Why is my Spaceplane Overheating?

[Zeiss Ikon]

Okay, so in pursuit of orbital tourism contracts in my first career game, I designed my first spaceplane.  Not a HOTOL SSTO, just a fancy-looking Dynasoar equivalent -- boosted vertically, maneuvers in orbit with a rocket motor on the aft end, reenters and lands more or less like a Space Shuttle (except parachute recovery at the very end, because my R&D hasn't produced retractable landing gear yet -- hey, give me a minute!).  A little "simulation" testing (launch and revert) verified that the orbiter handles well enough in both boost and glide, and even reentry -- and a little tweaking on the booster got it under control during the early parts of boost, when there's a lot of aero force available on those wings (swivel engines and fins on the boosters FTW).

Everything was great until I'd been reentering for five minutes or so -- starting from a 75 km orbit, I set up a periapsis of 48 km.  I held approximately 30 degrees above prograde through reentry, though that tended to drift in a pitch-positive direction (because I was using heading hold while flying around Kerbin).  When I was getting down a bit, I started to run low on battery, and since I had way too much fuel on board anyway, I ran the main (Swivel) engine at low throttle to get some power from the alternator, but I don't think the thrust was enough to contribute to the heating issue (well below what was needed for level flight in earlier atmospheric testing) -- but by the time I was down to 39 km, both the Mk. 1 Cockpit and the RCS sphere in the chin position were very close to their temperature limit (but the Crew Cabin, main wing, canard, and 800 fuel tank weren't even showing thermometers).  That's the point at which I reverted the flight, hoping the problem was simply one of reentry profile.  I'm aware that reentering too high will burn away an ablator with little braking effect, while going too low will overheat things and/or kill Kerbals due to excessive G loads; I'm sure there's a similar tradeoff in reentering a spaceplane.  I just don't know what it is.

Here's a VAB image of the vessel as flown.  I've since pulled half the Lf/O out of the 800 tank (might replace it with a 400 tank, if that proves helpful, though I really need the length to keep the COM from moving too much as main fuel burns off), moved the monopropellant sphere inside a service bay (between the crew cabin and reaction wheel) and mounted a pair of batteries inside the service bay as well -- but I doubt that'll have much effect, other than improving my ability to make a couple orbits and still have electricity to operate the SAS and reaction wheels.

Booster has Swivels on the core and the two boosters that also have fins, Reliants on the other boosters.  Asparagus staging dumps the Reliants first, keeping the Swivel boosters until atmosphere is thin enough the wing doesn't want to make the stack flip end over end.  The 10+ T orbiter reaches orbit with a full tank, if flown optimally, which means if I can lick the reentry this Taxicab generation could also fly tourists on Munar flyby missions.

Would I be ahead to reenter higher, lower, or at a higher or lower angle of attack, or am I just approaching the whole spaceplane thing wrong?

 

Edited October 1, 2017 by Zeiss Ikon

[OHara]

The 2000-K temperature limit of the Mk1 cockpit makes it difficult to use at the front of a spaceplane, where it takes the shock heating.  It is easier to reenter using the inline Mk1 cockpit, with some higher-temperature part at the nose.  The 2600-K RCS port on the nose of an Mk1 nose-cockpit can help a lot, by making a shock cone that clears the cockpit, but only on the way up where the angle of attack is small.

Aerodynamic surfaces, tilted a large angle to the airflow, help a lot to lose energy in something other than heat.  I was happy once to arrange an Mk1 spaceplane with tail-fins that would deploy so far that the plane was stable at 120° angle of attack, coming down somewhat tail first.  I think an angle of attack greater than 45° might be necessary to get a typical Mk1 plane to slow down without overheating.

You might move the canard further forward, until the plane is near-neutral stability, and then use its 'deploy' button to deflect it 150% on re-entry.  Splitting the vertical stabilizer into two fins, usually passive but deployable in opposite directions, could make a nice upper-atmosphere airbrake.

[bewing]

As said above, the MK1 cockpit has a low temp limit. It is always rather difficult to reenter that part for that reason.

In general for spaceplanes, it is much better to come in with a very high Pe. 59km to 63km. Come in with an AoA of 90 degrees for maximum braking. At those altitudes, you still have a good amount of cooling on your craft -- you can easily bleed off a couple hundred m/s before you descend to the point where things start to get really hot.

However, in your specific case, in early career mode with low-temp parts and a lot of fuel remaining:

1) Get yourself down to a Pe around 60km.

2) Set your navball to surface mode.

3) Turn around retrograde.

4) Burn until your surface velocity is at or below 1500 m/s.

At that speed, you can reenter anything -- no matter how delicate.

 

[invision]

i found the best way to kill velocity with space planes is using radial out on SAS options this will flip your plane upright causing the wings drag really hard below you as it enters  the atmosphere. try it yourself you may be very surprised how fast you can kill your speed. then go manual once speed is killed.

[invision]

okay so i made a small plane and placed the little heat shield under the crew cabin

l use the mun so we have way more speed then needed to blow us up

i clicked radial out, the craft struggled to stay still as the atmosphere meets the wings ( use reaction wheels for stability )  

after that single pass we killed off a bunch of speed and we can just repeat this using no fuel what so ever until kerblin pulls you in or you can retrograde and use fuel for precise landing

[Zeiss Ikon]

1 hour ago, bewing said:

As said above, the MK1 cockpit has a low temp limit. It is always rather difficult to reenter that part for that reason.

In general for spaceplanes, it is much better to come in with a very high Pe. 59km to 63km. Come in with an AoA of 90 degrees for maximum braking. At those altitudes, you still have a good amount of cooling on your craft -- you can easily bleed off a couple hundred m/s before you descend to the point where things start to get really hot.

However, in your specific case, in early career mode with low-temp parts and a lot of fuel remaining:

1) Get yourself down to a Pe around 60km.

2) Set your navball to surface mode.

3) Turn around retrograde.

4) Burn until your surface velocity is at or below 1500 m/s.

At that speed, you can reenter anything -- no matter how delicate.

 

Ah, nice.  Now I'm going to fill the tank back up.  With a full 800 tank and a Swivel, I should be able to almost stop a 10 T orbiter.  Tourists, queue up on the left.

[AeroGav]

Looks like you're in career mode,  but as @OHara says you want to swap to an inline cockpit ASAP.

It's a higher tech node so for my very early career spaceplanes (I have a Juno/Terrier SSTO!) I use the space capsule instead with either a mk1 parachute or a small nose cone on the front (you need a 0.625m pointy part on the very front or it suffers excess drag).

It's actually the low internal temp limit of mk1 crew parts that are the problem, not the skin temp of 2k which is similar to most other mk1 stuff.     The Gemini capsule has a slightly higher max internal temp and because it has a nose cone / parachute on the very front, is somewhat protected.

Once you get an inline cockpit you need to find ways to put things between it and the front of the ship.   Reaction wheels, service bays, air intakes (if applicable).   Fuel tanks are an obvious one, but not always possible because if you have the cockpit and engines at the back and fuel at the front, there's a huge cg shift to the rear when the tank is empty.

Having larger wings helps a bit too, because you can make more lift (stay up higher till you are going slower) or more drag.    Lowering the landing gear/opening service bays creates extra drag and slows you down quicker too.

 

[Spricigo]

4 hours ago, Zeiss Ikon said:

or am I just approaching the whole spaceplane thing wrong?

Due to lack  of experience a few mistakes are expected but nothing serious.

Looking at your design it don't seem to be capable to hold high AoA for much long. Changing the Swivel for a terrier and rearranging the fuel tanks can help to have a CoM that 1) don't shift and 2 ) is not so far behind. 

Alternatively a more controllable plane can dive faster into the  low atmosphere before building up much heat and then assume a high AoA to slowdown. 

Finally,  notice that for what you are doing you don't need a good plane, just a mediocre glider.  In fact, a 'better' plane may have worse performance for the task at hand.

[Zeiss Ikon]

Just finished testing @bewing's suggestion -- if I have a lot of fuel on board, I can just about stop the orbiter -- with careful piloting on the test flight (which carried tourists -- this is a career game, after all), I was able to get to orbit with ~10% remaining in the core booster, then used a fraction of that fuel to deorbit after a couple trips around; that left the orbiter's tank full (and I could have done another burn of 100 m/s or so with the booster core to slow things down in the upper atmosphere).  At about 61 km, I burned about 2/3 of the orbiter's fuel retrograde, leaving me just about 1500 m/s (forgot to change the ball to surface, so that was actually 1200 m/s surface).  Had I burned all the fuel, rather than saving some to stretch my glide once I was low and (relatively) slow, I could have gotten below 1000 m/s, possibly as slow as 800 m/s.

Since I have a second contract active for two more Kerbals (one suborbital, one orbital -- somebody gets a bonus) I'll try flying with the 800 tank almost empty to simulate a Mun flyby return and see whether the high angle of attack gets me what I need.  @Spricigo, you're correct, I don't have enough pitch authority (even with RCS on) to keep the nose in a deep stall attitude in dense air; I'm not sure I really need that as long as I can kill my speed up high where the air is thin.  I'd be delighted if Taxicab III is able to keep carrying tourists to Munar flyby and even Munar orbit destinations.  Every space program needs a cash cow.

BTW, this orbiter will maintain a controllable glide at least down to 140 m/s, which is plenty slow for parachute recovery (though the vessel broke up behind the reaction wheel on water impact; I've added another parachute to reduce descent rate).  The COM is a little far forward for easy flying when the 800 tank is empty (ship wants to nose down, more so as it slows and the canard loses authority); I may rearrange the tank, service bay, and reaction wheel to reduce the COM travel as the fuel burns off.  Then again, I may not; until I have retractable landing gear, this is perfectly adequate.  Updated pictures of the as-flown version coming up, probably tomorrow.  Bedtime now...

[bewing]

If you are going to land in the water, you don't need no stinking parachutes.

But you need to redesign a little to get your empty gliding speed down to 60 m/s, before ditching in the water becomes trivially easy.

As you said, the key is getting your CoM and CoL much closer together.

 

[OHara]

2 hours ago, Zeiss Ikon said:

if I have a lot of fuel on board, I can just about stop the orbiter

Good.  With the canard on the cockpit, a similar layout has enough aerodynamic authority to hold 45° on re-entry, so you can use your fuel to get to Minmus. 

[Spricigo]

3 hours ago, Zeiss Ikon said:

@Spricigo, you're correct, I don't have enough pitch authority (even with RCS on) to keep the nose in a deep stall attitude in dense air; I'm not sure I really need that as long as I can kill my speed up high where the air is thin.

No you don't need it if you can still avoid overheating and crashing. Its even inadvisable in some sense, given the chance that this 'improvement' end up causing trouble in the launch/orbit.

That said, being able to keep high AoA is  very useful to avoid overheating and crashing

 

4 hours ago, Zeiss Ikon said:

I'd be delighted if Taxicab III is able to keep carrying tourists to Munar flyby and even Munar orbit destinations.  Every space program needs a cash cow.

In part motivated by this thread, in part because it was in my to do list for a awhile I decide to design something similar. If you want some ideas take a look at Oroshi and Selene.

 

 

[Zeiss Ikon]

@bewing's method worked well, but requires having enough delta-V available to kill roundly half the reentry velocity.  Last night, I tried @Spricigo's recommendation, combined with @invision's multi-pass aerobrake method.  Since this is a career game, the test flight once again carried tourists, and since I "forgot" to nearly empty the main engine's tank before launch, I made it a test of future "Munar flyby" tourist launches.  I did swap the main tank as far forward as possible (directly behind the crew cabin), then adjusted the wing and canard positions to get nearly neutral COM/COL balance with the empty tank.  As previously, I launched on a near-optimum profile, reaching orbit with ~10% remaining in the booster core tanks.

In operational tourism launches to the Mun or Minmus, that fuel would normally have been used during insertion, but for the test, I staged away the booster core, leaving the orbiter on its own, with full tanks.  I then burned for the Mun, setting up a free return orbit that required only a minor periapsis adjustment.  After passing the Mun, I adjusted my Kerbin periapsis to about 60 km (a tiny burn, a few m/s only, could have been done with RCS).  On the first pass,  I burned most of the fuel left from the Munar flyby to lower apoapsis, then turned belly-forward, but got little braking effect (too high).  Next pass, I lowered my periapsis to (iirc) 48 km, burned the last bit of fuel retrograde at about 62 km, then turned belly forward once more.  I was able to easily maintain an attitude near "radial out" using only aero controls and reaction wheel(s) until I was well below orbital speed and around 40 km altitude, and with RCS turned on, down to 30 km.  The only thermometer bars appeared intermittently on the Crew Cabin when I let the nose come down too much, and raising AoA above 60 degrees kept it from getting excessive.

At the end of flight, with main tanks empty and RCS tank nearly so, I was able to slow the glide below 75 m/s during final flare; a little better flare timing might have gotten me below 60 m/s, but this velocity worked okay.  I thought I saw two parts shed just as the vessel got into full water contact, but I couldn't spot anything missing later; it might just have been splash pattern from the join between the wing and tank/service bay.

 

Edited October 3, 2017 by Zeiss Ikon

[bewing]

congrats

[Spricigo]

6 hours ago, Zeiss Ikon said:

After passing the Mun, I adjusted my Kerbin periapsis to about 60 km

Judging by my results  with similar craft, probably can be lowered to ~55km. Maybe lower since your craft have more wings and thus may be  callable of stronger aerobrake.

6 hours ago, Zeiss Ikon said:

was able to slow the glide below 75 m/s during final flare; a little better flare timing might have gotten me below 60 m/s, but this velocity worked okay.  I thought I saw two parts shed just as the vessel got into full water contact

You may find out what get destroyed in the flight report (F3 to access). You may try control surfaces acting as flaps to get that tiny bit of extra sustentaion needed for a slower glide or just use a parachute (less glamorous but functional).

[Zeiss Ikon]

7 hours ago, Spricigo said:

You may find out what get destroyed in the flight report (F3 to access). You may try control surfaces acting as flaps to get that tiny bit of extra sustentaion needed for a slower glide or just use a parachute (less glamorous but functional).

I started with parachutes.  Three, then four.  Landing in water with four parachutes on this craft resulted in killing everyone on board as the craft broke up on impact.  Apparently, 6+ m/s is too hard, and it takes a lot of parachutes to get a 10+ T craft to fall slower than that -- and given the craft is long and slender (at least compared to things like a command pod) I think it'll tend to break at the joints even at slower speeds.  Ditching apparently works better.

It was a little puzzling: after the ditch, I couldn't recover the craft.  Tried to switch to Tracking to do it from there, and was warned that switching away with the craft "moving on the surface" would result in resetting to the last save point.  Finally realized that, most of two minutes after setting down in water, the orbiter was still moving forward at around 3 m/s.  Fortunately, I had enough RCS fuel left to use the thrusters to stop the forward motion, allowing me to recover.  I might still be waiting for the vessel to stop otherwise.

Here's the final version, lifting off (I apparently forgot to take another VAB photo after finishing the mods).

Edited October 3, 2017 by Zeiss Ikon
Add photo

[Spricigo]

1 hour ago, Zeiss Ikon said:

Apparently, 6+ m/s is too hard

Odd thing is that the crew parts are supposed to resist 40m/s, same for the service bay, 12-15 for the wings, winglets and parachutes. Kerbals shouldn't really use sodium for parts that may get in contact with the water.

 

In any case my idae was to k maintain the ditching technique for landing just adding a parachute rigth before splashing down.

[bewing]

2 hours ago, Zeiss Ikon said:

after the ditch, I couldn't recover the craft.  Tried to switch to Tracking to do it from there, and was warned that switching away with the craft "moving on the surface" would result in resetting to the last save point.  Finally realized that, most of two minutes after setting down in water, the orbiter was still moving forward at around 3 m/s.  Fortunately, I had enough RCS fuel left to use the thrusters to stop the forward motion, allowing me to recover.  I might still be waiting for the vessel to stop otherwise.

An airplane has very low drag in the forward direction, and very high drag sideways. All you need to do, once you are successfully ditched, is to turn sideways in the water. Yes it takes a little while to slow down still. But once you get below 1.5 m/s, there is a little gimmick in the water physics that always makes you stop quickly after that. So no, you don't have to wait forever.

 

[Zeiss Ikon]

7 hours ago, Spricigo said:

Odd thing is that the crew parts are supposed to resist 40m/s, same for the service bay, 12-15 for the wings, winglets and parachutes. Kerbals shouldn't really use sodium for parts that may get in contact with the water.

Might have been a little Kraken influence -- one of the components (crew cabin, I think)  jumped out of the water and flew some distance after the water parachute landing.