Temstar

Nova's part count went down by 30 from 256 to 226 due to both new clustering method as well as better Sepratron placement allowing me to halve the number of Sepratrons. Zenith V, VII and IX benefited from Skipper and the part count decreases are pretty dramatic: Zenith V: 159 -> 115 Zenith VII: 208 -> 168 Zenith IX: 256 -> 212

A modernised set of Zenith rockets have been released. Head to the new thread to get the craft files: http://forum.kerbalspaceprogram.com/showthread.php/33381-0-20-2-Zenith-rocket-family-%28modernised-for-0-20-x-with-perfect-subassembly%29 Also included is a brand new method of setting up subassembly that overcomes the major problem of Subassembly Saver/Loader - missing struts and fuel lines.

Perfect Subassembly After posting the original set of Zenith rockets, by far the most common negative feedback I got was "your rocket doesn't work properly! I run out of fuel way before reaching orbit and the rocket was pretty wobbly". 90% of the time I don't need to ask for a screenshot or more information, I simply tell them "check the fuel lines and struts against the proofing payload copy, you'll missing some due to Subassembly Saver/Loader". If you've played with Subassembly Saver/Load you will know what I'm talking about. One of the big bugs holding this amazing tool back is its inability accurately render struts and fuel lines. You take off the booster section of a Zenith rocket, save it as a subassembly and then load it up under an payload and chances are you'll get this: See that, the fuel lines are all messed up. To get the rocket to work properly they will have to be recreated. And while that's fairly easy for me since I know the ins and outs of my rockets like the back of my hand I can imagine this is a mighty annoying task for someone who's using a Zenith for the first time. So for the modernisation I decided to tackle this issue. I've heard that this issue was due to the order the parts are rendered in - if the destination of the strut/fuel lines are not yet rendered by the game's engine due to the order they appear in in the subassembly file then the part goes missing. So the first thing I tried was opening up those .craft files in the subassembly folder and try to manually edit the file and put all the struts and fuel lines at the bottom. Save and load up the subassembly - no dice, still missing fuel lines. But from scrolling through that craft file I can see what the problem is. See when you save a subassembly you create a .craft file, only it doesn't resemble the original .craft file generated by KSP at all - it's all in some messed up order. No wonder when you then load up that subassembly some stuff goes missing But while I have the .craft file open, I noticed something: "Hmmmm, looks like this thing has the decoupler as the root part, that looks kind of like a KSP 0.20 .craft file now that we can have almost any part as a root part." And then the apple falls from the tree, the bathtub overflows and I had a moment of epiphany: So if a subassembly is a 0.20 craft file, except with all the parts in messed up order, wouldn't that mean a KSP 0.20 craft file, generated by KSP and able to be correctly loaded by KSP is actually a subassembly with the parts in the correct order? To the Vehicle Assembly building! So we start off with Zenith VII We remove the proofing payload and use the new "delete root part" trick to make a new decoupler root part and fit it to the rocket and save the craft. Copy the .craft file from save folder into the subassembly folder under lifters Start a new craft and upload up Zenith VII that we've just saved Hallelujah! It works! So there you have it. To have the perfect Zenith rocket subassemblies just download that subassembly file, extract all the .craft files to your: KSP\GameData\subassemblyLoader\subassemblies\Lifters Folder and there you have it. All Zenith rockets as perfect subassemblies. No more dealing with missing struts and fuel lines.

Some time ago in 0.19.x days I made the Zenith rocket family and it was good. But time has moved on since then and with 0.20.2 now out with the new Skipper engine I thought I'll do an upgrade to the line up. The common features of the Zenith rocket family remain the same: >15% payload fraction asparagus staging with reliable staging sequence clustered engine core for sustained high Isp RCS system for on orbit attitude control, in case payload has low maxRot probe core and electrical system (including small solar panels) for space loitering and self-deorbiting So what are the upgrades? Supernova SHLLV now officially added and recognised as part of the Zenith family, bringing the max lift capacity to 160 tons Skipper engine replacing some of the cluster engines on Zenith V, VII and IX, resulting in dramatic part count decrease for a small decrease in payload fraction No more hiding parts in the decoupler - I find that many payload often need to take advantage of that space in the decoupler, so that space have been cleared out for payload use Sepratron placement optimised - now all Zenith rockets have powered separation including Zenith II other minor improvements mostly to optimise part count. Notably Zenith IX and Nova's double cluster now constructed with a new, better method perfected subassembly files, no more missing struts or fuel lines. See the second post for more details So as with before, the common shared design of the Zenith family results in a similar ascent profile: 1. If you are launching a Zenith rocket with big gimbling booster engines (ie, Zenith V, Zenith VII, Zenith IX, Nova and Supernova), double tap 0 to lock gimbal. Then turn on ASAS, throttle up to max and lift off. 2. Ascend directly up and stage asparagus pairs once they run out of fuel. If you fly with either the proofing payload included in the original .craft files or ones with similar mass then all but one pair of booster should run out just after 10k altitude (that's two pairs jettisoned for all except Zenith III, which will lose it's 3rd pair of boosters at 10k). Once you're down on your last pair of asparagus booster pitch down slowly to 45 degrees for gravity turn. 3. By the time your last pair of boosters run out you should be nearly horizontal. Use the core stage to continue to build your orbital velocity. 4. Upon reaching stable orbit. Decouple the rocket from the payload, use [] to change control to rocket, flip it around with RCS and perform an deorbit burn with the remaining fuel to ensure no space junk. All Zenith rockets have been tested to have enough fuel for deorbit burn while using the proofing payload. How to put payload onto of a Zenith rocket: The best way is to install Subassembly Saver/Loader. Once you're done follow these steps: 1. First you have to figure out how heavy of a payload you have: Here we have a giant spacecraft designed to house 134 kerbals in orbit. At a hair over 110 tons we can see from our table of rockets at the top that this is just under the maximum payload of Nova. 2. So open up your subassembly loader, find the Nova subassembly 3. Hit load and you'll have the Nova booster under your mouse, So simply attach it under your payload. As with other parts in KSP sometimes the parts won't attach properly in one go, so just attach it to somewhere else, pick it up again and place it where you want it to go. There, 4627m/s coming out of Nova to send our spacecraft to orbit. Adding three struts between our unwieldy payload and Nova to stiffen up the connection and we have the craft I flew successful above. But wait Temstar! you say, you didn't explain how to actually get those subassembly installed into Subassembly Saver/Load for loading up under payloads. Aww see here is something different from 0.19, while upgrading the Zenith rockets I've discovered a major breakthrough to how to use Subassembly Saver/Load which overcomes an old problem with struts and fuel lines. For exactly how that's done you'll just have to read on. Craft files: Zenith rocket family - http://www./download/3xteo1dm7j5okba/Zenith2.rar Zenith subassemblies - http://www./download/9wmk6o78o4hcnge/Zenith2_subassembly.rar

I was going to release it after I figure out all the landers and launch vehicles and all that kind of stuff to actually get it all to Minmus and assembled there. Would you rather wait for that or do you just want the raw pieces now?

Pressure feed cycle engines require extremely heavy tanks to withstand that pressure. When you think about it, a burning rocket engine has inside its combustion chamber very high pressures which is what shoots the exhauster out the back at high velocity. When you have a pressure feed cycle you basically have two unobstructed openings in that combustion chamber - one in the back where the nozzle is, and one in front where the fuel and oxidiser are injected into the chamber (kinda sorta, fuel and oxidiser injector are pretty complicated and usually look more like a shower head rather than a single hole, but you get the idea). The hot gases is going to try to push out in both directions so in order to force the fuel and oxidiser under pressure into the combustion chamber you have to have all of your plumbing and fuel tanks able to withstand an even higher pressure than what's inside the combustion chamber. This results in very heavy tanks and rocket hardware compared to turbopump feed engines. Pressure feed cycle engines have poor thrust to weight ratio due to their heavy hardware, however they are mechanically simple and reliable (as opposed to rocket turbopumps which are some of the most difficult to build and expensive pieces of machinery in all of mankind's history) so is generally only used for small orbital manoeuvring thrusters such as shuttle OMS engines and Apollo's descent, ascent and service module engines.

Okay this is how you do double clusters: First you do the inner cluster using the conventional cubic strut method. Remember if the game doesn't allow you to place the four engines in one go you can always place them one by one without symmetry and they will work. Now then, to set up the outer cluster, grab a tail section and stick it on the side of the rocket. Put an engine on the button of that tail section. Grab the tail section and turn on symmetry. You will be able to stick them on as an outer cluster up to 6x symmetry However if you specifically want to do the 8x outer cluster like Nova it won't let you to do it nicely in one single step due to clipping. So instead... First you install 4 engines via 4x symmetry Then make another copy of your tail section + engine (what I do is I take the current set off the rocket to the side and hit ctrl+z) and stick them on as 4x symmetry in the gaps There you go, 12 engines. Nova uses an older, less part efficient way of doing this which is obsolete. I'm in the process of converting Nova and Zenith IX to this new method to save part count.

The idea with the little pieces is so that the base is modular and can be rearranged should I even change my mind. I'm also a bit weary of having to pick up huge pieces with the rover since it will make driving really difficult. To land them on Minmus I'll probably use landers with docking ports on the side. Dock the pieces to them in VAB and land them, then undock and drop them down one by one. For the bigger pieces like the modules I'll probably attach jettison-able engines radially to them, land and then ditch the engines.

Final engineering problem for my Minmum Base resolve. Behold: Crawler-transporter The completed base: Notice the base has two launch complexes, shown docked with a tanker rocket and a crew shuttle. To service the two complexes three crawler-transporters are used. It takes two crawlers to move a ship but only one to dock a craft to the base. The base at night: Now I just got to figure out a way to get all that to Minmus.

No if you get the height down correct you should be able to dock the module to the base directly while it's still being carried by the rover. Of course there is usually a small amount of height mismatch (usually the module being carried by the rover will be slightly lower). What I do is I will switch to the base and lower the nearest pair of legs on the end of the thing I'm docking with. This causes that part to droop down slightly due to gravity and allow a smooth docking. One other trick is just to bump the two ends together with some force. The magnetic attraction will be able to over come very small vertical misalignments when you bump. This is the rover I used for base construction: This is the result:

Eve is also fascinating because its mining process is necessarily involved due to the logistics to carrying the finished product to orbit. On Duna distilling rocket fuel will be a simple matter of drilling down the surface (or better the ice cap) to reach water ice. melt it, siphon it up and split it into fuel and oxidiser via electrolysis and you're ready to pipe it up to tanker rockets. But on Eve your rocket for the final product is going to be sitting at a launch complex on top of a 6.5km mountain top to minimise that delta-V to orbit while your ocean is at... well sea level. That means some kind of rover is needed to cart material in between the two areas. You have the economic question of where is the best place to build the Blutonium refinery (seaside vs mountain top vs mobile) to minimise tanker trip, balanced against hardware required to send to Eve. Then there is the actual problem of refining. You have to siphon the water out of the ocean, dry out the liquids and separate out the Blutonium. On earth nuclear reactors are generally designed to run on natural or low-enriched uranium. But since the actual energy only comes from the fissile isotope (Uranium-235 on Earth, Blutonium-238 on Kerbin) it becomes economical to only lift off Eve with a final product of highly-enriched, weapons grade Blutonium with very high concentration of Biutonium-238 isotope. That means you also need a large and complex centrifuge facility on Eve to enrich the raw Blutonium. Man I sure hope the Kerbals know their nuclear physics, because you can't just put a 500kg lump of pure Biutonium-238 on top of a rocket - you'll exceed the critical mass and either cause a meltdown or an explosion.

Unity's heat transfer from part to part depends on the distance between COM of the parts. There's a maximum distance for part to part heat transfer. Because the orange tank is so long it's COM is far away enough from COM of the engine that the engine cannot transfer heat to the tank, hence why hot engines (mainsail, LV-N) directly connected to orange tank tends to overheat. The trick is to stick a grey tank underneath. If your only need about an orange tank worth of fuel then use the smallest 800L pancake tank.

You're refining blutonium from Eve's ocean. Blutonium is a fissile material so you only need a tiny amount of it for enormous reactors. Nuclear bombs for example only use a sphere of plutonium about the size of a baseball.

I did a two men Eve landing and return mission using this 50 ton lander: Capable of returning two guys to orbit if it landed at at least 5km altitude (highest peak on Eve is about 6.5km). I hear in 0.20 an EVA Kerbal now weigh three times as much (up from 32km to 90kg) so for the same lander to take up three guys some modification (or just landing at a higher altitude) may be required, but the basic idea is sound. Replace ladders with the seats. Of course if you land the thing on a 6.5km mountain top you'll still need to drive your Kerbals to the lander, so you might also want to bring a rover on the same trip. Drop the rover separately from the rocket as close as possible to your downed crew. Get them to pile in and then drive to the lander. The craft files are in that thread if you'll like to play with them.

They spin down the wheel. Of course spinning down cause the craft to rotate in the direction of the spin. The RCS system is then used to counter this spin.

In versions 0.17 and earlier, SAS was indeed a reaction wheel as well as a PID controller (but without the I, so it won't lock you into a heading just like how it works now). When SAS is engaged it's controller will dampen rotation using P and D and the onboard reaction wheel. Back then in the staging list each SAS will actually show up with a "torque" bar underneath to show how much force each SAS's reaction wheel is exerting at any moment, like wise the manned pods also have a torque bar. When you steer you can see all of those torque bars fill to maximum as all the wheels engage. When ASAS is engaged on a ship the ASAS's PID controller overrides any SAS + command pod PD controller and issues commands to all the reaction wheels to keep the ship's heading. When it does this you can see the torque bar fluctuate as ASAS issues turning commands this way and that. Fast forward to 0.18 and Squad decided to decoupler the SAS reaction wheel control from manual control. Thus nowadays SAS's reaction wheel will only respond to commands issues by SAS or ASAS, it will no longer contribute its torque to manual control input. Why was it done this way? If I were to guess I reckon Squad thinks that SAS and its powerful (relative to real life) reaction wheel makes it too easy to steer automated crafts if you could just slap some on and use it for manual steering. Since in real life most altitude control are done via RCS systems they wanted us to follow suit and mimic that. The comprise is that for the larger manned crafts one would assume there's enough room to install big reaction wheels (ala ISS) so they didn't touch command pod's reaction wheel and still allow players to steer using those. Keep in mind almost all spacecraft every flown in real life used RCS system for altitude control. The only exception I can think of is ISS with it's control moment gyroscopes. Even then the ISS CMGs imparts very little torque to the ISS and is only used for holding the space station at a fixed attitude relative to the surface of the Earth. The CMGs also become saturated with stored momentum and must be dumped regularly with the help of RCS system. Thus when you think about spacecrafts in KSP you should really think "why are manned spacecraft with their magical reaction wheels so easy to steer?" rather than "why are unmanned crafts so hard to steer?"

It's kind of annoying that people refer to LV-N as NERVA though. LV-N is a NTR - a Nuclear Thermal Rocket. NERVA is a AEC & NASA joint project to develop a NTR. A NERVA rocket would be referring specifically to a rocket engine (eg NERVA XE) that was developed by the NERVA program. The Soviet Union also had program to develop NTR technology and it would be wrong to refer to their engine as a NERVA.

So I did some further work on my modular base:

Tylo needs about 3000m/s to land, then 2500m/s to take off again. The tricky thing is this has to be done with pretty high TWR, else gravity drag is going to further eat into your delta-V. Trouble with an universal lander, Eve aside is Moho. Moho requires up to 7000m/s to reach from LKO. It will be exceedingly difficult to bring a Tylo-capable lander to Moho consider the size of a single stage Tylo lander. To actually land on Moho and return to orbit is about as difficult as Vall - ie not very difficult. So to get that 7000m/s with a reasonable sized tug you want your Moho lander to be small.

Ion engines are also excellent for kerbolar observing probes intended for low orbit: Circularising at low Kerbolar orbit takes extraordinary amount of delta-V and an ion stage can give you 10km/s fairly easily. Close proximity to the sun also boosts the output of your solar panels by many times their normal Kerbin value so you don't need giganator panels to further mess up your TWR.

I'm working on a Minmus base prototype: Construction is done by this rover: More buildings to come.

It's better to shift a lot of those delta-V into lander stages instead of keeping it all in your mothership. If I were to do this mission this is what I would do: 1. Mothership with three landers docked arrive in the Jool system, aerocapture around Jool to Laythe orbit altitude, aerocapture around Laythe to brake into Medium Laythe Orbit 2. Undock the three landers, each lander perform a moon-to-moon transfer to their target. Land and then get back to low orbit of each moon. 3. All landers perform another moon-to-moon transfer back to Laythe to dock back with the mothership, taking advantage of aerocapture at Laythe 4. Mothership depart Laythe during a Kerbin transfer window to come back. You'll have to make sure you hit Laythe ejection angle, Jool ejection angle and Kerbin phase angle in one burn. For the landers, you may want to consider a lander/tug model where before their deorbit burn, each lander leaves an orbital stage in orbit with the NTR and fuel needed to get back to Laythe and land only the gear required to land and take off back to orbit to meet the tug stage. If done this way you can actually do both Bop and Pol in one go with a single lander. In fact if you go for the lander/tug model you can actually visit all three of your target moon AS WELL as Laythe with a single lander: 1. Mothership arrive at Jool for aerocapture, arrive at Laythe for aerocapture and insert into LLO 2. Laythe lander undock and makes a parachute assisted landing on Laythe 3. Mothership undocks two simple small fuel tankers, the tankers make a moon-to-moon transfer to Vall and Bop 4. Laythe lander takes off to redock with the mothership, the top stage of the Laythe lander is now your airless moon lander 5. lander refuel from Mothership and is docked with a tug stage. Tug stage fire to send the lander to Vall 6. at LVO, lander undock from tug and make a Vall powered landing, then return to LVO to dock with the tug stage 7. Tug stage and lander dock with tanker 1 to refuel 8. Lander and tug make a moon-to-moon transfer to Bop 9. Lander undocks, makes a landing, get back to orbit and dock with the tug stage. The two then dock with tanker 2 to refuel 10. Lander and tug make a moon-to-moon transfer to Pol 11. Lander undocks, makes a landing, get back to orbit and dock with the tug stage. 12. Tug stage fire to send the lander back to Laythe. Tug is allowed to crash into Laythe while lander aerocapture into LLO to dock back with the mothership. To give an example of this kind of lander + tug stage, here is my Gilly lander which I used as part of my manned Eve landing mission: Lander and Tug undocking from mothership Lander + Tug assembled in LEO Lander undocking in Low Gilly Orbit Eve landing Re-docking with tug stage Re-assembling with the mothership. For Bop and Pol your lander could be as small as my Gilly lander here. For Vall you will need something more beefy but you might also want to consider the option of building the Vall landing stage as another separate craft to be sent to Vall ahead of the lander. The lander then just docks onto of the Vall landing stage and ride on it for that leg of the mission. I once did a manned landing mission to all four Jool moons (Pol wasn't available in 0.17 when I did it) using a fairly similar architecture as above. The Laythe lander was reused for Vall by refuelling with the mothership waiting in Laythe orbit. Unfortunately that mission report got eaten by the forum purge. Here is a shot of the mothership and three landers I used for that mission:

Huh? You're not trying to dock with the Nova core stage still attached to the payload are you? Why would you need translate in a booster rocket?

I think the Skipper is very well balanced and Skipper+3/4/6 LV-T30 clusters are pretty good. However I'm of the opinion that for the core stage of an asparagus the traditional LV-T45 + LV-T30 clusters are still a bit better - more part count for a bit better Isp. Since the core stage burns for a very long time the higher Isp is more important. However I think Skipper is an excellent choice for boosters around the core in cutting part count since the boosters burn for a shorter time so Isp is less important. Where as saving say 6-9 parts by replacing three LV-Txx engines with an Skipper gets magnified by symmetry and results in significant part count saving while getting you roughly the same performance. Here for example is the new Zenith IX I'm working on: So still the same nine engine core cluster, but the boosters each have three engines replaced by a Skipper for a significant lower part count. Payload fraction is about 0.4% lower than the original but it's mostly to be blamed on the new tail sections I added for the two LV-T30 on each booster and not the Skipper. I did try clustering with Radial Attachment Point and the performance is excellent but the jarring look around the engines was bothering me too much.

To get back to Low Eve Orbit from the surface requires a craft that can roughly go from surface of Kerbin to LKO, then aerobrake down to land, then take off and get back to LKO again without refuelling. It's about 4500m/s to get to LKO. To get off Eve to 100km you need a minimum of: 7500m/s from 6.5km high mountain top 9000m/s from 5km high mountain top 12000m/s from sea level So generally you need a combination of very clever engineering, landing at high altitudes and brutal power to get back. I managed to do a 2 man landing and return mission to Eve in 0.19, this is my return craft: It's an open-air rocket with the Kerbals hanging onto ladders on their way back. Full details on the mission and craft files here: http://forum.kerbalspaceprogram.com/showthread.php/30909-Two-men-Eve-landing-and-return-%28image-heavy%29