Jump to content

Stars to stones: my nanocristalline diamond caveman attempt

king of nowhere

Recommended Posts

 If I try a challenge, I try it at the highest level.

One year ago, i considered the caveman challenge - and of course, I had to do it at nanocristalline diamond level. I gave up soon, stating "too much grinding, I don't like it".

Instead, I went on and did a bunch of large scale missions with kerbalism. The last one was 320-year long. Greenhouses require manual harvesting twice every year, and I had 19, so i clicked "harvest" over 12000 times. The mission also required manually servicing the nuclear reactors every three years, again, that's over 100 times where I had to go EVA and fly around the ship. Afer that, I felt I really can't complain about grinding.

So here I am, trying the nanocristalline diamond caveman.

Part 1: the low-hanging fruits

Getting initial science from Kerbin and its orbit


Normal beginning with a roller to get science from the ksc biomes. I got 5 science, I teched engineering 101, to get the thermometer. Made another roller with the thermometer, got another 5 science.

At this point, I took basic rocketry. Maybe it would be possible and effective to harvest more science and get the barometer too, but with basic rocketry i can get to orbit. and in orbit I can at least get my 5 science back. Got some early funding mostly by testing stuff on the ground.

So this is my first orbiter: a tested and true design.


I'm playing it safe here. It is possible to reach orbit - even high space - with only 14 fuel tanks, but I put in 15 because I know I am unlikely to hit an optimal launch profile the first time. And it is possible to avoid the decoupler, keep the whole rocket together, and recover it all. But a small mistake in reentry profile would flip the rocket, which would then have too little drag to brake properly, and it would kill Jeb. I can make more money with some time, better to avoid unnecessary risks.

So, this orbiter went to orbit, as it was supposed. I keep launching a bit too vertical, but I did include some extra fuel for that. unfortunately, upon reentry the goo canisters and thermometer melted for the heat. I still got plenty of money for world first, though

At this point I realized I am treating this as a no contract career challenge, which is not. I don't need to actually get to orbit. i only need to get to space to conduct some science. So I made a cheaper, weaker rocket for suborbital jumps


It did reach space, and it was slow enough that reentry wasn't dangerous. With this I got the low space experiments.

Then I decided to try and do something adventurous: test a thermal shield at a certain altitude and speed. I did regret it, because I kept missing the altitude or the speed, by very narrow margins.

And at some point I tried to keep the spent booster attached to recover it, messed the reentry, and exploded Jeb. Had to restart the career; fortunately, it's still early, and I learned the lesson. Never try to recover heavy machinery on a manned vessel.

I ended up launching half a dozen suborbiters. on the plus side, they fall down at random, and i got ground science from a bunch of biomes. With that, I could unlock survivability.

So I picked the barometer, I made another roller, and I finished mining the ksc for early science. At this point I'm still missing a handful of experiments, but the ksc is mostly spent. I've still got a lot of kerbin biomes, but going there would require launching an orbiter (which is relatively expensive) and trying to deorbit it in the right place. I'm not looking forward to that. If possible, I'd rather tour the kerbin biomes in an airplane. So I try to move away from Kerbin, to avoid the risk of locking myself.

I have a rich contract for making a Mun flyby, but it would require reaching orbit with 850 m/s. Make it 1000, for safety. I tried some designs, but ultimately I am unable to make it without the terrier; having to use a swivel for space manuevering increased my dry mass too much. Unlocking the terrier itself is out of the question, but if I unlock general rocketry I'll eventually get some contract to test the terrier.

I need a few more research points, which I can get from high orbit. Just strap a booster on my regular orbiter


this rocket got space high science and brought it back to kerbin. Reentry was adventurous. I forgot the cargo bay open, and the thermometer exploded for the heat. then the capsule flipped, and the crew pod has really low drag and was plummeting to its doom very fast. Fortunately, leaving the cargo bay open generated just enough drag that the parachute could open at 1500 meters from the sea. if I had reentered over mountains, I'd have been toast. I was aiming for desert, but I missed it by a good 1000 kilometers. i really do not want to be stuck mining the kerbin biomes by deorbiting over them.

Anyway, I recovered enough science to unlock general rocketry.

Then some more money grinding on easy contracts - included sending a roller a few kilometers away from the ksc to take ground measurements. Those things hare hard to control, the first one run out of electricity before reaching all the targets. but the contract paid 8000, so it was worth the effort.


Once I got enough money, I started looking for a contract that would give me a terrier to test. Sure enough, I got one soon.

Now I'll try to make a mun flyby and collect some science there. I'm confident I can at least land on Minmus with those techs.

Part 2: flyby season

Getting more critical science from Mun and Minmus flyby


Introducing the moonshooter. It does exactly what it says: shoot for the moons


the first stage launching


and the second stage, aiming for mun

this rocket has 1500-2000 m/s left once it reaches orbit, which are enough to get a flyby with a lot of extra. i do need that extra, because there's no way to plan a return trajectory


On the first try, I hit Mun square on. I needed some course corrections. And then I was on a course to exit kerbin's SOI entirely, and had to burn retrograde with everything I had to avoid losing Jeb in interplanetary space. After that, I was in a 70 km apoapsis; I was about to fall back on Kerbin, but what if the moonshooter came too close to Mun along the way and got steered off course? I gave moonshooter an inclination, so that it would not cross Mun's orbit. And it used the last of the fuel. I had 50 m/s left when i jettisoned the rocket for reentry.

But the mission was successful, and I got 17 science back home. And an additional 4 science granted by the contract "explore Mun". At this difficulty level, everything counts.

Unfortunately, all the biome-specific experiments are stuff I can't run in caveman, which greatly limits the science value of space. Can't do anything more on Mun for now. I must go for Minmus too.


Minmus is a lot more difficult to pinpoint accurately, so I made myself a tool. I put in apoapsis and periapsis of my orbit, and the tool returns me the orbital period. In this case, I figured since Minmus orbit is 50 days, and it has to make about 1/4 of it, it will be crossing Moonshooter's path in 12 days. And if I raise apoapsis enough that I have an orbital period of 22 days, then I will reach apoapsis in 11 days, just before minmus. which will let me catch the moon on the way back.

Everything was correct, except the last sentence. Moonshooter spent longer at apoapsis than I estimated, and Minmus was starting to slip away. So I did the only thing I could try: I pointed the rocket more or less in front of Minmus - without even being able to make a targeting to align properly - and fired.

Unfortunately, I forgot to take screenshots of that part. Of course, it was a very difficult manuever to pull off, mostly coming down to random luck. On the plus side, orbital velocities that far from Kerbin are slow, and Moonshooter still had 800 m/s, so by trial and error  I had enough fuel to eventually succeed.

I entered Minmus orbit with 250 m/s left. I had 30 when i reentered on Kerbin.

but again, mission was a success, and it got a lot of science. Enough to unlock basic science.

Basic science gives a couple of fundamental parts. One is the science jr, the most valuable science experiment. So I went back to mining the KSC for everything that was missing before

Unfortunately, the science jr is also extremely fragile. Putting it on a roller is unsafe. I tried many ways to mitigate this fragility


Project 1: the onion pods are keeping the science jr away from the ground. It failed, because this thing can't steer


Project 2: the science jr is very well protected. But this roller doesn't roll very well, and it can't go uphill

i tried many other projects to keep the fragile material bay away from the ground, but none worked. In the end I had to do with a conventional roller, and try to drive it very carefully. Never exceeding 4 m/s.

I still broke no less than 5 science jr, maybe a few more. Those things are expensive. Ouch.


To compensate, I took this contract that required somebody to walk 3 kilometers to collect eva reports. To my shame, I didn't think to use 4x time warp until nearly the end. It paid 8000, though

Once more, I run out of easy science, and I barely got 30 science. Can't make any more breakthrough.

Good thing is, there is another part unlocked with basic science that will allow me to land: the stayputnik. By cutting down on weight, it will let me increase deltaV. Equally important, it allows me to take risks I couldn't afford to take with a pilot. I took the contract to land on Mun; it's worth some more science when completed. And now I can't take any more contracts, and must hurry up with this landing.

In retrospect, it will be shown to be a terrible idea.

It's been quite a change of style. From managing ships weighting thousands of tons, and costing tens of millions, I'm reduced to flying some awful junk with punitive mass and parts limitations. Most troubling of all, I can't reload games; I am prone to taking risks and try dangerous stuff, that's been my main bane in my previous caveman attempt.

But I've adapted surprisingly fast. Already flying a ship without SAS feels natural, and I learned to avoid any dangerous manuever with my manned ships.

Edited by king of nowhere
Link to comment
Share on other sites

Part 3: the first landings

Landing the first time on Mun and Minmus


We choose to go to Mun in this decade, with this limited technology and preparation, because we underestimated the difficulty of the task

Now that I can throw light, disposable probes at the problem, landing on the moons of Kerbin will be within my grasp.

It would make sense to start with Minmus, because it requires less deltaV. However, there is a juicy contract asking me to explore Mun, and it pays 3 science points if I can land on it and return. And it expires in a few days... free science...

So I took it. But landing on Mun was a lot harder than planned.

First things first; if I have to go around Mun, I want some relays to keep control


Those probes have a double function. First, they are relays. But second, they have some science instruments and a lot of electric charge. I can leave them in orbit in hybernation, and whenever I get a contract "return/trasmit science from the space around X", I can go to one of those probes, take a temperature reading, and transmit it to Kerbin. It's the fastest, safest way to farm money.

I started building a constellation. Then I found the juicy Mun landing contract, and took it. And then I realized I can't take any more contracts: I have a limit of 2, one is needed to keep the terrier engine unlocked, the other requires the Mun landing. So I stopped after only 3 satellites. And only one made it as far as Mun. Turns out, the stayputnik has terrible aerodinamics, and the rocket keep trying to flip. It still has enough deltaV to reach orbit even after flipping, but it's not efficient.

Now for the actual lander. Once in orbit, I want 3000 m/s. In a normal scenario, 2600 m/s would suffice, but the lack of manuever nodes makes manuevering inaccurate. Putting together a final stage with 3000 m/s is easy, and it's light enough; more or less 3 tons. Now I have 15 tons available to send this payload to orbit. I can make a two stage mechanism; second stage with a terrier, and lift it with a first stage with a reliant engine. Except that no, I cannot, because I lack reaction wheels. And they are too expensive to unlock.

Without reaction wheels, I have no control authority. But it's ok, I can still control the ship - with engine gimbaling. Turn on the engine at minimum thrust, use it to set the ship spinning slowly until it points in the right direction, then make the real ignition. It's actually a lot more efficient than it looks like, you lose less than 1 m/s every time you make a manuever. And when you're not manuevering, you don't need attitude control. This goes for the relay probes too, and they work just fine.

But the reliant engine does not have gimbaling. Without control authority, and with the kind of awful drag the stayputnik generates in front - really, it's crazy - the rocket will quickly spin out of control, and I have to manually adjust it constantly. There's no way it can be lifted with a reliant. So I'm forced to use a swivel. But the swivel is heavier and doesn't have enough sea level thrust to lift 18 tons.


First failed Mun lander

So I had to complement it with some lateral boosters. But boosters are a lot less efficient than LFO engines. Factor in the terrible aerodinamic losses, and I can't get enough deltaV for a return trip. Not even close.

After crashing three or four of those rockets, I realized I was running out of funds, so I took the save out, imported it into another profile, and did more experimenting there. I also wanted to make sure that space stage can actually land on Mun without reaction wheels - keeping it pointed correctly is a real chore, but feasible.

To make things more efficient, I took out an old idea I used long ago: instead of a normal staging mechanism, drop tanks from the top of the rocket.


Second failed Mun lander

The advantage is that, instead of needing 2 terriers, I can use a single terrier. I deemed the manuever too risky to try in a no-save career, but this is an unmanned probe, I can risk. I also gave up on full science returns, I removed the heavy science jr and the experiment storage unit than came with it (the science jr is too fragile to survive reentry, and too big to fit in the cargo bay. The experiment storage unit can collect the data from the material bay and keep it safe for return). I can't afford those 250 kg, nor the 2 parts.

Anyway, this came a lot closer to the target, but still it's missing a few hundred m/s.

Note: the second attempt has radial decouplers. I had 30 science left, when I ported the save I tried to see if by unlocking those decouplers I could actually tackle the problems. But it didn't help enough, and later I managed to make the rocket work without needing them. So, when I went back to the actual career, I did not unlock those parts. And this explains why, later, I'll post a screenshot of the tech tree with the stability tech still not researched

The breakthrough came when I figured out that, by doing something counterintuitive, I could kill two birds with a stone. So I put a (empty) crew pod on top of the rocket


The successful Mun lander

The whole point of using a probe was to avoid the heavy crew cabin. And now I put it back again? May as well do a manned mission - and no, I don't have the deltaV to safely do that. But the (empty) crew pod has reaction wheels. Their control authority allows me to use the more powerful - and lighter - reliant engine, removing the need for the inefficient boosters.

Equally important - what actually convinced me to do it - the rocket was having huge aerodinamic losses for its flat top. After detaching the boosters, even though the sviwel had twr over 1.5, at 300 m/s the rocket stopped accelerating entirely. All that thrust was eaten by drag. The crew cabin is a lot more aerodinamic.

And after consuming the first stage, I can drop it. So I can still have a light, 3-ton probe.

I was ready to return to the regular career and launch this thing.


The first stage. The reliant still is insufficient for 18 tons - TWR around 1.2 - so it consumed most of its fuel just to pick up speed


Second stage. The terrier also has low thrust for this task.

The launch profile is mostly vertical, both to ensure that the terrier will be in near vacuum when it starts, and because this rocket is still draggy, and hard to control in atmosphere. Still, with later launches I learned that I could make a bit more of a gravity turn.


At 30 km drag is no longer a problem, jettisoning the crew pod


Now a lot lighter, the second stage keeps going


In space, jettisoning the spent drop tanks. In later launches I learned to squeeze some more deltaV out of the initial stages


And the probe has finished circularizing, and it's still got 3000 m/s! Just as required


Eyeballing the Mun transfer. Luckily Mun is big, it's easy to hit


Straight on target. But I'd land on the dark side, where I'd lose contact


So I turned the probe to the side and spent some deltaV to ensure a landing on the Kerbin side


No need to circularize, I'll go for a direct landing. With high thrust I could have waited a bit longer to start the suicide burn, but in a no save career I won't take risks


I got lucky and found some flat ground, will it stay upright?


YES! Landing successful! With 1100 m/s left


In theory, 850 m/s should suffice to return from Mun surface to Kerbin. Without manuever node it was a lot less efficient, though, the probe almost didn't make it


Finally, jettisoning the engine to prepare for atmospheric reentry


I didn't actually test the reentry phase. I was expecting the cargo bay to shield the stayputnik, but the craft started spinning, and the probe core got destroyed. Remember, I have 0 control over this probe


the parachute still activated, though, bringing something back

So I did got a goo experiment from Mun after all. But the game didn't consider that part as "returned from Mun". I suppose it wants a functional craft.

So, for the sake of the contract, I had to try again.

Figuring out how to survive the reentry was trivial. In hindsight, it was a rookie mistake. Here I have a perfectly serviceable return vehicle, with a rocket on one side and a draggy stayputnik and cargo bay on the other. It will keep the delicate stuff in the back by itself. By dropping the rocket part, I instead got a smaller vehicle with a draggy stayputnik and a draggy cargo bay, and it starts tumbling. All I had to do was not drop the spent rocket. On the plus side, I could remove one part from the rocket.

Second lander was sent up. Landed on a slope. For several tense seconds, it spinned precariously on a single landing leg. Then it tilted. I reignitied the engine, but at this point the rocket was spinning a lot. After a couple more aborted landings, I saw I wouldn't have enough fuel left to return to Kerbin anyway, and I let the thing die. Well, can't always be lucky.


Though the intact core and cargo bay could at least transmit a science experiment before the end

The third lander made it as far as the Mun flyby. Once out of time warp, it wasn't responding to commands. I forgot to hybernate the probe core, and it had drained the battery completely.

The fourth lander hit another slope, and it folded softly on the ground. But it was still intact, and with nothing to lose I tried something desperate


The fourth lander, seemingly dead


Skidding on the ground!


Until the rocket is fast enough to upturn itself. Or until it explodes, whichever would happen first


Got to orbit with 340 m/s. I don't have much extra, I must get the return burn right at the first try


I did. Almost perfect, only need an additional 40 km periapsis lowering


This lander made it back with 40 m/s. And indeed, without decoupling, it flies straight


The landing legs burned in the atmosphere, as intended. It reduces drag in front, making the rocket even more stable

With the science I got, I unlocked Advanced Rocketry. Not much for the engines, but for the bigger fuel tanks. Part count was a big deal, swapping out two small tanks for a larger tanks fixes most of it.

For a while I gather some more money. The relay sats are doing their jobs giving funds almost for free. I rarely get those contracts, though, I keep getting "return science from the surface of Mun". So I sent a modified lander there, without any of the trappings needed to return - without even landing legs, it can rest on the ground and I can put in three more batteries - just to transmit science from the surface. And guess what, now that it's in place, I get a lot less of those contracts. I also did drive more rollers around the ksc to get ground science - those contracts take some minutes to complete, but they are worth over 10000 funds. And I sent some missions to collect missing science, but without much success; the science jr tends to explode on landing far too often for comfort.

Finally, it was time to land on Minmus. Landing on Minmus is a lot easier than on Mun, but I want to use a crewed capsule, and to bring a science jr, to get the full science return. Ideally, I'd like to land on Minmus 9 times, get all the biomes, and then forget about the place entirely.

With the additional weight, landing on Minmus is not easy. First, I get another test contract to gain the spark engine; it saves some mass on the final stage. But then the rocket is wobbly, leading to inefficiency. And the reliant is barely fit to lift the first stage; it has a twr of 1.18, and gravity losses are huge. Attempts to use two thuds instead were also unsuccessful. Once more, I need another test engine, the bobcat, to increase efficiency on the first stage and get enough deltaV.


The Minmus lander


Second stage, inserting into orbit


I sent two separate capsules, on two different trajectories, to increase chances to hit Minmus at least once


Once more, both capsules missed, and I again have to resort to pointing the ship in the general direction of Minmus and burn


Both Jeb and Val made it to Minmus, after extensive manuevering

Having a crew, I have once more to think of safety. In theory I only need 150 m/s to return to Kerbin from Minmus orbit, but I decided I want at least twice as much to make up for any inconvenience. This requires that a lander in low Minmus orbit has at least 700 m/s to attempt landing. Jeb didn't, but Val had a better Kerbin ascent, a bit more fuel left, and she could land


Val landing on Minmus

Jeb wasn't useless, though. First, his science jr could be used to get some orbital science. And second, I got a contract (one of the cushy ones rewarding with science) asking for a rendez-vous in Minmus orbit, and with the landing aborted Jeb had a few hundred m/s to spare for the task.

Then both pods made it back, with some 200 m/s leftover. Jeb got close enough to Mun to be blown off course, but it only took a small correction to fix that.


A nice pic of a landing in the mountains


And my current tech tree

Landing on Minmus gave 32 science from experiments, so I'm going to get an additional 250 from the other biomes. Next thing to unlock is the docking port, with that I'll be able to land a crew on Mun.

Now the real grinding starts. I have to make a lot more Minmus landings.


Edited by king of nowhere
Link to comment
Share on other sites

Posted (edited)

Part 4: an end, and a beginning

Getting all the Minmus biomes. Val is lost in an accident, Liliana takes her place


Jeb and Val weren't done with their debriefing yet, that the engineering team was already trying to improve the Minmus landers.

First, it was realized they didn't need a spark engine. Sure, a spark engine is lighter than a terrier, but no engine is even more light. So I removed the spark and used a staging mechanism expelling tanks from the top, like I did for the Mun probes. Put some more fuel in its place, so the overall performance is increased. Also, more thrust during landing is always welcome.


The improved Minmus lander

Removing the narrow point was supposed to also fix the wobbling problem, but it didn't. So I realized, it's not a problem with the rocket, just with the engine; the bobcat is very powerful, when it tries to turn it just twists the rocket. So, I limited rocket gimbaling to 25%, and it doesn't wobble anymore.

As the science from Minmus kept coming, I did unlock Miniaturization, allowing the use of the small stage separator between the crew pod and the upper tanks, and Electrics, whence I put a solar panel in place of the battery.

Meanwhile, I didn't need the spark engine anymore, so I went to fulfill the contract to test it. And here, in the least suspecting circumstance, tragedy struck.

The contract required to test the spark in very specific circumstances: high altitude and low speed. Those are the kind of conditions you'll never get in a real flight, unless you are making a suborbital jump specifically for it. So I went and did exactly that. And since I was missing some science jr experiments, I strapped them on the rocket and tried to land on the cursed mountains west of the ksc. The mountains where I tried to land a science jr many times, without success.

First part of the mission was fine. Rocket went up, tested the spark. But during reentry, the crew pod flipped.

No panic, it's not the first time a kerbal survived that. Open the cargo bay, it will slow down the pod enough to open the parachute. This should be even safer than regular pods, because I took advantage of the low mission requirements to put in an additional parabrake. Parabrakes should open in harsher conditions than parachutes, and be a further insurance against accidents.

Well, neither the parachute nor the parabrake opened. Don't know why. The pod didn't even fall that high up to call thin atmosphere into account: it slammed into foothills at 1100 m altitude (yes, to add insult to injury, not only I lost a pilot trying to land a whole science jr on mountains, I didn't even hit the biome).

Of course, during the event I was too busy trying to do something, so there are no screenshots. The best I could take was the report on Valentina's death


The report on Val death

For a moment I was very tempted to cheat. But I still have 3 kerbals left. And then I can hire others.

Right, I can hire others. This is not the no contract challenge, this is the caveman challenge, I have basically unlimited money as long as I have patience. I have unlocked solar panels and the okto probe core, which comes with (very weak) reaction wheels.  I updated my relay fleet, putting a probe with antennas and a thermometer in Kerbin, Mun and Minmus orbit. I took advantage of contracts asking that to even get paid! Then I landed other similar probes on Mun and Minmus. Now I can complete easily any contract of "get science from X location", in a completely free and renewable manner. I have officially unlocked infinite money.


On the right, the orbital relay with the science instrument. The probe on the left will land on Minmus to provide a fixed science base. An identical probe went to Mun

At some point, to my surprise, I even see a regular contract offering science reward! Those weren't supposed to happen in nanodiamond level


This contract to place a satellite in retrograde orbit offers 1 science as reward

It starts dawning on me that I will really win this challenge, no matter the lack of saved games, no matter how badly I can screw up. I can make infinite money and reputation with the probes I already have in place. With that money, I can take contracts granting science. I could technically finish the challenge already, without exploring further. Well, I could use the administration building to convert money into science, either.

But it's not in the spirit of the challenge, and it would not be fun. At this point I decide I'm not going to try to launch a kerbal stuck on a ladder to get eva reports all around and finish the challenge without ever leaving kerbin, either; I'll explore Mun, and finish with a big interplanetary mission, regardless of whatever else I could do.

Anyway, now I want to replace Val. I have 200k funds, and I need 520k. It took two or three hours. At first it was slower; I'd find a few contracts "return science from X place", but after a while they stopped, until I went and did something else. Then I figured maybe the game is designed to prevent the player from only picking one type of contract, and I should differentiate. So I started picking also "test stuff on the launchpad" contracts - even though they don't pay enough to be worth the time - and this seems to have worked, because now I keep getting "return science" too.

Maybe that's not in the spirit of the challenge, but it was already a very boring couple of hours. Actually fulfilling regular contracts would have taken 0 effort, it would have been extremely time consuming, and mind-numbingly boring. It would have entailed 0 risk (unmanned launches only), and would have been just as much interesting. No, thanks.

Finally, I had enough to hire another female pilot.


Welcome to the rooster, Liliana

I hired her before the final two Minmus landing, because I wanted to give her a landing (for XP), and I wanted to make a docking and crew transfer with her and Jeb in Minmus orbit (for a contract).

Meanwhile, I got better at hitting Minmus. I realized, instead of shooting wildly in roughly the right direction and hope, it would be better to treat it as a rendez-vous. Minmus is overtaking my pod? Raise periapsis, get slightly lower than Minmus, wait to overtake it.



A couple of very successful Minmus insertions

As a result, I started having additional fuel left. Not that I could do much with it; I could jump to a different biome, but the science jr cannot be refreshed anyway. Though it did help the time my biome map turned out to be inaccurate.

Since I had free deltaV that would be discarded anyway, I tried to do something nice by taking some Minmus exploration contracts. Those tend to pay very well, and are easy enough with suborbital jump. There I made the mistake of picking up a contract that would spawn multiple targets in a close area. I run out of fuel after two, and needed two more.

I laid the rocket on the ground and used it as a roller. It worked.



The Minmus lander used as a roller

And so, all Minmus biomes were explored easily. Finally I won't have to shoot in the dark anymore. I used the additional tech to also get Heavy Rocketry; the bobcat engine is fundamental for efficient launches, I want it unlocked for good


Tech tree at this point

Edited by king of nowhere
Link to comment
Share on other sites

Part 5: if you believe, they put a kerbal on Mun

First manned Mun landing


Actually, I've done this part in the middle of the previous one, because - once more - a contract offered science for it. But narratively, it fits in a separate chapter.

So, need a Mun lander. A single launch to Mun with a level 1 launchpad is feasible - indeed, I've done it for an old challenge - but it doesn't leave any safety margin. A Mun landing is nominally slightly more than 5000 m/s from Kerbin, but with the perspective of inefficient injection and being unable to reload, I want at least 6000. The Minmus lander is already pretty well optimized, and I can't do much better, and it's not even close to 6000. But the Minmus lander would make for an excellent descent and ascent vehicle, and it conveniently reaches LKO fully fueled atop its launch vehicle. So I keep it as the base, and dock it to a space tug for the Mun transfer.


The Mun lander pod; basically a Minmus lander modified to have a larger base


Here docking with the space tug; basically a glorified fuel tank with a probe core on it.

I'm sorry those designs can't be as beautiful as my grand tour megaships


The tug does the orbital manuevering up to Mun, with its own fuel


And leaves the pod in low munar orbit. It's mostly spent now


I came in perfectly equatorial, so I went for the canyon


Perfect landing

Now I'll have to start mining the 17 Mun biomes for science too...


Link to comment
Share on other sites

Part 6: Just to show that I can

I decide that I will not try to finish the challenge by launching a kerbal on a ladder and collecting EVA reports.

However, I still land on Mun a kerbal on a ladder, to make a point that I could have completed the challenge thusly, had I wanted.


With the reliable capability to make manned Mun landings, I should be able to finish this challenge handily. That won't do.

Gathering all science from around Kerbin is an uninspired way to win. I want to run a full interplanetary missions, and I settle for a Jool 5. In part because it looks very difficult. In part because I'm actually a lot more worried about the perspective of pinpointing an interplanetary transfer blindly than I am at the idea of planning yet another Jool 5.

I also decided I won't use EVA construction, because for a long while people have been making caveman challenges without this resource. EVA construction makes the game a lot easier.

So here are the additional rules I gave myself for this challenge:

- No EVA construction

- No sending a kerbal on a ladder to get EVA science - else I could finish caveman with just the Kerbin system

- Aim for a level 3 Jool 5: five different kerbals, one for each moon

- bring a full suite of science instruments to every surface

- put some beautification on the mothership

The "beautification" part refers to the fact that I want to give every kerbal his own personal hitchhicker container to live in. After one year playing with kerbalism, it just feels wrong to send a crew in a long mission without some extra luxury. So the first technology I unlock with the science from Mun landings is Space Exploration, for the hitchhicker container.

This tech also unlocks ladders, bringing in an additional challenge. So far I stated several times that I won't bring a kerbal around on a ladder. But I won't, or I can't? To show that it's a self-imposed restriction, I decide to bring a kerbal on Mun in EVA.

It was a lot harder than I anticipated.

The main problem is that a Kerbal on a ladder will face aerodinamic resistance and will be blown off of the ladder. Strong accelerations can break a kerbal loose also in vacuum. So I want to bring a kerbal inside a cargo bay, so that he's protected. So I took the Mun lander, moved stuff inside the cargo bay to make room for a ladder, and devised a vehicle to bring the kerbonaut there


I'd have called it the Girapphe, but this is caveman, so I called it the Brontosaur instead


The Brontosaur lines up with the rocket


And delivers Liliana inside the cargo bay

Actually, it's a lot more complicated than that. It's very hard to grab the ladder, and very easy to make Liliana fall to the ground. She'll survive, of course - those little green men are made of iron - but then there's no way to bring her back to the cargo bay. No, I could not afford to put a ladder to let her climb, part count is the limit. And I kept changing the geometry of the Brontosaur and cargo bay to try and make it easier to enter.

Here I went to the alternate career where I test things without fear. Mostly because I could use Alt-f12 to move the Brontosaur instantly from the runway to the launchpad without losing time driving the 2 km distance.

Anyway, the cargo bay is slightly smaller than a kerbal's comically-oversized helmet, and this creates some clipping problems. Since I was in test mode, I decided to check what would happen if I remove the helmet to avoid the clipping. Yes, the kerbal dies during launch, of course. But not for lack of air: for overheating.

Which brings to the second, worse problem. The kerbal in the cargo bay should be protected, but she isn't. She's taking full heat as if she was directly exposed to the outside. Which, more than anything, is an aerodinamic problem


See that huge red arrow of drag? The one originating from the upper part of the rocket? That's Liliana inside the cargo bay. Notice the navball, how the rocket is twisting and turning around

By the way, the resistance value given by the aerodinamic overlay menu does not take that into account

Still, even though the rocket was difficult to control, the Mun lander pod has enough extra thrust and fuel that it did reach orbit easily. Though that fuel was supposed to be used for the Mun landing... But the main problem is, if the cargo bay does not protect during ascent, it's unlikely to protect during reentry either. Again, since I was in the test career, I tried it, and of course, Liliana died. Good thing there are no kerbal feminists.

I also discovered that a kerbal on a ladder inside the cargo bay - or stuck on a ladder in other awkward ways - will produce thrust. I guess I discovered what's the whole "kraken ladder engine" is about; put a kerbal on a ladder in a way that's microclipping into some other part, and the microclipping will shake the ship, producing some thrust.

Finally, I discovered that I cannot time warp with a kerbal on a ladder. Sure, you're not supposed to, but in the past I did do it by changing vehicles and time warping from the tracking station. Can't be done anymore, apparently.

I realized that the kerbal needs a dedicated launch vehicle. The lander pod doesn't have enough extra deltaV to deal with the huge drag. The kerbonaut can then latch to the ladder of the command pod, where it won't cause krakens. It will need another pod to get in before reentering atmosphere. And I will time warp with the kerbal detached from the ladder; it will slowly drift away, but a small push with the jetpack will suffice to return.

All this is the result of recent (1.8.1, I'm told) changes; in the past, it was a lot easier to send kerbals around on a ladder. Still, there's people who did a Jool 5 in caveman and did send Jeb all the way to Jool in EVA; and apparently they faced my same limitations - except that they did disable commnet, so they could get away with moving probes.

Anyway, for the actual mission I used Bob.


I devised this way of putting the kerbal in the crew pod. It's a lot more efficient than the Brontosaur


The truss, of course, gets jettisoned before launch



Bob in the cargo bay experiencing extra drag. But this time the ship is only meant to carry him to orbit, it's got a lot of extra juice


Once in orbit, Bob lets go of the crew transport, to avoid kraken drive issues


And he rendez-vous with a fully assembled Mun lander, prepared previously


The Mun lander goes to Mun. Thrust is not too much, and Bob can stay on the ladder easily. After the transfer burn, Bob leaves the ladder and floats


Arriving on Mun, straight on target. Bob only drifted 1600 m from the craft. He's got plenty of eva propellant still to return on the ladder


Circularized around Mun, and left the transfer stage in orbit. The lander is much smaller, and can get a lot more acceleration. To avoid kicking Bob free of the ladder, I reduce the thrust to 30%


Landed on Mun. A bit more expensive for the lower thrust, but no issues


Back to orbit. I can't aerobrake, so I rejoin the transfer stage to recover its fuel and rocket brake around Kerbin


Here the transfer stage was spent and jettisoned, the lander was mostly spent, the orbit was lowered to 180 km apoapsis. Now Bob leaves the lander


And the crew transport, which brought Bob in orbit, comes back to grab him and deorbit

In this, I took advantage of Bob's ability to refresh the science jr to get more measures in orbit without a dedicated mission. I pointedly did not collect eva reports, except for one on the surface, just to leave it on the record.

Now it's time to start planning the Jool 5 mission

Link to comment
Share on other sites

Posted (edited)

Part 7: it's not a King of Nowhere's mission without some massive mothership!

Preparing for the Jool 5, I begin assembly of the Navis Sideralis Neanderthalensis


Back when I was playing the no contract career challenge, I did end up assembling a large mothership for a Jool 5 (actually, it came close enough to a grand tour) with the limitations of a level 1 VAB and level 2 launchpad.

Now I will have to do the same, but with a level 1 launchpad too.

Also, can't transfer fuel between tanks, so I can't send up a lander and refill it; all fuel tanks must be detachable. Can't use autostruts. And a bunch of other limitations.


Tech tree for the task; I've also researched Propulsion Systems and Flight Control. I've landed on 9 Mun biomes, I can afford to get a couple more tech if needed

As usual, I must name the ship. Being a caveman challenge, I want to stick "neanderthal" somewhere.  I decide to call it neanderthal spaceship, but in latin, so that it sounds like the name of an extinct specimen. So I look up "spaceship in latin... What? Yes, of course there is a latin expression for spaceship. Latin is the official language of the church, they've been updating it. There is a whole list of modern words translated in latin.

So here it is: Navis Sideralis Neandethalensis.

The core of it all is going to be a bunch of 5 hitchhicker containers, that will be the living quarters. To them I'll strap the main engines. Then the drop tanks. Landers and other bits and odds stuck around hapazardly.

With the launchpad limitations, I was able to send 4 tons in orbit on a single launch. The hitchhicker container is only 2 and a half tons, and I wasn't expecting to have many problems launching it, even though it's aerodinamic problems incarnate.

After some rough calculations, I decide I want 1 cheetah engine in the central core, and 4 poodles in the lateral ones - detachable, to save weight after I discard most of the drop tanks. The cheetah, of course, comes from another test contract


Navis Sideralis Neanderthalensis, core stage

The core stage, which also includes a probe core and some antennas (no, I don't know why I put them since Jool is way too far), barely made it to orbit. Aerodinamics is terrible.

The lateral modules have an additional docking port between the engine and the crew. Which exhacerbates aerodinamic problems. I tried a half dozen launches, but it always falls short of orbit. It may have been smarter to just change design, but I grew fond of this concept. So I set to strap on additional fuel tanks on the launchpads. I keep using the cheetah engine, it's very useful for my needs here.


The core launcher for the additional 4 crew cabins. With the docking ports in the bottom for the strap on tanks

This thing's aerodinamics are so ugly, the launch profile entails flying perfectly straight up until 200 m/s, then throttling down and going like that all the way until the first stage runs out. A launch profile normally reserved for Eve. In this configuration the first stage runs out around 9-10 km altitude, the cheetah is already fairly effective but there's still too much drag. With the extra tanks the first stage can last to 15-20 km


And the tanker to load four more tons of fuel in the first stage


The tanker goes to the launchpad. That black line is a terrain glitch that will break your wheels if you're not careful


Docking the fuel tanks


Unfortunately, docking stuff on the ground tend to cause microclipping issue. In this case, they scrambled my poor rocket badly, tilting the point


The rocket starts flying askew. To try and save it, I pull up the engine gimbaling (it's normally set at 25%) to steer the rocket upwards


Ascent was a lot inefficient, but I at least managed to not flip the rocket and keep it pointed vaguely upwards. At 15 km the first stage runs out, the second stage tries to restore an equatorial profile


The second stage barely managed to reach orbit, it run out of fuel during rendez-vous manuevers. I rescued it with a Mun transfer stage (which was later recycled for a Mun landing)


The transfer stage brings the crew pod to the growing spaceship.

The core stage, the one with less aerodinamic problems, I got it flying at the second or third attempt, in regular career. The other modules, after blowing up a couple of them, I went back to the test career to get the whole tanker mechanism right. I went then back to the caveman, and I lost one of them. The second, for which I showed the gallery, was the most adventurous launch. Another launch tilted too much and was lost in the ocean, the others went with less drama - though all of them had some hiccups.

In the end I also sent the Service Probe. It's a constant fixture of my motherships, but here it will be specially needed to move around all tanks. I can't refuel my landers, swapping tanks will be needed.

I was about to launch two, but then I remembered this is not kerbalism, there's no risk of one breaking up. I launched it last because, before the last launch, I was still missing a crucial tech node.


The Service Probe in front of the station


The service probe helping with improving the alignment of the modules. Yes, I had to undock and re-dock most of them many times to get them properly aligned

The Service Probe includes an additional fuel tank for itself. But the flight got wrong, tilted the rocket too much and it started flipping, couldn't recover it until it was at 20 km height. At least I managed to shut down the engine when it was pointing downwards, and turn it on again when it was pointing upwards, and so I could save the probe - but I had to spend all the fuel in the extra canister to compensate for the awful launch.

Too bad I didn't think to take screenshots while there.

That's it for now. Navis Sideralis Neanderthalensis will keep growing. It's currently at 15 tons. Estimating 50-60 tons of payload for the landers, and calculating very generous amounts of fuel for interplanetary transfers in the blind, I want some 300 tons of fuel. It will take some time.


I read some stuff in another thread that gave me the doubt that maybe testing stuff in an alternate career like I've done may be disallowed. I asked the challenge keepers for legality (I already asked about it earlier, but I got a partial answer and I took it as an ok). I'll see if I can test properly the landers, or if I am supposed to eyeball them. Meanwhile, I'll send up fuel tanks, those require no testing.


Edited by king of nowhere
Link to comment
Share on other sites

Part 8: ad memoriam (and updates)

I make a memorial for Valentina. Meanwhile, construction of the Navis Sideralis Neanderthalensis proceeds and a fifth kerbonaut is hired.


Ever since Val crashed, I've been wanting to leave some kind of memorial. I finally decided on it. So I built a statue; for its pose, I called it the Valentina redentora


Here is Val (well, as much of a likeness as I could do with those parts) standing in front of the mountains, arms outstretched, palms pointing upwards to the sky, urging kerbalkind to claim the stars

Val died while trying to land with a science jr on a mountain biome, so I figured using a science jr as the base and placing it on the mountains was fitting.

I would have liked to carry her on top of a mountain. But the rover I used to carry the monument was very crappy, limited by part count and the awkward payload


This thing has very few solar panels. Even on flat terrain, winning air drag consumes more energy than it's produced. To send this contraption to the mountains west of the KSC, I neeed 4 days of brief traveling and long waiting to recharge the battery. It could move uphill pretty well thanks to its many wheels, except that it would drain the battery in 20 seconds. So, I just left the memorial as soon as I got into mountains.


The plan for a Jool 5 level 2 requires hiring another kerbal. Farming money is extremely boring, but I hit a good pace with doing a flight, mining 50k money, and flying again. This way I could slowly hoard enough to hire another kerbal. Say welcome to Quidatta!


Quidatta has an excellent curriculum with a lot of bravery and very little stupidity. An excellent addition for the kerbal space program

I wanted to get a level 3 pilot to track manuevers, it's a lot more convenient than doing it manually, so I sent the newly hired Quidatta in solar orbit


A 100Mm apoapsis is outside Kerbin's SOI, but it will entail very little speed compared to the planet


Moving slowly compared to Kerbin ensures that I can return to it just by pointing the rocket in its direction and burning. Perfectly successful manuever, with enough fuel left to land on Minmus

Regarding the Navis Sideralis Neanderthalensis (NSN for short), I developed the ecape pods. They are not a luxury; they could save a lot of weight on the return trip if I don't make it with the whole ship.

By tradition, I call escape pods "Dolphin". But this is caveman, so I googled up the evolutionary history of dolphins and picked one of its prehistoric ancestors. So I named the pods Cetotherium.


Cetotherium with launcher

The Cetotherium uses the Pea pod because it's the only 2-person pod I have available. It uses two thermal shields, because last time heat transfer from shield to pod was almost killing my ship, and I don't want to risk it. By the way, getting a test contract to access the 1.875 m thermal shield was a nightmare, it took hours, and I only got one after I finally gave up and assumed it was impossible to get such a test contract.

The Cetotherium, with its heavy thermal shield and its fuel reserve, is rather heavy, and has awful aerodinamics. Just like everything else I'm launching for the NSN, really. To get it to orbit, I had to use six extra drop tanks, delivered by the fuel shuttle


The second stage is powered by a poodle engine. It's attached with a docking port because I want to also recover the engine

I was surprised by how powerful the poodle is. I got used to launching the kind of ships that required dozens of wolfhounds and still had low thrust. It gives some perspective to how incredibly huge were my motherships.

Recovering the poodle and attacking it to the station was difficult. It required grabbing the free-floating engine with the service probe, aligning it perfectly with the docking port, giving it a push. Then turning around 180° so that the docking port of the engine is pointing in the right direction, undocking the service probe from the engine - which is now without control - and moving it away. Doing it too fast entails getting the service probe stuck on the docking port where the engine is supposed to go. Doing it too slow entails the engine drifting too much and missing the docking port. It took some trial and error, but I only have to do it twice.  Four times if I send the other two poodles the same way.


Detaching Service Probe from the engine...


And getting it out of the way so the engine can dock properly

In addition to the two Cetotheriums, carrying 2 people each, there is the command pod. That one is supposed to be docked to all the landers, and to also return to Kerbin - this way, I'll get the "return from the surface of X" bonus. It also has the science container to grab all the science and return it.


Navis Sideralis Neanderthalensis so far

In the above picture, it's the one in the middle. The long fuel tanks on it are the reserve fuel for the Service Probe - which already ate a full canister going back and forth tugging extra modules who ran out of their own fuel to a rendez-vous.

The ship is 31 tons at this point, and it's already starting to experience the first stability problems. If it gets much worse, I may have to reevaluate the "no eva construction" clause, and use it to add struts everywhere.


Edited by king of nowhere
Link to comment
Share on other sites

Part 9: the price of science

Navis Sideralis Neanderthalensis gets its supply of science jr and goo containers


So many things to add to NSN, in no particular order. Here I am sending up the last two poodles. I realize there is no chance to have a probe control point with my technology, so I also send an additional command pod to stick to the service probe, it will be needed when too far from Kerbin for remote control. The poodles are quite heavy, so extra tanks are used. I take the chance to send up Bob; being a crewed mission, I add a parachute for safety in case the launch fails.


The launch was about to fail - those poodles aren't just heavy, they are also aerodinamically challenging. The problem is that all is stuck together with size 0.5 docking ports, because they are the only ones I have, and the rocket keeps changing diameters, which of course is bad. This launcher was about to fail, but I could rescue it by jettisoning a poodle. A second launch sent up the last poodle - and a new crew pod, because I realized this one is lacking a docking port.

Then I send up the expendable science. I want to mine science from everywhere I go, and unfortunately science jr and goo containers cannot be restored with caveman limitations. This means sending up one for each measure. I will land on 5 moonss, for each of them I'll take a reading on ground, low space, high space, so 15. Plus there is Laythe's low and high atmosphere, and Jool low and high space, total 19. For a moment I consider adding a skim on Jool's upper atmosphere; it would be feasible, but I'm not in the mood for it.

Sending up the science containers turns out to be extremely challenging, as they make a ludicrous amount of drag. Also, they have high part count, leaving few for the actual rocket. After the first flipped and exploded already at low speed, I went and tested. I had at least half a dozen failed launches of the 5-container rocket before accepting it would be easier to launch 4 science containers in five different launches


Those things flex and flip so easily. Removing the front fuel tank helped only marginally

The payload mass is very small, and even the aerodinamic gain is not easily noticeable. But the major gain of using 4 science containers instead of 5 is that it frees up 4 parts. With them, I could afford extra fuel tanks and a reaction wheel.

With the extra fuel, I tackled this aerodinamically challenging launch by brute force: fly straight up until 150 m/s, then throttle down, and keep flying straight up at 150 m/s until you're well above 15 km. Horrible gravity drag, but I have the extra deltaV to make up for it. The second stage alone has 2800 m/s vacuum, it can reach orbit alone after the first stage cleared the atmosphere. Even this way, three launches out of 5 flipped. But the rocket has enough extra deltaV that it can reach orbit regardless.


The science canisters about to be docked to NSN


Current status of NSN: 50 tons, of which 15 fuel, and 231 parts so far


And a schematic breakdown of the budding mothership

As it is, the station can turn around easily. But if I try to hold it still with SAS, it will start wobbling. I did not try to discover what happens if I let the wobble build up; deactivating SAS and time warping is enough to stabilize the station, and without SAS it will stand still no problem. I'm still hoping I won't need to add struts everywhere.

Now I will start working on the landers; I expect to need 30 tons for a Tylo lander, 10 tons for a Laythe lander, 5 tons for Vall, and 5 tons for the last two moons. Then I will want at least 200 tons of fuel tanks, for an estimated 300 tons of additional mass - and no less than 300 additional parts - I need to launch.


Edited by king of nowhere
Link to comment
Share on other sites

Part 10: Landers ahoy!

Developing and launching the landers for Navis Sideralis Neanderthalensis

Being unable to transfer fuel poses some unique challenges, as fuel tanks have to be swapped every time. Launch mass limitation also forces a modular design

After reading another report where people consider cheating the testing of stuff in sandbox (the challenge rules only specify mods), I asked challenge keeper JAXO if I was fine. He told me to go ahead as I was, but I'm still uncertain on whether I was supposed to do it or not. So I decided I'll keep testing in sandbox, but I'll limit myself to tests I could conceivably run in the career. So, I'll test everything in the Kerbin system, and won't try the landers on Jool.

10.1) Tylo lander


Landing on Tylo is expensive. You need at least 4500 m/s, if you have high thrust and perform all manuevers perfectly. 5000 m/s is a good target for a normal mission. Being unable to save scum for a good trajectory and having to eyeball everything, I aim at 5500 m/s. I also want a very wide base, because I won't have the luxury of trying again if the lander flips on the ground. I can't launch a wide base with my limitations, this requires a central tank with docking ports on the side on which to attach lateral connector tanks.

It is possible to make very compact landers by reducing mass; people launching a kerbal on a seat have managed with around 1 ton of mass. But I can't do that either. I need to carry my lander pod, which includes the heat shield for its eventual reentry, and I have to carry science instruments.

My first creation had the tanks attached on top, with long tanks descending in columns to act as landing legs. It was quite peculiar (pity that I didn't take pics of it) but not very effective, it weighted 33 tons. Realizing a single poodle would be more effective than 4 terriers, I based a new model on that and eventually managed to fit my requirements in under 20 tons.


Tylo lander, final model (almost)

The whole thing has a nominal deltaV of 5200 m/s vacuum, but it gets better because the science jr and goo can be discarded after landing and various spent tanks can also be discarded easily. First the lateral drop tanks are spent, for 1300 m/s. They get dropped just before the last part of the descent, granting maximum thrust to weight when it's needed most. After landing I can probably already discard the landing legs, too. The fuel for the poodle will last for the first part of the ascent, where a high thrust is important. Then the poodle is discarded, and a last stage with a spark engine will finish circularization. The whole thing for 18.5 tons, including the lander pod - which is already docked on the NSN - and the disposable science.

I tested this thing in sandbox mode around Kerbin. Mostly I wanted to make sure that its joints would survive some jolting. Tests performed included

- in space, thrusting at maximum power. The leg connectors and lateral drop tanks flex in a scary way, but nothing gets broken.

- on Kerbin, detaching the science jr. It falls down after inducing some oscillation with SAS, it can explode, but in five trials it never destroyed pieces of the lander. It could have been a funny accident

- on Kerbin, detaching the drop tanks and dropping the lander from 15 meters (deactivating physical easing). The lander slams on the ground at a bit over 10 m/s, testing a hard landing. Result, the landing legs flex very precariously, but they hold. However, the poodle comes into contact with the ground and explodes. I fixed it by moving down the docking ports, to get more clearance between the poodle and the ground.

First part to be launched is the core, comprising the poodle engine. Using a poodle engine for a 2-ton payload feels a waste, so I send up the whole central part, including the last stage with the spark engine. This way I also save on 2 docking ports to connect it to the poodle stage.

Of course, the launcher needed additional tanks for this - 4 of them, to be exact - but by now, adding fuel tanks is routine. I forgot to take pics of the launcher, though it was another one of those aerodinamic monsters meant to be launched slowly and vertically.


Tylo lander, core stage, third stage of the launcher. Here with some drop tanks in front, though it already dropped more tanks


It's quite uncomfortable to dock, because it had the probe core attached on its docking port. So to dock it has to expel its probe core and something else has to grab it like a piece of junk


Here the Service Probe grabbed the lander pod and docked it to the Tylo lander


The lander pod is then reattached to the NSN. I plan on landing at Tylo first anyway, may as well have the lander prepared

Then I take the other parts of the lander and launch them one by one. The side connectors, which include the drop tanks, weight around 3 tons each, and aerodinamic is not too bad. A comfortable launch


Dropping tanks from the top of the rocket is a constant fixture in this challenge


This launcher expelled the landing leg attachment due to a staging error. Whooops!


This visual perspective is used during approach to try and align the lateral docking port pointing downwards, as precisely as possible


The landing legs are light, I could launch all four in one go except for part count. Limited at 2, I use a simple, light rocket with luxuries like fins


And I dock the first two landing legs

At this point I realize the mistake. The landing legs are not giving the poodle much clearance. But I specifically moved everything down to avoid the Poodle slamming on the ground on a hard landing. And then I saved the craft, and I took its parts when I made the various launchers... ah, crap, now I see what I did wrong.

I made the first model of the lander in the career. Then I opened it in sandbox, and tested it there. Then I saved the modified version, but it got saved in sandbox. And then I went back to the career, and opened the old model.


The actual Tylo lander. Notice how the lateral docking ports and the landing legs were moved down, to give the Poodle additional clearance

After a couple hours of work, I will have to launch the lander again. Do I really have to? The lander will work just fine, as long as I slow down properly before touchdown.

Yes, and if my finger slip for half a second, I will crash the lander and lose after all this work. Better to spend some time to remake the lander.

On the plus side, I can recycle the side connectors.


Moving the faulty tylo lander to a lateral docking port


Using the service probe to transfer a side connector


Attaching a landing leg, careful about alignment


Finally, the Tylo lander is complete

With this inclusion, Navis Sideralis Neanderthalensis has 299 parts for 66.5 tons


And after writing all of this, I remember that my original idea was to use one of the poodle engines from the mothership for the lander. The plan had always been to drop a couple of poodles as I drop most fuel tanks, so I could have recycled one for this lander. But this time it's really not something to be worth making the lander again; I'll just load one extra fuel tank to make up for the extra mass.

10.2) Laythe lander


Landing on Laythe is a lot less expensive than on Tylo, but a lot more complex. On Tylo there is only the matter of deltaV and thrust; on Laythe you have to take care of the atmosphere during reentry, and you need to worry about whether you'll land on solid ground or water, and you need to factor in drag...

You can land a rocket or you can land a plane. So far, whenever I did land a rocket on Laythe, I used save scumming to make sure to hit an island. Of course, this won't be an option here, so a plane seems the better way. So the plan is, make a plane to land on water, take off from water, reach maximum altitude, then ditch the wings and keep going as rocket. My first spaceplane, the Not!Albatross (used in the Jool 5 science run), managed to take off from water on propeller power alone; how difficult can it be with jet engines?

As it turns out, a lot.


Basic spaceplane model

This is my core model. It flies really well (and by that I mean, I intentionally tried to flip and stall it, and it never had a problem), it lands on water easily, and if I stick some wheels to it, it also takes off from the runway all right. The juno engines should be able to push the plane to around 15 km altitude, at which the terrier rocket can already work at full efficiency and make a low-angle ascent to orbit after the wings are jettisoned. Orbital velocity around Laythe is 2000 m/s, the rocket part of the plane has an extra 700 m/s which should be able to cover any remaining drag and inefficiency. This neat little flyer weights less than 7 tons, including the command pod

It only has one problem: it can't take off from water. The plane barely reaches 20 m/s before drag with water stops it. It would need at least three times as much speed to take off; even adding two more junos doesn't help. Here I try a lot of other options


Floaters, to reduce drag with water. But the drag and thrust are now misaligned, and they create a torque that flips the plane nose-first into the water


A flea booster, to give a strong push out of the water; but not even the 160 kN are enough to accelerate significantly; this plane sinks a lot


Same boosters, but angled to force the nose upwards out of the water. Of course, once that happens, the nose will keep being pushed upwards, until it makes a full circle and crashes back in the water

Hours of experimenting, and the closes I came to getting out of the water was with a swivel engine - and the rocket-plane spiraled out of control and crashed soon after. So much for taking off from water. Not!Albatross was actually pretty well optimized for it, it had low drag, it floated a lot, it had very strong propellers and a lot of wings for its size, it had the center of mass way backwards to help turning the nose high; even then, it barely succeeded. There's no way I'll be able to replicate the feat on a plane made for maximum savings out of scrap recovered from the junkyard.

Change of plans! The jet engines let me move around; I can find an island to land on.


Plane project, now with wheels for ground landing

I tested this model, took off from the runway, tried to land on the flat ground surrounding the ksc, exploded. The LY-01 is supposedly rated for 125 m/s impact speed, but it's actually very frail. Or maybe it's some of the other joints that break. Regardless, I've never been able to make reliable ground landings anywhere, but especially on Laythe, whose islands have a lot of cliffs. I always saved before landing, and though some planes were more reliable than others, I've never mastered the art. Not with Not!Albatross, not with Arrowhead, not with Craplane. If I run a no-reload mission, ground landings are not an option.

Still, there was one last option: land on sea (where I can always do it reliably) but close to a shore, use the jets to move on water over the shore, take off from dry land. Testing was successful, and I did consider it for maybe a couple minutes. But in hours of testing planes, I became aware of many other problems

- I'll need to bring down three science jr; there's no way to dock them without them burning up in the atmosphere

- the wheels will also burn, because all the models I have available have low thermal resistance

- taking off from the ground can still end up with a broken plane, if only I hit a bump in the road

- this plane requires a very long runway for the juno's slow acceleration, hard to find such a place on Laythe. It increases the chance of finding the aforementioned bump

- moving on water is extremely expensive, it took me 1000 m/s to move 500 m to the nearest beach; if I fail to pinpoint the landing, I have a serious risk of running out of fuel

- all this testing was done on Kerbin; planes are generally harder to fly on Laythe, the latest self-imposed rules forbid testing on Laythe, and I have no idea how to compare performance

All the above factors, added together, result in a big "screw this" for the idea of flying a plane on Laythe. In a normal mission with reloads and in situ testing I could certainly make the idea work, but in this case there are too many unknown, too many potential failures, too many risks.

It will be a rocket, then.

A rocket is going to be a lot heavier and less efficient than a plane; but hey, I was factoring in 30 tons for the Tylo lander and I got away with 18, I can spend some extra for Laythe and still be on budget.

To land on water and stay upright, a rocket needs some kind of floaters in its high portion. Those can act as additional drop tanks, they will be terrible aerodinamically, but they will be discarded soon, and they are practically free fuel. I can tie the parachutes there, so they won't weight the ship during ascent.

I start with a swivel because it has good performance at Laythe's sea level, but I eventually jettison it for a terrier. The rocket's main body can't be made any longer than this, or the floaters don't work anymore.


Laythe lander project

Finally, this model has the added advantage of being very stable during reentry, the floaters also act as aerobrakes, and they keep the fragile science jr shielded. I tested a Kerbin reentry, and they didn't even start to overheat.

The rocket has the disadvantage that it cannot land on the ground; after considering some very impractical ideas to make it hybrid, I decided I'd just deorbit over DeGrasse sea, where there is half a planetary circumference without any island; I'll sink periapsis down enough to ensure I land in it.

Now for launching; the rocket is 10 tons heavy, even without the crew pod. Yet I must launch it as a whole piece; joints based on the small clamp-o-tron are terrible for aerodinamics.


Trying to launch the Laythe lander in a single launch

Besides mass, the lander is also very draggy, with those floaters in front. Six additional tanks strapped to the booster won't be enough. So I devise a new modular sustem to strap on additional tanks


The projected extra tanks, complete with engine

I modified the drop tanks, gave them an additional docking port on the back so that moar drop tanks could be latched on, and added a spark engine on the bottom because the bobcat does not have unlimited thrust.

Before putting in all the work to strap on a dozen boosters on this rocket, without even any guarantee that it would work, I had a spark of wisdom and went again to test in sandbox


Sandbox test of the infernal contraption I considered to launch the Laythe lander

The launch fails. I tried a fourth line of tanks (for a total of 16 drop tanks), it still failed. Hard.

It was an easy decision then to remove the floaters and put them on later. The main body of the rocket is still heavy, but it's reasonably aerodinamic, and launching it was mostly a matter of using enough drop tanks.


Laythe lander core launcher

This launch would barely be worth mentioning, except that I experimented an innovation; notice the probe core stuck on a lateral docking port.

Placing the probe core is always a problem in this career. It is a small part, and it must be removed after docking, so it will always leave a gap into a rocket's aerodinamic profile. To minimize this, I sometimes placed the probe core on the front docking port, where the rocket would already become narrow and there's nothing to be done. But having a probe core there prevents docking, so I have to rendez-vous with NSN, eject the probe core, go grab the part with the Service Probe, and perform a docking with the probe in front - the kind when I have to detach the probe and get it out of the way and hope that the part will fly straight without control - and those dockings are very difficult.

So instead I stick the probe core to a lateral docking port, so that it won't force a narrowing of the rocket, but it also won't be in the way of the frontal docking port. The two small fuel tanks on the other docking ports are there for balance.

And how is sticking three parts out from the front of your rocket an aerodinamic improvement?

This is a very smart question, one that I myself haven't been able to answer. But this model did fly straight, so it worked well enough.


The Laythe lander core rendez-vous with NSN


And docks to one of the free docking ports

Now the floaters. After a failed attempt to launch all three, I go to sleep, and while I sleep I am filled with worry.

The Laythe rocket has 4000 m/s deltaV, vacuum. In every stage it has TWR >1, Kerbin. Orbiting Laythe nominally takes 2900 m/s; even with an aerodinamically challenged lander, I should be good. Should I?

Back when I did the Bolt mission, I had similar problems with aerodinamics. In that case the lander was good, but it was struck by a bug. Whatever the reason, aerodinamics was a major problem, and made a sound lander fail. So I went back to check on that mission report - keeping those diaries is useful! - and what I read did not comfort me.


The Heavy Lander had 3400 m/s, yet it failed to reach orbit by at least 400 m/s

3800 m/s would have been barely enough to orbit. And it was launching from a few hundred meters of altitude too, I'm stuck at sea level! I also have an aerodinamically challenged launcher; if it is only slightly worse than the (kraken-struck) Heavy Lander, or if I perform a suboptimal ascent profile, then I won't be able to reach orbit. And, as I keep repeating, I can't take risks. Everything must work at the first try. An additional fear came from Kerbin testing; I failed to reach orbit by well over 500 m/s. Laythe requires 500 m/s less than Kerbin, if I fail Kerbin orbit by that much, it stands to reason I should fail at Laythe too. Aerodinamic data also worries me; MaxQ is reached around 10 km altitude, but drag remains high until well past 20 km. And Laythe actually has higher atmospheric pressure than Kerbin at those altitudes; at 20 km the pressure is twice that of Kerbin. So I'll get a lot of drag at a time when I'm using the weaker terrier engine.

I'm probably overreacting. I have 4000 m/s, I should be able to push through any aerodinamic inefficiency. But for good measure, I added an additional half a ton of fuel to each of the drop tanks. Plus an aerodinamic nose cone.


Floater tank launcher

The floater tanks have a peculiar design, and I had to figure out something creative to dock them on the side; in particular, while the easier way is to stick a second docking port for the Service Probe to grab, I don't want to leave extra mass and draggy parts on the rocket. So I did develop this system; there is a cargo bay underneath the floater tank, which contains the probe core and two ant engines for orbital manuevers. The second docking port is there (in the image, covered by the open cargo bay). I grab this docking port with the Service Probe, dock the floater tank in place - I may even be able to do it with its own ant engines - and then I jettison the cargo bay, leaving the floater tank clean.


Ejecting the second stage just before orbit, for once I won't have to clear junk manually. Activating the ants, which turn out to work even inside the cargo bay


Trying to dock with the core stage of the lander. It's very crowded here


Unfortunately, it's a matter of spacing, there's no room for docking


So I stick the tanks whenever there is free space, for later assembly

I've never had a ship with so much stuff stuck hapazardly around it. Even the DREAM BIG didn't ever came close to this level of encumbrance. The original ship core, made by the five hitchhicker containers, had 17 docking ports, not counting those for the engines; yet I used them all, and then I docked stuff on top of the docked stuff!

Navis Sideralis Neanderthalensis is now at 353 parts, 79.5 tons. I have 12.5 tons of drop tanks, the rest is all material that must reach Jool. The remaining landers should be easy and light, but still, by my estimation - considering a large extra because I won't be able to make accurate manuevers - I'll need at least 400 tons of fuel tanks. I'll get to 1000 parts again! :o

10.3) Vall, Bop and Pol


The remaining moons don't require any special challenge. Vall is slightly bigger than Mun, a normal fuel tank with a normal engine is enough


The Vall lander

The Vall lander only weights 3 tons without the common lander pod, and it includes a very generous amount of extra fuel.


And its launcher

When the bad aerodinamics are factored in, 3 tons are the limit of what I can launch without extra tanks.

Notice the ant engine stuck laterally on the probe core. It's yet another experiment to make an easier docking without using the Service Probe; the docking port is aligned with that on NSN, the final push is given, then the probe core is detached (always the problem when putting the probe core on the front docking port) and the lateral engine helps moving it out of the way. I was too slow with it, though, and still had to chase the lander with the Service Probe.


Final lander for Bop and Pol, with launcher

Landing on Bop and Pol is even easier. A single large fuel tank with a single spark engine already provided 1500 m/s and TWR 0.7 (Kerbin); both are more than enough to land on both moons. I could make an even smaller lander, but I don't see much point in saving half a ton. It's all fuel anyway, I'm sure it won't get wasted.

Since this piece was light, I also sent the last part I needed: the command pod for the taxi. After some effort at imagining how all the various parts will be docked, I decided it's better to leave an engine on it permanently. Drop tanks will be stuck on top of it, and the landers on top of the drop tanks.


The Service Probe with its own manned module rejoining the station


I just love to break down a complex ship into schematics

With this, NSN is at 369 parts, 85 tons. From now on, it's all drop tanks.


Edited by king of nowhere
Link to comment
Share on other sites

Part 11: Fuel for the fuel god!

Beginning the biggest part of the Navis Sideralis Neanderthalensis: the fuel tanks. Also, trying to tackle the strutting problem


NSN is loaded with everything it needs at Jool. Now the problem is bringing it there.

For a normal mission, 2000 m/s are enough to reach Jool. Jool is an easy target to reach, especially with the moons to provide free insertion by gravity assist. Unfortunately, I can't do any of that. Reaching Jool and its massive SoI shouldn't be too hard, but there's no way I can hit a gravity assist precisely - even a moon would be challenging! Worse, I am likely to end up in a very high periapsis, possibly with high inclination too. 2000 m/s are still a reasonable estimate for getting to Jool, but I may have to spend the same amount just to enter a stable orbit. Then I will need a fair amount of fuel to wander around the moons, and then more fuel to come back - though NSN will have greatly reduced mass by then.

All in all, I'll need a lot of deltaV.

It's virtually impossible to calculate exactly how much deltaV drop tanks are worth, but I have devised a reasonable approximation: calculating the tank mass as fuel with Isp=0. This way I can just use the rocket equation with the full mass of the ship, and then multiply the end result by 8/9. So NSN weights 70 tons without the fuel, if I load 400 tons of fuel I have 400+50(fuel tank mass)+70=520 tons, and I estimate ln(520/70)*3500*8/9=6200 m/s. Or, I could estimate very roughly that after attaining Jool orbit I may need 80 tons of fuel (not including that already in the landers), and so I want to reach Jool with a 160 ton ship. The estimate is based on 40 tons to bring around the landers, which themselves have 40 tons of mass, and 40 tons to return, with an NSN stripped to basic and reduced to less than 30 tons; it's probably an overestimation, but better to be safe. In that case, with 400 tons of fuel, I'd get ln(520/160)*3500*8/9=3700 m/s.

Well, either way, you understand that I decided on 400 tons of fuel as an amount that should suffice even under pessimistic estimates.

In my previous missions, 400 tons were pocket change for my motherships. But this time, I have a limitation of 18 tons on the launchpad. I may need well over 100 flights for this.

There are no particular requirements to drop tanks; they must have docking ports to be stuck to other drop tanks, they must have fuel, they must be droppable. However, the huge number I want to launch require some conditions:

- I want big tanks, to reduce part count

- I want to send as much fuel as possible on a single flight, to reduce the number of clamp-o-trons used

- I want them to be thick, because thick, short lines of drop tanks will have less structural problems than long, thin lines of drop tanks

Drop tanks will need to have a docking port on either side. On the rocket, a thin docking port interrupting the radial size is a big aerodinamic and stability problem, so I resolved to put docking ports on the sides, as shown in the picture


Optimal location of docking ports

The biggest tank I have is the FL-T400. It's thin and narrow, so it could work very well with docking ports on the short side. I toyed with the idea of making a sandwhich-like structure with them


Project of drop tanks

That drop tank setup would have been wonderful. Six tons of fuel, only 5 parts, and thin. Unfortunately, I couldn't launch that kind of payload; too heavy, too draggy. Using two tanks side by side would have created all kinds of symmetry problems, using two tanks linearly would have made them too long and cumbersome, with high chance that they would interfere with the engines, and using only one would not be worth it.

So I went back and did a more conservative model based on the Mk 1.5 size


Drop tanks, actual launcher. It's funny that with a 18-tons mass limit I am routinely launhing 32-tons rockets

The actual drop tank section is made by the three FL-TX220 tanks and two adapters, for better aerodinamics during flight. Total fuel mass, 4900 kg. I have four binding points for tanks, I can get by with 20 of them on each one, for a total of 80. Of course, launching that payload still requires the maximum amount of lateral extra tanks I can bring. To reduce the amount of trips with the fuel tanker, I designed a new version with 3 tanks instead of two.

I did show in part 7 what can happen with docking on the ground, but I found a better way: get close, then undock the fuel tank, then cover the last meter or so until the clamp-o-trons stick.


It's not a 100% failproof system, but it does have an almost 0 chance of ruining the rocket, which is the most important thing

A nuisance of launching with lateral drop tanks is that the game does not detach the docking ports on staging. So I must select them all manually and manually detach the tanks. I learned to open all the relevant docking ports preemptively, during flight I am too busy to go chase them


Now I am ready to fly


Starting TWR is low, the first couple of tanks are spent just to pick up some speed


This rocket has decent aerodinamics, it can do a normal gravity turn


The bobcat is jettisoned, the second stage is based on a cheetah (got a test contract for it, of course)


Once more, spent tanks are jettisoned from the front

I want to spend a few words on this practice of jettisoning tanks from the front I'm using all the time in this challenge. In the real world, you never see it. The main reason for it is, in the real world drop tanks are a lot less convenient. In the real world, tanks have a fuel ratio of something like 50:1, so you gain very little in jettisoning them. Until you have a lot of spent tanks, and then you use normal staging. In KSP, tanks have a lot more dry mass, with a 8:1 ratio. So it becomes a lot more convenient to get rid of them one by one. Those two tanks are 500 kg of weight, plus 30 kg for the nose cone. Furthermore, in KSP rockets can turn around a lot faster in vacuum, so you can point the nose away from your trajectory before jettisoning; in real life, you need RCS and a lot of work for that. And finally, it would require fuel lines bypassing the payload, adding a lot of complexity. But in KSP dropping spent tanks in small batches is convenient; and then you can't stick them on the bottom, or you'd have to jettison the engine too; you can't put them on the sides, for aerodinamics. And so you put them on front.


The drop tanks reach orbit with the last of its launcher fuel


There, to further save weight, I drop the cheetah and use an ant engine for orbital manuever

Actually, I realized later that this is pointless. I have a 5-ton payload, so dropping 1 ton of engine does not make a huge difference; going from 355 s Isp to 315 s makes up most of it. I only gain, like, 8 m/s by pulling this stunt, so I'll remove the ant engine for the future. It is conveniently oriented for docking, though.


Docking to NSN. The remains of the launcher are then jettisoned and deorbited

Testing things in sandbox - with proper decouplers and probes that can hold all directions - was a lot more efficient, I managed to dock with less than 50 m/s remaining. Sure, as long as the tanks are in orbit, I can send the Service Probe to grab them- I had to do it on the second flight - but it's time consuming, and it also eats up a lot of fuel from the probe; I also had to send up extra canisters for it afterwards. I'll need to improve the launcher.

Meanwhile, I also finally face another problem I've been postponing: structural stability. NSN wobbles worse than ever, and I finally decide I'll have to get struts; if not now, before the end.

So I make another Mun landing with Liliana, whom I left on the ground in case of need. Since it only took a little extra, I also sent her in solar orbit for level 3.


Here the manuever to return to Kerbin

With the last batch of Mun science, I get General Construction.


Here's how my tech tree looks at the moment

I have still 5 Mun biomes left, plus one missing the science jr and goo experiments, for a total of 134 science points still easily achievable. I can get one more tech if needed, probably two if I complete the exploration of Kerbin.

This got me struts. I sent Bill back on Kerbin, then I sent him up in a cargo bay (See part 6) to place struts. Then I discovered with struts unlocked, I can autostrut stuff, so the whole mission turned out a waste of time.

I also got a larger decoupler, improving aerodinamics and stability on the launcher.

After the last missions I wanted some extra deltaV, especially more thrust during liftoff. I further modified the drop tanks to include a couple of flea boosters. This is actually the first time I find a practical use for them outside of the first launch.


New fuel tanker. Placing one tank in the bottom also improves a lot stability while driving, and makes docking easier


New Drop Tanks launcher


The fleas push it fast to 100 m/s. Afterwards, it's ok to have lower thrust

The modification earns a good 100-150 m/s, which make the difference between needing to call in a Service Probe at the slightest mistake, and being comfortable.

The launcher in its current configuration is a good, reliable vehicle. I launched 6 successfully and lost 2, which does not seem reliable at all, but both those I lost were for very stupid mistakes: the first got a clipping accident while loading the side tanks on the launchpad, bending the rocket and causing it to fly askew. It was an easily seen flaw, I did see it, and I decided to fly anyway. I should have recovered the rocket and start a new one. The second, I thought I could save fuel by dropping the cheetah too early; with an impressive TWR of 0.03, the ant engine wasn't enough to circularize before the drop tanks got back in the atmosphere.

So, the launcher is realible, unless I do something real stooopid.

A whole launch sequence takes from 40 to 60 minutes, depending on whether I have accidents on the fuel tanker, how close to NSN I get with the launch, and how much I get distracted doing other stuff. So I'm looking forward to 70 hours of work to load all the needed tanks on NSN. It will take approximately a month, maybe less.

at 520 tons, NSN won't have enough reaction wheels, so I want to include some in the drop tanks. Fortunately, the latest launcher iteration is good enough that I can add one wheel and it still reaches orbit, if less comfortably


The drop tank version with the reaction wheel at the bottom


Seen on the launcher, it showcases the aerodinamic problems it creates

The problem with adding a wheel is not in mass - the launcher itself had a wheel in the cargo bay, I just moved it. But, as you can see from the picture, putting the wheel someplace it would stay attached to the tanks creates aerodinamic problems. Or, I could have stuck it to a docking port on a side, which would have added mass and docking ports (and drag, anyway).

I'm planning to send one wheel for every three drop tanks, but I'll keep seeing how NSN turns around and will base my decisions on that; if it becomes sluggish, send up wheels.

Finally, NSN still wobbles badly, so I learned I could check autostrutting with the alt-f12 menu (it's not in the cheat section, so I assumed it to be fair game, just like the aerodinamic display I've shown in a few screenshots). The results are... troubling


There isn't a single autostrut in a useful place!

I would like autostruts to stabilize the weak and wobbly small docking ports, but there isn't a single one of those struts that's doing it. And apparently, strutting to grandparent or to root causes no effect at all.

I may need that EVA mission with manual struts after all.

This is probably the last update for a few weeks; I have to launch 74 more drop tanks, not much to report.

Navis Sideralis Neanderthalensis currently at 421 parts and 119.5 tons

Edited by king of nowhere
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Create New...