Using Laythe for Jool aerocapture

[cephalo]

On my lunch hour I tried a couple of times to do an aerocapture on Laythe. This is a large ship with some heat shields but not covering the whole thing. I'm moving about 8000 ms, and at both 25k and 45k altitiudes it was like hitting a brick wall. Is this possible to do in 1.05 or do I need a different way into Jool? I'm thinking that above 45k, there will be no aerocapture.

The log shows everything exploding in the same instant, though there are items missing from the list. There was nothing left of my ship. I think I overloaded the capacity for logging explosions. I win.

EDIT: Do explosions heat other parts? I did accidentally leave some nuke generators exposed, but they have already served their purpose and aren't strictly necessary. I wonder if their exploding cause a chain reaction of overheated parts. One of them was the first part to explode. Normally on Kerbin, I have been losing these without consequence to other parts.

Edited December 7, 2015 by cephalo

[FancyMouse]

Yup it does sound chain reaction. From my experience, it's also existent on Kerbin reentry, and depends on the size of the exploding part. Solar panels are usually not a problem, but if engine explodes, it almost certainly heat the part next to it really fast and cause the next explosion.

Nowadays I don't do interplanetary aerobrake without extensive testing (which is what's supposed to be). 

And 8km/s speed sound scary - it's almost direct Jool atmo entry speed. It's going to rip anything unprotected apart. Are you entering the wrong direction into Laythe atmo, or are you even aware that the direction matters?

[cephalo]

I am NOT aware the direction matters. This is my first go at the Jool system. Please explain! I have a save to go back to before such decisions are made, but from a 150 days out, it seemed that my options were limited. I first wanted to try a Tylo grav assist but, Laythe was really getting in the way, so I thought I'd try Laythe instead. I guess I'm catching it coming at me instead of going away. I wonder if I can correct that.

EDIT: It was very scary. I panicked, thinking I had left the time warp on, then realizing it was at x1... This is gonna hurt... and not slow me down.

Edited December 7, 2015 by cephalo

[FancyMouse]

I can't draw, but imagine - if you enter Laythe in the same orbital velocity direction of Laythe, then Laythe's orbital velocity will cancel your orbital velocity in Jool reference frame. Instead if you go in wrong direction, it won't cancel but rather add. Considering Laythe's orbital velocity is 3km/s, that's not negligible.

Regarding how - it should be some adjustment prior to entry the Jool system. Play with the radial/anti-radial to enter at the right position, and play with prograde/retrograde to adjust your entry datetime so that Laythe appear at that right position.

[Valerian]

1 hour ago, FancyMouse said:

I can't draw, but imagine - if you enter Laythe in the same orbital velocity direction of Laythe, then Laythe's orbital velocity will cancel your orbital velocity in Jool reference frame. Instead if you go in wrong direction, it won't cancel but rather add. Considering Laythe's orbital velocity is 3km/s, that's not negligible.

Regarding how - it should be some adjustment prior to entry the Jool system. Play with the radial/anti-radial to enter at the right position, and play with prograde/retrograde to adjust your entry datetime so that Laythe appear at that right position.

You just made me understand what an orbital slingshot was ... why didn't I get that earlier?

[Teilnehmer]

Consider using Tylo’s gravity assistance to colse your Jool orbit.

[RexKramer]

43 minutes ago, Teilnehmer said:

Consider using Tylo’s gravity assistance to colse your Jool orbit.

Using Tylo is probably the best tool for getting anywhere in the Jool system.  Use Tylo to close your orbit around Jool.  With a little tweaking, you can use Tylo to put your orbit just about anywhere around Jool.

By doing this, you can get your velocity relative to Laythe to a much more reasonable amount without spending much DV.  Relying on aero captures is now very difficult in 1.0.5, capturing at 8,000m/s is really not a good approach to the problem, although in earlier versions of KSP it worked great.  

 

[Old Foxboy]

Just for your information,

the Space Shuttle hits Earth's atmosphere at about the same speed. well, 7000 m/s, but however, it won't "rip everything apart", because the reentry starts high up at 100km above the surface while the main heat is down below 50km.

I suggest to enter upper parts of the atmosphere - and don't enter it streamlined, but at a high AOA, so you make maximum use of the little aerobraking.

You won't get a capture, but you will save fuel for the orbital insertion. You might also burn before the aerobraking so you can go deeper into the atmosphere without braking anything.

 

[cephalo]

I've discovered that my ship is just not cut out for aerobraking, and the heat shields I brought are just dead weight. However, I'm finding that the Jool system is a gravity assist paradise. I started with a capture around Tylo, then with the help of both Jool and Laythe, I will arrive back at Tylo again, nice and slow at apoapsis, which is where I wanted to start anyway, for around 115 dv. Not too bad.

The aerobreaking thing at 8000 ms is ridiculous. It is in no way feasible unless everything is well behind a heat shield and also inside a cargo bay, which pretty much rules out any large craft with multiple components. Even then, you would gain nothing because even fully shielded, you can't survive an atmosphere thick enough to slow you down. Being in an atmosphere at all at that velocity causes your craft to glow red, even before you see those red streaks that you would see on a Kerbin reentry.

It seems odd to me that you would be generating that much friction in an atmosphere that can't even change your velocity. Is that broken, or realistic in some way? Shouldn't the change in velocity be directly related to the increase in temperature? It is friction after all that slows you down at the same time it generates heat. Do we have a math guy around here to figure this out for us?

EDIT: For example, while I was at about 193km altitude on Jool, my whole craft was glowing bright red, and my apoapsis didn't change at all. If I burn my engines just outside of the atmosphere, it costs about 200 dv to cut the apoapsis down by 50 percent. That scenario seems very wrong.

Edited December 8, 2015 by cephalo

[Warzouz]

I've learnt that most of the heating doesn't come from friction, but from gas compression in front of the ship. So Heating and drag aren't directly related (but I can be wrong...)

[FancyMouse]

RL space shuttle analogy doesn't quite work in KSP in terms of exact numbers because physics constants have been scaled so that Kerbin is simulating Earth.

A normal Laythe aerocapture should not even exceed 4km/s. 8km/s problem is not on KSP. It's on your trajectory which is explained above.

Jool is a separate problem... it has not worked in 1.0.4 and since dev claims something is fixed in 1.0.5, we need some test.

Edited December 8, 2015 by FancyMouse

[cephalo]

15 minutes ago, FancyMouse said:

RL space shuttle analogy doesn't quite work in KSP in terms of exact numbers because physics constants have been scaled so that Kerbin is simulating Earth.

A normal Laythe aerocapture should not even exceed 4km/s. 8km/s problem is not on KSP. It's on your trajectory which is explained above.

Jool is a separate problem... it has not worked in 1.0.4 and since dev claims something is fixed in 1.0.5, we need some test.

Right, I get that now. However, even at 8000ms either on Jool or Laythe, there should be some velocity change any time there is heat correct? My problem is that I'm getting catastrophic heat in extremely thin atmospheres and no associated velocity change.

EDIT: Can anyone explain to me why temperature change should not have a linear relationship to velocity change? Gas compression is also the result of friction, and should not cause added energy into the system.

Edited December 8, 2015 by cephalo

[FancyMouse]

Most likely gravity is overcoming the tiny drag from thin atmo. That may explain why you see no speed change. Keep in mind that drag~v^2 but heat~v^x where x is something roughly 2.5 to 3 depending on what model used. That will mean even drag is tiny, heat can still be nontrivial.

EDIT: seems I remember the wrong exponent... the funky 0.5 thing goes to density not velocity.

Edited December 8, 2015 by FancyMouse

[cephalo]

1 hour ago, FancyMouse said:

Most likely gravity is overcoming the tiny drag from thin atmo. That may explain why you see no speed change. Keep in mind that drag~v^2 but heat~v^x where x is something roughly 2.5 to 3 depending on what model used. That will mean even drag is tiny, heat can still be nontrivial.

I still question whether this model is correct. For one thing, even a small change in velocity at periapsis has a huge change in the apoapsis for a highly elliptical orbit. Like I mentioned, even 200 dv, a very small change, brings the apoapsis down to half altitude (when apoapsis around Jool is initiailly near SoI.) Gravity creates the orbit and keeps it stable. Any changes, even tiny ones, ought to be visible in such an extreme circumstance. TIny drag should equal tiny heat.

[FancyMouse]

(Re)entry heat is a complex topic, but I think some of your claims are wrong by some simple arguments.

> has a huge change in the apoapsis for a highly elliptical orbit

Apoapsis is a misleading measure to discuss aerobrake efficiency. Take an example of what you said, "half altitude", when that altitude is really high, it's actually a tiny difference between those periapsis speeds. All my point is, "half altitude" is not that a big difference as you think.

> TIny drag should equal tiny heat.

But power equals force multiplied by speed - that's actually how you derive heat~v^3 from drag~v^2. There are some weird stuff going on (e.g. you don't know how much heat is actually transferred to your ship etc.), but still, tiny drag=tiny heat just sounds to me a bad intuition (as rocket science has always been).

See the thread I initiated a while ago and excellent info from one of the devs.

Edited December 8, 2015 by FancyMouse

[OhioBob]

7 hours ago, FancyMouse said:

Jool is a separate problem... it has not worked in 1.0.4 and since dev claims something is fixed in 1.0.5, we need some test.

Right after 1.05 was released, I ran a large number of tests at Jool to see how suitable it is for aerocapture.  Really high ballistic coefficients don't work too well,  but anything with a low to mid BC is pretty easy to aerocapture.  Part of the reason for the improvement is that Jool's upper atmosphere thickens with increasing depth more gradually than it did prior 1.0.5.  However, I think what is even more important is the change to the heat shields, which now ablate much more slowly than before.  The exact periapsis needed for aerocapture is going to depend on the drag characteristics of the vehicle and the entry velocity.  In most cases I found that an aerocapture could be achieved with a periapsis somewhere in the range of 155-165 km.  For a high BC vehicle and/or a high entry velocity, you should be closer to 155 km.  For a low BC vehicle and/or a low entry velocity, you should be closer to 165 km.  You'll have to experiment to know for sure.  In my experiments, I found that my test vehicles wanted to flip around when I got down to an altitude of about 155 km, which did not end well.  If you're going to have to dip that deep into the atmosphere, be sure to take steps to make your ship aerodynamically stable in its heat shield forward attitude.

[cephalo]

15 hours ago, OhioBob said:

Right after 1.05 was released, I ran a large number of tests at Jool to see how suitable it is for aerocapture.  Really high ballistic coefficients don't work too well,  but anything with a low to mid BC is pretty easy to aerocapture.  Part of the reason for the improvement is that Jool's upper atmosphere thickens with increasing depth more gradually than it did prior 1.0.5.  However, I think what is even more important is the change to the heat shields, which now ablate much more slowly than before.  The exact periapsis needed for aerocapture is going to depend on the drag characteristics of the vehicle and the entry velocity.  In most cases I found that an aerocapture could be achieved with a periapsis somewhere in the range of 155-165 km.  For a high BC vehicle and/or a high entry velocity, you should be closer to 155 km.  For a low BC vehicle and/or a low entry velocity, you should be closer to 165 km.  You'll have to experiment to know for sure.  In my experiments, I found that my test vehicles wanted to flip around when I got down to an altitude of about 155 km, which did not end well.  If you're going to have to dip that deep into the atmosphere, be sure to take steps to make your ship aerodynamically stable in its heat shield forward attitude.

This is completely unlike my recent experience. I was burning up at 195 km at 8000 ms. I had some vulnerable parts sticking out, but my whole craft was very quickly becoming red hot, no matter if it was behind the shield or not. I guess I had 1.02 on the brain when designing this ship, but I was surprised to find the very cusp of the atmosphere to destroy things.

What kind of ships are you testing with?

Edited December 9, 2015 by cephalo

[cephalo]

16 hours ago, FancyMouse said:

See the thread I initiated a while ago and excellent info from one of the devs.

So I read the thread, and I'm not sure I understood the whole thing but I think I gleaned that there is a 'correct' re-entry angle for a particular ship. Too steep and you will get a destructive heat spike, too shallow and you will get much more heat for the amount of velocity change.

What that means for a Jool aerocapture is, everything must be in a cargo bay that isn't fuel tanks or a heat shield. 

[cephalo]

I'm aware that intuition is not a good thing to guide your rocket science, but when I read about this sort of thing on Wikipedia, there should also be some cooling along with the heating. Friction creates heat, but also there is heat transfer between the gas and the surface of the craft. Once that craft gets hotter than the air, it should be losing some heat too as new cold air is coming in contact with the ship. Perhaps this factor is missing from the calculations. So a high velocity, low drag situation wouldn't just add heat as it currently does, but would be counter acted somewhat by the cooling.

[Old Foxboy]

7 minutes ago, cephalo said:

I'm aware that intuition is not a good thing to guide your rocket science, but when I read about this sort of thing on Wikipedia, there should also be some cooling along with the heating. Friction creates heat, but also there is heat transfer between the gas and the surface of the craft. Once that craft gets hotter than the air, it should be losing some heat too as new cold air is coming in contact with the ship. Perhaps this factor is missing from the calculations. So a high velocity, low drag situation wouldn't just add heat as it currently does, but would be counter acted somewhat by the cooling.

That's like putting a pizza in a hot oven expecting the air in the oven to cool the pizza as well as it heats it...

 

[cephalo]

51 minutes ago, Old Foxboy said:

That's like putting a pizza in a hot oven expecting the air in the oven to cool the pizza as well as it heats it...

 

Jool is a very cold oven. Why is it cooking my pizza?

EDIT: The air that touches your spacecraft is heated by your velocity, but this takes some small amount of time. Before it gets hotter than your craft, it is transferring heat away from it. If you only account for the heat added, you will get inaccurate results.

I have not seen evidence of any 'high speed cooling' in this game, which should happen at subsonic speeds etc. I admit I haven't actually looked for it. Does it happen?

Edited December 9, 2015 by cephalo

[OhioBob]

7 hours ago, cephalo said:

This is completely unlike my recent experience. I was burning up at 195 km at 8000 ms. I had some vulnerable parts sticking out, but my whole craft was very quickly becoming red hot, no matter if it was behind the shield or not. I guess I had 1.02 on the brain when designing this ship, but I was surprised to find the very cusp of the atmosphere to destroy things.

What kind of ships are you testing with?

I am surprised our results are so different. In my tests the heat shield began to glow orange shortly after encountering the atmosphere, but everything behind it seemed fine.  I didn't experience any significant overheating problems.

I tested three different ships, which I hyperedited into a hyperbolic encounter with Jool.  Each ship was tested using two different approach trajectories, which I thought represented the likely minimum and maximum velocities for a ship arriving from Kerbin.  The velocities at atmospheric entry were about 9660 m/s and 9970 m/s.  I tested a range of periapsis to find the minimum and maximum that would result in an aerocapture.  I considered an aerocapture anything that resulted in an apoapsis that was (1) above Jool's atmosphere and (2) inside Jool's SOI.

The three ships were just something that I threw together to test the feasibility of aerocapture, i.e. they weren't equipped to do anything more than the aerocapture tests.  All three used a single 2.5m heat shield, behind which I stacked varying amounts of mass to alter the ballistic coefficient.  Below are brief descriptions:

Ship #1 - A very small probe consisting of 0.625m and 1.25m parts tucked in behind an oversized 2.5m heat shield.  Total mass at entry: 2915 kg.

Ship #2 - Consisted of an in-line stack of 2.5m parts.  From front to back:  heat shield, RC-L01 remote guidance unit, service bay (with RTG and batteries), reaction wheels, X200-32 fuel tank (with fuel only), and a LV-N engine.  Total mass at entry: 14,825 kg.

Ship #3 - Consisted of an in-line stack of 2.5m parts.  From front to back:  heat shield, RC-L01 remote guidance unit, Z-4k battery bank, reaction wheels, Jumbo-64 fuel tank, and a Skipper engine.  Radially attached to the fuel tank were six SP-L 1x6 solar panels.  Total mass at entry: 41,350 kg.

Ship #1 is something that I originally tested in either version 1.0.2 or 1.0.4 (don't remember which).  I was actually able to achieve an aerocapture with it back in the old version, through just barely and with an extremely small margin of error.  Under version 1.0.5, Ship #1 worked like a charm.  I had absolutely no heating or stability problems.  The periapsis sweet spot was about 165-170 km.

For the most part, tests with Ship #2 went pretty smoothly as well.  I didn't have any heating issues, but there were some stability problems.  Above an altitude of about 153 km it worked fine, but once I got below 153 km the ship wanted to flip around.  Once it flipped, it was destroyed in a matter of seconds.  The periapsis sweet spot was about 155-160 km.

Ship #3 was problematic, not because of heating but because of instability.  Because of the ship's high ballistic coefficient, a low periapsis was needed to achieve an aerocapture.  The lowest periapsis I could attain without the ship flipping around and destroying itself was 155 km.  At the low entry velocity, aerocapture was possible with a 155-160 km periapsis; however, at the high entry velocity aerocapture was extremely difficult.  I found a very narrow corridor just above 155 km where I could aerocapture without the ship flipping, though the small margin makes it unpractical.  With a more aerodynamically stable ship, aerocapture at high BC/velocity might be possible, probably with a periapsis in the 150-155 km range.

 

Edited December 9, 2015 by OhioBob

[FancyMouse]

6 hours ago, cephalo said:

I'm aware that intuition is not a good thing to guide your rocket science, but when I read about this sort of thing on Wikipedia, there should also be some cooling along with the heating. Friction creates heat, but also there is heat transfer between the gas and the surface of the craft. Once that craft gets hotter than the air, it should be losing some heat too as new cold air is coming in contact with the ship. Perhaps this factor is missing from the calculations. So a high velocity, low drag situation wouldn't just add heat as it currently does, but would be counter acted somewhat by the cooling.

But reentry heat has nothing to do with friction. Quoting from NASA (with my own annotation):

"...but tremendous heat, minimally due to friction between the air and surface of the spaceship as it moves at mind-boggling speed. (Most of the generated heat is due to rapid air compression without time for the air to cool off.)"

Edited December 9, 2015 by FancyMouse

[Old Foxboy]

On 9. Dezember 2015 um 17:09:26, cephalo said:

Jool is a very cold oven. Why is it cooking my pizza?

EDIT: The air that touches your spacecraft is heated by your velocity, but this takes some small amount of time. Before it gets hotter than your craft, it is transferring heat away from it. If you only account for the heat added, you will get inaccurate results.

I have not seen evidence of any 'high speed cooling' in this game, which should happen at subsonic speeds etc. I admit I haven't actually looked for it. Does it happen?

You are captured in a wrong paradigm. The question is not how cold the surrounding air is, but what actually happens at the surface of your craft. The air 1m away from your craft might be like minus 150°C, but your craft squeezes the air in front of it so much that when the air molecules finally hit the surface of your craft, they are compressed so insanely high that the whole front (!) is covered in hot ionized gas. No cooling at all.

There is a significant heat decay when the air flows around your craft and it's pressure drops instantly, creating a zone where excess heat may or may not be taken away. However, since the pressure of the "rear end" gas is so low again, it hasn't any significance.

A fly sitting on an iron girder makes it bend. But it's neglectable.

Flying subsonic and in higher pressure lower atmosphere is a totally different thing and the numbers game changes a lot. But don't forget the difference between linear and exponential. Simple logic only works if you stay in the same sphere of numbers influence. Exponential changes can throw your logic very fast.

[cephalo]

6 hours ago, Old Foxboy said:

You are captured in a wrong paradigm. The question is not how cold the surrounding air is, but what actually happens at the surface of your craft. The air 1m away from your craft might be like minus 150°C, but your craft squeezes the air in front of it so much that when the air molecules finally hit the surface of your craft, they are compressed so insanely high that the whole front (!) is covered in hot ionized gas.

But this gas begins in an extremely rarified state. There's almost nothing for your craft to compress against. In any case, as your craft presses against the gas, the gas is also pressing against the craft. For every action, there is an equal and opposite reaction. This model seems to defy the conservation of energy. Setting up a high orbit around Jool with a periapsis of 190k would give you a massive source of power that would last millions of years. 2000k of heat can be useful if you're prepared to harness it! It sure seems wrong though.