EPTKerbal Space Program - Page 44
| Forum Index > General Games |
|
Ljas
Finland725 Posts
| ||
|
m4ini
4215 Posts
edit, finally 1080p now. Forgot to turn on sound recording, hmph. | ||
|
Broetchenholer
Germany1901 Posts
On December 20 2015 07:39 m4ini wrote: It still should work though. Dumb question: what exactly comes down for you, just the capsule or a whole rocket? Because even if you go straight up, and then straight down, the capsule itself should easily be able to apply enough drag to be at safe speeds around 2500m. I'm guessing now, but a capsule from 110k straight down reaches around 1700-1800ms, that should easily be stopped by drag (if you keep the capsule retrograde, or "nose pointing up"). edit: actually gonna try now, to make sure i don't say something wrong. edit2: yup, capsule itself slows down to safe speeds (280ish ms) at 2400m. Shot a Mk1 straight up to 111000m, decoupled and let it fall straight down. Only 1x Mk16 on top of the Mk1. It's just a capsule, a heat shield and the chute (or 2 additional radial chutes, which does not work). The ship i used had a hammer as first stage and then as 2nd stage a reliant with two 200 tanks, then third stage heatshield with MK1. It goes up to 140 km i think, then falls and dies. I had tried to get some more science by doing biome hops, but i don't have the manoevering capabilities to actually steer anything bigger then the smallest ships. It has become really hard  Edit: On December 20 2015 08:09 m4ini wrote: Video on last page, edited in. Quality still sucks, but yeah. edit, finally 1080p now. Forgot to turn on sound recording, hmph. Guess i just need to park the capsule lower in orbit, my fall seems to be too much. There is no need for that altitude, i just experimented with my first staged rockets in the new version and wondered how to solve this. Thanks for the help. | ||
|
m4ini
4215 Posts
 Stages before that are not important for re-entry, that's why i just used a booster and infinite fuel to check. At 140km it might not be possible anymore, that's true. Biome hopping never really works, all you can do is basically build a small "lander", then basically ballistically shoot it at the biome you wanna check, land and retrieve. Easier when you get an actual working plane, but that's so far down the line that for the same amount of research needed, you could just fly to duna. At the start, all i do basically is get all the science from launch pad and runway (they count as different things), also important: if your capsule lands, and you go EVA for an EVA report, do two. One while hanging on the capsule (counts as "flying over x"), and one of the ground. Then just shoot stuff randomly in all directions, one into water (again, one hanging on ladder, one in water), etc etc. That'll get you going. edit: by small lander i mean just the capsule, one goo and later one science bay (then you need more than the Mk16 chute though, otherwise the bay will blow up). edit2: also, of course, see to it that you get at least the science bay and thermometer quick. It's easy and "light" science that you can take most places. Also, upgrade, uhm.. I think it's the radar station? To be able to go EVA in flight? I can't remember anymore, fact is though, there's two buildings, one upgrade gives you the ability to go EVA in flight (two more options per planet/satellite to get science, low orbit and high orbit/space near x), and the other one enables you to get ground samples. That's the more important one, that's extra science in every biome (120 science extra on mun every landing). edit3: sigh, should stop "updating post" before finish thinking. You could also use the rocket you're using right now, fly straight up until on the mapview it shows your apoapsis (highest point on the curve) at 110km (or 100km, for that matter), and then flop the rocket just 90 degrees sideways, picking up horizontal speed. Gives you longer time on descent for drag to work. | ||
|
Broetchenholer
Germany1901 Posts
Edit: And i am rambling again, the vehicle assembling building is only responsible for the number of parts, the weight limitating come from the launchpad, which is way less expensive to upgrade | ||
|
m4ini
4215 Posts
And, it's not really hard. The only problem i have with the difficulty is that it's somehow inverted. It's hard-ish at start, and once you get your first mun-flyby, the game's pretty much over. From there on, science for days (because landing is easy on the mun, each flight nets you around 300 science), and money never was an issue in the first place. Only difficult thing from there is basically "spotting" the transfer windows for Duna etc. | ||
|
felisconcolori
United States6168 Posts
On December 21 2015 06:52 m4ini wrote: Yep, these are pretty much two of the most important upgrades for going to space. After these are upgraded, you can basically brute-force your ships into orbit with boosters. And, it's not really hard. The only problem i have with the difficulty is that it's somehow inverted. It's hard-ish at start, and once you get your first mun-flyby, the game's pretty much over. From there on, science for days (because landing is easy on the mun, each flight nets you around 300 science), and money never was an issue in the first place. Only difficult thing from there is basically "spotting" the transfer windows for Duna etc. Eh. Throw enough money and dV at it and transfer windows aren't that important. | ||
|
m4ini
4215 Posts
On December 21 2015 08:43 felisconcolori wrote: Eh. Throw enough money and dV at it and transfer windows aren't that important. While true, i really dislike orbital construction, so single launch to duna (with return) doesn't really work in 1.0.5 if you don't hit a decent window. At least i couldn't do it. | ||
|
Broetchenholer
Germany1901 Posts
Now i got the next tech and tried to implement the new generation of stuff. The thing is though, the physics don't make any sense any more. the bigger jet engine should have more thrust then 4 junos together, yet, the plane just is not faster, it reaches less altitude and it behaves worse with a lighter construction... I every sort of air intake to provide the engine with air, they all did basically nothing. One small circular intake seems to suffice for the whole new big engine. If i install more, i don't get more thrust. So, i checked the thruster and saw the thrust decline with every meter higher above sea. It amkes no sense. The Engine should not lose thrust the higher i am if i provide it with enough air which does not seem to cap. At the same time, i should get faster if i am higher due to less drag. The drag never is a problem though, i am just getting slower and slower. Is the engine limited to height? The engine is barely better on sea level compared to the 4 junos, do these jet engines have different ISPs on different heights as well? | ||
|
stenole
Norway868 Posts
This varying thrust is also normal in a lot of real life engines which dictates their normal operating speed and altitude. When the air is thinner, you can't just push more air into the engine. The more air you compress the more heat is created. An engine despite being designed to handle heat, has its limits. Why your more advanced jet engine is underperforming can't be well explained from what you've written in your post. A reason a plane would be going slower could be drag and/or mass. No matter what, there will always be a bump in drag around mach 1, the sound barrier. Engines like the Juno, whesley and goliath will have trouble exceeding this speed. | ||
|
Cyro
United Kingdom20285 Posts
On December 20 2015 08:01 Ljas wrote: I swear I've killed a kerbal like that by bringing just a capsule down from a suborbital flight. When that happens, it's due to re-entering too steeply. You need to do suborbital more sideways than up - if stuff goes really wrong, you can also jump out of the capsule with your kerbal and it'll slow down to about 50m/s by itself. If you burn upwards with RCS (shift hotkey i think) you can slow it down to about 37m/s and i have not died hitting water at that speed since the 1.0.5 water rework! ----- While true, i really dislike orbital construction, so single launch to duna (with return) doesn't really work in 1.0.5 if you don't hit a decent window. At least i couldn't do it. Study the rocket equation and don't be afraid to build large rockets (: ----- The thing is though, the physics don't make any sense any more. the bigger jet engine should have more thrust then 4 junos together, yet, the plane just is not faster, it reaches less altitude and it behaves worse with a lighter construction... I every sort of air intake to provide the engine with air, they all did basically nothing. One small circular intake seems to suffice for the whole new big engine. If i install more, i don't get more thrust. So, i checked the thruster and saw the thrust decline with every meter higher above sea. It amkes no sense. The Engine should not lose thrust the higher i am if i provide it with enough air which does not seem to cap. At the same time, i should get faster if i am higher due to less drag. All air breathing engines have thrust curves against two things: Speed and Altitude. The basic engines have most of their thrust below mach 1 and have a lot of trouble exceeding it, but can get through it and cruise around mach 1.4 if you have a lot of TWR or if you dive a bit to pick up speed to get from ~300 to ~380m/s. The other engines like different speeds. For the Whiplash and Rapier this speed is around mach 3.2 and 3.8 IIRC, but you have to get up to a decent speed before the real thrust starts to kick in, especially for the Rapier. The other engine with the thrust vectoring is a middle ground, not super fast but better than the basic engines and can use afterburner to break mach 1 easily. As you gain altitude, thrust decreases. The amount of drag on the aircraft decreases faster than the thrust drops off up to a certain point - the peak thrust to drag. That happens around 7km for the wheesley and about 15km and 18km for the Whiplash and Rapier. That's the altitude where peak speed is reached, though it's important to note that acceleration is easier at lower altitudes due to the higher thrust. Intake air is not important as long as you have enough. If you're not flaming out, don't worry about it. To help visualize some of the stuff said, here's a few old pics that may no longer by entirely accurate: ![[image loading]](http://i.imgur.com/yciepgi.png) ![[image loading]](http://i.imgur.com/Cp3Hk6P.png) The basic jet performs well up to like ~mach 1.7 actually, but it has a very hard time accelerating through the ~300-380m/s barrier because of the increased drag there. ----------- I'm thinking of putting one of these things into low kerbin orbit - beautiful :D ![[image loading]](http://i.imgur.com/xlBsqS5.png) can fly into the middle and dock, then it has ~4km/s of delta-v at high thrust. Plus with a mass like that, carrying the fuel to refuel one of these planes once or twice wouldn't impact the delta-v that much. This is the type of plane that can technically almost certainly do a laythe return by itself, but you'll have to do 25 gravity assists. That's true for any laythe and back SSTO really - long burns, terrible TWR's and gravity assists again and again and again. Having some lovely engines to dock to really makes the experience a whole lot simpler and more enjoyable! :D | ||
|
Cyro
United Kingdom20285 Posts
| ||
|
m4ini
4215 Posts
Study the rocket equation and don't be afraid to build large rockets (: Yeah, no. Has pretty much zero to do with "study the rocket equation". Has to do with the fact that i think building rockets wider than the space center is kinda stupid. Launching 20k-30k (even more actually) delta-v isn't fun to me. | ||
|
nimbim
Germany983 Posts
| ||
|
stenole
Norway868 Posts
On March 06 2016 13:51 m4ini wrote: Yeah, no. Has pretty much zero to do with "study the rocket equation". Has to do with the fact that i think building rockets wider than the space center is kinda stupid. Launching 20k-30k (even more actually) delta-v isn't fun to me. So if you don't like launching monster rockets that make the game run at 0.1 fps and you don't like orbital construction, you do things in the game you think is fun instead. Designing each stage for what it needs to do instead of overengineering is one way of reducing size, Making sure your manouvers are are close to optimal is another. Sometimes mods can help make previously tedious tasks in the game less tedious. Orbital construction might be more fun if an autopilot did 90% of the work for you. | ||
|
m4ini
4215 Posts
Well, you have to settle for something if you want to play this game at all. If we are talking just about Duna, it's really not that hard to get there and back in a single launch without building a monstrosity, but if you look at Jool and Moho the delta-v requirements are simply too high. Either you construct in orbit or launch gigantic stuff from the launchpad, there is no alternative except maybe timewarp for 200years to get a perfect gravity assist transfer. I do, i wouldn't have 600 hours on steam otherwise - plus countless hours before that. And no, obviously it's not hard to get to Duna and back without building monstrosities - if you wait for a decent transfer window. So if you don't like launching monster rockets that make the game run at 0.1 fps and you don't like orbital construction, you do things in the game you think is fun instead. Designing each stage for what it needs to do instead of overengineering is one way of reducing size, Making sure your manouvers are are close to optimal is another. Sometimes mods can help make previously tedious tasks in the game less tedious. Orbital construction might be more fun if an autopilot did 90% of the work for you. The performance partially is the reason why i don't build "big things". Might change with 1.1, which i hope fixes all the things. And i certainly do things that i think are fun, even though nowadays it's mostly dicking around with weird plane designs. I am certainly in the "optimal", "streamlined" and "precise" camp though. Autopilot, well i used mechjeb before but i think it starts to get boring rather quick - only using Kerbal Engi Redux nowadays for the data. Btw i'm not complaining, i just said it's not possible to go to the "more advanced planets" without a decent transfer window. Maybe i should've added "without launching 50k dV". | ||
|
Cyro
United Kingdom20285 Posts
On March 06 2016 13:51 m4ini wrote: Yeah, no. Has pretty much zero to do with "study the rocket equation". Has to do with the fact that i think building rockets wider than the space center is kinda stupid. Launching 20k-30k (even more actually) delta-v isn't fun to me. I have no idea why you would want 20-30k delta-v for a duna return, even if you have no idea what a transfer window is and completely screw it up. LKO is 3100m/s Duna transfer is 1040m/s aerobraking and landing without parachute is 350m/s return to orbit and to kerbin is 2000m/s that's 6.5km/s. If you take a completely horrible launch time, you may have to add 1 or 2 km/s onto there but unless you're launching completely at random, that much shouldn't be neccesary - a direct transfer to Jool takes 1950m/s, why would Duna take 3000m/s? You can timewarp to a vaguely correct window and wing it without checking the angles at all by that point. You can even do some of those more direct transfers to get a fraction of the travel time that it'd take with a hohmann transfer. Increasing payload mass requires a linear increase in rocket mass while increasing delta-v requires an exponential increase - that's why it's legit to carry heavy stuff around, but crazy to try to get 20-30km/s of delta-v for no reason with chemical engines. In KSP balance, that takes about ten stages to do "properly", each one multiplying your mass. For a sane duna return, you can go orbit on one stage, transfer + land on another, orbit+return on third. For an insane one, you can dedicate two stages to transferring and landing - then you're done in 4 relatively simple stages. I made a rocket to demonstrate this. Quick and simple 3 minute build, no solar panels etc but i overbudgeted the delta-v in transfer and landing stage by 3km/s, 100% stock no scaling. I didn't have to go beyond a relatively simple 2.5m rocket in order to achieve a single launch duna return with such huge margins left. ![[image loading]](http://i.imgur.com/2XzbAph.png) ![[image loading]](http://i.imgur.com/5Yr1pOQ.png) ^I launched with a crazy phase angle so that hohmann transfer was inviable. A single-orbit bi-elliptical transfer cost about 1km/s more, as shown in this picture - still 2km/s reserved in the tanks. Might use a little bit of them capturing at duna, i don't know how tricky that will be. --------- The performance partially is the reason why i don't build "big things". I launch absolutely huge things all of the time and run at triple digit FPS. The key is to understand that you can make a perfectly good stage with a fuel tank, an engine and maybe a couple SRB's; you can use one of them instead of using 15 smaller ones and they'll give the same weight, thrust etc but not break the physics system or your CPU. My 10 meter rockets run just as well as my 1.25 meter rockets because of that, but they carry 512x as much payload. | ||
|
m4ini
4215 Posts
I launch absolutely huge things all of the time and run at triple digit FPS. The key is to understand that you can make a perfectly good stage with a fuel tank, an engine and maybe a couple SRB's; you can use one of them instead of using 15 smaller ones and they'll give the same weight, thrust etc but not break the physics system or your CPU. My 10 meter rockets run just as well as my 1.25 meter rockets because of that, but they carry 512x as much payload. It's not the size, but the partcount that screws the physics. And i'm certainly not going to argue with you there, it's a well known fact through the whole of the KSP community that the performance tanks horrendous as soon as your partcount goes to 200 and god forbid above. You won't build interesting, or even working rockets with 20 parts if you have a couple of difficulty-mods installed (TAC, DR, ISA, KAS/KIS, IR, RTech). DasValdez (one of the best builders in KSP) constantly struggles to "obey" the "soft" partlimit. On pure stock KSP. The only way to get around it is to either use plenty of physics-less parts (not really working anymore), or to just stay small. And no, before you start it: it has nothing to do with configs, settings or anything but a bad engine. Which squad sees the same way, hence the 1.1 engine update, possibly (finally) with a (edit: working) 64bit client. But i stand corrected on the duna issue, you can go whenever if you don't play with mods like TAC. In other news, 1.1 actually went experimental this week, which (that's me guessing) means that if nothing goes wrong, in around 4 weeks they'll throw it out. http://forum.kerbalspaceprogram.com/index.php?/developerarticles.html/opt-in-prerelease-for-11/ You can actually opt in for a pre-release. | ||
|
Cyro
United Kingdom20285 Posts
100 parts runs fairly well, 200-300 does not (CPU demands are nonlinear with part count, more parts add disproportionately more load) This rocket that i linked you is 20 parts, but i can do it in 10 or so if i'm scaling parts. The ground-to-LKO stage is 8 parts (3 decouplers - 2 fuel tanks - 1 engine - 2 srb's) when it could easily be 2 (decoupler, booster) | ||
|
m4ini
4215 Posts
On March 07 2016 02:30 Cyro wrote: I'm well aware of the partcount limitations. You can do this fine with TAC and similar mods, i assure you Sure you can. I do it all the time. It's just not fun to launch at 15 fps because there's stuff in physic loading range. This rocket that i linked you is 20 parts, but i can do it in 12 or so if i'm scaling parts. Yes. And i said, if you don't play with mods, that works. Clearly. | ||
| ||
