Vector

Aerobraking and gravity assist are two separate things entirely. But yeah MechJeb is great for predicting results from aerobraking.

The closest thing I've done is used gravity assist from Eve to change planes and then with a tiny correction get a second encounter with Eve to enter an equatorial orbit. The first encounter was highly inclined because I didn't match Eve's orbital plane (the purpose of the exercise), but the second encounter was equatorial. The key is that small adjustments prior to the fly-by produce large changes after the fly-by, so I think the general method would be to 1. plan first encounter 2. half-orbit or quarter orbit prior to first encounter, play with maneuver nodes to see what you can reach via slingshot, just trying to hit the SOI for a second encounter or get in the neighborhood 3. After entering first SOI but prior to periapsis, further adjust maneuver nodes and orbit to see if you can hit the SOI for the second encounter 4. After first fly-by, after leaving SOI, further adjust maneuver nodes to see if you can minimize periapsis for your second encounter. 5. Half-orbit or quarter orbit prior to your second encounter, play with maneuver nodes to see what options you have for a third encounter. Repeat from step 2.

Very cool idea. May be handy for searching for anomalies or something.

Are you not reducing throttle below 100%? It is very important to reduce throttle and go past the flameout altitude because you can get a lot more power out of those engines at partial throttle until the edge of the atmosphere. Either use MechJeb or place one jet in the center and deactivate the jets symmetrically until you have only one in the center. Even when the air is so thin that you have to throttle down to 5% you can still raise your apoapsis and periapsis.

Have you tried this before jumping to these conclusions? 1) If you do it properly, you can get apoapsis outside the atmosphere with purely jet engines, and then you don't need high rocket TWR to bring your periapsis above the atmosphere. If you start your rocket burn low/slow, then yeah you are fighting a ton of drag and you need high TWR to even maintain the same speed. 2) If you get going fast enough, the centrifugal force counteracts gravity so you don't even need wings at all. I'll admit most of my SSTOs are vertical launch instead of horizontal launch, but even so, with enough intakes it's not a big deal to get apoapsis outside the atmosphere with no wings and no rockets.

If you want to get near orbit with jets, there is a fundamental limit with number of intakes vs. weight. When you're leaving the ground, TWR is what matters, but once you start feathering the engines to prevent flameout, the ratio that matters is intakes to weight. Thrust is proportional to intakes x air density, and drag is proportional to mass * velocity^2 * air density. Your peak velocity will be proportional to velocity^2 = intakes/mass. I can appreciate your desire to keep some realism and not just cram overlapping intakes into the same space. Even with your self-imposed constraint, the intakes/mass ratio still holds, meaning if you want to limit to 4 intakes per engine, you may need to add extra engines (not necessarily extra fuel) to get the intakes/mass ratio up. Your TWR will be crazy high at takeoff, which seems like a waste, but it's worth it because overall payload fraction will be much better if you are able to reach about 30k @ 2100m/s and don't have to burn a ton of rocket fuel to reach orbit.

The rotation of the planet only makes a difference of about 174 m/s, so going east it helps you that much, going west it hurts that much. It's a relatively small difference but unless you have a good reason to go the other way, you might as well go east.

Sure, here's the "Half Can" fueler: https://dl.dropboxusercontent.com/u/25339567/KSP%20Craft%20Files/SSTO%20Half%20Can%20Fueler.craft And I also managed to get a whole orange tank to orbit, though the craft is a bit overkill (405 parts). The engine kept coming off on landing so I went nuts with parachutes. https://dl.dropboxusercontent.com/u/25339567/KSP%20Craft%20Files/SSTO%20Whole%20Can%20Fueler.craft Feel free to modify or optimize, especially the latter could really use it.

I too have been working on SSTO reusable craft to haul fuel to orbit. I'm focusing more on the reusable aspect and not so much on trying to make a space plane. It's quite different, as a vertical launch SSTO coming down on parachutes is really a whole different game. I also allow myself MechJeb, which really only matters for two things: preventing flameout on the way up and making sure I don't land in the ocean on the way down. As you can see I also allow myself 'unrealistic' intake spamming. I had a few attempts that got close to getting an orange tank into orbit (none succeeded). They were beastly with part count, and were extremely difficult to get near-orbital, even when they weren't spontaneously exploding. I scaled back to a half-orange (18t) tank and it's very doable. This one here is 153 parts. Next I might try a "3/4 orange" with one 18t and one 9t stuck together.

I'm late to the conversation and most things have been said already, so I'll speak in terms of two gameplay elements: 1. The discontinuity at the SOI boundaries sometimes bothers me, most often because KSP is sometimes flaky about predicting whether I will intercept the SOI or not. The discontinuity means it matters a great deal whether I touch the SOI or not. 2. And this is tangentially related but has to do with rails vs. physics is that from a practical standpoint it would be really nice to be able to use ion engines with high time warp. If one were to ditch the patched conics for a model that uses numerous time-varying sources of acceleration, it would seem less of a big deal to allow an additional acceleration source. This would open up a whole new class of possible missions using ion engines that are currently infeasible. Also, I did a thought experiment on your Lagrange points, and I think L4 and L5 do not exist because Kerbol does not orbit the mutual center of mass, as you suggested. Under real-life conditions, when the Sun orbits the center of mass, then when you look at the two components of Solar gravity and centrifugal force in a rotating frame, it's shaped like a ring that is slightly higher on one side. The Earth then punches a hole on the high side of the ring, leaving the two remaining peaks. In contrast, since Kerbol is concentric with the rotating frame, the ring would have exactly the same potential all the way around, and when Kerbin punches a hole, it would not leave adjacent high points. However, following this logic, the L3 point should be stable instead of unstable as in real life. I would be interested to see if your model would produce a stable orbit around L3.

Great diagrams. Something else I'll point out, which took me a bit of figuring out, is camber can become tow-out or tow-in if your craft doesn't sit at exactly the same attitude as in the SPH. Even a seemingly mild camber with a small attitude difference can create a terrible tow-out or tow-in that tosses you off the runway and makes great explosions.

I have noticed large variations in the orbit when a craft rotates, as it seems the velocity of the command module (or whichever is 'control from here') is extrapolated to calculate the orbit. Then when physics is switched off, whatever orbit your command module happened to have at that moment is locked in, and that becomes your rails orbit. This can be used to cheat... er... to change your orbit without using any fuel. When physics is turned on, I have the feeling nothing is really stationary. Everything "vibrates" perhaps due to roundoff or perhaps on purpose to give a seed for dynamic instabilities. I believe SAS can even contribute to the vibrations if it is trying to counteract the small perturbations and creating perturbations of its own. I'm sure you have seen the AP/PE markers wobble when very nearly circular, and AN/DN nodes wobble with very low inclinations. These are the vibrations I am talking about. I would say try building a craft with the command module or remote guidance unit as close to the center of mass as you can possibly manage. See if that helps. Edit: I tried this myself, with the center of mass very close to the control, and I still saw apoapsis/periapsis jump when I left time-warp, so it's not that.

Right, this is why I am skeptical of the claim. But separate from whether planets could be inclined, I believe they are not.

Sidereal rotation period determines the speed of rotation but direction is implicit. All planets and bodies rotate about the same axis and all rotate to the east. None have an inclined axis of rotation or rotate to the west. Edit: nearest thing I know of to actual documentation is this: http://forum.kerbalspaceprogram.com/threads/45519 wherein someone claimed "Unity does not support tilted planets." (not sure if I believe it, but it was claimed).

Even better: 434 m/s By moving the inline reaction wheel to the staged part, the TWR on the final stage increases considerably. The peak speed is reached fast enough that there is not enough time to crash. I also tried using MechJeb to more accurately time the staging, but it won't do auto-staging if you haven't left the ground.

418 m/s Observation: it's all about TWR, not because of acceleration, but because air resistance is proportional to mass. The little radial guys are not as good as a SRB when nearly empty, but the tanks can be dropped along the way, unlike SRBs.

Keep in mind things are reversed when going from higher-to-lower orbits (say Kerbin to Eve) as opposed to lower-to-higher orbits (Kerbin to Mun or Kerbin to Duna). I made this error myself and ended up going the wrong way even though I thought I was going to end up prograde. From lower-to-higher orbits with Hohmann transfer (i.e. intersect is tangent to target's orbit), you are going slow and the target is going fast, so the target is approaching you from behind. Prograde and retrograde can both potentially sling you forward, but prograde will sling you outward while retrograde will sling you inward. Look to see which way your orbit is going after the encounter, if it has bent to the left or bent to the right. For higher-to-lower orbits like Kerbin to Eve, the situation is reversed because you are going fast and the target is going slow. You are approaching the target from behind. Prograde or retrograde can both slow you, but prograde will sling you inward while retrograde will sling you outward. If your intersection is not tangent to the target's orbit, then you have some radial velocity which makes it more complicated because you very likely have both radial and tangential velocity. For the radial component alone, if you are moving radially outward relative to your target, then prograde around your target (say Mun) will send you prograde around the parent body (Kerbin), meaning higher Kerbin orbit around after the encounter. Retrograde around your target will send you retrograde around the parent body. But again if you are going from a higher to lower orbit, moving radially inward, then the situation is reversed. Prograde around the target will send you retrograde around the parent body, and retrograde around the target will send you prograde around the parent body. I believe there is also a way to see directly using settings to change how conics are rendered, but I haven't ever used it.

I find this idea interesting, that the path should be the same regardless of the craft. It sounds reasonable that something should be invariant to rocket design and it is definitely not the center/nose marker as a function of altitude. An invariant prograde angle as a function of altitude sounds plausible but I can't prove it. Keeping the prograde marker along a fixed path would mean that for underpowered rockets, the nose would be in between prograde and zenith, with some extra thrust spent on gravity. For a rocket with plenty of thrust, the nose would either be between prograde and horizontal, or it would be precisely at prograde with less than 100% throttle, truly a gravity turn.

Wow, you're right of course. I just tried what you suggested and on the first try, I made it to Mun and almost but not quite all the way back (Kerbin periapsis at 229 km). Reaching orbit was much easier than any of my previous SSTO attempts for some reason. The fuel tank and jet engine, plus four intakes and 28 (4x7) cubic struts all came to just 4.47 tons. I may slap on a bit more than 1.5 t of the bipropellant and the trip should be a breeze. Maybe also an okto2 and it could be a rescue craft or make a nice shuttle.

Hats off to tavert, the best I could do is 14.2, and that's using jets. With rockets I couldn't get under 20 tons. The 48-7S are very good in terms of specific power, so I just stacked lots of them (18 total) using wings as scaffolding to hold them. Flying mostly manually (except for MechJeb "prevent flameout"), I ended up with about 600 m/s to spare, so there's room to make this lighter, esp with an expert pilot. I think I need some more struts too, because this thing spins like crazy just before the jets quit.

Sorry to revive such an old thread (though I think this one is worthy) but is the physics model similar enough for 0.21 for results to be comparable? I am trying to use a different method of optimization and I am unsure of whether to expect the optimum to be the same. I am also having trouble downloading the craft file (if the parts still exist) so if anyone can help out with that it would be much appreciated. If not then maybe there needs to be a new 1D Goddard challenge (my program is not yet able to handle the 2D case... just a simple extension haha).

I try to minimize garbage, usually ditching the last stage just before the periapsis rises above 20km. Even so, I once separated a stage, then went to map view and planned the next part of my mission, and when I burned, I crashed right into the stage I had just dropped. The only other times I have come close to a collision has been when docking carelessly. I aim for as close an encounter as possible and sometimes I have a weak TWR and don't realize it until it's too late.

Here's the double-encounter method, with the goal of minimizing fuel to get roughly equatorial around Eve. It's hard to explain but maybe someone will be able to decipher it. Note that the fuel in the final shot is only very slightly less than after the burn to encounter Eve (slide 8). The slingshot inclination correction is very nearly free if you don't mind a extra trip around Kerbol. The slideshow doesn't show circularizing around Eve, but with the inclination already close to equatorial and the periapsis near the aerobrake range, it should take very little fuel to capture and circularize around Eve.

Sirine, in your 6th image it shows your inclination matches well and you have a nearly equatorial position relative to eve. Your inclination matching in the first image is accurate. If you are heading straight toward Eve, you don't usually want to use the pink nodes. In these situations I fly manually without maneuver nodes. Your prograde indicator will be straight toward Eve, very near the center of the brown area on the nav ball. Orient yourself so that you are aiming prograde and north is up. Then yaw 90 degrees to the right. You should be pointing near the horizon on the nav ball and the nav ball should be telling you 90 degrees "longitude" or whatever. Then slowly burn. This will push you to the right of the planet and the direction of your exit will naturally swing around to point to the left after you pass eve. Since you are lined up straight toward Eve, it should not take much, and this will bring you to be equatorial. Once your orbit is more rounded and not so severely hyperbolic, then the pink nodes are more useful to adjust your inclination.

1. Start off in LKO with full fuel. 2. Leave Kerbin SOI in retrograde direction (relative to Kerbin's orbit around Kerbol) 3. Inclination is 2.1 degrees relative to Eve. Luckily I don't have to wait long to reach AN/DN 4. Adjust normal (pink arrows) until AN/DN nodes get "wobbly". The dashed trajectory will be parallel to Eve's plane. The AN/DN indicators going wobbly are much more accurate than eyeballing the plane. 5. After burning, inclination is zero (pic is actually before I finished matching inclination) 6. With inclination matched, you don't have to aim for AN/DN, you can do Hohmann Transfer anywhere. In this case it's 1 and 2/3 trips around Kerbol and I can get an intercept. 7. After starting Hohmann Transfer 8. Fine-tuning intercept to get periapsis down to 100 km or so. Maneuver node closer to intercept helps me know which side of the planet I'm arriving on. By tweaking radial and normal I can arrive on the right side (prograde) and level with the planet. Tweak pink arrows until periapsis is at a minimum, and you'll be sure to arrive in the plane of orbit. You can lower your periapsis down to 10 km and then fine-tune with the pink normal arrows very accurately, then raise your periapsis back up. In this case I'm aiming for 70km periapsis. If the adjustments are too sensitive, move the maneuver node closer to the intercept. If you are farther away it's more efficient but can be touchy to make small adjustments. 9. Manuever node costs only 0.2 m/s for the fine adjustment 10. I did my burn early. It's not a problem because you can watch the actual periapsis to guide you. 11. Approaching Eve SOI 12. Crossing SOI boundary at warp 13. Periapsis changed from 70km to 46km! 14. Even with warp disturbance, my orbit (hyperbolic) is still precisely in the orbital plane. Cost about 330 m/s to match inclination and 665 for Hohmann transfer. Fine-tune corrections were very very small. Double-intercept (or a slingshot around a moon, if there were a suitable one) can greatly reduce the inclination cost. Without some sort of gravitational help, you are going to have to pay the 330 m/s one way or another, either in deep space by matching Eve's inclination, or pay a lot more once you arrive in Eve's SOI.