Streetwind

Less thrust means you spend more time fighting gravity, means you need more dV to get out of the gravity well. Super simplified fantasy example: - You can spend 200 seconds gaining 10 m/s per second, or 100 seconds gaining 20 m/s per second, for the same fuel cost. - In both cases, you could reach a theoretical speed of 2,000 m/s. Thus your rocket has a dV of 2,000 m/s, regardless of which configuration you choose. - However, every second spent ascending, you lose 5 m/s of your gain to gravity pulling downwards on your rocket. - In case 1, you spend 200 seconds ascending, thus you lose 200 * 5 = 1,000 m/s to gravity, leaving you with 1,000 m/s final velocity. - In case 2, you spend 100 seconds ascending, thus you lose 100 * 5 = 500 m/s to gravity, leaving you with 1,500 m/s final velocity. Clearly, the high thrust option is better. It produces 50% more speed out of the same dV. In fact, even if the specific impulse of the high thrust engines was worse, and thus there was less dV available to the high thrust option, it would probably still end up with more final velocity. That is why the Mainsail engine is great despite its poor Isp rating. It simply catapults you out of the gravity well so much faster than other engines that you save more fuel than you lose. Of course this comes with two caveats: - If you launch out of an atmosphere, then you will incur atmospheric drag, which grows exponentially with speed. Ostensibly the fastest launch is the best launch, but go too fast in an atmosphere and you spend more fuel fighting air resistance than you would spend fighting gravity while going slower. - If there is less gravity, then there is less to be gained by leaving it behind fast. And in orbit, there is no gravity at all (because the orbit itself counteracts it). So for orbital maneuvers, TWR is almost entirely irrelevant, and Isp is everything that matters. Of course, you might still want some thrust in order to avoid having half-hour burns for your maneuver nodes, and to be able to slingshot effectively. The takeaway is that you want to optimize your launch stages for TWR (and potential landers as well, especially if they need to relaunch), and your orbital/transit stages for dV. Just don't optimize them too much, or you end up with a rocket that burns itself up in the atmosphere and later can't maneuver accurately.

Thanks for the info! I'm planning to go easy on mods for starters - especially nothing that changes vanilla stuff (so definitely no Ferram Aerospace). SCANsat looks like a very fun and non-intrusive thing though. I reckon that the next time KSP updates in such a way that savegames will need to get reset, then I'll do a completely new playthrough with lots of mods. For now I'm just testing that waters and learning vanilla mechanics.

Hello all, I hope this doesn't get buried in the flood of bug reports, this thread is certainly moving fast when an update breaks something I've only been playing KSP for a few weeks (it was a christmas present), and I have never installed a mod before. Kethane seems like a great first candidate, since it doesn't have any prerequisites, is focused on doing just one thing, and adds a great and logical next step to vanilla gameplay progression. Now while the install instructions are clear enough, there's something I wanted to ask before letting things loose on the one single savegame I have and have ever had. I am playing in career mode. How does Kethane integrate itself with career mode, if at all? Will the parts just pop up in my VAB regardless of science, or will they be available through existing tech tree nodes which I may or may not have unlocked, or are there new nodes for them?

Thanks for the answers!

Nah. Mostly I was wondering if an ion thruster could lift itself under ideal conditions. But with 2 kN per ton of thruster, not ever happening Oooh, so a 0.5 torque reaction wheel in a 1-ton part will exert a stronger turning force than the same 0.5 torque reaction wheel in a 0.2 ton part? That's kind of a little silly, considering how light all the inline reaction wheel / SAS parts are... Another thread said 9.82, specifically. Right, another question: 3.) Do large extended solar panels have a g-force tolerance limit? I.e. will I tear them off if I fire up the engine without retracting them first?

...before I go sleep! 1.) What's the minimum thrust in kN for one ton worth of rocket to achieve liftoff on Kerbin? Plain mass, not payload. Is it 9.82, because gravity? Or is it something else? (Sorry my physics is bad ) 2.) Are there any differences between probe cores that I am missing? There's so many of them, and they all seem to have largely the same stats. I am looking at weight, at reaction wheel torque, and at power consumption (both probe itself and reaction wheels). Is there something else to them that gives a good reason for this large a variety?

But then how am I going to get something geostationary, if I making the orbit circular is such an issue? As far as I understand it, that needs to be done super precisely, but the second apsis at least will be jittering so much that it's impossible to see its value within a large margin of error. Is being in time warp precise enough to set it up? And now for something completely different: if I import a craft file of a vessel, and I am in career mode, and that vessel uses one or more parts I have not yet unlocked... what happens?

I've read somewhere that you can change the physics simulation timestep in the KSP config file. Usually people do this to unlock extra performance, but I have a very good computer, and I'm curious: are there any advantages to doing more calculations as opposed to less? I'm also somewhat annoyed by the way my apoapsis and periapsis points start jittering like mad as soon as I get them under 2 km of each other. I kind of like my near-perfectly circular 72km orbits around Kerbin, but all I can see is the markers spazzing all over the place. Is there something I can set in the config to maybe increase the resolution or whatever it takes to make the markers more precise?

Well, I would offer you my latest Munraker VII rocket that carries a really good lander, but I fear that as a one-seater it's not what you need right now... =/

Do the tanks have to be full at all costs? You can make the upper stage a lot lighter if you remove some fuel from the tanks. Over 90% of their weight is fuel. If you plan to mount these tanks to a station instead of using them to fuel something, then it might be worth launching them empty (or at least reduced) and then fill them up with a separate refueling run or two.

If I learned one thing, then it is that piloting can make or break a design. I read up on physics on wikipedia instead of browsing KSP gameplay guides, and I found some things in that that blew my mind as I was able to replicate the effect exactly ingame. For example, you can reliably run almost any lander out of fuel trying to land on the Mun if you simply do it wrong enough. The basic gist is, your engines should be on as rarely as possible - and if they are, they should be on full throttle. Every moment your engines aren't on full throttle, you're wasting fuel. That is because you need a certain percentage of your engine power just to cancel out "gravity drag", leaving the rest to do actual deceleration work. And if you throttle your engines down enough, then all you are doing is maintaining your status quo while not actually doing anything useful, while fuel is still being consumed. Maintaining the status quo is the worst possible thing you can do. In a practical example, I took a lander to the Mun and, after the initial intercept burn, left my engines on just a little bit, really low, enough to slowly float gently down to the surface. They took almost no fuel, and I was never in any danger of crash landing - until I ran out of fuel before even making it down to the surface simply because it took me so long. Then I took the exact same lander (loaded from a quicksave), and just let it freefall straight down to the Mun as fast as it wanted to go after intercepting. Once or twice, I did short, hard burns at full throttle to decelerate a bit and get a feel for how much height I would lose while canceling out a certain velocity, but even then, the less of that you do the better. Afterwards, I let it freefall again for a while. The landing made me a bit nervous, braking down from 400 m/s only a few thousand meters before touchdown, but then I afforded myself the luxury of floating the last 30-50 meters on low engine power. I landed, did my EVA, boarded again, hopped the lander a kilometer or two to a neighboring biome, landed again, did a second EVA, boarded again, launched back up from the Mun, escaped to Kerbin, decelerated to aerobraking height and decoupled the final fuel tank still half full before re-entering. That's the difference between how you do your landings. It is perfectly possible to do so badly that even a completely overprovisioned design fails to go half the way, just because a law of physics you didn't know or remember at the time eats all your fuel like an insidious little gremlin hiding in your tank. Now obviously this is an extreme example, which I engineered to visualize the difference to myself, but the point is: piloting skill ultimately equals delta-v. It applies to knowing how to launch while minimizing the sum of aerodynamic drag and gravity drag, knowing how to get the orbit you want through cheap mid-flight course corrections instead of expensive inclination tilts, knowing how (and when!) to try for an intercept, knowing how to land without burning more fuel than necessary... every single thing has room for error. Because of this, I used to have severe troubles building rockets that got anything useful at all into orbit. I solved this problem not by building better rockets, but by learning how to fly them. Oh, and another trick: go for Minmus first. Costs you maybe 5%-10% extra delta-v to intercept, while costing way, way less to land on and launch from. And in 0.23, it now has biomes as well for juicy science.

Torque from the engines? I had no idea that was even a thing, shows what I know I also spent a good 5-10 minutes simply trying to figure out how one could possibly shoulder-mount wings. They don't attach that way! Until I remembered the rotation keys. Aircraft construction sure is an advanced topic. Anyway, I now have this. And to my great surprise, it actually worked. It handles a bit like a highly caffeinated rodent, despite already using the smallest control surfaces, but I actually managed to take off (barely) and land (was actually easier than taking off). I'm still not quite sure how everything interacts, but thanks anyway - this helped a great deal!

So this is my first real, proper jetplane. Yes, it mostly involved sticking parts together at random. Considering that, it actually flies quite well. What it doesn't do, however, is launch off the runway. If I launch the plane off of a clamp high up and let the turbines spin up a little, I can get enough speed to stay airborne and fly. But when I am on the runway, on my wheels, and try to accelerate I invariably see my plane make a very sudden and very sharp turn to the side at some point. Which side varies, sometimes left, sometimes right. If I have no supporting wheels on the wingtips, the plane immediately crashes and explodes; if I do, it just drivers straight off the runway. It doesn't matter if I gently increase my thrust or if I go full throttle. And it happens before I have enough speed to take off. I saw the same phenomenon on an earlier test run with rocket parts. I assumed it was because I was not using proper airplane parts, but now it just feels like the physics simulation is giving me the middle finger for no reason I can discern. Why is this happening, and how can I stop it from happening? EDIT: I should maybe link the screenshot before posting...

Yeah, making rockets shorter and wider seems to work, as long as you liberally apply struts. Kind of counterintuitive to RL, but I suppose that might have to do with KSP's simplified physics simulation. Building handmade launch clamps with decouplers and girders was something I also thought about briefly. Might be a fun project to mess around with sometime

Interestigng idea with the support from below. @Seret: Yeah, I did that. Adding struts to the rocket itself helped. The reaction wheels are of course always on, and help stability in flight, but don't really do anything at all about wobbling onthe launchpad. And no, that rocket cannot lose any boosters, and doesn't stand to gain anything from making them taller Mind you, I have other rockets too that suffer from the wobbling, but I randomly grabbed this one for the screenshot because it was the one I worked on last. It's a silly experimental design that I wanted to do just because I could. The boosters are neither solid nor liquid engines... they're jet engines, six times three of them. And because they start to flame-out north of 10 kilometers, they hardly need any fuel. Those already tiny tanks carry only 10% of their maximum filling amount. All in all that gives me a ridiculous thrust-to-weight ratio of over 100 kN per ton of booster. No liquid or solid fuel engine comes even close to that. Here's another screenshot of the boosters without all the hubbub around them. They're kind of iffy to handle overall though and cut out just a bit too early for my tastes, so I probably won't be using that design for regular carrier duty. Maybe when I unlock some of the better jet engines and intakes, these are just basic ones.

Right, I tried some of these suggestions. For starters, I was able to firm up my rocket by using struts on the rocket itself, so that it wasn't quite as "bendy" anymore. I noticed that it especially helped to support the decouplers that held the boosters... because the boosters were held by launch clamps, and therefore the entire rocket was effectively suspended by the set of decouplers. Then I went and built girders up on the launch clamps, and conencted struts to the top stage. This actually made things massively worse, because the tall towers of girders wobbled about in an extreme fashion. I tried to solve this by attaching extra launch clamps to the top of the girders, and strutting them among each other. Ultimately I ended up with this. Which sadly still jittering about (you can see the 0.1 m/s velocity display even though the rocket is just sitting on the launchpad).

Won't that make the lower stages collide with the girders and/or struts as they fly up, though?

I'm currently in the process of designing various rockets. Since I'm still new, I don't really push for any particular goal (despite playing career mode), I just stick stuff together and launch it to see how it behaves, what leaves me with the most fuel in orbit, what makes rockets easy or hard to control and such. It also helps me practice controls... At this point I can launch directly into a perfectly equatorial 70km orbit with periapsis and apoapsis within 250 meters fo each other... and reproduce the feat at will. I have never installed any autopilot mods. However, there's something that's seriously cramping my style, and that's the way launch clamps work. Rockets are naturally slimmer up top, because later stages are smaller and lighter... and as a result, the wide lower stages (especially with boosters) keep me from attaching launch clamps to the upper stages. This, in turn, results in the upper part of any rocket higher than 2-3 fuel tanks wobbling about erratically on the launch pad, and eventually settling into a minimal lean off to one side. The result is that even with perfectly symmetrical rockets fitted with gimbaling engines, extra SAS modules and control winglets, the craft just won't fly straight. At this point I am still using less than a quarter of the VAB's vertical space, and I honestly have no idea how I could utilize more of it. I've tried slimming down the lower stages or fattening the upper ones, but I always end up in a situation where I can't fit enough engines on the first stage to launch to any meaningful speed. I need the wide base for engine space, and the slim top to save weight. So how can I properly stabilize the top of a tall rocket?

Ahaha... pushing with the jetpack? I am not nearly confident enough to handle myself in EVA to dare something that crazy. I'll end up spinning off into deep space or something Even the transfer from one craft to another is something I'd approach with great, great trepidation. Besides, I got over 3 tons of equipment to deorbit... how much delta-V can a personal jetpack supply? According to this maneuver node I need at least 18, better 18.5 in order to get into the outer atmosphere. I suppose though I can keep the option in mind for the future. After all... Bob isn't going anywhere in a hurry!

Well, after a failed attempt at the first Mun landing, Bob ran out of fuel on his way home. He managed to come close, so very close... Kerbin periapsis at 180,000m. Now, while I can effortlessly send a vessel into a similar orbit, the question is, does it matter? Bob's craft has no docking collar. Can I even attach another spaceship to his, and either a.) transfer fuel or b.) transfer Bob or c.) deorbit Bob's vessel? What would you recommend I should do?

Many thanks for your comments! Based on what you told me, I was able to come up with a few new test flights, and now I understand a lot better how different parts and engines affect rocket stability at launch, which was an issue I was encountering. My newest design is rock stable even with the reaction wheel disabled, and looks good while at it I'll mark this answered now, and go see what mischief I can make with my newfound knowledge.

Greetings, I'm new to Kerbal Space Program. I'm someone who is all "oooooh space, cool stuff!", but my science background is a little lacking. Been at it for just two days, slowly figuring things out by myself in career mode. But after so long, there are a few questions I haven't been able to answer for myself through ingame actions. I would greatly appreciate the input of the more experienced community on these - Do I need to pay attention to aerodynamics? One of the first things I did after unlocking my first research was take a really simple rocket - three newbie boosters around a standard fuel tank + command module for central fuselage - and launch it straight up, noting the height at which the boosters cut out. Then I reverted to the lab, slapped shiny new aerodynamic nosecones on the three boosters' flat tops, and launched that. Again straight up, until the boosters cut out. Against my expectation, the more aerodynamic variant actually achieved *less* height (16.5 km vs. 16.6 km). So what is the point of the nosecones? Should I be using them or do they actually interfere with the rocket? And in a similar vein, does sticking winglets to my rocket do anything other than add weight? I mean, obviously you need wings for spaceplanes, but on a rocket going straight up...? - What are the advantages of having RCS? I have the option to research it right now, and I know what it is - little maneuvering thrusters. But I am making do quite nicely with reaction wheels, and it costs a lot of science. What is the selling point of RCS over reaction wheels? - In a similar vein: what does gimbal / vectoring do on engines? The game doesn't do a good job at explaining what practical advantages it gets me. Is that a steering aid? If so, why is it useful in the presence of reaction wheels? - How can I see the stats of the rocket I am building? Right now, I build a rocket, save it, click launch, wait for a loading screen to the launchpad, then go into solar system map mode (of all things!) and there, finally, is a little info button that shows me at least the weight of my rocket, if nothing else. There's *got* to be an easier/more informative way to do this... - What can I do with a Stayputnik in career mode? I have googled up a thread that shows you how you can make a great unmanned probe with it, and cites the main advantage as being available early. But frankly, I am 2-3 tech levels past the Stayputnik, have unlocked everything on those levels, could research the next probe command module right now if I wanted to, and I still don't have access to any parts that are actually usable for probing. For the purposes of generating science, I have mystery goo, the junior lab, crew reports, EVA reports and surface samples. The first two lose 80% of their data value on transmission, and the latter three cannot be done by an unmanned probe at all. Am I missing something here? Why would I want to take advantage of the early availability of of an unmanned probe command module when there's no science to be generated with unmanned probes until much later tech levels? I just want to make sure that the idea of probe construction didn't completely go over my head and I missed a cool feature because I didn't understand it.