Starman4308

You can make a spreadsheet, but to save you the hassle of manually recalculating delta-V each time you change something, I strong suggest either the Kerbal Engineer or Mechjeb mods, which will give you TWR and delta-V as you're building the rocket. Really, once you think about it, KSP kind of is rocket science, with some of the fiddly engineering details abstracted away. One suggestion: make your lander wider. Put some fuel tanks radially around the bottom (connected with fuel lines) or something, because the wider the base of your lander, the more stable it is and more likely to not tip over. One little trick if your lander needs a lot of delta-V* is to put those radial fuel cans and the lander legs on radial decouplers, so that when it's time to leave, you can shuck the mass of both the empty fuel cans and the lander legs. *And if you don't plan on re-using the lander. That's a rather important consideration if you planned multiple landings. EDIT: Also, I think Pecan may have messed up a little bit in his explanation: for the purposes of calculating delta-V, G is always Kerbin gravity (9.82 m/s^2). Isp is exhaust velocity, and the form reported in KSP is exhaust velocity divided by surface gravity. That was done way back in the day to facilitate communication between those using SI and Imperial units, because an Isp reported in seconds is independent of whether you're using meters, feet, parsecs, etc. One way to think about Isps given in units of time is that it would take Earth/Kerbin surface gravity this much time to cancel the velocity of the exhaust gases.

I've enjoyed the use of Real Fuels with the stockalike configs: the decreased weight of fuel tanks and engines really helps you get to orbit without excessively ballooning the size of your rockets. About the one comment I have is that you will probably need to increase TWR faster than you used to as you go down stages (which I think has to do with the fact that the ratio of per-stage wet/dry masses is steeper with Real Fuels). Also, will probably use this repackaging the next time I start a new career. Thanks for this: 6.4x Kerbin is one of my favorite mods.

Pretty sure that thing is only telling you about TWR (Thrust-to-Weight Ratio), not delta-V. You can stick an Oscar-B on top of a Mainsail and get a TWR of 25.14, but the fuel won't last even a tenth of a second, and you will have an apogee of ~80 meters. You need enough TWR to resist gravity (for stock, around 1.6-2.0 launch TWR is the sweet spot for efficiency), plus enough delta-V to reach orbital velocity (for stock, around 4400 m/s, around half of which is spent just fighting gravity and atmosphere). I strongly recommend using Kerbal Engineer or MechJeb, which will give you combined TWR/dV readouts. You can let TWR slip a bit as you go upwards: it is plenty possible to circularize on TWR < 1, but launch absolutely requires you fight the full force of gravity and still accelerate upwards. EDIT: Just to be thorough, delta-V is how much you can change your velocity, and is related mostly to fuel mass vs. dry mass (look up the Tsiolkovsky rocket equation), while TWR is how quickly you can change your velocity. In vacuum, there is little reason to have much TWR: while it's nice to not have to burn so long, and the Oberth effect favors short burns, most of the time, you get more overall efficiency by having lighter and higher Isp engines.

I would find that one unlikely: the obvious question is "why?" If your astronauts aren't going to go out, collect samples, and do other things you need a human for, why not stick the same instruments on an unmanned probe and reduce the cost ~95%? Ceres and the moons of Jupiter are much more likely candidates: there is an actual purpose to sending people there beyond "we can".

There is a wonderful guide to building spaceplanes out there. Specifically for your craft, though: wheels should be straight, rear wheels should be just behind center of mass (as close as you dare*), and center of lift should be behind, and potentially a bit above, center of mass. *The big danger is your CoM shifting backwards as you burn fuel. I use RCS Build Aid to check where my dry CoM is. Yes, it is a pain to balance it out: if you ever happen to pick up Real Fuels, I think you can add lead ballast to help even things out.

Linux can be a bit wonky to set up: look at the Linux thread in the unmodded support forum to figure out how to get set up, particularly if you're using Steam. Windows x64 is so unstable that many mod authors have entirely disabled their mods for it, because they are fed up supporting such an unstable build. The only way for you to get the mods to work for Windows x64 is, depending on the mod author, either ask them in a PM (acknowledging that they won't help you if it breaks), or to delete the check from the source code and recompile (again, expect no support).

After many trials, tribulations, and added struts, I got my first manned mission sent to Minmus in RSS. My grand plan of recovering my first stage was foiled by the first stage having too much dV and burning up on reentry, but otherwise, I launched beautifully into a 130x150 km orbit with little wasted dV, and set up my Minmus transfer. I had made but a single mission-killing miscalculation. In stock, it takes just over a week to get to Minmus; a single small TAC life support canister is more than enough to see a Kerbal to and from there. In 6.4x RSS, however, the transit time each way is ~22 days, and at least two cans would be needed. Bill is safe, however, thanks to having more than enough delta-V to cancel the Minmus transfer the power of revert-to-VAB.

I'm not sure which thread to stick it in, but whenever I combine Engine Ignitor, Real Fuels, and the RF Stockalike config, I am getting tremendous slowdown in flight. I haven't checked if it's raw part count or just number of engines, but any large rocket chugs the computer so badly KSP is essentially unplayable. It's easy enough to reproduce it: I stick a Stayputnik on top of an FL-T800, an LV-909, a decoupler, and below that, a barrel of 19 LV-T45 engines each fed by 4-5 FL-T800 tanks. If it helps, out of sheer habit, I set the tech level to 3 (it's where my career save was when the lag monster hit). OS: Linux x64 Minimum tested mod set: Engine Ignitor, Real Fuels, Real Fuels Stockalike config (all latest versions, 3.4.1.1, 8.1, and 1.0.3 respectively). Engine Ignitor won't lag by itself, and RF/RF Stockalike won't lag by themselves, but the combination causes hideous chugging (CPU is running at 100%, and it is very very slow). Log file Last comment: I'm not sure how significant this might be, this occurs regardless of whether I extract the pressurized fuel tank config, the engine ignitor configs, or both.

Bug report: for duplicated tanks, I can add infinite fuel. Instead of getting an option to remove all fuel tanks, I can click the same fuel mix repeatedly and just add fuel forever. OS: Linux x64 Reproduced using only RealFuels 8.1 and RF Stockalike 1.0.3 Reproduction steps: take any craft, plunk down a tank, duplicate it (potentially repeatedly and recursively: I was duplicating 1 to get 2, 2 to get 4, 4 to get 8, etc), add an engine, and then start adding fuel to the tanks. Log file Possibly relevant mods: RealFuels 8.1, RF Stockalike 1.0.3, FAR 0.14.3.2, Engine Ignitor 3.4.1.1, DRE 6.2.1, Kerbal Joint Reinforcement 2.4.4, MechJeb 2.4.0, Procedural Parts 0.9.19, Procedural Fairings 3.10, RSS 8.1.2 (6.4x config 1.1.1) Probably not relevant: Trajectories, TAC Fuel Balancer and Life Support, Chatterer, Kerbal Alarm Clock, NavHud, RCS BuildAid, Sane Strategies, Stage Recovery, Zero Point fairings Definitely not the cause: KW Rocketry (I had experienced this prior to installing KW Rocketry)

There is a wiki, you know.

It's more like a tight window of success sandwiched by a vast sea of failure (started too low or lost control) and a somewhat less vast sea of mediocrity (started too high). Granted, your main point is correct: don't be afraid to revert to launch until you figure out the right gravity turn for your rocket's TWR profile. Anyways, quick question: after I've punched through the lower atmosphere, should I scramble to get to vacuum, or loiter a bit in upper atmosphere to build up some horizontal velocity while abusing body lift for everything it's worth? I've had issues efficiently circularizing, and I'm wondering if it's because I have an overly steep ascent (usually ~45 degrees until I hit a 110-120 km apoapsis) through upper atmosphere. Note that I am using the 6.4x RSS config, so I need to build up ~6 km/s of orbital velocity.

I'd look up various online guides to orbital maneuvers, planetary transfers, etc, and additionally watch how MechJeb does things like plan a planetary transfer. Other than that, the stock solar system is pretty easy to navigate: all you need to do is a bit of proper staging. Live by the rocket equation (delta-V = Isp * Gm * ln(mw/md)), use MechJeb or Kerbal Engineer to calculate how much delta-V your rocket has, look up the Kerbol system delta-V map on the wiki, etc. What might work just to get you started is to build a small, simple rocket with a MechJeb AR202 case (plus batteries, plus solar panels, plus parachutes, plus lander legs, etc), and then light up the debug menu (alt-F12 on Windows, right shift-F12 on Linux, something else on OS X), click infinite fuel, and, as mentioned before, try to figure out how MechJeb does ascents and plans maneuvers.

Right-click, "run test". A few part test contracts aren't completed via staging.

Almost any contract in Kerbin's atmosphere is trivial: just stick the part on a liquid engine (or stick liquid engines onto the part), boost straight up from the launchpad to the desired speed/altitude regime, trigger the contract, pop some parachutes, and recover. Because you can get to basically any height in the atmosphere (or poke your head just above the atmosphere for suborbital) on a single stage without even trying very hard, you can have 100% (possibly 98% if you land outside KSC) recovery, and the fuel costs are basically pennies. There are also SSTO designs (particularly spaceplanes) where you can trigger orbital contracts for pennies' worth of fuel, and skilled players can get to ridiculous places and back on a single stage. Of course, once you're outside Kerbin orbit, the payouts tend to be more than sufficient for a conventional, several-stage rocket design. There are also the aforementioned tricks of leaving little probes in orbit around Kerbin (and all sorts of other bodies) with a thermometer, and leaving Kerbals landed on each planet so you can trigger plant-flag contracts with ease.

Congratulations: you've re-invented Isp*. *In order to simplify things between metric and Imperial units, rocket scientists threw up their hands and divided by gravitational acceleration, getting a measure independent of metric/Imperial units. Isp as listed in the game is related to exhaust velocity in this fashion: "it would take Earth/Kerbin gravity this many seconds to slow the exhaust back down to 0 relative to where it started". Anyways, for the OP's question: efficiency is a combination of mass and Isp. For the rocket equation, Isp is literally there in the equation as a component of exhaust velocity, and mass of the engine is part of your dry mass. If you're just trying to scoot a little 50 kg Stayputnik probe around, a Poodle is a bad choice, because the Poodle engine is 40x heavier than your payload, so you spend most of your fuel moving the Poodle. Unless you have an absurdly enormous fuel tank to go with it, the lesser mass of an engine like the Rockomax 48-7S or LV-1 will result in more overall efficiency*. *Plus, you need to lift the engine up there in the first place, which is a second factor strongly favoring lightweight engines. The only two theoretical reasons to go bigger are to get higher Isp engines (such as the LV-N) or to improve your TWR, to better exploit the Oberth effect, not have to spend 30 minutes of real time on your burns, improve landing/ascent speed and efficiency, etc. There are practical reasons, such as "I find clusters of tiny engines ugly" and "Sure, the 48-7S is perfect, but I haven't unlocked it yet" as well.

How do you manage that? Orbital velocity is almost 6 km/s for LKO, so either you have the world's most perfect ascent profile, your 6.4x is bugged, or you mis-calculated your delta-V expenditure. I usually use around 8 km/s, though I will freely admit I'm nowhere near good.

RSS has very few technical dependencies. For any sort of good gameplay, however, you're going to want FAR or NEAR (preferably FAR) and probably RealFuels*, Procedural Parts, Kerbal Engineer or Mechjeb, and Procedural Fairings (possibly Zero Point fairings). Without those, your rockets will balloon to absurd sizes to fight through the souposphere with the absurdly overweight stock tanks and engines. I am personally enjoying the 6.4x config with all the above and the stockalike Real Fuels config. *If you don't want to screw around with worrying about all the different fuel types, you might instead use the Isp balancer mod to improve the performance of stock parts. Granted, your rockets will still be much larger than they used to be, but it's not as steep as going for the full RO experience right off the bat. For example, I'm working on an unmanned mission to bring 2 Science Jrs and 4 mystery goo pods to the Mun and back: the rocket currently has a mass of 1000 tons, and I'm still 1 km/s short of my delta-V target of 17.7 km/s. I shudder to think of what monster I would have to design with the full 10x scale RSS.

As with many: much abuse of F5/F9. I downloaded Realism Overhaul and then uninstalled it. I once got impatient and infinite-fueled my way to a rendezvous. I subjected my Kerbals to explosive decoupling. Sadly, either FAR or DRE is keeping me from doing it anymore. Now that I've got 6.4x RSS, I'm too afraid to send live Kerbals anywhere, and am doing a 99% unmanned space program. I never re-use booster designs, resulting in explosions (escape, revert-to-VAB). In sandbox mode, I have slain literally hundreds of Kerbals trying to build spaceplanes which don't simply slam into the water past the edge of the runway.

If it winds up becoming a mod, I'd probably try to have an "easy config" where you set a multiplier for getting to LKO (it takes X more dV, thus your rocket will be a flat e^Y=X times larger), and a general scale value which gets plugged into a function to estimate generically how much more difficult it is (so, for the 6.4x RSS config I use, you would plug 6.4 into the config or GUI window, and it would automatically figure out which multipliers to set). My idea was to set a flag variable on each Kerbal: "I've planted a flag for a contract, I can't do that any more until I've flown back to Kerbin and been recovered". While "instrument can be re-used in multiple biomes for multiple contracts" is probably better for gameplay, that's a lot of flags to set for each instrument: having just a single "I've been used for a contract" would shrink the complexity of it all. It could also be generically replaced with "take a rocket with a size 3 black box module to point X, activate" type of contract.

Mostly what Northstar said. The 6.4x config is just about right for getting a "stuff is big" experience without requiring absurdly gigantic rockets to get anywhere. FAR helps you get to orbit and is less ridiculous than stock aero, Procedural Parts helps you tweak fuel tanks to size without stacking up too many stock cans, Procedural Fairings helps keep your payloads from playing havoc with FAR, and Real Fuels with the stockalike engine config lets you build engines a bit more efficiently (much lighter tanks, and hydrogen-oxygen stages get quite good Isp). In all, what I consider my "core mods": the 6.4x RSS config, Deadly Reentry, FAR, Real Fuels with stockalike engine config, RealChutes, Procedural Parts/Fairings, TAC Life Support, and Mechjeb. If you use DRE, you can either play with the config file to get parameters which work with the stock heatshields, or download the Realism Overhaul config, and potentially adjust upwards to account for slightly easier re-entry.

First stack decoupler comes with the first node (5 science), and the first radial decoupler I think is in Stability, which is another 18 science. If you're wondering what the "simulate decoupling" thing is: it's abusing a quirk of overheating mechanics. If you overheat an SRB (Solid Rocket Booster), it explodes, so if you deliberately design your rocket so that you can overheat RT-10s the moment before they run out of fuel, you can eliminate the weight associated with them. One of Scott Manley's videos features this. For now, though, I would suggest sticking with regular decoupling: it's not too hard to get a suborbital or maybe even orbital flight without decoupling. Be sure to abuse science: get a crew report on the launch pad, EVA and get a report "flying" over the coast, get another crew report shortly after liftoff, get a crew and EVA report from upper atmosphere, get a crew report from space and as many EVA reports from over different biomes as possible, and do as much as you can on re-entry. Remember that, with starter tech, the only way to regenerate electricity is by firing rockets, so there's a limit to how much you can transmit after you finish getting to orbit. Different EVA reports can just be stored infinitely, crew reports take 30 electricity to transmit if I remember right.

Small quibble: I'm pretty sure that, for your typical FAR ascent, you are never even close to terminal velocity. If you're going up, you pretty much by definition need a TWR >= 2 to hit terminal velocity. The issue is that flying at terminal velocity is a very bad idea in FAR, because in FAR, the primary limitation is aerodynamic stress: go too fast, and the atmosphere will be very unforgiving of even the slightest deviation from prograde, and if you have DRE, your rocket is also likely to burn up. You build for max-Q, not for terminal velocity. Ferram wrote a couple good articles, originally for RSS, but the theory is the same for any FAR installation. The big deal is that, with FAR, flying anywhere except directly prograde will cause aerodynamic wobbles, and if you go too far, your rocket will flip over with usually-catastrophic results. As such, your gravity turn is designed to gradually go down to a reasonable angle once you hit upper atmosphere, where you can start to have your own ideas about where your rocket should be going without worrying about aerodynamics.

If you're using RemoteTech, I would advise that you ensure your heatshield is oriented the proper way significantly before you hit mid-atmosphere: your antenna might break off in reentry. It also helps to have your chutes armed well ahead of time*. Otherwise, it's the same as ever: re-enter at a shallow angle, perform a lifting re-entry if you so wish, and double-check that your heat shield is facing the right way. *If using stock, just tweak your chutes at the VAB to only open at high pressure; if using RealChutes, you can set low deployment altitude. Otherwise, you risk either not having contact with the probe to deploy chutes manually, or you will have your chutes burn up because they decided that the middle of re-entry was a perfect time to deploy. I've only ever done one RT re-entry: it was a suborbital flight with no available comm links past apoapsis. Fortunately, aerodynamics corrected the alignment of the probe from about 30 degrees off to face re-entry directly, and my parachutes deployed exactly as planned.

I'd like to chime in here a bit: for those of us who use RSS or other scaling mods, it would be nice to have easy-ish places to set the scaling of rewards. Testing parts in atmosphere is 100% as easy as it is in stock, suborbital almost as easy, anything past that much more difficult because you need hugely more delta-V to get anywhere. If I have to figure out how to write the mod myself, I'll do it, but I'd need to know where to look to modify it. Agreed on that. Funds -> science is absurdly broken. Maybe it would be possible to, once you've used an instrument to transmit science, render it unable to do so for any more contracts, and have Kerbals unable to plant flags for contracts until they've been recovered? It would at least force you to send duplicate probes or have duplicate experiments (preferably only for Science Jrs and Mystery Goo, as otherwise you can spam physics-less thermometers all day).

This might be a silly question, but do you still have ablative material on the heatshield? It's possible your Jool aerobrake is wearing the heatshield down to its structural base, at which point the shield is just another part, unable to really prevent heat from transferring into other parts of your craft.