My First Eve ascension suggestions

[Nich]

I have only landed on Duna but the Eve window is 60 days away in my current career and I want to maximize my opportunity. I will probably also send down a rover so I can hit a couple more biomes so I may remove the science equipment.

[Foxster]

What's the question?

One thought right away: Get rid of the extra crew cabin unless you are desperate for the extra crew. It will make it much much harder to get to orbit with it. 

Also - what are the decouplers for and where are the engines?

[Rhomphaia]

50 minutes ago, Foxster said:

What's the question?

One thought right away: Get rid of the extra crew cabin unless you are desperate for the extra crew. It will make it much much harder to get to orbit with it. 

Also - what are the decouplers for and where are the engines?

Guessing the engines are above the decouplers, which are there to ditch the landing gear on take-off. 

[Plusck]

I must admit I haven't got a huge amount of experience landing on Eve, but that ship looks like it will be a pain to heatshield properly.

You'll basically need to put 2.5m heatshields on each of those 9 segments of your lander. Since the bottom isn't flat, you might have clipping issues. You'll definitely have clipping issues with the landing legs. However, on the plus side with all that heatshield you probably won't have any need for the airbrakes...

And it looks like you need to find another 4000 m/s dv from somewhere...

[Nich]

Each 1.25m stack has a 1.25m heat shield on the bottom of it that will be decoupled before takeoff.  Above those decouplers I have Aerospikes.  I thought I saw people getting to eve orbit with 2900 ASL dv.  It has a vacuum dv of about 5300. And in initial twr of 1.47

I was worried the landing gear would blow up so I changed the outside heat shields to 2.5m in hopes of protecting them

Edited March 1, 2016 by Nich

[Plusck]

I'm pretty sure you need the full 8000 m/s to get up off Eve. It's known to be the single hardest thing to do in the game.

I've landed probes a few times. First attempt simply blew up in a millisecond due to poor understanding of exactly how deadly the atmosphere is. All the rest have been fine (including where landing legs were only just covered by the heatshield). But as long as you can keep pointed retrograde, it's really easy to land. Even parachutes seem a bit of a waste considering how slow you are by the time you get near the surface.

However, the gravity is crushing. A plane might be able to get to orbit on much less fuel, since it can avoid most gravity losses, but a rocket cannot.

Personally, if I were you I'd go to Gilly with the crew and drop off a probe or two on Eve on the way. That'll give you eyes on the ground and some idea of the challenge you're setting yourself

[Plusck]

Also, I wouldn't be surprised if those landing legs were crushed on landing. And 8000 m/s is apparently an optimal burn from a high altitude start.

[Nich]

So I did some spreadsheeting and found for my payload optimal dv/ton for the last stage terrier is 6 FL-T100 tanks (of merged into there larger counter parts).  And for the second to last it was 18 FL-T100 tanks.  These 2 stages alone got me close to 5km dv.  I now have 8160 vacuum dv but I have some concerns.  

Is .67 TWR too low for a second to last stage?  I have gotten some Kerbin craft orbiting with .5 TWR in a middle stage but how different is eve?  

Is 1.33 initial TWR too low?

How fast do I need to be going before I need nosecones?  The first 3 stages only last 20 second and 480 dv

Should I invest in tail cones or rapier spikes for the middle stages?

 

[Snark]

You'll likely need a much higher TWR, something closer to 2 (relative to Eve gravity).  That's because gravity losses on Eve are excruciating, and you have to climb really high (like 15-20 km) before starting your gravity turn.

You need much stronger landing gear. Use the heavy ones, and you probably need twice as many of them. Also you're going to need struts to reinforce the ship, if you don't want to snap apart on landing. A potential problem with your design is that there's a big lever arm with multiple joints between the landing legs and the central core, which can lead to a lot of mechanical stress. Consider having all eight directly connect to the center, instead of in onion layers. An 8-way asparagus can work well.

 

[KerikBalm]

Well, this is what I used to get 3 kerbals down to eve sea level and back up...

The bottom 2.5m docking port is to allow it to be towed by a tug to eve. The docking port burns up during re-entry, but who cares

This is it in ascent configuration:

Two of the stack have no engines, they are drop tanks because the vectors are really potent... a 2nd version I made has aerospikes underneath them because its starting TWR is a little low, I have the impression that this extra aerospike pair helped it get to orbit with more dV.

After jettisoning the first side stacks:

 

and the next staging event (the core of the asparagus cluster):

Final stage that gets to orbit:

It seems to be quite a bit bigger than yours, but its got good margins, and will take off from sea level.

Unfortunately, I don't have any screenshots with the dV or TWR stats... I'll try and get those later today.

[Snark]

On 3/1/2016 at 4:22 PM, Plusck said:

However, the gravity is crushing. A plane might be able to get to orbit on much less fuel, since it can avoid most gravity losses, but a rocket cannot.

I can't speak with authority, since the last time I tried an airplane-style ascent from Eve was before 1.0.  However, I'd have to say that generally no, a plane won't help you-- I expect it would actually be considerably worse than a vertical-ascent rocket.  (Also, anecdotally:  every time in the last year I've seen someone post pictures of their successful stock Eve ascent vehicle, it's been a vertical lifter.)

The problem is that on Eve, the soup is just so darn thick.  You have to get up over 20km or so before you can really start to stretch your legs, horizontally.  Until then it's just a grind, trying to balance "get me out of here fast" against "don't lose too much to drag."

The plane just doesn't win.  The math cancels out.  Having a more-horizontal path means less gravity loss per second, yes... but it also means you take more seconds to get up into thin air, by the exact same factor.  If your ascent angle is 30 degrees, for example, your gravity loss per second will be half what you'd have on vertical ascent... but you'll have to spend twice as many seconds climbing.

There are other factors that stack the deck in vertical ascent's favor, too.

First:  One way to reduce gravity losses during vertical ascent is to boost TWR.  So why don't we do that on Kerbin?  Well, because it doesn't make much sense-- on Kerbin, we're not actually climbing straight up for very long.  The air thins out fast with altitude, and by the time we're at 10km (less than a minute from takeoff), we're already tipped over 45 degrees or more.  Most of the dV to orbit is spent traveling mostly horizontally, which is a much more effective way to reduce gravity loss.  But on Eve, you pretty much have to just grind, grind grind for the first 20 km; you can't even think about starting your gravity turn before then.  So "boost TWR to cut gravity loss" suddenly makes a lot more sense.

Next:  On Eve, spaceplanes lose a major benefit that they have on worlds (like Kerbin) that have thinner atmospheres.  On thin-atmosphere worlds, a plane is very quickly going a significant fraction of orbital velocity.  Kerbin spaceplanes, for example, hit Mach 2 when they're still fairly low.  That's already a third of the way to orbital speed!  Not only have they just knocked off a big chunk of the work needed to get to orbit, but they also get a nice side benefit that they've effectively reduced the strength of gravity because of traveling a large fraction of orbital speed.  But on Eve, you don't get that benefit.  Orbital velocity is a lot higher than on Kerbin, and your plane has to travel much slower while it's going through the "soup" up to 20 km or so.  Which means that you lose basically all of the benefit.

Finally:  On Eve, drag is king.  Ships need to be super-streamlined; anything even a little draggy is very expensive.  Wings = drag.  Therefore, they're best avoided unless they provide some significant benefit that can outweigh the drag penalty.  (Plus there's the practical matter that reentry heat on Eve is just brutal.  Heat shields are pretty much mandatory, and wings have an unfortunate habit of poking out in places that are hard to heat-shield.)

Now, when an airplane ascent does make sense is if you can find some way to free yourself from the tyranny of the rocket equation, typically with a mod.  For example, RoverDude's Karbonite has a jet engine that works in Eve's atmosphere, so the huge boost in jet Isp over rocket power pretty much dominates all other considerations.  Ditto Kethane.  Another example is Porkjet's Atomic Age, which has a nifty nuclear turbojet that runs in atmosphere without fuel at all.  Eve airplanes work great with all of those, because you can build a plane that climbs very cheaply and efficiently to well above 20km before you have to start spending precious rocket fuel.

58 minutes ago, KerikBalm said:

Well, this is what I used to get 3 kerbals down to eve sea level and back up...

Nice ship! 

One of the things I love about KSP is that there's so much flexibility, and I always enjoy seeing the various ways that different players solve the same problem.  However, I gotta say that for Eve ascent, I suspect that there must be a certain amount of convergent evolution going on:  most of the screenshots I've seen for successful Eve ascent vehicles, post 1.0, look very similar.  Just about all the ones I've seen consist of a tall skinny central stack (either a lone Big Orange, or a couple of stacked FL-T800's), with 4 to 6 radial boosters consisting of stacks of two FL-T800, and a final stage with 2-ton LFO & Terrier perched on top.  Or something very close to that.

That's not to say that you couldn't possibly solve Eve ascent a different way... but it sure looks like that must be pretty close to optimal, given how many designs trend in that direction.  (My current favorite ascent vehicle for my own career play-through looks a lot like Kerik's, but I pack it into a slightly tighter footprint, and fit the whole shebang behind a single big 5m heatshield from SpaceY.)

All of which is to say @Nich, it's worth taking note of Kerik's design. 

Edited March 2, 2016 by Snark

[Gojira1000]

You'll want to be on the ground with about 9K dV just in case. Mind you, I returned from sea-level with that (well, 200 meters), so if you can land on that high super-plateau your mileage may vary.

[KerikBalm]

yea, converegence... aerospikes, mammoths, and vectors are the only engines that are suited for the lower stages of eve.

The mammoth... lives up to its name... its massive, its only for massive eve ascent vehicles, not simply getting some kerbals up.

Aerodynamics dictate tall and skinny. How tall and skinny is dictated by TWR that a single stack can achieve... I chose the vector because it lets one build really tall and skinny... but then its too tall and skinny, and too floppy, also, staging is generally a good idea.

I'm not super happy with the design though, because I need to throttle back soon after dropping the first stage... and that means that I'm hauling up useless dry mass in the form of heavier than needed engines. I'm not sure if the extra engine mass is enough to justify the added drag from another stack with less massive engines.

I'd like to invert the drop order of the revised design with an extra aerospike pair so that the vector pair drops , leaving 3 aerospikes propelling it (instead of 2 vectors and an aerospike)... but then its under powered... maybe if I made the first stage with vectors even taller and skinnier, so that the stages with 3 aerospikes is lighter... but then there's also the launch vehicle constraints... and my launch vehicle for that design is not very convergent with other people's launch vehicles (hint: look at my avatar).

[Nich]

hhhmmm... I cant seem to see the post with kerbalms ship.  

I plan on landing the ship empty and filling it up with an ISRU.  Current landing gear work on kerbin with the ship full and I have not unlocked the heavy landing gear yet  That is about 2 years away when the Duna Expedition returns.  Parachutes get it down to 7.5 m/s full as well.  I was having trouble getting TWR up with just aero spikes and was considering vectors.  Most likely replace the aero spikes on the 3rd stage with vectors and change the first, second and third stages to the largest 1.25m tank.  Most of my Kerbin rockets start about 1.7-1.9 anyways

I was thinking about adding struts but I was worried about drag.  It can be a limp noodle if you go off prograde but as long as you maintain prograde it is fine.  I have no problems quick saving and taking 20-30 attempts to get off eve.  I will call those simulations   I just want to bring the right craft.

[Snark]

9 hours ago, Nich said:

hhhmmm... I cant seem to see the post with kerbalms ship.

That's because it got voted up and is now up at the top of the page, right under your original post.  Welcome to one of the more confusing features of the Q&A forum feature. 

9 hours ago, Nich said:

I was thinking about adding struts but I was worried about drag.

A small number of struts are okay on a big ship, just don't go nuts with them.  But if you can get by without them, great!

If you can land the ship empty, that will help-- one of the big problems with Eve ships is that they have to be really massive, and they have a tendency to break apart on landing due to the mechanical shock-- even at low speeds.  If you're landing it empty, it has a whole lot less inertia and should be much better able to withstand landing shock.

[Superfluous J]

If you want to actually watch a gravity turn on Eve, I start testing an Eve launcher right at the start of the below video.

The "successful" launch that gets into orbit starts at 7:30. From what I've read above I was too aggressive so may have been able to keep going straight up to 20km. I may try that and see if it's better.

Note: This is a Twitch stream, so some of my random things I say are responses to chat.

Also Note: I don't actually make orbit, but started much lower than expected and got really, really close. I'm confident the ship would work from anywhere I'm likely to land.

 

Edited March 3, 2016 by 5thHorseman

[Nich]

@KerikBalm

Ok i see it now I was actually working on a config very close to that but I just could not get 8000 dv.  I rebuilt your ship and it only has an initial TWR of 1.09 with only 7280 dv.  It does jump up to 1.91 after dropping the supply tanks.  I can see how 2 extra aerospikes would help significantly.  All said if you add a little fuel to you terrier and another big tank to your final aerospike you would significantly increase your dv/ton on your final 2 stages assuming they have enough TWR.

 

[Nich]

@5thHorseman ha ha thanks I totally forgot that you could jetpack.

 

So I think I have settled on a ship.  I am really thankful for all your help.  My testing ship was named Eve lander v13 however this is Eve.  This ship gets a REAL name.  I was thinking Wall-E because it is chasing Eve.

[KerikBalm]

Yea, I think you're right about it having less than 8km/s vacuum dV...

But its fairly streamlined, and the atmospheric Isp:vacuum Isp ratio is better than for most other engines... like a swivel for instance, so the "real dV" you get can be higher even if vacuum dV is a little low.

As to your suggestions: yea, I thought about packing on more fuel to later stages, but as you observed, the TWR of those later stages is a bit low.

The starting TWR without the extra aerospike pair is also really low... my philosophy was to just pack on more and more fuel until it became pointless... obviously, adding fuel once TWR is 1.0, just means burning fuel and not going anywhere until TWR is over 1.0 again... Without actually doing much math at all, I just decided that my limit would be a TWR of 1.1 ... and since I had to pick fixed size tanks, I ended up with 1.09... knowing that in that case... 90% of the dV spent is initially lost to gravity, as opposed to about 50% later (assuming a TWR of 2:1, which is what you need to travel at terminal velocity... though in practice, you'd need more than this as terminal velocity increases as you get higher... and then it gets more complicated as you start climbing at an angle) But then again, almost none of that initial dV spent is lost to air resistance at the lower speeds. Climbing slower than terminal velocity wastes dV to gravity drag, but less to air resistance. So while its not optimal, a lower TWR is not as harsh of a penalty as it initially appears I think.

That said, I think 1.2 or 1.3 are probably better "minimum acceptable initial TWRs"

There's other things to worry about... 

As eve has a thick atmosphere, and strong gravity. When travelling close to terminal velocity (to minimize dV loses), the aerodynamic forces are quite high. To minimize dV losses, one wants a tall skinny rocket.

This results in prime conditions for rocket flipping.

On top of that, most designs enter engines first... so your design needs to be aerodynamically stable travelling in two directions. Fins to stabilize it on the ascent make it unstable on the descent. The tanks draining top down are a problem for stability, I had my fuel lines set up to drain the reverse direction from the side tanks (and yes, my final tweaked design did add aerospikes to the side stacks), but no such tweaking is very effective for the vector stacks because the central aerospike drains so little fuel compared to the vectors... between the first and second staging event is when my craft would flip the most.

To make it stable on entry into eve's atmosphere... thats what those side pods at the top with airbrakes and fins are for... they also double as science packages and points of attachment for the parachutes. Additionally, as others have noted, even landings slowly with a big massive craft in eve's 1.7g gravity field... can break your craft. I use a quick retroburn just before touchdown, and the fuel tanks in the side pods supply that fuel without cutting into my ascent fuel

Heating is also a major concern. If I start turning too soon, the pod explodes, if I start turning too late, its too inefficient and I may no make orbit. If I wait a little long and make a more aggressive turn, it risks flipping (the strong vector gimbal can help there though, but I fly with their gimbal limited for most of the ascent... I wish I could change the gimbal limit with an action group.. just like I wish I could enable/disable fuel flow from a tank with an action group to make the top tanks drain last)

Now... landing it empty or nearly empty and filling it with ISRU can be a good idea. You could leave the top tanks empty, and the bottom tanks full, just to help make it tail heavy to make it stable during the descent. Its something I considered, but I wanted to make a lander that could land anywhere on eve regardless of the ore concentration... and be a "standalone" craft, rather than one that would need surface rendezvous to fuel it (in which case you could send it down empty, and then send down separate fuel tanks to fuel it if there's no ore nearby). But... if you've got a landing site in mind that has a good ore concentration, go for it, the ISRU is a good solution.

Eve is much tougher, with many more problems to overcome with a design, than before 1.0.

People said it would be too easy when the dV requirements were coming in at 8km/s... I think its great now. A complex challenge to tackle and be proud of conquering.

Edited March 3, 2016 by KerikBalm

[Nich]

Eve mission is on the way.  Just wanted to thank everyone for all the help.

 

[Candlelight]

Eve is the only planet where MOAR BOOSTERS is the ONLY solution.

[Plusck]

Well, I also decided to try, using a virtual copy of KerikBalm's design.

It didn't go too well, because:

a) my interplanetary/capture stage was about 600m/s short on dv, so I had to do a lot of high passes through the atmosphere to get the speed down. This rubbed off all of the ablator on two of my heatshields.

Spoiler

b) I didn't include enough torque authority, so when it got down to about 62km, it lost retrograde hold. It was slow enough not to burn up, but all the science gear was lost.

c) the ship was spinning too much when I tried to jettison the heatshields, and they took bits of the rocket with them.

d) I didn't pack enough parachutes - landing velocity (even with bits of ship missing) was about 12 m/s.

Spoiler

e) trying to cushion the landing with engines, the whole thing fell apart. Phosy, amazingly, survived in the Mk1 pod which was one of about 5 bits that didn't explode.

 

Phosy was happy to win the world first for landing, but shocked and dazed by the destruction and the feeling of being very, very alone with little hope for rescue.

Luckily for her she awoke a few minutes later, still in orbit around Eve in her underpowered, under-torqued, insufficiently strutted ship. She got on to mission control to tell them not to go through with the mission.