Okay, about aerocapture . . .

[CobraA1]

Okay - I'm finally getting to Eve and Jool, and I'm hitting a wall with aerocapture. I'm having ships simply explode. I've resorted to having ships just use lots of fuel, but that's making them large and unwieldy. Would be nice to have some ideas on how to best accomplish aerocapture, as well as good altitudes - is the KSP wiki still accurate on this? Thanks for any help you can give.

[bewing]

At Jool, you can do gravity maneuvers with the large moons to slow you down enough that the standard heatshields will allow you to survive an aerocapture maneuver (at Laythe, for example).

To survive a high-speed aerocapture (Eve is a good example), your only option is the 10m inflatable heatshield. But it can be a little tricky to use. It produces a lot of drag, and you need to build a rocket that can stay behind it without flipping.

Edited July 26, 2016 by bewing

[Blaarkies]

Aerocapture is dangerous and should be planned more rigorously than before re-entry heating was added. Follow what you saw in The Martian, big interplanetary motherships do not aerobrake...they use extra fuel of gravity assists to slow down. Like bewing said, this is easy at Jool by using Tylo(bigger SOI) or even Laythe(tiny SOI). A single slingshot from Tylo(as you approach Jool for the first time) can slow you down so much that you fall straight down into Jool...now that I think about it, there never was any reason to aerocapture at Jool, ever.

You probably want Jool atmospheric science? Just adjust your slingshot to get a Jool Pe of about 199 999m...
Jool has a really, really dangerous atmosphere, dont go deeper than 195km without a heatshield.

Afterwards, time a Tylo encounter to slingshot yourself back up to the orbital height you desire(keep an eye on the Laythe encounters while you are down there)

[CobraA1]

8 minutes ago, bewing said:

At Jool, you can do gravity maneuvers with the large moons to slow you down enough that the standard heatshields will allow you to survive an aerocapture maneuver (at Laythe, for example).

The trick is going to be making sure I can actually be in a position to do so. 

39 minutes ago, bewing said:

To survive a high-speed aerocapture (Eve is a good example), your only option is the 10m inflatable heatshield. But it can be a little tricky to use. It produces a lot of drag, and you need to build a rocket that can stay behind it without flipping.

Returning from Jool to Kerbin is pretty high speed I've found as well. That heatshield is pretty massive for smaller craft, especially if all I have is the part of the ship that's meant to return to Kerbin.

[bewing]

27 minutes ago, CobraA1 said:

The trick is going to be making sure I can actually be in a position to do so. 

AFAIK, it's always pretty trivial to adjust your arrival with a mid-course correction to achieve a Tylo flyby. And to adjust that flyby to suck out a lot of your speed.

[Superfluous J]

37 minutes ago, bewing said:

AFAIK, it's always pretty trivial to adjust your arrival with a mid-course correction to achieve a Tylo flyby. And to adjust that flyby to suck out a lot of your speed.

It is very simple, once you get it. I personally got it by just playing with maneuver nodes until I got a connection.

Use radial and prograde, and do it right after your AN/DN alignment burn. When you understand the concepts involved, you can be a baws and incorporate it into your AN/DN burn. Just get the maneuver node projected orbit to touch Tylo's orbit on the right side of Jool (looking down on North so the moons are going counter clockwise) and keep raising one maneuver node gizmo (-grade or radial) while lowering the other to keep your trajectory passing by Tylo. Eventually you'll get an encounter.

And as is pretty much everything in the game, this is MUCH MUCH easier with PreciseNode installed.

Edited July 26, 2016 by 5thHorseman

[Blaarkies]

1 hour ago, CobraA1 said:

Returning from Jool to Kerbin is pretty high speed I've found as well. That heatshield is pretty massive for smaller craft, especially if all I have is the part of the ship that's meant to return to Kerbin.

Eve -> Kerbin arrives at Kerbin Pe at about 3200m/s. Any heatshield on a balanced craft will survive, just dont go too deep first time...but when arriving at Eve, it is like +-4200m/s. That is where fully filled(max out the ablator slider) heatshield are necesary

[mk1980]

eve is fairly close to kerbin so a capture burn (as opposed to aerocapture) really isn't that expensive when you arrive from a hohmann transfer. something like 300 m/s to get into a highly eccentric orbit. the extra fuel for that burn probaby doesn't weigh much more than the heatshield you'd need for an aerocapture.

of course it also depends on what you actually want to do around eve. if you want to land on gilly, an elliptic eve orbit might even be beneficial, since the correction to intercept gilly should be quite cheap when you do it on the "slow" side of the eve orbit (near AP). if you want to land on eve, you'll probably need a heatshield anyway, even after you're already in orbit (much denser atmo and higher orbital speed than kerbin).

[OhioBob]

Aerocapture at Eve and Jool are definitely possible, but there is a fairly narrow window that you must hit.  There are two main factors that come into play when selecting the correct periapsis altitude for aerocapture - entry velocity and ballistic coefficient. 

I performed a bunch of aerocapture tests last December to try to determine the range of periapsis altitude that will work.  I used test vehicles with a range of ballistic coefficients and tested them at different entry velocities.  Generally speaking, the higher the ballistic coefficient the lower the target periapsis.  Similarly, the higher the entry velocity, the lower the periapsis.  Therefore, if you are approach very fast with a high ballistic coefficient, then you want to set the periapsis on the lowest end of the range.  On the other hand, if you have a realtively low speed encounter and you have a low ballistic coefficient, then you set the periapsis on the highest end of the range.

The combination of high velocity and high ballistic coefficient is the most dangerous and the least likely to succeed.  The tolerances are small, and in some extreme cases I couldn't achieve an aerocapture at all without either burning up or losing aerodynamic control.  Much greater success is possible if you keep the ballistic coefficient as low as practical.

All of my tests at Eve fell within the range of 53 km to 75 km; however, for any particular combination of velocity and BC, the range is much narrower.  For example, for the lowest velocity/BC scenario, the range was about 64-75 km, while for the highest velocity/BC scenario, the range was about 55-57 km.  The median periapsis altitude was about 62 km.

At Jool all of my tests fell within the range of 144 km to 171 km, with a median of about 155 km.  For the lowest velocity/BC scenario, the range was abut 155-171 km, while for the highest velocity/BC scenario, the range was about 145-155 km.  As you can see, a periapsis altitude of 155 km works for almost every scenario, though it's definitely not ideal for the extreme cases.

The above are just some general guidelines.  The best periapsis altitude is determined experimentally.  Just be sure to do a quicksave in case you get it wrong and have to try again.

(edit) By the way, the descriptions in KSP Wiki are based on my experiments.
 

Edited July 27, 2016 by OhioBob

[DaMachinator]

The other option is to install the Inline Ballutes mod and go much higher than you ordinarily would, as they have a lot of drag. I aerocaptured to suborbital trajectory on Duna with one of those from a flyby trajectory with periapse 12km. The craft was of a design such that aerobraking was possible without the 10m inflatable or the ballute, instead attaching heatshields with integrated decouplers to the bottom of the main stack and each of the four radially attached engine+fueltank pods. 

[OhioBob]

2 hours ago, DaMachinator said:

The other option is to install the Inline Ballutes mod and go much higher than you ordinarily would, as they have a lot of drag. I aerocaptured to suborbital trajectory on Duna with one of those from a flyby trajectory with periapse 12km. The craft was of a design such that aerobraking was possible without the 10m inflatable or the ballute, instead attaching heatshields with integrated decouplers to the bottom of the main stack and each of the four radially attached engine+fueltank pods. 

All the tests that I performed were before the inflatable heat shield was available.  My test vehicle consisted of a 2.5m heat shield with varying amounts of fuel tanks stacked up behind it to vary the ballistic coefficient.  I haven't tried the inflatable heat shield, but with that big surface area it should definitely lower the ballistic coefficient and make aerocapture easier and safer.  Of course it is still necessary to target the correct periapsis altitude, which is still best determined experimentally.

[CobraA1]

On lunch break at work, so this will be short. My current biggest aerocapture issue is with Jool to Kerbin. Thanks for all of the advice, it will be helpful. Is there a way to know what the ballistic coefficient is in game?

[OhioBob]

2 hours ago, CobraA1 said:

On lunch break at work, so this will be short. My current biggest aerocapture issue is with Jool to Kerbin. Thanks for all of the advice, it will be helpful. Is there a way to know what the ballistic coefficient is in game?

I didn't test Kerbin aerocapture in the game, but I did play around with some computer simulations.  I came up with periapsis altitudes ranging anywhere from 20 km to 40 km depending on the entry velocity and the BC, though 25-35 km is probably more realistic for most common scenarios.

I don't know of anyway to get the ballistic coefficient prior to flight, but you can get it during flight.  Just open the Debug Toolbar (Alt+F12), click the Physics tab, and check "Display Aero Data GUI".  This will open a panel that displays a bunch of aerodynamic data, including ballistic coefficient.  I usually just run a test flight (Hyperedit is your friend), get the aero info I need, and then revert back to the VAB.

[OhioBob]

Quote

My current biggest aerocapture issue is with Jool to Kerbin.

If you return from Jool at one of the ideal launch windows, you should arrive at Kerbin with a hyperbolic excess velocity of about 2800 m/s.  I didn't simulate that scenario exactly, but I did run some simulations for V∞ = 3000 m/s.  I used three different ballistic coefficients:  2000, 4000 and 6000 kg/m² (measured at Mach 10).  I determined a periapsis range of 27-29 km for BC = 2000, 23-25 km for BC = 4000, and 21-23 km for BC = 6000.  Like I said, these numbers were determined by computer simulation, not by in game testing.  I can't guarantee how well it will work in the game.

(edit)  My recollection is that a Mk1-2 command pod has a BC of about 2000 kg/m² .
 

Edited July 27, 2016 by OhioBob

[CobraA1]

14 hours ago, OhioBob said:

I didn't test Kerbin aerocapture in the game, but I did play around with some computer simulations. 

Simulations, eh? Did you use anything that's publicly available?

14 hours ago, OhioBob said:

I usually just run a test flight (Hyperedit is your friend), get the aero info I need, and then revert back to the VAB.

Well, I'm trying to run a legit campaign here, so I prefer to minimize actual cheating, and building ships does cost money. Although right now, money isn't much of an object, as I have plenty to spend.

I don't consider getting information such as ballistic coefficients to be cheating, though. I'm sure NASA crunches all of the numbers they can for every new spacecraft they build.

21 hours ago, DaMachinator said:

The other option is to install the Inline Ballutes mod and go much higher than you ordinarily would, as they have a lot of drag.

I'm also shying away from most parts mods, because I want to stick close to the intended stock balance.

On 7/26/2016 at 3:45 AM, 5thHorseman said:

And as is pretty much everything in the game, this is MUCH MUCH easier with PreciseNode installed.

Humm, CKAN has Precise Maneuver. I will try that. MechJeb's (yes, I know . . .) built-in maneuver editor seems a finicky.

Lastly, I should note that my current destination is actually Bop, as I have a mission for that. Hopefully all this will help :).

[OhioBob]

2 minutes ago, CobraA1 said:

Simulations, eh? Did you use anything that's publicly available?

No, it's just something that I created in Excel.
 

Quote

 

Well, I'm trying to run a legit campaign here, so I prefer to minimize actual cheating, and building ships does cost money. Although right now, money isn't much of an object, as I have plenty to spend.

I don't consider getting information such as ballistic coefficients to be cheating, though. I'm sure NASA crunches all of the numbers they can for every new spacecraft they build.

 

If you revert to the VAB, it doesn't cost you any money.  I also usually have a sandbox game that I use for testing and experimentation.

I also don't consider running a test to get something like ballistic coefficient to be cheating.  Surely Kerbals have ways to figure this out without having to launch a costly mission, such as wind tunnel tests.  I just consider my hyperedit/revert to be a wind tunnel test, or whatever, using a low cost model rather than an actual spacecraft.  I also figure that Kerbals have computers that can run simulations, so I consider running something like aerocapture tests to determine the correct periapsis altitude just part of the simulations that are run in preparation for an actual mission.  I don't like to cheat, but I'm also not a purist who thinks that any revert/quicksave is inexcusable.  Sometimes it just makes sense that certain data should be attainable without having to fly a full scale mission.

[CobraA1]

2 hours ago, OhioBob said:

I also usually have a sandbox game that I use for testing and experimentation.

Humm, is there an easy way to transfer ships between saves?

[Superfluous J]

38 minutes ago, CobraA1 said:

Humm, is there an easy way to transfer ships between saves?

Yeah all the ships are just files in \<ksp>\saves\<savename>\ships\vab.

Once I even soft linked the directories between a career and sandbox game, though that was more of a pain than it was worth.

[Spricigo]

3 hours ago, CobraA1 said:

 

Well, I'm trying to run a legit campaign here, so I prefer to minimize actual cheating, and building ships does cost money. Although right now, money isn't much of an object, as I have plenty to spend.

I don't consider getting information such as ballistic coefficients to be cheating, though. I'm sure NASA crunches all of the numbers they can for every new spacecraft they build.

Maybe you consider KRASH clean enough.  (However it only allow simulation starting from launchpad/runaway,  autor sugest hyperedit to simulate test from different conditions) 

[CobraA1]

3 hours ago, 5thHorseman said:

Once I even soft linked the directories between a career and sandbox game, though that was more of a pain than it was worth.

I have an app called GoodSync that can copy a bunch of files manually. I'll just do that with a sandbox save.

[DaMachinator]

19 hours ago, OhioBob said:

No, it's just something that I created in Excel.
 

If you revert to the VAB, it doesn't cost you any money.  I also usually have a sandbox game that I use for testing and experimentation.

I also don't consider running a test to get something like ballistic coefficient to be cheating.  Surely Kerbals have ways to figure this out without having to launch a costly mission, such as wind tunnel tests.  I just consider my hyperedit/revert to be a wind tunnel test, or whatever, using a low cost model rather than an actual spacecraft.  I also figure that Kerbals have computers that can run simulations, so I consider running something like aerocapture tests to determine the correct periapsis altitude just part of the simulations that are run in preparation for an actual mission.  I don't like to cheat, but I'm also not a purist who thinks that any revert/quicksave is inexcusable.  Sometimes it just makes sense that certain data should be attainable without having to fly a full scale mission.

There are mods that allow you to run simulation flights, like K.R.A.S.H. 

[CobraA1]

Thanks for all of the advice, everybody! 

Using Tylo for a gravity assist worked. Got my mission completed. 

Still managed to mess up, though.  I designed the craft to only return a single Kerbin, but I sent two out. So one of them had to be rescued. Oh well. I believe that's the first successful mission I've ever had to a moon of Jool and back.  I believe I'm at the point where I'm comfortable with interplanetary missions. Again - thanks!