Temstar

For recoverable rocket, can the recovery of the launch vehicle be conducted in multiple phases instead of in one go? As in, instead of a SSTO rocket that goes into orbit, release payload then come back all in one piece. Can you have a mission architecture where a SSTO rocket goes into orbit, release the payload then separate into multiple sections where each section then makes its own way back to KSC? I have a design like that that's nearly completed.

I don't really like the funding structure in Career where all your funds come from contracts, I would prefer a more NASA approach where you are given a yearly budget and contracts then become icing on the cake if you need more. That said I like Career more simply because money is a factor. It encourages clever engineering like making payload as light as possible and making launch vehicles that aim to achieve low cost per ton to orbit. With Sandbox and Science you could always just overbuild your way out of a problem. The Vector engine for example is pretty balanced in Career because of its high cost. If money wasn't an issue then Vector and the Mammoth would be the only engines I would use on my launch vehicles.

Wait what? That's not suppose to happen, you must be executing your burn incorrectly to be dropping your PE. For a long burn you're suppose to burn prograde all the way regardless and start your burn earlier than the node, so that when the burn is completed your ejection angle is exactly the same as if you had infinite TRW and executed your burn on the dot at the node. As for burning straight at the target then cancelling out the relative speed once you are beside it, if you think about it that's basically the same as raising your orbit to meet it the next time around, only instead of meeting it after a whole orbit, you're meeting the target after only an section of one orbit (in others words an arc).

Those type of contracts have extremely long duration though, something like 20 years to complete. I find that quite realistic - when Kennedy did his "we choose to go to the moon" speech he basically forced a contract just like this on NASA with a duration of 8 years, at a time when NASA could barely launch a single person into earth orbit on a spacecraft best described as "worn" rather than "piloted". Plus if you don't like a contract just don't agree to it and let it expire, plenty more where they came from.

This, this is good design. Rule of thumb in multi-stage rocket design is that for the best efficiency each stage should contribute about equal amount of dela-V to the mission. It's easier to do in real life since stages are made to order, in KSP it harder since we only have limited range of parts to build our rockets from. For example: This two stage LV has a all solid first stage which contributes about 1100m/s of delta-V. If this was real life you'll tell the SRB maker to make you a longer, fatter SRB that gives you 1750m/s of delta-V in a single tube, and while they're at it setup proper grain profile so the thrust reduces during the burn and your vehicle TWR doesn't go up too much. You could simulate this in KSP by bundling even more SRBs than I have here and reduce their thrust to control the TWR, but it's probably easier to just make the upper stage bigger to pick up the slack.

I prefer 75km orbit, yes delta-V to higher orbit isn't much, but being that the rocket equation is logarithmic it does matter when that extra delta-V filter downwards to the launch vehicle, Particularly if the launch vehicle IS the spacecraft (ie SSTO). I don't find rendezvous much harder. Yes if you find yourself behind your target it will take longer to catch up. However instead of targetting the station exactly at your AP you could just launch slightly ahead of it so when you get into circular orbit you are ahead of the station. Then it's a matter of raising your orbit slightly so you meet up in next orbit.

Veering off during take off is usually caused by too much lift at the back of the plane, this almost always happen with delta wing planes with tricycle landing gear. The lift generated at the back causes the back wheels to lift off slightly, decreasing your AoA and forcing your front wheel into the ground. At this point your plane becomes a wheelbarrow hence why it starts snaking. To counter this what you can do is hold the joystick fully back before you start your engine. Once you start moving this lifts the front wheel and force down the back wheels

They actually do do this in real life: http://www.nasa.gov/topics/aeronautics/features/hifire.html

Woah that's massive. Tha's like a sixfold increase in diameter after inflating. It would be very useful for interplanetary motherships for aerocapture.

Nah its not a matter of resource, at the end of the movie they had baby Rangers coming out of their ears, presumably each equipped with one of those magical fusion engines. Single stage delta-v to change your velocity by two digit percentage the speed of light and then back is an unimaginable amount of energy. I'll say it again if you could make that engine in mass (and they did with those tiny rangers) you wouldn't need to manipulate gravity to lift space colonies - you'll just attach some of those engines to the bottom and shoot it up like a giant rocket.

The decoupler won't add much, it's mostly the guidance (read flywheel) and the stuff that you need for reentry (heat shield, airbrake, hovering fuel, landing legs, parachutes etc etc) that will eat into payload fraction I find. But fine I'll give aerospike another try.

It's to do with maturing as a player as well I think. When I first started my average part count was in the 400s, and that was a time when Mun was just added and the physics engine was MUCH worse than it is now. You just add all these stuff to the outside of your ship because they are nice to have and real spacecrafts probably had them. I recall when I designed my first propellant depot when docking mods came about it had more than 500 parts because I had those little ladder things all over it since I was still green at the whole EVA control thing and I reasoned real spacecrafts where you expect to do a lot of EVA would also have handholds all over. Nowadays I don't stress about the little things so much and let my mind fill in the details, so average part count is a more reasonable 200 including the launch vehicle. Stronger stock joints helped a lot, I recall back in the days you would have something like 30% of your part count being struts just holding the damn thing together during launch.

If performance does improve, people will probably complain that KSP runs like a dog with 1000 part rockets and wondered how they ever reached orbit with 200 parts. Still, improved performance would be nice.

That can't be right: Aerospike has a vac thrust of 180kN. This is an SSTO, so assuming a low lift off TWR of 1.3, one aerospike can power a rocket that's a hair over 14 tons. 20% payload means 2.8 tons of payload Aerospike is 1 ton, turboramjets are 1.8 ton each. So that's 6.4 ton dry weight not including tankage. Assuming the other 7.6 ton is all bipropellent fuel tank, that should get you 6.76 ton of fuel and 0.84 ton for tankage. So wet weight 14 ton, dry weight 7.24 ton. We'll use vac Isp 340 for rocket equation. That only gets you 2200m/s of delta-V. You will save some delta-V by using just the jet engines and liquid fuel, but you increase dry weight again from all that air plumbing. I'm not so sure without wings two jets can support a vertical climb for very long. Can you show us your test vehicle?

Seeing Rapier being so dominant as "best jet engine" even for aircrafts not intended to go into orbit feels wrong. Perhaps what's needed is an air-breathing only mode version of the Rapier. Call it the S.T.I.L.E.T.T.O engine or something.

SST = supersonic transport, eg Concorde and Tu-114 It's a huge buzz word back in the Concorde days before it was proven uneconomical.

I did actually try to make it aerospike with the same sort of thinking that if you have a good Isp all the way up it would greatly help a SSTO spacecraft. Turns out that KSP's aerospike TWR is so poor that coupled with the lack of control authority in KSP via differential throttling an aerospiked rocket SSTO is only able to lift itself and tiny payloads, making the whole thing rather moot. Funnily enough that's also one of the main reason why NASA rejected it. NASA knew that Chrysler's numbers were pretty aggressive and if critical components like the fuel tanks or the aerospikes engine cannot hit those performance specification then this being a SSTO the SERV might not reach orbit at all. If that ended up being the case NASA could well have a two decade hole in its manned spaceflight capability. That and NASA really wanted a spaceplane.

If you could make a SSTO like that you wouldn't make chemical rockets any more. It would be a bit like car makers hauling cars to the dealership using horses instead of trucks to save petrol.

Having designed a low cost per ton launch vehicle for the Light Lift category and another one for Ultra Heavy Lift, it's now time to fill the big gap in the middle. Being in the middle can be a bit tricky though - they're too big for monolithic SSTO flyback booster and too small for highly optimised asparagus giants. Some new thinking is needed here. Of course to make the upper stage reusable is easy - Nova II already demonstrated this feasibility with its Reusable High Energy Upper Stage. The problem is the larger lower stage - they get left behind in the atmosphere and their tend to cost a bomb due to their sheer size. Getting rid of them by making a single stage to orbit rocket also becomes increasingly difficult once your payload get bigger. Now SRBs are cheap, and if you look at the list of launchers in real life that use them: space shuttle, Ariane 5, Atlas V, Delta IV, etc etc I'm by no means the first person to think of them as a way to reduce the 1st stage price. But if you have played with SRB first stages in early career you would know that they are quite fiddly beasts - you either get too low TWR on lift off, or too high TWR on stage burn out or rocket losing control as SRBs have no gimbling and fins lose control authority. Then it occurred to me: why not just use MOAR BOOSTERS on a bigger rocket? Having a big payload on top means TWR swing won't be too crazy and the all solid stage will burn out while still in lower atmosphere, where fins still work. Thus I present you: Honest Jeb Medium Lift Launch Vehicle Craft file: http://kerbalx.com/Temstar/Honest-Jeb-MLLV (Craft file contains proofing payload) Specifications: dry weight: 59 tons wet weight: 268 tons cost: √88,646 part count: 99 payload: 45.5 tons to 75km x 75km orbit payload fraction: 14.51% Cost per ton to LKO: without eRHEUS recovery: √1948 95% average eRHEUS recovery rate: √770 90% average eRHEUS recovery rate: √832 From this you can tell how cheap the first stage is, considering it contributes 1/3 of the delta-V from the bottom. I must say too, that bottom stage coupled with that upper stage and lofting that payload really makes a nice looking rocket. Typical mission profile

I did originally have some vague plans to redesign the others but I'm not sure if it's all that relevant in this day and age. Back in those 0.20-ish days the only metric to measure how good a rocket is designed was payload fraction and part count per ton. Zenith LVs were pretty good in payload fraction but were never that good in parts per ton. Nowadays the picture is more complicated. Cost per ton is obviously very important now and that sort of work against asparagus design since you're throwing away a lot of engines. Similarly one of the original impetus for making Zenith was that LV-T30 and LV-T45 has better ISP than the 2.5m engines and cheaper per unit of thrust to boot, so it made sense to cluster them for core stage of an asparagus. These days 1.25m engines are worse than 2.5m engines so the main reason for clustering is gone. I did try to make a clustered core stage using a few Swivel and lots of aerospike but aerospikes are really expensive for the amount of thrust they give so it wasn't really worth it. That said I'm not setting around either. The current plan is to make a new group of launch vehicles focused on low cost per ton to orbit. It won't be a rocket family as such since partially/fully reusable launch vehicles of different size need different methods of approaching that problem. For example the smallest one: Is a SSTO rocket booster with jet engines to allow atmospheric flight to land back at KSC. The next biggest one can put 45 tons into LKO, it uses all solid first stage (7 kickbacks! It's pretty exciting to look at when launching) with a stretched RHEUS as upper stage. Currently working on this guy: Which is three winged boosters pushing an engine-less fuel tank and payload. Still having trouble to get those boosters to fly straight though as they are extremely ass heavy. Once the whole set is ready I'll release them as a pack and do a tutorial to talk about reusability.

I too find myself wanting to use fairing to "bundle up" multiple 1.25m SRBs into a large first stage and become disappointed that you can't do this. Perhaps instead of fiddling with fairing, a better solution is to have a procedural fuselage part. Works the same way as fairing does now, except lighter, cheaper, cannot be jettisoned and does not prevent actions of parts stored inside, except for extending solar panels and radiators.

Anyone seen that formula one movie "Rush"? I imagine Jeb as someone like James Hunt while Val is a more diplomatic Nikki Lauda.

Why not just leave the SAS on while translating with RCS? That way the reaction wheel contract any yaw from RCS and you're left with only translation.

You know how the space shuttle replica builders are always complaining that the SRBs we have are not powerful enough and the KS-25 Vector is too powerful and so the STS stack won't work properly in KSP? Well that gave me an idea - instead of trying to surround an ET with SRBs, why not surround an ET with robotic shuttles powered by these overpowered Vectors? The result is this - a launch vehicle consisting of a giant fuel tank in the middle powered by three winged liquid fuel boosters. This baby has already been proven to be able to achieve single stage to orbit with 81 ton payload. Once I prove that reenty and recovery of the winged boosters work it will be ready for release. The winged boosters have already been proven safe for landing by doing a lap around the runway.

Long time ago I built up this large modular Mun base which I named Von Braun City, built up of components over something like 30 launches.