Ninety-Three

KSP's heating system is pretty basic, heat is shared and divided to each neighboring part. If you're having overheating problems, they can be solved by sticking on a few of those super light octo-strut pieces as heatsinks. A further explanation of drag: Your ship simply behaves as though it has a drag coefficient equal to the weighted (by mass) average of the drag coefficients of its parts. And nose cones, for some reason, are given the highest drag value that parts get, making them worse than dead weight.

As mentioned in the imgur, there are 248 large solids, 46 small ones, and six tons of parachutes. Neglecting clamps and struts, it comes out to 2140 tons. I'm afraid that Flight Engineer would die a horrible death if I tried to make it read the ship, so that's as close as I'll get. I have literally never had the collision problem. Dropping burning SRBs is actually very safe, as it's only in vacuum, and the ship is no longer under thrust when it drops. So long as you don't do anything silly like start spinning around before you drop them, I can't imagine it going badly. Greater than 15? Judging by the re-entry screenshots, I'm losing about five hundred m/s per second. Kerbals are made of some strong stuff.

Short version: I'm not quite sure. Long version: The front of the ship was _covered_ in RCS thruster ports that were there to reduce drag. My best guess is that KSP's weird water physics responded well to a low drag, high-speed impact. I hit at well over 1000 m/s, it's also possible that I was moving so fast that I clipped through the water's collision frame. Sadly KSP's acceleration measuring thing is really non-functional, but based on the timestamps of some of the screenshots I got, I lost somewhere in the vicinity of 2000 m/s in a second. Kerbals are made of strong stuff.

I've been told that the link to the craft file was broken, here's a Mediafire link which should work.

Here's my entry. Depending on how you count it, I was 13 seconds from atmosphere to water, or 38 seconds from atmosphere to complete rest. The imgur has slightly more pics, but here's the three main confirmation ones. If we're being scored on coming to a complete rest, I could re-run it a couple times trying to time my parachutes better and probably get a much better score. For those interested, my technique for getting so much speed was to fling a mainsail and one hundred-some tons of fuel nearly out of Kerbin's SOI, then at AP, point them at Kerbin and burn. The launcher was an ugly as sin lightly-asparagus-staged mainsail monstrosity that totaled about 2000 tons of fuel. Someone with more patience could get a higher re-entry speed using an atomic engine, but really the big place to save time is on surfacing from diving underwater. Also, I'll say it here for those who didn't check the imgur: The secret is covering the command pod with a few tons of RCS thrusters (drag coefficient = 0.001), lowering its overall drag coefficient to 0.048 Entering the atmosphere 200 meters underwater, in just 13 seconds. Using parachutes to not end up quite so deep underwater, coming to surface rest 38s after atmosphere.

14000 is about what you get out of a single garden variety SRB with a probe core strapped to it. My approach for getting crazy altitudes was asparagus-staged Mainsails with 4 tons of fuel each, keeping TWR at a constant value equivalent to one mainsail and its 4 tons of fuel. The approach that the highest-scoring people are using is partially-burned SRBs. Consider a craft with three layers of SRBs, the top layer with 1, the middle layer with 10, and the bottom with 100. At T minus 20, you ignite the lower stage. At T minus 10 you ignite the middle stage. At T-0 you ignite the upper stage, release the docking clamps and take off. Your TWR is mainly set by 100 SRBs which, for 10 seconds, will have between 1/3 and 0 of their fuel left, making them very light and thus, very powerful. At +10 you stage out, dropping the lower layer that has just run out of fuel. Now the middle layer is determining most of your thrust, and its SRBs have ten seconds of life left; you can see where this is going. The actual ships use more stages with less extreme ratios than 10-1, but this should illustrate the principle well enough.

I'm a little baffled by this. The train must both be on rails, and move freely? And how exactly are we to construct 1800 m of rails out of stock parts?

I don't think I'll be doing this, as I'm terrible at anything involving lift, but I've always theorized that a "ship" which was 80% control surfaces (and 20% ion drive for circularization) should be able to get to orbit completely stock. I've seen silly stuff done, like leaving Kerbin's atmosphere in half a minute, and my understanding is that control surface flight should be even more powerful on Eve, since despite the gravity being higher, the atmosphere is way thicker, multiplying the control surface power. Would anyone who knows more about spaceplanes care to comment on this theory? If it really is plausible, I might have to teach myself some plane design and put it to work.

During the building of my 48 stage monster, when it was only 20ish stages and less, I was waiting between stages a bit to try to keep my speed at the optimal ascent profile (going by this handy post). As the ship grew, it ended up wanting to fire each stage the moment the last one depleted. For a really large heavy ship, you should be able to end up pretty close to optimal ascent with the heavy early stages, which means more or less immediate separation, and by the time you reach the light upper stages, you'll be in high atmosphere where it's always optimal to burn everything as quickly as possible.

So my dreams of leaving Kerbin's SOI turned out to be a bit farfetched, but I beat Kolt by an order of magnitude. The craft is a little silly. Here's the breakdown on it, for anyone that's curious: 1192 separatrons, divided into 48 stages. Stages, topdown: 8*2, 6*4, 4*6, 3*8, 2*10, 2*12, 3*16, 2*20, 3*24, 1*28, 2*32, 2*40, 1*48, 2*56, 2*64, 2*80, 2*88, 1*104. I launched it every couple stages as I was building, getting the following heights: 2: 5200 3: 7200 5: 10800 8: 17000 10: 29000 12: 43500 14: 63000 16: 80000 20: 124 km 25: 197 km 28: 299 km 33: 499 km 36: 650 km 38: 838 km 43: 1486 km 48: 2387 km It also came out a bit slower than the optimal ascent profile, but by the time that was showing, I was rather committed to it. A redesign with slightly higher TWR could do better.

People in this thread seem to be making the mistake of using way too many separatrons per stage, and way too few stages. With about 300 separatrons divided into 33 stages (the top 8 stages having only 2 separatrons each), I've managed to hit 500 km, and the ship's not even starting to lag. I'm continuing to build, I wouldn't be surprised if this thing could leave Kerbin's SOI before it gets CPU-killing big. I'll post something soon.

I'm familiar with the basic principle behind infinigliders: control surfaces that make up a significant percentage of your plane's weight can lift it off without thrust. I'm trying to make an infiniglider that can put itself into orbit (with a little bit of ion assistance at the end), but I'm not very good with even regular planes. Is there anyone who has ever done a serious investigation of how to build efficient infinigliders that could offer me some advice?

Per request, here is the craft file. I put a bunch of mission photos up in the imgur linked in my first post, I had some more that I'll put up tomorrow which mostly show more detail on the burns I performed. Computer Surgeon General's Warning: This is a 1284 part craft, with about 400 large parts to it. Side effects of excessive SRBs include lag, lag, and OH GOD THE LAG. Professional computer on a closed course, do not attempt. Edit: I've been told the above link to the craft file is broken. Try this Mediafire link instead.

That SRB Mania thing is a great idea. I took it as an excuse to do something very silly. This is my ship. It's going to Duna. And back. Yes it's all solid. I've written up the journey here. I wasn't really trying for max points, which would have probably involved visiting Minmus and the Mun in one trip, but I sure had fun.

After a slight paperwork mixup, KSC high command accidentally approved one of the ships Jeb designed himself. He was pretty ecstatic. Here he is on Duna. And here's the return, the last stage of the rocket having accidentally exploded upon impact. I documented and wrote up the whole journey here, check it out! Edit: Per request, here is the craft file. Computer Surgeon General's Warning: This is a 1284 part craft, with about 400 large parts to it. Side effects of excessive SRBs include lag, lag, and OH GOD THE LAG. Professional computer on a closed course, do not attempt. Edit 2: I've been told the above link to the craft file is broken. Try this Mediafire link instead.

Are there any goals to this landing, like "Land at a survivable speed" or "Land something that can return to Kerbin"? I'm interested, but "Spend fifteen minutes burning, half an hour later slam into the Mun at ten thousand m/s" seems a little dull even accounting for all the waiting involved.

I assume the main method of reaching the Mun will be to wait until the Mun is directly above the KSC, then burn straight towards it, reaching the Mun in around an hour (less if you go really crazy on your launcher).

I'm about to try my first mission that involves aerobraking on Duna, and since I don't want to figure out how to do the terminal velocity calculations myself, I thought I'd ask the forums: About how much parachute drag/ton do I need to safely land something on Duna? If someone has a rough sense, but no numbers, I'd appreciate even just a description like "A little bit less than on Kerbin". Also, because Duna's atmosphere is thinner, am I right in thinking that abruptly deploying parachutes are much less likely to tear themselves off the ship than they are on Kerbin?

You may want to specify that this has to take place on Kerbin, as other bodies have varying amounts of friction. I think the Minmus salt flats are actually frictionless (if they're not, they're close enough that I mistook them for it).

By "Small and cheap", do you mean part count, physical volume, or the literal (and extremely arbitrary) cost numbers for parts?

Oh, we can get points for multiple landings? Sign me up for a 0 point trip to the Jool system then. If we dock with something in orbit, do we pay the part count for just the orbital stage, or do we also get points for its launcher? And, are we allowed only one docking, or several?

Where's your Jeb spirit? You tell me to get to the Mun as fast as possible and I immediately picture slamming into it at 10 kilometers a second.

That's six orange fuel tanks, three mainsails, and a decoupler. No asparagus staging there, can't afford the parts! It took the payload to within 100 m/s of Kerbin orbit.

Your second post says there's no prize for splatting on the surface, but that doesn't seem to be anywhere in the main post. If that's the case, can you add it?

I'm not familiar with the Apollo mission plan, but since I ditched the stage with landing legs before I had my Minmus intercept, I'm pretty sure I didn't follow it. That said, I did manage to get to Minmus and back with a 10-part launcher. Score: -5 (0) A full accounting of my trip can be found here. I managed to make the 920 dV return trip on just 240 m/s thanks to Munar assists. I'm not sure if that's an impressive number or not, but I'm sure proud. I made a slight change to the launcher, giving triplicate symmetry to the ladder and probe core because I'm OCD about center of mass, and then switching the generator over to symmetrical solar panels. The overall weight got slightly larger, so it seems like a fair modification. Doing a trip to Minmus is actually easier than the Mun: It's only slightly more dV to reach the SOI, and way less to land due to the lower gravity. You might want to swap their point values to reflect this. Edited to add: This was done completely stock. I'm morbidly curious as to how an autopilot mod would handle that wobblemonster.