lose a lot of their effect at speeds above about 200m/s, even at lower altitudes.

WHEEL BRAKES: When you combine high amounts of lift, low stall speeds, and low gravity, you lose most of your braking power. Brakes only work when the wheel is on the ground and low gravity plus high lift make the plane float long distances with the wheels just above the ground. Thus, retro thrust is very desirable to keep stopping distances short. But OTOH, because brakes have little effect at all, they're not likely to flip you over, either, unless you get really sideways due to a bounce.

III. Common Problem Areas

PROPULSION:

With stock, about the only viable option is ion engines powered by 6-Pack fuel cells. Because Duna's air is so thin, ion engines have reasonably good performance in its atmosphere, enough to keep the plane flying anyway. However, they lack the oomph to take off in reasonable distances so pretty much require assistance by small rockets to get off the ground.

This leaves mods. There are several options there. Both KAX and Firespitter have electric props which work quite well. The large amount of electricity they require, however, means you should probably go with fuel cells to power them, given the weak performance of solar panels on Duna. The plane below, powered by a single Firespitter electric prop, can achieve suborbital trajectories. It has an FL-T400 and 2 6-Pack fuel cells in the cargo bay, which is enough juice to circumnavigate Duna several times. Note the large wing area. When the pic was taken, the plane was doing about 900m/s at 22km.

Another option is nuclear thermal jet engines. Both Atomic Age and Mk2 Extensions have some of these. Their virtue is that they work in any atmosphere and don't need fuel tanks, although some of them need more nuclear fuel eventually. However, they're also rather heavy, which means lots more wing needed, or accepting dangerously high landing speeds. Personally, I haven't yet used these on Duna but they work quite well on Eve.

Finally, Kethane (if that's still a thing) and Karbonite have jet and propjet engines that burn the respective resource and work in any atmosphere. These are typically lighter than nuclear thermal engines but make up for this by consuming fuel in larger quantities than electric motors powered by fuel cells, so weight is still something of an issue.

Thus, you have tradeoffs no matter which way you go. Personally, I prefer the electric prop route as the lightest overall while still being plenty powerful, and also because (at least with Firespitter), you can reverse the thrust so don't need a separate retro thrust system (see below).

TAKEOFFS and LANDINGS

The whole point of having a plane on Duna is to go everywhere. Problem is, except in the basins and canyons, Duna has nothing resembling the KSC runway. It's all dunes, hills, and mountains, meaning all potential LZs are pretty short and rather bumpy, just small flat areas scattered between obstacles. To be able to operate safely, therefore, short-field performance and rugged construction are essential. Landings are typically more difficult than takeoffs because of the difficulty of slowing down (see below), whereas you can usually use a hill as a jump ramp on takeoff. The main problem with takeoffs is picking a good path so you don't drop off a hill before you've got flying speed.

There are several ways to provide short-field performance: VTOL, STOL, and STOVL. VTOL can be done with an electric helicopter instead of a plane, but othewise will need rockets or Kethane/Karbonite lift jets. STOVL usually means parachutes, which these days means the crew must include an Engineer to repack them. The least-complicated option is STOL, but it puts a high premium on your flying skills and giving the plane enough control authority to steer it to the desire LZ. I prefer STOL myself as I suck at VTOLs and still don't trust repacked chutes. And by STOL, I mean stall speeds below 50m/s. You'll still roll a long way after touching down unless you've got some good way to slow down, however (see below).

But except in a well-flown VTOL, both the landing and takeoff are probably going to be rough due to the bumpy terrain. Therefore, your plane must be strong enough to survive this. You also want high ground clearance, small safety wheels under the wintips to keep them from breaking off when striking the ground, and a general lack of flex in the structure. Kerbal Joint Reinforcement highly recommended.

SLOWING DOWN and STOPPING:

Because of the lack of drag, a Duna plane can build up lots of speed relatively quickly. However, this also means that airbrakes aren't very effective. A single pair of stock airbrakes will slow you down about 1m/s per 2-3 seconds of elapsed time at "low" altitudes (4km and below), and do rather less well at higher altitudes. And you have to kill off most of your horizontal speed to land safely, no matter what method you use.

Therefore, unless you want to carry a huge number of airbrakes, some form of retro trust is virtually essential. Firespitter electric props are great because they allow thrust reversal, so you can use the same power plant for 2 purposes. If you go with ion engines, you'll need some small, forward-facing rocket engines. AFAIK, the other mod engines lack thrust reversal, too.

The problem of slowing down doesn't stop once you're on the ground, either. Unless you can stop quickly, a STOL plane will bounce and float along forever, and will frequently meet a fatal terrain obstacle or cartwheel the plane. Wheel brakes, as mentioned above, aren't very effective at all. Therefore, if you touch down with even 30m/s horizontal speed, you still need reverse thrust.

LAUNCHING FROM KERBIN

OK, you've designed and tested a plane that flies to your satisfaction on Duna. Now, how do you get it off the ground on Kerbin? This is one of the more interesting challenges in the whole program. The basic options are either either to fly it off the runway or launch it vertically, and both have their problems (or you can use Extraplanetary Launchpads or UKS to build the thing at Duna, but what's the fun of that?).

Runway takeoffs are desirable, especially if the plane is a Duna SSTO. It's actually extremely easy to make a Duna SSTO spaceplane. However, as mentioned above, Duna planes typically have so much drag from lift, and so little thrust, and so little fuel, that they're utterly incapable to SSTOing on Kerbin, assuming they can even fly there at all. So if you want to go that way, expect to need some huge takeoff stage (or reuseable spaceplane) that actually carries your Duna plane to LKO.

The above pic dates from 0.21 but you get the idea. The actual Duna plane is the small biplane in front, and the part with the big wings is the SSTO lifter stage for it. Then most of the lifter stage went away and I docked a transfer tug to it once in orbit.

If that's not your thing, then you're stuck with using a rocket. But the huge wing of anything capable of STOL/STOVL flight on Duna will do its best to flip the rocket out. If the plane has a small enough wingspan, you can maybe surround it with a fairing, but big fairings present their own problems on launch. Besides, if it's that small a plane, you might be able to "Dreamchaser" it with sufficiently huge tail feathers on the rocket.

A less orthodox, but totally workable, alternative is to use the airplane's wings as the rocket's tailfins. IOW, attach the plane(s) at the bottom of the rocket. The most effective way to do this is to send out 2 identical planes, one on each side of the central rocket, so that the lift of their wings cancel out with no net force (except lots of drag) on the rocket.

NOTE: It's a good idea to disable the plane(s)' control surfaces during launch and any aerocapture passes at Duna. Also, the rocket will need gobs of control athority from torque, RCS, and control surfaces.

The 2 planes above were going to Laythe but the concept is the same for a pair of Duna planes.

If, however, your plane is rather large, you can just lift 1 at a time and deal with the asymmetrical lift during launch. Such as this actual Duna plane:

In this case, the top of the plane faced south on the launchpad so pulled the rocket a few degrees that way during ascent. I had intended to rotate the whole rocket to keep this from happening but forgot. So, about halfway up, I simply rolled the whole thing 180^ so the lift pulled the rocket back to the north. The end result was an inclination of only 2 degrees, close enough.

IV. Final Thoughts

Don't sweat aerocapturing your Duna plane. Duna's atmosphere is completely harmless and won't throw things off course. And then during the initial landing from space, you won't see any flames. Just be sure you're well-versed and on your landing proceedure (such as how and when to use retro thrust), and have practiced it enough in your test game to feel confident. Once you're down that 1st time, pat yourself on the back and go conquer Duna .

Edited August 19, 2015 by Geschosskopf

[Geschosskopf]

First off, thanks to those who repped me for this. Now I'll try to answer the questions in the rep comments. It would be easier to do this if you asked the questions in the thread, though .

@ B-STRK:

I did some very limited testing in 1.0.2 (see 2nd to last page of the old "Flying Duna AGAIN" challenge thread) with what turned out to be the prototype of my current 1.0.4 DORK. As you can see, things were pretty much the same. In fact, so similar that when I decided to do a Duna airplane in 1.0.4, all I had to do was refine the plane's design slightly. It actually had too much control authority to begin with.

But seriously, why stick with 1.0.2? That version's heat model was totally bogus and unrealistic, not to mention very bug-ridden, and much more prone to spontaneous combustion than what we have in 1.0.4.

@ Cirocco:

Yup, if you want to use chutes, they should be balanced around the CoM so the plane will hit the ground more or less level and on its landing gear. If the chutes are more on 1 side of the CoM than the other, the plane will land at a steep angle and might break off part of the nose or tail. However, this also depends on the terrain where you land. It's quite possible the chutes will make you come down on the side of a steep slope, in which case a level plane might break a wingtip (or nose or tail) on the hillside before the wheels hit.

Also, chutes should be viewed as the final step in landing, a substitute for a conventional, controlled landing. Because Duna is so lumpy, you have to spot and maneuver over a good landing zone. You have no control once you pop the chutes and the higher and faster you are when you pop them, the less likely you are to hit the spot you're aiming for. Besides, IIRC it's "risky" (per the right-click menu) to open chutes between about 350-500m/s and Duna planes can easily exceed such speeds. Therefore, even with chutes, you still need some other means of killing off most of your horizontal velocity, preferably to just above stall speed. So what you do is, while still at normal flying speed, pick out a good spot and line up on it, then glide down towards it while slowing down a lot. As you pass slowly just over the target, pop the chutes.

----------------------------

On a related note, it's a VERY good idea if your plane has some powered wheels for taxiing. There are powered landing gear available in several mods, or you can use ModuleManager to add power to stock wheels. It's quite likely that, due to terrain issues, you'll have to land a few clicks away from your real target. It's a LOT easier to drive the remaining distance in rover mode than trying to taxi there using airplane engines. Besides, powered wheels have a reverse gear which is quite useful for maneuvering in tight spaces. And even if all your destinations are on relatively smooth ground, you'll probably have to turn around to take off again once you've landed there.

Edited August 19, 2015 by Geschosskopf

[Moonk]

Great thread, thank you

[Geschosskopf]

You're welcome. Post up your Duna planes when you get them working.

[Moonk]

You're welcome. Post up your Duna planes when you get them working.

With pleasure ))

I use 8 Spark Engines, she can easely went to orbit after landing. But landing.... it is really hard, going to add parachutes tomorrow.

Edited August 19, 2015 by Moonk

[Geschosskopf]

I use 8 Spark Engines, she can easely went to orbit after landing. But landing.... it is really hard, going to add parachutes tomorrow.

Nice functional plane. I told ya landing was the hard part . What was the lowest speed you were able to land at?

[Moonk]

About 40-60 m/s and something about ground i used 2 Drague Chutes and when touched ground 2 Mk-2R Chutes - was easy landing less then 10 miter to the target )

Screens befor launching to Kerbin orbit, it was my 1st ship from Runway to Orbit. Probably there are to much air intakes.

Edited August 20, 2015 by Moonk

[Hagen von Tronje]

I had a plane built and wanted to send it to Duna for testing before reading this, to see how well I could design something for a planet I've never visted. Now that I've read your guide, I see my guesses were not too far off the mark!

Javascript is disabled. View full album

I sent it glued to the side of a big rocket with a radial decoupler (the ugly bits after decoupling are removed by save/reload) then flew it in like a spaceplane. It uses two Firespitter electric props and performs like a stuntplane on Kerbin, but is a bit floaty on Duna. However, this was anticipated - I tested by flying into the upper limits of Kerbin's navigable atmosphere on the assumption that Duna's thin atmosphere would not be terribly unlike those conditions.

Flying is different, feels almost like a hybrid between atmospheric flight and orbit, because you can pretty well flip around and do straight retrothrust maneuvers and actually get results, especially at higher altitudes.

I should also mention that I find the electric props get hot, especially as atmosphere thins, which is why I used Randazzo's fantastic active heat management system, giving me actively pumping radiators that won't break off in atmospheric flight or create needless drag. You can really get going fast though, Duna's thin atmosphere lets you go supersonic on props!

But landing is definitely the real challenge IMO. Dunatian terrain is very unkind to planes, the low gravity means less downforce for the wheels to engage the surface (probably a major reason why wheel brakes feel so worthless), and just like you say, low drag means lengthy stops. My design packs a parachute in the rear shroud (not shown deployed because this was sent unmanned just in case it was a total failure, it can now fly to wherever my kerbals land and pick up a pilot), but reverse thrust on the props seems to work well, although using that to stop with precision is not easy. I'll probably go with the chute once I have kerbals to pack it.

Still getting the hang of it, though. Really great guide for an environment that is hard to test for!

[Geschosskopf]

Still getting the hang of it, though. Really great guide for an environment that is hard to test for!

Thanks . And that's a nice-looking plane. Glad you've got something that works and good luck in your future adventures with it.

[Edax]

 

My advise?  Aim for the lowerest possible point on Duna, as that'll give you the most atmosphere to work with.  You'll want to glide with almost no vertical speed during the last few seconds, and you should aim for a slight upwards incline to minimize bounce.

 

 

 

 

 

Edited January 30, 2016 by Edax

[MacLuky]

Thanks for this thread, it still holds many truths in 1.3.1 and led to the sparkle: