Kulebron

Two -45 and two -30 in the main stage would probably work, but this adds 500 kg, and reduces thrust (and I run on a blade here ). A small RCS tank, FL-R10, weighs 250, and 4 extra thrusters weigh 200 kg in total, so I won some weight here and kept TWR. Although, if one chooses a more conservative design, the whole RCS deal indeed becomes irrelevant. The spacecraft had some struts, but was still wobbly, because the heavy command module was far, and between it and the bottom modules there were the tiny tanks. (again, because of weight saving). I never liked when guys have obsession with light-weight parts for mountain bikes (i ride a stock one), but here I do the same thing.

I'm learning from your corrections.

Another chapter, the last for this week. How to build and pilot a ship into orbit.

Here's my Duna rocket. 2 landers and the whole ship in 140 tonnes. (But I did not make a rover.) For Ike I use the small "space bicycle", a tank with 2 "saddles". It's weight is balanced, but drag is not, so you have to countersteer in Kerbin ascent a lot. download

You need 2000-2100 m/s for the lander, because apart from getting back into orbit you'll need to rendezvous, and before that to deorbit, soft land (can easily cost 200 m/s even with parachutes) and break to save parachutes too. I was descending with them at 200 m/s when they expanded and one of them tore apart the decoupler it was attached to. I landed on the side. Also you can aerobrake with aerobraking calculator to have apoapsis at Ike's orbit and not need to burn 250 m/s for it. Saves a lot of fuel.

I was trying this "legitimate" option to put two cans side by side, but they were clipping with neighbors, and eventually I chose metal plates and trusses.

I see Tour Eiffel! Très bien! Admirablement!

i always made my landers flat to not tip over on slopes.

A new chapter is out. Click the image to read.

I found figures of construction and fuel costs, not a hard source, but some enthusiast forum. They found the fuel cost about $1 per kg and 1 Soyuz rocket about $17M. (Couldn't find what portion it makes.) So it might be worthy to save about $10M in each launch. The question is at which price does this saving come? What payload can it carry and how much useful delta-V can it have? Also, NASA built custom designs that optimized the hardware a lot (lightweight, high pressure), for that they had to make rovers to carry an upright rocket to the pad. Soviet space program was underfinanced, that's why Soyuz was reused a lot, a bit crude, not the most efficient, but turned out cheap and robust. An empty rocket is supported in 3 points if you watch the launch preparation videos, which was impossible with Saturn or Atlas. And it is debugged pretty well: had only 4 crashes in 1800 flights. As I read the specs of SpaceX rockets, I see they use the same approach to cost reduction and took it further: only 2 stages, less fuel efficient, but I guess, cheap as well.

I don't see any cheating in Engineer plugin, because it eases the routine calculations, and gives some readouts which are not in a convenient shape together. (like apoapsis is seen in the map, but height is shown in camera view) We could have calculated everything by hand, but eventually would have assembled spreadsheets. It does not give you any supernatural abilities. That said, I had to arrange a big spreadsheet because Engineer for me isn't enough! It can't guess all my dockings and undockings. To send a mission to Duna & Ike, with two separate landers, I needed to check if this was better than 1 lander, and that I had enough fuel. Had to make a whole balance sheet for that. After some flying and knowing tricks I don't need Engineer in game, but, hey, Apollo missions already had Flight Engineer in the landing can!

I managed to shoot from Laythe to Kerbin in 1750 m/s, from a highly inclined 60km orbit, then corrected at the cost of 200, but here the forecast is 2160, which is much more.

Kerolyov, that's what I thought of drawing myself! Great map! And I also knew the launch window calculator, a great tool, was going to link it.

Yes, please PM me, because I can't find a proper spellchecker in SublimeText, and English in not my native language. Thanks!

Thanks! RCS lander would be a good idea for Ike, indeed.

Ok, that's clear now. I still have to assemble it from another part, but this is easier, thanks, UbioZur!

Added the Mün mission chapter. RSS feed is working now.

No, I pulled the ship by the root part into drop zone (I have all sorts of things as root part ) and KSP rejected it. If you pull the rest of the ship, yes, it accepts the subassembly. But I use to make a central metal plate as a root part of lander (because I don't know what I'll put around it) with parts on 4 sides of it. It's impractical do dismantle it this way.

Subassembly becomes pretty useless if you can save only a subpart of a ship. It's much move convenient to assemble a rover alone, test and then attach it to a ship. But impossible now. It's much more convenient to assemble a ship and then attach it to a modular expedition. Currently I have to put a stray root part only for the purpose of saving a subassembly. And if I make a ship, I'm in trouble: I have to tear it down, make a new root part and assemble stuff from another point somehow, to make it a subassembly. Makes no sense. I understand it can be a limitation on the logic, but then reorder parts (there are those that have attachment nodes sticking out, make those usable, for instance.)

I'd say there's almost no room for optimization there now. Lander can't get smaller, unless you use EVA pads The narrow 215 kN engine in the middle is better than Poodle. Quads of these engines on the sides of bottom stages are better than Skippers.

I'll write this in my tutorial soon. I flew the coke-can probes to make sure I can do these orbits. I used Node Editor to hit Laythe right on arrival, without orbiting Jool. Then I just made a rocket for all these maneuvers. 2200 for return. (insertion and corrections) 3500 for landing, take off and rendezvous. I don't remember exactly the away trip, but it took about 1950 to shoot to Jool and about 200 for all corrections with the ascent stage (plane matching, then fine-tuninng) I fine-tuned to hit Jool, then dropped the ascent stage and corrected the orbit to intercept Laythe. Then airbraked and circularized. So this big tank on the command pod was enough for corrections on arrival and for return. And then the takeoff stages were built to make about 4500 m/s, to be dropped back and not leave debris. (This is almost exactly the way Apollo missions flew) An important thing was to return directly from Laythe and also to encounter Laythe directly as well. And I got into an inclined orbit aligned with islands. If you correct your path just after Jool encounter, it will cost you 5 m/s (you guess it costs thousands to change inclination when you've arrived). If you want to know if this is doable without plugins, no, you need to edit nodes precisely. But I flew everything manually, from kerbin takeoff to Laythe landing (that's why I aligned it with islands). Getting into orbit or landing with parachutes is easy. Manual Kerbin takeoff is even easier than with MechJeb, because it shakes the rocket like crazy.

Jesus Christ, why make such big ships? Here's my Laythe lander. And the rocket for the whole round trip. https://dl.dropboxusercontent.com/u/6721960/2013/10-15-laythe/Laythe Expedition 3.craft The trickiest thing is to shoot back to Kerbin from Laythe orbit, without descending to Jool. But it pays off, you need just 1800 m/s. forgot the ladder though, had to stand on the engine If you want to know the mission limits and maneuvers, do a reconaissance mission with a probe with 6 coke-cans in Asparagus staging and small engines.

Do you plan to use them without recovering? I saw a video with an SSTO spaceplane, with plugins. Not sure if I have power to do it. I thought of doing an Eve land/takeoff plane, because it's a very challenging task, but I tested some options, and don't see the solution so far. I'm pretty much exhausted with KSP and this tutorial, and the most rewarding part now is publishing it, and I'll take a break for a while. Even when the screenshots are there, and the texts are written, it takes hours to arrange everything, do many minor corrections and find and put images in place.

Thanks! Fixed one, and changed a sentence: "any [landing] can can do" is too tricky

Good points, thanks!