Spaceplane Re-entry Shenanigans

[Ozzallos]

So I seem to consistently encounter two major problems when operating my spaceplanes, both manifesting themselves in reentry.

1. Burnination. The cockpit can't seem to take the heat; though realistically I'm probably not using the correct decent profile? I've also been experimenting with mounting radiators, but not entirely certain they make a difference or even if they applicable to this application.

2. It's really hard to keep the tail from swinging out from behind me basically flat spinning my way down to around 2000m where the air is finally thick enough to allow me to pull out of this. It's tough to maintain stability through rentry and one wrong move during the decent turns my wonderful plane into a paper weight. Most of my weight is predictably aft (once the fuel is used), so I understand this to some degree, but it seems rather obnoxious.

Any tips somebody can toss out?

[dafidge9898]

Make sure you have the center of lift behind the center of mass, and as few stabilizers as possible in front of the center of mass. That will keep your craft from flipping and possibly prevent it from spinning. It's also because your center of mass moves aft with fuel burned. As for the heat problem, make sure you're up at 15k or so. When you start overheating at 15k, it's time to ascend higher to get the apoapsis above 70k. If you wan't to go extra fast, put a heat shield or shielded docking port on the nose instead of a pointy cone; it has something to do with the shockwave that reduces heat with these blunt noses.

[TheDestroyer111]

27 minutes ago, dafidge9898 said:

Make sure you have the center of lift behind the center of mass, and as few stabilizers as possible in front of the center of mass. That will keep your craft from flipping and possibly prevent it from spinning. It's also because your center of mass moves aft with fuel burned. As for the heat problem, make sure you're up at 15k or so. When you start overheating at 15k, it's time to ascend higher to get the apoapsis above 70k. If you wan't to go extra fast, put a heat shield or shielded docking port on the nose instead of a pointy cone; it has something to do with the shockwave that reduces heat with these blunt noses.

15k? r u mad? that's not a good periapsis for reentry, it can quickly burn you up. The CoM (center of mass) doesn't always move back with fuel use, be sure to check that! Also, KSP uses a simple heating and drag model and shockwaves have nothing to do with it. It's just the increased drag of a blunt heat shield.

The probably best way to deal with heating is facing perpendicular to prograde on reentry. This way, drag is maximized which can make your speed drop faster than heat builds up, allowing to survive reentry.

Radiators are commonly attached to spaceplanes and are sometimes effective, but usually, if a spaceplane explodes after removing radiators, it is either a bad plane or someone flies it badly.

As for reentry profiles, lower the periapsis to somewhere between 25-40km, and if you keep exploding, you can go for 35-50km periapsis and reenter several times to gradually bleed off speed. If applicable, a good method can also be adding excess deltaV to a spaceplane to decelerate to a completely suborbital trajectory, which, coupled with facing perpendicular to prograde, can give a very safe reentry. That's what I do on a spaceplane I extensively use in career mode right now. Despite being theoretically capable of doing a Mun flyby, I instead use the additional fuel to lower reentry speed.

Oh, one more thing. For reasons unknown, replicas of the NASA space shuttle in KSP are very hard to built, as some unknown force makes many of them unstable on reentry even if CoL is behind CoM and they are stable in pure low atmospheric flight.

Edited July 22, 2016 by TheDestroyer111

[juanml82]

Which cockpít are you suing? IIRC the mk1 cockpit has lower heat tolerance than the MK2 parts

Fuel should have been burnt even while using airbreathing, but IIRC, closed cycle engines can burn it unevenly from the tanks, so you may end up with the front tanks empty and the ones at the bottom (where you also have the engines) almost full. It's a good idea to use the TAC Fuel Balancer mod to balance the fuel in all tanks before reentry.

I'm under the impression a lot of issues around 20,000 meters are due the horizontal speed becoming too slow. Therefore, the plane stalls. Worse, if the plane is tilted towards the airstream, I think it gets into an asymmetric stall, ie, a spin.

So, if you're reentring at a high AoA (which is correct) it's a good idea to monitor the horizontal speed. If it begins to fall below 300 m/s, point the nose down and turn on the engines, so you descend while keeping your horizontal speed between 100-300 m/s. I may be entirely wrong about this, though.

[TheDestroyer111]

4 minutes ago, juanml82 said:

Which cockpít are you suing? IIRC the mk1 cockpit has lower heat tolerance than the MK2 parts

Fuel should have been burnt even while using airbreathing, but IIRC, closed cycle engines can burn it unevenly from the tanks, so you may end up with the front tanks empty and the ones at the bottom (where you also have the engines) almost full. It's a good idea to use the TAC Fuel Balancer mod to balance the fuel in all tanks before reentry.

I'm under the impression a lot of issues around 20,000 meters are due the horizontal speed becoming too slow. Therefore, the plane stalls. Worse, if the plane is tilted towards the airstream, I think it gets into an asymmetric stall, ie, a spin.

So, if you're reentring at a high AoA (which is correct) it's a good idea to monitor the horizontal speed. If it begins to fall below 300 m/s, point the nose down and turn on the engines, so you descend while keeping your horizontal speed between 100-300 m/s. I may be entirely wrong about this, though.

You are not entirely wrong. However, optimum speed depends on altitude. At 20km and 300m/s, wings don't do much unless they are of huge area relative to the plane's mass.

In theory, Mk2 cockpits are more resistant to reentry due to higher temperature tolerance and the cockpits themselves generating quite a bit of lift. Some people may think otherwise than I do, but from my experience however the Mk2 cockpits are very vulnerable to exploding.

Edited July 22, 2016 by TheDestroyer111

[foamyesque]

Aim your Pe at 45km, put your AoA at 40-50 degrees, build your ship so that your dry/near-dry CoM is dead centered on your CoL, have a large vertical stabilizer, and turn on RCS during your descent to give you useful attitude control while your control surfaces are stalled out.

[Empress Neptune]

On the stability issue, keep in mind that it's not just the CoL/CoM relationship that matters. Your rudder deals with sideforce, not lift. The fuselage will produce its own sideforce just as it produces lift and drag. Sideforce ahead of the CoM will make the plane unstable while force behind the CoM will help stability. Fuselage aero forces in KSP can be be very large, so if your rudder is closer to the CoM than the nose, the nose sometimes wins and overpowers the rudder. To fix this, try moving the CoM and CoL forward together while keeping the tail where it is. Alternatively, you can move the tail backwards on a boom.

[Ozzallos]

Thanks peeps.

My re-entry profile has basically been like a capsule-- Ride it to the ground. Instead, it sounds like I should be using the atmosphere to brake my craft naturally. Mistakes were made. Might help with the heat and controllability at the same time by keeping speed up.

(using the Mk1 cockpit)

Edited July 22, 2016 by Ozzallos

[bewing]

A Pe as high as 63km can sometimes be the answer. As you say, use the atmosphere to brake. Holding an AoA of 90 degrees as long as you can is usually a good idea. -- And the MK1 cockpit is thermally the weakest of all possible noses.

An MK1 command pod with a small nosecone on it weighs less, costs less, and performs much better. An MK1 inline cockpit can perform even better, depending on what you stick on the front of it.

[dafidge9898]

4 hours ago, TheDestroyer111 said:

15k? r u mad? that's not a good periapsis for reentry, it can quickly burn you up. The CoM (center of mass) doesn't always move back with fuel use, be sure to check that! Also, KSP uses a simple heating and drag model and shockwaves have nothing to do with it. It's just the increased drag of a blunt heat shield.

I must not have been clear enough. I didn't mean a 15km periapsis for reentry, I meant an altitude of 15km during ascent, where he should start accelerating to a good speed for orbit, but not too fast to burn up. And in any case, it's good to have the CoL behind the CoM, whether it moves forward or back with fuel use. If it does happen to move back, then it's even more important to have the CoL behind the CoM. His case the CoM moved back.  I thought they added shockwaves with the new aerodynamics a while ago? I must be wrong, then. 

Edited July 22, 2016 by dafidge9898
Addition

[foamyesque]

If you put your CoL behind your CoM for reentry on a spaceplane, it will try to lock you nose-forward. You want, if at all possible, your CoL square on your CoM. I favour canard designs over tails for that reason, ideally using the Standard Canard for its maximal control surface area and good deflection angles. The trick's in setting things up so your CoM shift during ascent is as minimal as possible. Skylon-style designs, with the engines close to midbody, work great for large ships; for smaller ones you probably want a rear-mount engine to counterbalance the weight of your pod.

[Ozzallos]

 

Your reward for teh helpfulness.
SSTO Vortex Mod B

[TheDestroyer111]

15 hours ago, foamyesque said:

If you put your CoL behind your CoM for reentry on a spaceplane, it will try to lock you nose-forward. You want, if at all possible, your CoL square on your CoM. I favour canard designs over tails for that reason, ideally using the Standard Canard for its maximal control surface area and good deflection angles. The trick's in setting things up so your CoM shift during ascent is as minimal as possible. Skylon-style designs, with the engines close to midbody, work great for large ships; for smaller ones you probably want a rear-mount engine to counterbalance the weight of your pod.

NO, NO, NO!!!! Have you ever tried maneuvering something with CoL exactly inside CoM? And even if you didn't crash while doing so, It is good that a plane tries to "lock itself nose-forward" without input. This means that it wants to point in whichever direction it moves. A typical powered aircraft is nearly impossible to recover from a spin if it has CoL in front of CoM or very close to it (such planes do exist, but they have large, strongly and quickly deflecting control surfaces coupled with redundant computerised fly by wire. If it is even possible to turn off the SAS system on such planes, they become either unflyable or overresponsive to controls.)

 

15 hours ago, Ozzallos said:

 

Your reward for teh helpfulness.
SSTO Vortex Mod B

Looks cool! Does it fly well, or is it the craft that used to have trouble with reentry?

[Ozzallos]

2 hours ago, TheDestroyer111 said:

NO, NO, NO!!!! Have you ever tried maneuvering something with CoL exactly inside CoM? And even if you didn't crash while doing so, It is good that a plane tries to "lock itself nose-forward" without input. This means that it wants to point in whichever direction it moves. A typical powered aircraft is nearly impossible to recover from a spin if it has CoL in front of CoM or very close to it (such planes do exist, but they have large, strongly and quickly deflecting control surfaces coupled with redundant computerised fly by wire. If it is even possible to turn off the SAS system on such planes, they become either unflyable or overresponsive to controls.)

 

Looks cool! Does it fly well, or is it the craft that used to have trouble with reentry?

She rocks the ascent and tops out at 1250m/s @ 40' before stalling for air and switching over. With satillite payload she can still clear 1150 to ceiling with enough left to circularize and maneuver

Deorbiting was where i was having the issues and it looks like my descent profile was flawed per this thread. She's hard to handle without power above 10km but is a dream under that. im considering a couple dawn engines to suppliment orbital manuvers so i can keep gas for the atmo.

[TheDestroyer111]

46 minutes ago, Ozzallos said:

She rocks the ascent and tops out at 1250m/s @ 40' before stalling for air and switching over. With satillite payload she can still clear 1150 to ceiling with enough left to circularize and maneuver

Deorbiting was where i was having the issues and it looks like my descent profile was flawed per this thread. She's hard to handle without power above 10km but is a dream under that. im considering a couple dawn engines to suppliment orbital manuvers so i can keep gas for the atmo.

So, I can mark your question as answered?

BTW, you can consider using shock cone intakes. Reason: http://forum.kerbalspaceprogram.com/index.php?/topic/125015-105-intakes-lets-figure-them-out/ So basically, most intakes give maximum amounts of air at a certain speed, and then give less air at even higher speeds. The shock cone intake, however, gives air proportionally to speed. If your engines run out of air at 1250m/s, you could get 200-450m/s more speed (depending on TWR) and a bit of altitude in jet engine mode with shock cone intakes.

 

[AeroGav]

This is a career mode partially re-usable spaceplane.    It jettisons the jet engines to give it enough delta V to reach Minmus.

At that point, it becomes a car 

But I digress.  Re-entry from Minmus in the Mark 1 cockpit was surprisingly comfortable.   Referring to the first picture -

1.  The cockpit is not at the front.   Ahead of it lies a service bay, then FT800 Fuel tank,  NCS adapter, then small nose cone.   At the very tip of the nose is the antenna.

2.  Lift ! You can see those blue lines off the wings.   Together with moderate pitch up, it keeps us from dropping down into the thicker atmosphere until much of the speed has bled off.

3.   Drag ! Open service bay. Landing gear down.

4. Radiator.    Simple radiator panels only take heat from the part they directly attach to.    Deployable ones like this work on the whole ship.   Simple,  radially attached radiators increase drag,  but i have a small cylindrical battery & reaction wheel in the service bay, which I've put my radiator system on.   Open the bay and deploy the rad when you care about heat more than drag.  Note,  the radiator will disintegrate if kept deployed much below 38km on re-entry.

 

On re-entry, if you dive straight down into the atmosphere as some folks recommend, you'll slow down quickly which may reduce total heat input.  However, the rate of heating will be very high.  Parts explode when either the core temperature or skin temperature exceeds their limit.  Re-entry heat is applied to the skin, which tries to get rid of it either to the environment or by conducting away to the core of the part.    A steep descent could cause the skin to overheat due to the rate of heating being too high even though the core temperature is nowhere near max.

On the other hand if you use lift to skim off the upper atmosphere and come down slowly, over time heat soak can make the internal temperature your limiting factor.  That's why i went with radiators.  The drag from the landing gear and open service bay also help to shorten the re-entry overall and reduce total heat input.

How I came back from Minmus -

Burned retrograde at Minmus to drop my Periapsis to 44km or so, our Apoapsis is still way out at Minmus altitude but this saves Delta V.    Each time we fall below 70km I raise the nose to 20 degrees above prograde.  The lift generated causes our actual low point to be at least 4km higher than the predicted Periapse.   After about 3 or 4 aerobraking passes our Apoapse has fallen below 70km too.

At this point I set the pitch angle to 5 degrees which is our "best glide" at supersonic speeds.   Lift keeps us up for a while, but when we start threatening to drop below 38 km,  the altitude at which we'll be forced to stow our radiator or have it break off,  start pitching up to generate additional lift and keep out of the lower atmosphere as long as you can.

Now, you could do this smoothly like when you're landing a plane you approach at gliding speed then start pulling the nose up just above ground level to delay touchdown until the plane is almost stalled.          Due to handling issues I ended up pitching up to 25 degrees immediately however which sent the plane ballooning up to 50km before sinking like a rock.  It worked pretty well however, because by the time we passed 37km we were down to just 1100m/s.    Peak temperature never went above 60%, nothing ended up glowing red.

[mk1980]

if you're just coming down from LKO (as opposed to some high speed reentry returning from minmus or duna or whatever), consider a MUCH steeper reentry. works pretty well and makes it a lot easier to aim for the space center. if you have enough fuel left, there's nothing wrong to burn your periaps down below ground level.

don't know if it will work for your specific design, but in general that tends to work better than those overly careful PE at 50km aerobrake pass approaches. the main advantage is of course that the landing zone is more predictable and the reentry is a lot faster. both are important factors when you actually use the SSTO for economic reasons in career mode.

can't hurt to try it. if it doesn't work you can still go the careful route. in my experience, the reckless approach tends to work just fine in KSP, though.

[dire]

If you are coming in from further away than LKO, drop your landing gear during aerobrakes and until you hit 30km, and then turn OFF your SAS. This will induce non-symmetric drag on your craft, causing it to freely tumble, which will frequently present a flat surface to your prograde direction (it cobra's for you). It will also prevent a low-drag flat spin, which OP mentioned was a problem (because your spin will be on all 3 axes).

I highly recommend fitting at least one parachute to any craft you try this on, because coming out of that tumble in time can be a bear.

Edited July 25, 2016 by dire

[foamyesque]

On 2016-07-23 at 7:51 AM, TheDestroyer111 said:

NO, NO, NO!!!! Have you ever tried maneuvering something with CoL exactly inside CoM?[

Of course. I design all my planes to the plan I laid out. Spins can be eliminated with reasonable v-stabilizer sizing and/or non-aerodynamic control inputs-- you'll note I endorse the use of RCS.

 

The problem with a natural tendency to go nose-prograde is that, on re-entry, if that is too strong for your control systems to compensate for (and it requires very little for that to be true), you get locked nose-on, overheat because your drag is way too low leading to high low-atmo speeds, and then explode. It also leads to difficulties in pitch up on final approach. Sometimes I even put my CoL in front of my CoM, though then the tendency to backflip becomes problematic. The absolute best option is for your zero-trim balance to be just a smidge behind your CoM-- the balls should be overlapping, the more the better-- so that when you add a command input your CoL moves to or ahead of your CoM. That gives you maximum agility while retaining passive stability. I believe this is known as "relaxed stability" design in real life, but it's the best way to build a KSP plane.

Edited July 27, 2016 by foamyesque

[fourfa]

I love using a fuel pump mod for spaceplane re-entry (for instance GPOSpeed). Say you have two tanks on your plane, front and rear. Set your CoL safely behind your CoM in the SPH. On reentry approach, moving the CoM backward is a simple matter of changing the front tank priority below the rear tank. Quick enough to do during flight, and can make huge shifts in attitude. 

In stock, it's easier in 1.1 than it used to be - just pin the relevant tank right-click menus, and pump fuel around as needed. Takes more clicks and attention, but it's doable. 

If you're reentering with no fuel, space shuttle style, this method is not going to work. You might try shifting payload around if you've got a payload bay?