Aegolius13

As an example, say you're coming out of a Tylo gravity assist and about to aerobrake at Laythe. Generally a radial burn would be the cheapest way to lower your peri to the desired altitude. But it has the side effect of increasing your velocity at peri, which can make the aerobrake dangerous. Or you can burn retrograde, which will bring down both ends of the orbit, but is not a very efficient way to bring down peri if you're close to peri. In that situation, seems like you can get a small advantage by adding a small retrograde burn to your radial burn. Not saying it's a big savings, but every little bit can help.

-Putting fuel in "hats." Sometimes you want to discard your spent fuel tanks, but don't want to have to include an extra engine, or mess around with radial tanks or engines. The solution: put a detachable tank on the top of your ship. You may have to stop your burn for second and turn a bit to detach the hat, but that's usually not a big deal. -On very large launches, I have taken to adjusting the SRBs so they burn out at the same time as the first stage of LF boosters. That way, you can just attach the SRBs right onto the booster, and forget about extra decouplers or struts. Of course, when doing this you lose a ton of thrust when the whole thing detaches. But it works well with Rhinos set to kick off on the next stage. -Whenever helpful, sprinkle in a little radial burn to a prograde/retrograde burn, or vice versa. It's practically free due to the Pythagorean nature of burn vectors. A good example is when you're adjusting your perisapsis as you enter a new SOI.

Chute deployment is a very non-intuitive three step process. First, you have to arm the chute through staging. You can do this manually, or in orbit. The important thing is that you don't stage while ascending, for obvious reasons. Next, the chutes partially deploy (visually pop out and slow the ship a bit). This is governed by the pressure tweakable. The default is very low, meaning if armed, the chutes will start deploying in upper atmosphere. This usually destroys them. Don't know why Squad chose such a bad default - it makes everyone ignore these settings. The solution is either to wait to arm, or turn the number up. Don't know if .65 would open too late; you'd have to test it. Depends in part on reentry speed and angle. But this setting is unimportant if you always wait to safe speeds to even arm. Third, the chutes fully deploy (spread out and slow you to landing speed). This is governed by the altitude tweakable. The default here is generally fine, though you can speed up landing by lowering it to 800m or so. Sounds like your problem is linked to reentry speed, which is not something these settings would probably help with.

I would double check the chutes are not staged earlier. The icon does not go away when they stage do you might not be noticing it. You can also try turning up the deploy pressure tweakable to .3 or so. This will keep the chute from starting to deploy until you're lower, and probably slower.

As I look at it, there are two kinds of solutions. As described above, you can burn pretty much anywhere to PUT an ascending our descending node right on the target. This means you'll hit the target mid-planet, but at an incline. Then you can either land, fix the plane in orbit, or just continue with an inclined orbit. Or, you can burn AT a node to match planes exactly. This is the more complete solution, but can be expensive. All things being equal, it's better to adjust a plane while moving slowly. For example, if you have to correct plane in orbit of a planet, you can burn at peri just enough to capture, then plane change at the far end of the orbit, then burn again at peri to enter your parking orbit. It's often possible to save some on the plane change cost by including it in your ejection burn. (Like the sides of a triangle, it's cheaper to go diagonally then straight in one direction and then straight in another). In an ideal situation, if you eject while Kerbin is at a node, you can plane match exactly for very cheap. One method for going to Moho calls for doing this, rather than using launch windows based on phase angles. But that's excessive for Duna. In some systems, including Duna, you can also use a trip around a moon to level out your plane.

Have you tried setting SAS to "radial out"? I don't do VTOLs, but that's my tool of choice when something needs to go perfectly straight up. EDIT: just remembered that only works if you have a control core facing straight up, which might not be ideal for VTOL. And would have to be at the COM... so probably not the best idea.

Not sure what you'd consider massive, but it's definitely bigger than Kerbin, owing to Eve's significantly stronger gravtiy. AlexMoon's launch window planner says about 1,800 ejection delta v from a 100km Eve orbit; that sounds right to me. Mining is definitely an option on Gilly. Since its gravity is practically non-existant, you can build a big honking mining ship without worrying much about thrust. And with the close distance to the sun, it's easy to power the mining equipment. However, it's pretty expensive to send a mining ship Even from Gilly's orbit down to low Eve orbit. Even more so if you can't aerobrake effectively (my Gilly mining ship can't, but it's probably possible to design one that can). Mining on Eve itself is, uh, not recommended -- unless you're seeking a particular challenge. Or, if you're planning to land anyway, it can make sense to bring mining equipment but ditch it for the ascent. Eve is definitely expensive enough it may make sense to go nuclear (at least on the return trip). You can do tugs, or just place nukes on the ship you're sending to and from Eve. But if you have a very small payload, chemical may still come out ahead. Other thoughts: Aerobraking will not help you LEAVE Eve, but it will make it cheaper to get there, and cheaper to get back to Kerbin. So with aerobraking, you can put most of your fuel supply towards this particular burn. However, Eve has a pretty unforgiving atmosphere, so a heat shield is recommended. What exactly are you planning on doing at Eve? You can potentially save quite a bit of fuel by not getting all that close to the planet in the first place. Of course, if you're aerobraking on the way in, at least your periapsis is going to be quite low. It should be manageable to get most any type of ship to Eve and back, but it takes some care in the design and staging. I would call Eve (orbitally, at least) a medium-difficulty planet: harder than Duna, easier than Moho. Jool is more complicated and generally more expensive, but gives you more options in terms of gravity assists or whatnot. Happy to provide some more specific ideas to extend range, if you want to provide some more mission details and/or pics of the craft in question. EDIT: Think I got the wrong number of the launch window planner. Also, my post was talking about returning from Eve Orbit... not Eve surface.

You can see your max acceleration on Kerbal Engineer Redux. Or, if something shows you TWR scaled to Kerbin surface, you can divide by ten to get an approximate number (since gravity is same as Earth, 9.81 m/s2). But I'm not sure why you'd really need to... maneuver nodes incorporate this into the burn times.

Nuclear engines are great for long-haul trips of this nature (just make sure to use LF only tanks). If you have a very small payload, you may want a stage or two of small chemical rockets before you get to the nukes. But if you're sending a sizeable base, nukes probably make sense for the top stage. It's also helpful to design so you can discard your empty tanks (and possibly some of your nukes) as you burn out the fuel. As far as flight plan, agree with Snark that you really need to nail the ejection burn. You want your transfer orbit to be just tangent with Jool's orbit at your apoasis. If you can set up your ejection so that Jool is at the new AN or DN, you may not need to do a plane change en route. Or sometimes you can do some of the plane change at the same time you eject, which can save a lot of fuel (this works best if Kerbin is at or near the AN or DN when you leave. Whether this is the case depends on what launch window you pick). To conserve fuel on the capture side, you'll probably want to do one (more more) of three things: aerobrake at Laythe, aerobrake at Jool itself, or get a gravity assist (Tylo works best). For the first two options, you'll likely need a heat shield. The inflatable shield is ridiculously good for things like this. Even if I plan to aerobrake at Lathe later, I generally start with the Tylo assist, since that won't kill you if you're going too fast. Hopefully that all makes sense - if not, happy to provide some screen shots or more info.

That's odd. Not sure if this is causing this particular issue, but you can only use one set of decouplers at a time to attach a particular part (this is a result of the logic KSP uses to connect part to part). So your dual decouplers are at best doing no good, and at worst could be playing a role in this problem. Generally you want to put the decouplers a little over halfway up the booster (this helps get a clean separation as the decoupler pushes the booster away). Your upper decoupler location is pretty good. If you want more rigidity in your connection, use struts for additional connection points. They are far smarter than decouplers and will disengage automatically when you drop the boosters. One strut per booster, mounted near the bottom, will likely to the trick. Also, maybe your boosters have attached to the core rocket, rather than the decouplers? Sometimes it take a little work to ensure that they're on right. This is one of several things I like better about the next-level decoupler, that provides a little space between parts. To address both these issues, I'd start with a clean rocket core, then attach one set of decouplers, then attach the boosters (making sure they attach right on the decouplers), then strut as necessary.

Fortunately my lander was so bottom-heavy that flipping was not really an issue. (It was catastrophic when I tried a top-mounted shield, though). I did have some wobbling problems at first, but was able to keep it nice and straight with some beefier reaction wheels (to be detached with the heat shield) plus some airbrakes on top.

That's a good idea. Maybe some day we will finally be able to do engine shrouds in other sizes. Every Rockomax 8 tank + Terrier lander setup would work so much better on launch. True for the ceramic heat shields, but the inflatable heatshield does not use ablator, so you can brake for as long as you want. That thing is pretty OP, though I guess the flipping issues it can create help even it out a bit. But I basically never use the ceramics anymore.

Yeah, it ended up having a lot left in my latest test, from around 500m altitude. Maybe I could lose the outermost set of boosters, or a little bit from the core stage. But leaving in a margin of error does not seem like a bad thing.

There is also a nice mod called "Science!" It gives you a little pop-up window that has similar data, is accessible when you're flying, and is searchable. If I'm not mistaken it also tells you if you've collected, but not yet recovered, a particular experiment.

Agree this could also be useful on staging. For example, I was recently considering using separatrons to make a heat shield flip. Would have been nice to delay the burn by a second or two, to give the shield some time to fall away from my ship.

Just spitballing, it might be nice to have a mechanic that provides for diminishing science returns for data from a given celestial body. For example, as you take a surface sample across various biomes on the Mun, the science output for each one goes down. This seems realistic, as we would learn a lot more new stuff about the Mun from the first sample than we would from the tenth sample, even if samples from other parts were a little different. This might also require you to go interplanetary to max out the tech tree, which is too easy to do now just grinding the moons. Alternatively, this might be an unpopular opinion, but maybe there should be fewer lunar biomes (especially on the Mun). It just promotes grinding by taking a science hopper from a station to the surface and back, which is not exactly exciting. Of course, science labs would probably need an overhaul as well. I think they're still way too powerful. Maybe rather than granting extra science (and a multiplier at that), they should just let to access the "recover" value of data without taking the data back to Kerbin. Or maybe the lab should be the only way to reset Mystery Goo and the Science Jr.. Or give the lab some unique experiment of its own, rather than multiplying data from other experiments.

That heat shield cannot be destroyed by any force known to man (or Kerbal). It is of the Kraken. It is eternal.

Interesting idea - this would definitely save weight for launch, and go a little easier on Eve entry. But also adds complexity and more parts to fit under that heat shielld. Yep, my revised version put all that stuff on decouplers. I ended up adding a lot more chutes (24 total). The original version might have been enough for an engine-assisted landing, but that's a little fussy given the demands of Eve. Definitely better to let the chutes do all the work. Seems like this would save a little mass on takeoff (when it's much more important to be light), but add again add complexity as well as mass for the reentry. Personally I think ladders are light enough to stick with here. I was more concerned about the lousy kerbal climbing skills not working, but they did, even on Eve. Anyhow, I got down to Eve and back to orbit with a slightly revised version of this design. Other than the parachutes and decouplers mentioned above, the other major change was to swap out two of the Darts for Vectors, and go with just one Vector rather than two under the core stage. Here's my mission log. Thanks everyone for your input!

Mission report posted at:

Hi folks, I thought I'd share a report from my first successful mission to plant a flag on Eve and get back to orbit (I skipped out on the Kerbin return part since this was a shakedown test for career mode). I didn't start with any particular objectives or constraints - just make a ship that can survive Eve without cherry-picking a mountaintop landing. But my design grew around the idea of using those heavy, expensive, efficient Eve ascent-engines for the entire trip. This lead to a somewhat ridiculous, heavily asparagus-staged launch vehicle, but it all worked quite well once I worked the bugs out. (Which admittedly took a while). Here's a link to the album with pics and narrative: http://imgur.com/a/tqh5M And here's a sample pic of the ship in action on Eve reentry: (Full disclosure: after doing the entire mission, I went back and stitched these screenshots together from my various save files. So everything might not flow together exactly. But it s a generally accurate account of how everything went down).

As other folks have said, it's a tradeoff. Build it in LKO, and will likely be very floppy (unless you reinforce with KAS struts or the like). Build it in Duna orbit, and you'll be managing a bunch injections at the same time. This can get tricky... you definitely don't want two ships going into atmosphere at the same time. Also, if you use aerobraking (generally a good idea), it tends to make docking a little messier. I'd strongly consider launching the whole thing from Kerbin in one go. No docking, much less floppiness, and as a plus, you get to use those enormous engines that are not practical for a lot of missions. In particular, I love the Rhino for moving space stations to the moons or nearby planets. But this might rule out some un-aerodynamic designs, like ISS-style right-angle-fests.

Funny you should mention that. I have some saves with chutes fully deployed, but it turns out that as soon as the physics kick in, an uber-Kraken instantly annihilates my entire ship and even sends the camera back to space. Luckily, I also had a save from just before chute deployment, I've done quite a bit more testing with that save, and, for posterity, here are the results: Trying to shimmy with my reaction wheels did not do much - my ship is pretty much locked in place. This might be because my parachutes are spread pretty broadly over the top half of my craft. I assume if they're closer to the CoM, it's easier to pivot. Releasing the shield just after my chutes deployed (1000 meters above ground) seemed to give the heat shield enough of a drop to blow up on impact. A couple of the other pieces of the drop rig survived, but my ship crushed them on landing without damage. This might work even better if the chutes were set to deploy earlier. I tried jettisoning the heat shield just as the chutes were slowing my ship down. (i.e., at 10-20 m/s). This worked: the shield dropped off safely, but started tumbling as it fell, moving it out of the landing zone. I think the bit of remaining horizontal velocity did the trick here. I also tried giving the shield a little push by lighting up one of my off-center engines at the same time I jettisoned. This worked very well: the shield tipped and started tumbling immediately. This is probably similar to what a Separatron would do, but simpler and only costs a tiny bit of fuel. One semi-cheaty option. First, dropping the heat shield at about 100 m/s velocity caused it to blow up, but (sometimes) did not destroy anything else on my ship. And one very cheaty option: if I dropped the shield as soon as I reloaded (i.e., before physics kicked in), the shield flew upward right through my ship. It might eventually fall down and hit me, though. So basically I guess I had some bad luck last night, when I could not get the shield to either budge or blow up in multiple attempts. I think I'm going to call this ship the Hex, since it (1) uses a lot of six-way symmetry, and (2) is CURSED. At any rate, I have now 100% successfully landed on Eve, and gotten back to orbit! Now it's time to do it for real in career mode. Thanks again for everyone's help!

At last... success (mostly)! The single-shield design worked great, and kept the ship straight and cool on descent. After my chutes deployed, I was down to about 7.5m/s. And the heat shield, weighted down by my extra reaction wheels and a decoupler, fell non-explosively to the surface. Only one more problem. The heat shield fell... straight down. And did not blow up on impact. So when my lander also fell straight down, it landed right on top of the accursed heat shield. That thing just will not leave me in peace. I guess I could try to use my main engines to angle away while I'm coasting down. Or attach a couple sideways separatrons to the heat shield to blow it in one direction or another. But these seem like manageable issues. Thanks again for y'all's help!

Yep, data is now consumed when loaded into a lab. To get around this, I try to leave one copy of the data in the instruments, and put one copy in the a command pod. That way you can processs one copy and send the other one back home. Of course, that does not work if you fo multiple experiments in one trip. Though you can add another command pod just to hold the second set of data.

I suddenly have a desire to do this at the same time I fire a KAS harpoon at the shield. Pop! Sounds like Kerbal engineering at its finest.