Early High Altitude aircraft

[se5a]

So I've been trying to do those "above 17km" contracts, and strugling to come up with a stable aircraft that'll get me up there.

my current design has two wheeslies and a reliant. the wheeslies get me to the position and to a reasonable altitude and the reliant gets me up that final altitude.

problem is I'm having problems with stability after I've burned most my fuel, I suspect the CoM changes enough to screw me over.

What designs have you guys used?

I'm also using a few mods, such as procedural fuel tanks, which I'm filling with oxygen, since I've got enough fuel. 

https://imgur.com/a/PylQTVG

[mk1980]

to be honest, i usually skip those contracts altogether :) anything involving science from kerbin, really

doing them with a capsule (if it's about crew reports) or a simple probe (other experiments) in a polar orbit is probably the least effort. takes some patience, though. may take a few roations until the capsule crosses over the target area at the right moment.

 

as for a plane based solution - i'd either wait for panther engines. they operate just fine up to 12+ km or something and have enough thrust with afterburners to do a steep climb to slingshot above the required ~18km before falling back down.

for a wheesley + rocket solution, i'd go with very little liquid fuel (wheesleys don't need much). don't know how much fuel you have in that plane, but the rocket fuel should easily be enough to push that bird up a few kilometers multiple times. unless you made the plane too heavy.

might also be worthwhile to try a terrier instead of reliant. wheesleys should get you high enough for the terrier to work efficiently, so you'd save some mass. a few sparks would also do the trick if you already got those.

 

i'd probably put the wings a bit further in the back. the engines should be roughly the same weight as the cockpit (engines probably a bit heavier) so if the wings are roughly in the middle of the plane, the dry and wet COM should be reasonably close to each other.

 

[bewing]

I like something like this:

Very low tech. It can hop 3 times to 18km+ in a single flight.

You are right that CoM shifts are a big problem on this kind of a flight. I find it useful to have a forward fuel tank and lock it, so that it gets used last. That keeps your CoM as far forward as long as possible -- but on a flight that comes close to running on fumes, the landing can still be very dicey.

 

[Freds]

28 minutes ago, bewing said:

I find it useful to have a forward fuel tank and lock it, so that it gets used last.

Another way is to change fuel priorities Instead of tank locking - this is better because fuel just get consumed in the right order, so there is no need manually unlock the tank while in flight and also eliminates the risk of a flameout due to fuel starvation while having fuel onboard.

 

Edited July 24, 2018 by Freds

[Snark]

4 hours ago, se5a said:

So I've been trying to do those "above 17km" contracts, and strugling to come up with a stable aircraft that'll get me up there.

my current design has two wheeslies and a reliant. the wheeslies get me to the position and to a reasonable altitude and the reliant gets me up that final altitude.

My solution is pretty much like @bewing's:

1 hour ago, bewing said:

I like something like this

^ This.  A simple, small Juno-powered plane can fly easily up to ~10 km, then a Terrier can pop it up high enough to snag the target zone.

I like the Terrier because it's small and lightweight.  Not only does that make it not bog the airplane down, but also, engines are heavy, and having a big heavy engine at the back end of an aircraft moves the CoM backwards and can reduce airplane stability.  The Terrier, being nicely lightweight for an engine, doesn't cause so much of a problem.

("But Terriers are vacuum engines!", I hear you cry.  Yes, but do bear in mind that at a mere 10 km altitude, you're already almost 90% of the way to a vacuum, and the Terrier gets very close to its excellent vacuum Isp.  You don't have to get very high above sea level for Terriers to become effective.)

So, the main thing to maintain aero stability is to design it so that shifting fuel loads don't muck around much with the CoM.  There are two main things I do to cope with this:

First, make sure the CoM isn't way in the back of the plane.  I do this by putting only one relatively lightweight engine in the back (a Terrier), and then mounting my Juno engines amidships rather than at the back end of the plane-- e.g. a pair of Junos, one on either side of the fuselage.

Second, mount the fuel tanks symmetrically fore/aft around the middle of the plane.  If your fuel tanks are mostly-fore or mostly-aft of the CoM, then your CoM will shift dramatically as they drain, which can kill your stability.  Whereas if the fuel tanks are in the middle, then the CoM doesn't move much as they drain.

I do this by putting the LF tanks (for the jet engines) amidships, with the jet engines themselves.  Then a small (say, 2-ton) LFO tank for the Terrier is also amidships.  If I need to make the airplane longer (to provide better lever arm for control surfaces, to adjust CoM position, etc.), the Structural Fuselage part works great-- it provides a lot of length without adding significant drag, and is very lightweight.  I may not need to add a Structural Fuselage, depending on the overall airplane design; when I do add one, most commonly I put it at the back end, right in front of the Terrier, so that it moves the CoM forward.

A final note, if all else fails:  If you've got a plane that works fine to accomplish the objective, and its only problem is "hard to land afterwards because it became unstable" ... well, you could always "cheat" and just tack on a radial parachute or two for landing it.    Perhaps not quite as satisfying as engineering a reliably stable plane, I suppose... but it gets the job done!

[AeroGav]

RCS build aid is really invaluable,  I suspect you are having a problem because your engines are too heavy and when there is no fuel aboard the weight of the cockpit alone is not enough.  The mod shows a red dot which is where your dry CoM is, so you can get that right first then place the fuel equally either side of it.   Also shows torque induced by thrust, so you can center the engine better with offset tool if needed.

As Snark said, try using the lighter Terrier, it sill has plenty of power for an airplane,   and try pushing the jet engines forward.    Again,  four or six Junos may be easier to place forward than a pair of heavy wheeleys.

[steuben]

I've got a _very_ early design.  It can do one hop up to 19km-ish. It requires, Aviation, Engineering 101, Basic Rocketry, and Stability. You can add in Survivability for the PresMat for a bit extra science.

You strap a pair of Junos either side of the Swivel, four FL-T100s, a pair of Mk-0 LF tanks, and enough bits and bobs to get it to fly and land... if you want to. I'll bang up a blueprint when I'm on my usual machine.

[se5a]

This little dude got me up there, though was a tad nose heavy (actually it was a bit heavy overall) on landing - well, 'landing' the pilots survived....

https://imgur.com/a/LU4YsjG

 

I could drop some unneeded stuff, like sci jr, and drop some of the fuel, but was trying to come up with a little multi-purpose shuttlecraft, not sure how it'll handle re-entry heat.

I guess I'll install tweakscale and increase the size of those wings. why'd squad not make winglets procedurally sized as stock? and wheels... and other things?

[Rocket In My Pocket]

CorrectCOL mod will show you a nice graph read out and tell you if your dry com vs. wet com is stable on top of just plain out improving a lot of over sights in the stock col marker.

I'd link it but I'm on my phone.

[se5a]

Did some science and built this: https://imgur.com/a/u4As5x3

flying it manualy is all good, fairly stable and not too bad to fly.

trying to get mechjeb to fly it though... it just goes crazy and I end up in a flat spin. seems like even if I turn mechjeb autopilot on then off again it stays completely screwed up and uncontrollable.

Since some of my contracts are a fair way around the planet flying them all manually is a bit tedious, a bit of straight and level autopilot is kind of nice. but, mechjeb really hates this plane. not sure why at all.

These engines seem to just sip at the fuel, I think I can get just about anywhere on Kerbin with this, and fairly quickly too. but I'm too impatient to just sit there staring at the screen keeping it at the same altitude for the duration of those sorts of flights.

 

Tried the CorrectCOL mod, which says the pitch is unstable, but I'm unsure why this is. does it just not like canard configurations?

 

Edit: tried a tail vs canard version of the above, and it was even worse, horrible to fly manualy and would stall at the slightest provocation.

Edited July 25, 2018 by se5a

[Freds]

MechJeb are nice for space piloting - i use smartASS all the time. For atmospheric flying i use AtmosphereAutopilot mod, it really helps piloting aircraft when using a Keyboard.  Just remember to disable mechjeb smartASS (or any other mechjeb functions) when you enable the mod.

 

[se5a]

Just tried AtmosphereAutopilot. it does seem to be way better than MechJeb. I got to the point where I felt just putting the mechjeb modual on the aircraft caused to be unstable. does mechjeb mess with trim at all? sometimes I felt like as soon as I turned it on, and turned it off sometimes itd screw everything up.

 

That being said, there must be something up with my designs that make them prone to high speed stalls.

Edited July 26, 2018 by se5a