How do you handle interplanetary aerobraking?

[Decoherent]

I haven't done any serious interplanetary missions since pre-1.0, so now that I'm moving through a new career (whee!), I'd really like to take advantage of aerobraking when heading to the bodies with atmospheres. I've also discovered that re-entry heat is exciting I'm wondering how people design their craft to successfully aerobrake. I was playing around in sandbox last night, and the only real design I could come up with was mounting a heat shield on the bottom, then using radial engines. I'm not really a fan of this- the radial engines are heavy & wimpy, and the landing legs burned off anyway. I suppose I could waste an engine/stage below the heat shield, which seems wasteful in terms of mass, or do something clever swapping parts in flight with docking ports, but I feel like I'm missing something crucial in my design. What's the current best-practice for these kinds of crafts?

[Plusck]

From the videos I've seen, mount the heatshields on the top, perhaps with a separator+nosecone for the ascent from Kerbin.

[Snark]

Plusck's suggestion of a heatshield on top, with nosecone, works well. You don't even really need a decoupler for the nosecone, just mount it directly to the heatshield and let it burn off when you hit atmosphere. I've done this and it works well.

For bottom-mounted heatshields, I often don't mind sacrificing the engine below the shield: if I'm going to be landing on wherever-it-is, there's a good chance that I'd want a different engine for ascent anyway. Ascent engine = high TWR, good atmospheric Isp. Interplanetary transfer engine = low TWR, high vacuum Isp. So I just design my mission so that the interplanetary transfer stage has just enough dV to get to the destination, and discard it as soon as I hit atmosphere. (Or I skip the mass of a decoupler and just let it burn off.) Then I let the heatshield slow me down to a safe speed, and then decouple it before landing.

If I have a somewhat awkward lander (e.g. with radial bits sticking out that are at risk of getting burned off), I may use a heatshield one size up (e.g. a 3.75m shield on a 2.5m stack) for a little extra "buffer zone".

Edited December 14, 2015 by Snark

[cephalo]

I've recently tried aerobraking on Laythe and Jool, and decided that gravity assist was the better choice. Next time I may try to put things like lander legs inside fairings. I have found that using a heatshield one size up is not effective at really high velocities. Things need to be totally covered up and stowed. Forget about airbrakes lol those pop off immediately.

[OhioBob]

2 hours ago, Snark said:

If I have a somewhat awkward lander (e.g. with radial bits sticking out that are at risk of getting burned off), I may use a heatshield one size up (e.g. a 3.75m shield on a 2.5m stack) for a little extra "buffer zone".

That also makes aerocapture more effective by lowering the ship's ballistic coefficient.  What you don't want to do is load up a heat shield with too much mass behind it.  For example, a 2.5m heat shield with 5, 10 or 20 tonnes behind it will probably aerocapture rather easily.  However, put 30 or 40 tonnes behind it and the aerocapture attempt will be more harrowing with a smaller margin of error.  At least that's the experience that I've had in a series of tests that I've performed.  The higher the loading on the shield, the deeper into the atmosphere you'll have to descend.

[Snark]

8 minutes ago, OhioBob said:

That also makes aerocapture more effective by lowering the ship's ballistic coefficient.  What you don't want to do is load up a heat shield with too much mass behind it.  For example, a 2.5m heat shield with 5, 10 or 20 tonnes behind it will probably aerocapture rather easily.  However, put 30 or 40 tonnes behind it and the aerocapture attempt will be more harrowing with a smaller margin of error.  At least that's the experience that I've had in a series of tests that I've performed.  The higher the loading on the shield, the deeper into the atmosphere you'll have to descend.

That's definitely true.  Also, the higher the loading on the shield, the more ablator you'd better have packed on it.  Ships with "normal" (i.e. not very high) shield loading usually don't need anywhere near a full ablator load, even at interplanetary capture speeds.  If you've gotten into the habit of lowering your ablator to save mass, be sure to crank it up if you have high shield loading.

Anecdote:  The only time I've actually fully used up a heatshield with a full ablator load was when I was deliberately using a screwy design just to see what would happen.  I had a contract to take four tourists out of Kerbin's SoI.  So I built a long, needle-skinny lawn-dart of a ship to do it.  Nosecone, 1.25m heatshield, probe core, two Mk1 crew cabins, two of the long skinny 4-ton LFO tanks, and a Viktor engine, with some AV-R8 winglets on the back for stability.  It came home at about 3500 m/s, all that mass stacked up behind a 1.25m shield.  Had to dive down to the low 20's altitude to capture, and it used up every last ounce of ablator.  Was a bit of a nailbiter towards the end.

[Decoherent]

4 hours ago, Plusck said:

From the videos I've seen, mount the heatshields on the top, perhaps with a separator+nosecone for the ascent from Kerbin.

Hm, I did try this, but I couldn't get it stable on re-entry. I tried adding some small fins, but they made the rocket unstable during ascent (I don't have fairings yet), and the little ones just burn off anyway. Well, maybe I'll just gravity assist anyway. Aerobraking is just so cool...

[OhioBob]

2 hours ago, Snark said:

Also, the higher the loading on the shield, the more ablator you'd better have packed on it.  Ships with "normal" (i.e. not very high) shield loading usually don't need anywhere near a full ablator load, even at interplanetary capture speeds.  If you've gotten into the habit of lowering your ablator to save mass, be sure to crank it up if you have high shield loading.

That's been my observation as well.  After the release of 1.0.5, I ran a bunch of aerocapture tests at Jool, Eve, and Duna to see how the new atmospheres behaved.  Unfortunately I didn't keep notes on how much ablator was used, but my recollection is that it wasn't much more than about 1/2 in the worst case.

 

Edited December 14, 2015 by OhioBob

[Technical Ben]

Now? I don't. I got for mainly powered or gravity assist captures. I've done a few aero captures in Duna etc in a sandbox save. But I needed lots of reloads to avoid overheating. Even with heat shields.

 

PS, though as above, if the speed is not too high, a heat shield will help just fine.

Edited December 14, 2015 by Technical Ben

[FancyMouse]

6 hours ago, Decoherent said:

I was playing around in sandbox last night, and the only real design I could come up with was mounting a heat shield on the bottom, then using radial engines. I'm not really a fan of this- the radial engines are heavy & wimpy, and the landing legs burned off anyway.

How about designing payload first, which has heat shield at top and your favorite engine (mine in most cases Poodle) at the bottom, and then attach the whole payload upside down to the rocket onto a decoupler?

[NathanKell]

I calibrated the shield such that a Mk1-2 pod with a chute and a 2.5m heatshield will just barely survive an entry->landing at Eve at 6km/sec surface speed. That does mean the shields are quite over-spec'd for Jool aerocaptures or Kerbin munar reentries.