RoboRay

"Best" being highly subjective.

You need a force to capture you into orbit. If the planet has a moon, the moon's gravity can pull you into orbit around the planet (under certain situations), but it's temporary... At some point in the future, you will re-encounter the moon and it will either collide with you, drop your orbit into an impact with the planet, or eject you right back out of the system.

Exactly. The free return actually costs slightly more delta V than a pure Hohmann transfer, but it's safer as you'll just get dumped back into the planet's atmosphere if your engine doesn't fire for orbital capture.

Total time spent at the controls is about the same, since you can timewarp around the various orbits. Your burn-length actually ends up being slightly shorter due to the efficiency gains, but you still have to add some time for the warping around. And you probably only need to make three or four kicks, even for massive vessels, unless you're using ion drives, so it's not as bad as those pictures (from Orbiter) make it look. Which reminds me of one more note to add... Pe kicks do take time. If you only push your Ap out to the Mun, the total time is only a day or so. But if you try to push it out to Minmus or to the fringes of the SOI, your final orbit will take a week. This may delay you past your ideal transfer window, which is why I suggest only pushing Ap out to the Mun... The delta V difference between rising to the Mun and rising to Minmus is tiny, but the much longer orbital period can cause you to miss your transfer. Well, unless you plan for it and start your kicks a week early.

Your burn time is too long for a single-burn injection. If your orbital period is, say 30 minutes, and you need a 20 minute burn to transfer, you're going to have problems. The solution is periapsis kicks, where you break your long burn up into a series of short ones on sequential orbits. Set up your transfer node as normal. When you're two or three minutes from reaching it, start your burn. When you are two or three minutes past it (be sure to burn exactly equal amounts on each side of the node), shut down. Delete the node. Each orbit, burn prograde for a few more minutes each time you pass periapsis (which marks where your maneuver node was). The times can vary orbit to orbit, but you need to do equal amounts on each side of the Pe marker every time. Your apoapsis will rise higher and higher each orbit. When your Ap is about to the Mun's orbit or beyond, set up a new maneuver node at your Pe to make the final transfer burn. Execute that maneuver normally. Your sequence of orbits will look something like this: Pe Kicks are the real best option. They are much more efficient than one long burn as all delta V is delivered right at the ejection angle, where all thrust is aligned with both your prograde vector and your desired trajectory and you maximize the Oberth effect by conducting more of the burn near Pe where you're moving faster.

Those are some nice w̶a̶r̶h̶e̶a̶d̶s̶ impact penetrator probes for, uh, science!

A folding rover is going to be difficult. I just settled for a design that snugs up against the side of the LM and has a small robotic arm to deploy it. Landing: Rover deployed: The descent-stage, with another look at the deployment system: The arm was three rotary motors that swung the rover out to the side of the LM then pivoted it down to the horizontal position before dropping it onto the ground. The extendable instrument platform on the other side was mainly to balance out the mass of the rover. And, just for completeness, the LM ascent stage:

An extremely high orbit does make docking easier because of the lower gravitational gradient (ships drift more slowly relative to each other because of their slightly different orbits), but it's really not worth going all the way out there because you have to double your distance from Kerbin's center to make a 50% difference. Going from a 100km to a 200km orbit does nothing... you'd have to go up to 800km to cut the drift-rate in half. You'd need to go all the way to 2200km to cut it by 75%, and maybe half-way to the Mun to almost eliminate it. But don't try to do it at 75km, either. In a super-low orbit, you'll find yourself dipping into the atmosphere during the rendezvous. Somewhere around 100-150km should be fine. The simplest way to make docking easier is to turn your craft so that the docking ports point Normal to your orbit. This is 90° left or right of your Prograde vector and along the horizon. So, if you're in an equatorial orbit, you turn North/South. If you're in a polar orbit, you'd turn East/West. This prevents the craft from appearing to rotate away from each other and circle around each other as a result of their curved path around the planet. Turning in-line with the Normal axis gives you a stationary target for docking.

Sure, but the point of safety systems is to eliminate or mitigate the possibility of something bad happening. When "it got hairy" it could have gotten bad. It didn't, this time, but that doesn't mean the collision won't ever happen. With an LES that can outrun every other engine on the ship, that risk is close to eliminated. I spent 16 years working in aviation safety (12 years as a part-time aspect of my job, and 4 years full-time), so risk-management techniques are just second nature to me. I treat KSP spacecraft with the same kind of concerns I became accustomed to applying to aircraft in the real world. Risk-management can be looked at as a pretty simple system... Consider what can go wrong. Determine the worst-possible outcome if it does go wrong. Devise a means of eliminating (or at least minimizing) that hazard. Consider any impairment of the mission by that means. Implement the means if impairment of the mission is less significant than the worst-possible outcome of the hazard. An LES tower, if it's lightweight and high-powered, is very challenging to disregard under that type of analysis. Is it really necessary for KSP? No. It's pretty hard to kill the crew in a launch-failure. But, in the real world, nobody would take that kind of chance. And I try to keep things realistic.

The Abort group can't shut down liquid engines on sections that have broken free of the spacecraft, so a high-gee powered escape (rather than tumbling through the debris cloud and hoping nothing hits the pod) is always desirable to me. My orbiters generally have very low TWRs to maximize delta V, pretty much always less than 0.5 gee, making an 8 gee LES option nice to have for outrunning lower stages (or even SRBs) when the fecal matter impacts the rotary air impeller.

Very true. I've actually done pad-aborts in KSP, when a craft started crumbling when a support stand broke as the physics loaded. But I do try to play with realistic safety concerns, so I include an LES on manned flights. I also do a roll and pitch maneuver right off the pad, for range-safety purposes (so any debris will fall into the ocean instead of back down onto the launch site).

No, not really, so long as your first stage doesn't tear itself apart with structural failures, you're probably going to be fine. I'm assuming at some point, though, in Career Mode, random equipment failures will become a possibility. That's when it's going to get interesting... deciding on the fly whether an engine cut-out leaves you with enough thrust to reach orbit, designing craft to remain in-balance during engine-out scenarios, and giving real value to including abort systems to keep your highly-trained crew (and Bob) alive. I use the NovaPunch LES tower, because nothing I've ever managed to cobble together provided sufficient TWR in a reasonable weight and part-count limit. The NovaPunch tower can pull a 3-man pod at 8 gees for one fairly light-weight part (two if you add a nosecone).

The really unfortunate thing would be discarding your abort option before you have another abort option available to replace it, and then needing an abort option. My LES weighs 0.3 tons. The delta V it costs me isn't of major significance. By that argument, it would be smarter to take no LES tower at all. If you take an LES, you're going for realistic safety options and/or historical accuracy. If your goal is simply maximizing efficiency at all expense because you're just playing a game and it doesn't really matter if anyone dies, leave the LES at home.

Don't jettison it based on your altitude... jettison it based on when you no longer need it as an abort option. Apollo dropped it much higher, during the second stage, once they were high and fast enough that the S-IVB and CSM's SPS motor could manage an Abort to Orbit if the S-II failed during the burn. In the event that the S-IVB also failed and the craft was still to low and slow for the CSM to reach orbit alone, the SPS would be fired to maneuver away from the S-IVB and the SM would be jettisoned from the CM for landing. If you're interested in the specifics on Apollo abort modes... http://en.wikipedia.org/wiki/Apollo_abort_modes I jettison my LES tower when the spacecraft has enough delta V to reach orbit on it's own. At that point, I can either bring the crew home in a more controlled and deliberate manner or send up a second launch to provide whatever they need to continue the mission.

I've only ever seen one report here of craft breaking apart, so I'm chalking that up to other problems with the craft design. There are a couple of comments on Spaceport about it, but they sound like the doors are colliding with something while moving.

The key to geostationary orbits isn't altitude, it's the orbital period (or the orbit's semi-major axis, depending on how you want to look at). A perfectly circular orbit at that altitude possess the proper period. However, creating a perfectly circular orbit simply isn't possible (and before somebody posts brag-shots of having it precise to the meter, you were off by centimeters that aren't displayed by the UI). And if you do get a really close one, once you timewarp it's going to change a little due to floating-point errors. Either way, it's not going to stay over the spot where you parked it. You'll have a lot easier time if you install Kerbal Engineer or MechJeb (or some other mod that displays your detailed orbital parameters) and adjust your Ap/Pe to have exactly a six hour period. It's ok if your Ap/Pe are a few kilometers above/below 2868.75km... your satellite will very slowly oscillate eastward and westward over the the same spot of the surface, by too small an amount to even be visible. Real geostationary satellites do the same thing, drifting back and forth across their designated "box" in the sky.

That raises a great point... if your orbit is so low you're skimming along right at the top of the atmosphere, you might find yourself dipping down into the goo during the approach if you deviate too far from a circular orbit. The drag won't be enough to pull you down, but it can be enough to throw off your intercept. But that's just a complication that comes with extremely low orbits. The delta v requirements for a higher orbit are pretty miniscule, and the shallower gravitational gradient at higher altitudes makes docking marginally easier (the craft drift more slowly relative to each other) as well as eliminating the low-altitude issues. The stop-and-go approach isn't wrong... there just aren't many situations where it's the best choice. But extremely low-altitude operations certainly can be one.

You don't need to "finish docking in a few minutes" and it certainly doesn't matter that you're only close to the target for a few minutes if your orbits aren't similar. "A few minutes" is more than enough time to match velocities as you get close to target. Once you match velocities, you've also matched orbits and have all the time in the world for docking. If two people are on opposite sides of a room and want to shake hands, one of them has to start walking. "Making a bulge on one side of your orbit" is just part of the approach. It's not a problem to avoid... it's an essential part of the rendezvous. From what I've seen, the people who struggle the most with rendezvous and docking are the ones who stare at the map and try to maintain perfectly matched orbits. It's completely impossible to approach the target if you have nearly identical orbits. Your goal is to put your closest point of approach as near to the target as you can get it. The shape of your orbit is absolutely irrelevant. Once you have an approach within a few kilometers, close the Map and do not open it again. Every bit of effort you expend trying to make perfect orbits on the Map is completely wasted, because you have to turn right back around and undo that perfect orbit to start moving toward the target again. Once you're within a few kilometers, bring your relative velocity to a comfortable value (depending on your distance, but slow enough that there's no danger of a collision) and keep your yellow prograde marker on the pink target marker. You can steer with lateral RCS or by turning the ship to the side and thrusting with the engines. The NavBall shows the direction of the target, your direction of motion and the relative speed, while the on-screen target marker shows the distance. That's all you need. The Map has nothing useful to you at this point and trying to make perfect orbits on it is not only a distraction from what you actually need to do, it's counterproductive because it's often driving you to do the opposite of what you really need (close the distance). An approach where you steer while drifting in rather than coming to a stop numerous times during the approach to correct is not "hurrying up." It can be just as slow and patient as the stop-and-go method, and it's a good bit more fuel efficient. The closure rate can be as low as you want it to be, and steering by correcting your velocity vector rather than zeroing it and and then building it right back up again can add up to a significant amount of propellant.

Your orbit is fine. To rendezvous, you don't need matched orbits... you just need a close encounter. And you're getting plenty close enough. Once you have an intercept within a few kilometers, close the map and quit even thinking about your orbits, because any differences there are now irrelevant. When you get to the minimum distance, make sure your NavBall is in Target mode and match velocities by aiming at the yellow retrograde marker then burning until your relative velocity reaches zero. From there, you can simply point directly at your target and thrust toward it to begin the docking approach.

"Master Yoda, moving rocks is one thing, but this is a little different."

If you want to argue with how more than a century of aeronautics and a half century of spaceflight has been conducted, that's your business. I'm going to go with "the rest of the world does it right."

Actually, as a pilot and also someone in the avionics business for more than 20 years, there is absolutely such a thing as negative pitch. If you put the nose of a craft 10 degrees below the horizon, you don't say you're in a 350 degree climb. You have -10 pitch. Even in spacecraft, pitch is referenced relative to a plane tangent to the nearest point on the surface of the body you're orbiting, creating an arbitrary "down" orientation as well as an imaginary horizon for the attitude indicator. Pitch is indicated as positive and negative values through 180 degrees and never as a 360 degree circle. That's actually how it works in the real world.

Oops, sorry about that... I must have miss-clicked. It should be working now.

I've been experimenting with ascent profiles on my Apollo/Saturn V and have it just about right where I want it... http://youtu.be/smuZsxxarhg Stage 3 shut-down is in a 100km by 60km orbit, with the spacecraft at periapsis. This requires only a mild 80m/sec or so circularization burn, while staying low (40-60km) for most of the horizontal acceleration to maximize efficiency. (I'm filming a movie that recreates the Apollo 15-17 "J-class" moon missions, which is why I had that video handy.)

Question: Can the mod load scripts from plain text files? I'd much rather generate longer scripts "offline" in Notepad++ than have to type them all up in-game.