Design Problem

[Zapo147]

First off, it's not my first rodeo in KSP. However, I found that one of my greatest challenges for interplanetary missions is designing a ship that is aerodynamically stable while performing an aerocapture maneuver. The primary reason for this is that rockets are naturally bottom-heavy, especially when launched from KSC. So while one can easily place an inflatable heat shield between the launch abort system and the command module, the center of mass will be so far back that the ship will tend to flip unless you add a ton of drag to the back end, in the form of literally tons of fins or more creative means.

That was my first instinct, and I ended up with various semi-successful designs including turbine-like fins and infernal robotics deployed drag-inducers (heat shields). This solution isn't consistent, so I started designing ships to aerobrake engine-side first. Designs, using KIS/KAS, included using winches to haul a massive heat shield in front of the engines, assembling a heat shield in situ, or robotic arms moving radial shields into a heat shield snow plow design. Then came my most successful design concept - separation. I started designing ships to separate into two equally massive halves and aerocapture separately. Then would then rendezvous and rejoin after they left the atmosphere. The two halves were always much more stable than attempting to aero break with my typical burj khalifa looking design.

My question is: what have been your successful solutions? I have a sneaking suspicion that there is a more elegant one out there.

[The Space Dino]

I find it a challenge to keep all my rockets stable, especially with the limited parts in career mode. Gotta blame my bad piloting skills for that. 

[DerekL1963]

4 hours ago, Zapo147 said:

My question is: what have been your successful solutions?

To not use aerobraking...

[HighFructose]

I've seen someone put an inflatable heat shield in the front, and have 3 or 4 inflatables in the back. The inflatables induce a lot of drag, so it keeps the ship stable. I've never actually aerobraked, because I always forget heat shields, but this is how I'll probably do it once I go do Jool-5 or a grand tour.

[Dark Lion]

@Zapo147 Your question may need to be a little more specific for best results. Firstly, when exactly is it so important to keep said burj khalifa stable during the aerobrake? Is the issue related to accuracy of trajectories or are you not satisfied with the deltaV equivalent of the aerobraking maneuver because of the tumble through atmo?

[qzgy]

My large interplanetary shuttle aerobrakes quite efficiently, long as you don't yaw to hard.

 

[Zapo147]

18 minutes ago, Dark Lion said:

@Zapo147 Your question may need to be a little more specific for best results. Firstly, when exactly is it so important to keep said burj khalifa stable during the aerobrake? Is the issue related to accuracy of trajectories or are you not satisfied with the deltaV equivalent of the aerobraking maneuver because of the tumble through atmo?

I'll try to clarify. Aerobreaking saves crazy amounts of dV, so I want to use it for capture, particularly when arriving in the Joolian system. The problem is to design a ship to be able to enter the atmosphere at very high speeds and not break apart. The way that's done is by placing the center of drag well behind the center of mass while only exposing heat shields, which is not normally the case for a large upper stage. There are many creative ways to accomplish this, some in the VAB, some in situ. My favorite method, after a lot of struggle, is to break the ship into two halves and each half is aerodynamically stable and protected from reentry heat. I'm wondering how other people accomplish stability while aerobreaking.

[qzgy]

1 hour ago, Zapo147 said:

I'll try to clarify. Aerobreaking saves crazy amounts of dV, so I want to use it for capture, particularly when arriving in the Joolian system. The problem is to design a ship to be able to enter the atmosphere at very high speeds and not break apart. The way that's done is by placing the center of drag well behind the center of mass while only exposing heat shields, which is not normally the case for a large upper stage. There are many creative ways to accomplish this, some in the VAB, some in situ. My favorite method, after a lot of struggle, is to break the ship into two halves and each half is aerodynamically stable and protected from reentry heat. I'm wondering how other people accomplish stability while aerobreaking.

Ahh, Jool aerobrake. Don't. Really, its a lot better to use a tylo gravity assist to capture into a jool orbit. Much easier and doesn't have a chance of explosions happening.

Tutorial from the KSP Wiki: https://wiki.kerbalspaceprogram.com/wiki/Tutorial:_Gravity_Assist

[Zapo147]

10 minutes ago, qzgy said:

Ahh, Jool aerobrake. Don't. Really, its a lot better to use a tylo gravity assist to capture into a jool orbit. Much easier and doesn't have a chance of explosions happening.

Ah yes! I meant to learn to do gravity assists for a long time now, and kept putting it off. I'm actually glad I've left myself something this important to learn still. It doesn't seem too difficult to learn to use. What about Duna? Do you typically gravity break with Ike or do you aerobrake or do you just brute force it?

Edited August 19, 2017 by Zapo147

[qzgy]

10 minutes ago, Zapo147 said:

Ah yes! I meant to learn to do gravity assists for a long time now, and kept putting it off. I'm actually glad I've left myself something this important to learn still. It doesn't seem too difficult to learn to use. What about Duna? Do you typically gravity break with Ike or do you aerobrake or do you just brute force it?

TBH, I think its easier to either just brute force and capture with a burn or to aerobrake. Both would work, but capturing with a burn is easier, and more accurate. With that shuttle I posted above, I actually managed to capture into a suborbital trajectory using an aerobrake. Which wasn't bad, but a bit annoying. And capturing with just brute force doesn't require anything too complex in terms of craft design. Just more fuel.

Edited August 19, 2017 by qzgy

[Dark Lion]

@Zapo147Now to answer much more directly:

My most successful solution is: Don't force efficiency before effectiveness. Consider orbital mechanics are actually much easier to calculate without aero-brake maneuver variables. More dependable physics can be calculated from KSP's "on-rails" solar system which only effects your ship one celestial body at a time, and thus, much more simplistic equations are readily available for your use. A tumbling spacecraft appears to be the least of your concerns, however...

3 hours ago, Zapo147 said:

Ah yes! I meant to learn to do gravity assists for a long time now, and kept putting it off

Some things to study up on:
~Orbital resonance (relevance: timing and precision for encounters)
~Phase angles (relevance: fuel-saving encounters)
~Gravity assist (relevance: economy interplanetary transfer/capture)
~Oberth effect (relevance: maximizing engine effectiveness)

I highly advocate the use of infinite fuel to practice some desirable captures and some note-taking to determine the most effective altitudes and such to capture from an interplanetary transfer. This would almost certainly help you retain what you've learned by turning theory into good solid practice and teach you which world offers the most deltaV savings, which worlds help the least and how to make the most of a bad situation when the fuel light does come back into play. Quite literally, get mad and throw your ship around Kerbol from world to world until you're confident you fully understand all you've read up on. By the time you're confident with your understanding of how to use gravity to your advantage, you will also very likely understand what the best work-around is for your craft, specific to its given objective, or if that craft is truly viable on said task.

3 hours ago, qzgy said:

Ahh, Jool aerobrake. Don't. Really, its a lot better to use a tylo gravity assist to capture into a jool orbit. Much easier and doesn't have a chance of explosions happening.

Tylo can be friend or foe. It depends entirely upon relative phase angles at the time of your encounter at Jool. One position of Tylo may send you interstellar, another may suck you right down into Jool's green death. The latter option still allows you an opportunity for that beloved aero-capture-- though it may require a correction burn first to keep you from diving too steeply (which may or may not turn out to be more efficient than brute force captures depending upon many variables like starting altitude and velocity.) The point being, while it is indeed some sound advice, I wouldn't follow @qzgy's advice to the letter, as sometimes an aero-brake at Jool can save you fuel, other times it can save your mission. But yes... if you come in straight for Jool aero-braking, you're hot on the heels of Death, not efficiency. You'd better have brought all your celestial armor (ablator) for that one. Even then, good luck on surviving the G-force if you delve too deeply into those toxic-looking Joolian clouds...

Wow, I really went on a ramble there, sorry. Hopefully, at least some of this has been helpful to you. Can you tell I once struggled with this as well?  Good luck!

[Zapo147]

@Dark LionWhat does orbital resonance have to do with this? I'm fully aware of the relevance of the other 3. Like I said, it's not my first rodeo. I've been everywhere in the stock system. I just have either used aerobreaking or used high efficiency probes and brute forced it on the far rocky worlds. I've got over 100 mods installed and play mostly in IVA. That's part of the reason I've never mastered gravity assists...I don't use maneuver nodes anymore! Not sure how I could plan a precise assist IVA.....I'll think on that. Meanwhile I've been practicing my Mun-Duna assist. Turns out a 3km high powered gravity assist saves a lot of dV thanks to the Oberth effect but ended in a quick mission LOL.

Edited August 19, 2017 by Zapo147

[qzgy]

50 minutes ago, Zapo147 said:

play mostly in IVA.

Whaa.... You madman!

No, its hard to do precise orbital mechanics,sch as gravity assists in IVA mode.

 

Also @Dark Lion very useful points and very good rant. You do make a point about not following my advice to the letter, there can be cases where its super annoying to use tylo to assist. I would still say that at least for likelihood of death, tylo is much lower than an interplanetary speed aerobrake at jool. Actually, I myself could still learn some stuff, especially orbital resonance.

[Zapo147]

6 minutes ago, qzgy said:

Whaa.... You madman!

Innit purrty? Forgive my use of mechjeb here, I'm testing the new assets for the creator.

[qzgy]

2 minutes ago, Zapo147 said:

Innit purrty? Forgive my use of mechjeb here, I'm testing the new assets for the creator.

Oh yes, very. RPM does add quite a lot to the functionality of IVA's, and the prettiness.

[Dark Lion]

I had to check and make sure I didn't mis-name it, but yes, orbital resonance is basically the timing in which celestial bodies pass one another by a certain distance, in any case that the two orbits intersect. In this case, for rendezvous/encounters in KSP, we'll consider the craft a celestial body, but one which cannot exert any gravitational force upon the encountered. For a very basic example: an orbital period which lasts twice the length of time to orbit Kerbol as Kerbin, yet still intersects Kerbin's sphere of influence would be considered a 1:2 orbital resonance at Kerbin, the moment that orbit is stable.

In order to best utilize every drop of fuel, sometimes it's necessary to use the same celestial body for gravity assists to cancel or add orbital velocity relative to Kerbol. When this is the case, you'll need to understand the appropriate resonance to both plan and time your desired course. And because KSP doesn't have the physics to support multiple bodies' gravitational effects across an entire star system, you won't have to factor in the effects upon anything more than your craft itself (or asteroids, if you're into that) as we should know by now, the "on-rails" system doesn't allow you to move any moon, planet or anything, even a little bit. Laws of Conservation of Energy do not apply to anything other than what you've put there yourself.

@Zapo147 @qzgy I hope this helps! 

[Zapo147]

3 hours ago, Dark Lion said:

I hope this helps!

It helps me understand that there's something I don't understand and should take the time to learn haha. So knowing what would constitute (the parameters of) a resonant orbit with the body I'm leaving can help me use that same body in a much later gravity assist to get to my final destination? So, for example, escaping Kerbin with a big burn and getting into that 1:2 orbital resonance, then intercepting Kerbin after exactly 2 years and using it to gravity assist me to all the way to, say, Moho? And all I need to know is the orbital period right? If I make sure it's 2 years then I'm guaranteed to be in resonance and intercept Kerbin again in 2 years right? Thanks for taking the time.

[septemberWaves]

If you're still wondering how to design interplanetary vehicles for effective aerobraking, my advice is to assemble them in orbit and put the propulsion module (usually the most massive part) at the front. If you look at the (now extremely outdated, but still relevant to this discussion) vehicle pictured below, you'll see that the engines and heat shields are at the same end of the vehicle, and the much-less-massive crew compartments are at the back behind the cargo bay. I designed it this way so that it's not likely to flip during aerobraking (this vehicle is designed for missions to Duna or a modded planet with a similarly-thin atmosphere, for gas planets I'd be using wider heat shields but the principle remains the same).

This sort of vehicle is incredibly difficult to assemble with a single launch. It can be done, of course, but it's not practical at all. If I recall correctly the full mass of this interplanetary transport (including the contents of the cargo bay) was around 250 tonnes (give or take 25 tonnes), and that's not an easy mass to launch. That on top of the fact that the engines must be on outriggers like that so that they can be protected by the heat shields while still being at the front of the vehicle means that it's far more practical to assemble a spacecraft like this in orbit. There are seven modules on this one, but I probably could've condensed that to be fewer modules if I'd really wanted to.

Edited August 21, 2017 by eloquentJane

[Dark Lion]

@Zapo147 I believe you've got the idea about orbital resonance!

@eloquentJane As aero-braking maneuvers are meant to conserve fuel for those lenghty missions, I have to wonder how that transpo of yours doesn't flip retro when the tanks are near empty and the CoM shifts back towards the lab module. It may be out-dated and impractical but it's still an interesting design nonetheless!

[septemberWaves]

12 minutes ago, Dark Lion said:

I have to wonder how that transpo of yours doesn't flip retro when the tanks are near empty and the CoM shifts back towards the lab module.

Actually it doesn't. The nuclear engines are the quad ones from Modular Rocket Systems, which have both four times the thrust and four times the mass of the stock equivalents (with the same I_sp). The fuel tanks are never entirely empty during aerobrake maneuvers; the lowest they go is about 10 to 15% before a Kerbin aerobrake (this is partly because there is some extra fuel required for circularizing back in Kerbin orbit, and partly because I always like to leave a reasonable margin in case anything goes wrong). And the payload bay is also rather massive even when its tanks (which are LFO because they're for refueling landers and other LFO payloads) are empty. And the crew modules are so much less massive by comparison to the rest of the vehicle that the center of mass will always be in the front half of the vehicle even when all tanks are empty.

 

16 minutes ago, Dark Lion said:

It may be out-dated and impractical but it's still an interesting design nonetheless!

When I talked about impracticality I was referring to the potential of launching a spacecraft of this size in a single launch. I was intending to contrast that with how it is actually launched in several separate modules and assembled in orbit, which is a very practical method of doing it - this being the suggestion for the OP of this thread for designing their own interplanetary vehicles capable of aerobraking. And the spacecraft I showed actually worked extremely well, it's one of my most successful designs for an all-in-one Duna exploration vehicle.

[Johnny Wishbone]

On 8/19/2017 at 6:14 PM, Zapo147 said:

Innit purrty? Forgive my use of mechjeb here, I'm testing the new assets for the creator.

I found it amusing you had your mouse over the abort switch pretty much the entire launch. Have a little faith, man!

[Zapo147]

2 hours ago, Johnny Wishbone said:

I found it amusing you had your mouse over the abort switch pretty much the entire launch. Have a little faith, man!

hahahaha you neever know. Not letting Jeb die on my watch!

20 hours ago, eloquentJane said:

If you look at the (now extremely outdated, but still relevant to this discussion) vehicle pictured below, you'll see that the engines and heat shields are at the same end of the vehicle

I think this is exactly the solution I was looking for, very elegant, thank you. What are your thoughts on aerobreaking vs gravity assist breaking from a moon?

[MR L A]

On 19/08/2017 at 11:14 PM, Zapo147 said:

Innit purrty? Forgive my use of mechjeb here, I'm testing the new assets for the creator.

This was cool! What mod is that?

Also, I'm sorry, but that ascent profile was pretty... horrifically steep  any particular reason for that? There's an amazing guide on here somewhere for maximum efficiency... the turn starts really early and you pretty much burn all the way to orbit (which is hella sexeh)  

[septemberWaves]

5 minutes ago, Zapo147 said:

I think this is exactly the solution I was looking for, very elegant, thank you. What are your thoughts on aerobreaking vs gravity assist breaking from a moon?

I hope it helps you out. It can be a pain to orient propulsion modules like that though; unless you have a mod like Mk2 Expansion (which is what I use) which adds angle-locked docking ports, you'd have to do multi-port docking to ensure their correct alignment, which is something I know many people struggle with.

 

As for using gravity assists vs aerobrakes, it depends on the situation. Obviously you're not going to be getting a gravity assist from a moon like Gilly, so in some cases aerobraking is your only option for saving fuel when capturing in a planet's orbit (and of course some places have neither an atmosphere nor moons to help out). When there is a moon which can help out and an atmosphere to brake with, I have a few preferences based on what's the most practical way I've found to do things.

At Jool, it's obviously best to use Tylo for a gravity assist because it's so massive; if you plan things well enough you can probably get several consecutive assists from it with minimal fuel expended, which is ideal. And plus, you don't have to design a craft for braking in the Joolian atmosphere (which is far harder than aerobraking at Duna or even Eve due to the pressure and the speeds involved) and nor do you have to expend additional fuel raising your orbit after aerobraking (though there often is a need to move it out of the way of Tylo). Plus, if you do the maneuver to set up the gravity capture a few years before entering Jool's SOI, it's easy to get an encounter with Tylo with only a few cm/s of delta-v provided by RCS (and of course if you need more precision you can do a correction closer to the destination).

At Duna, by comparison, I'm not entirely sure if Ike is massive enough to capture a vehicle in Duna's orbit, but in any case I think I'll always prefer to aerobrake at Duna. The reasoning for this is that Duna's atmosphere has such low pressure that it's not difficult at all to protect a spacecraft for aerobraking (the one I pictured earlier would probably have lost its solar panels and radiators if I tried to brake it in Jool's atmosphere, but it's fine at Duna). If you're at the stock scale you generally don't even need heat shields; I'd include them anyway though for aerobraking back at Kerbin. And the delta-v required to raise your orbit out of the atmosphere once captured is generally not even enough that you need to consider it if you give yourself a reasonable margin for error on delta-v for a mission.

In general (and this applies to modded planets and scaled-up versions of the stock system too), I'd say it's best to use a gravity assist to capture in orbit of a gas planet when possible (though this isn't possible for some modded ones, like Neidon from OPM) or to capture in orbit of planets with extremely large moons (I think it may be possible in some cases to capture in orbit of Plock (again from OPM) using its moon, but I haven't tested this). But otherwise, if the planet has an atmosphere, use that to aerobrake*.

*That said, sometimes an atmosphere is so thin that it's not really doable (like one of Neidon's moons) and other times even a terrestrial planet has such a ridiculously dense atmosphere that designing a craft for aerobraking becomes far more difficult.

[Zapo147]

9 minutes ago, eloquentJane said:

you'd have to do multi-port docking to ensure their correct alignment, which is something I know many people struggle with.

DPAI has an alignment gauge right at the top, and I also always install Thrust Controlled Avionics, so even if was way off it would only decrease efficiency.

25 minutes ago, MR L A said:

Also, I'm sorry, but that ascent profile was pretty... horrifically steep  any particular reason for that?

I wasn't flying it, it was stock mechjeb's ascent profile. I could edit the profile and see if I gain efficiency.