An Eve Lander/Ascent Vehicle

[AlphaMensae]

(Note, this is still in v 1.04)

Proving that what goes to Eve does not have to stay on Eve, I set about to make a not-entirely-stock Eve lander for my developing fleet of interplanetary ships.  It may not be the best design, but I did successfully make and test a (rocket) vehicle that can leave the Purple Ball O' Doom, without being too big or tall.  

This is a "final" revision of my third design, meant to take a Mk 1 lander can down and back up.  To make EVA easier without putting a zillion ladders on, I put a Mk 2 lander can at the very bottom to act as an EVA "elevator", i.e.transfer the Kerbal down and up.  Could have used another Mk 1, but I wanted a bigger attachment node for the 5m stack separator.  It also meant that the main stage is purely radial engines, with no central one; the central and outer tanks feed the inner engines.

The pics combine two separate tests (used HyperEdit), the entry phase from the previous revision and the landing/launch from the final version.  The only things different between the two ships was that the four Kiwi retro-burn engines (from Space Y) were replaced with a single Mainsail below the heat shield, a Mainsail replaced the 8 Swivels in the middle stage, plus 8 active winglets were added, and more fuel was added to the main stage.  I had thought the Kiwis could perform as descent engines to slow down the final few meters, but they did nothing and still ended up using 4 of the main engines, so I ditched them.  I made the changes to the middle stage after it kept flipping out after staging.  I wanted to keep both the middle and upper stage short, so either the Mainsail and/or winglets proved to be the solution.

The main engines are 8 Ratites from Space Y, a 2.5m engine with more thrust than a Rhino, especially at low altitudes (Rhinos proved to be useless on Eve).  Less efficient too, but that's not I needed for the first stage of an Eve ascent vehicle.

Most of the main stage tanks are from Fuel Tanks Plus, the 3/4 and 1.5x versions of the standard orange tank. Added some stock Rockomax tanks as needed.  Main chutes are 2.5m Mk16 XLs plus 8 radial drogue ones.  The big radial decouplers with built-in sepratons are also from Space Y (those things are so cool and useful!).  And the heat shield is a Tweakscaled 5m one from Space Y (not sure if it mattered) up to 20m size--no problems with heat there!

Middle stage as noted is a Mainsail on a Rockomax -32 tank with 8 radial FL-T400 tanks.  Upper stage is a FL-T400 with 4 -200s and 4 1.25m Poodles (used Tweakscale), they have a mass of .377 and 45 kN of thrust.  Thought about using aerospikes but wanted something with thrust vectoring and lower mass. The upper stage has an RCS fuel tank from FTP and 4 quad thrusters.  And of course I put MechJeb on it, though I only used it to hold pro/retrograde, not to land or launch.

The landing struts are LT-2s upsized 300% mounted on 200% pocket I-beams.  Several 2.5m reaction wheels are in the ship, with main power from 4 200% size RTGs

Not sure what was going on with the landing struts on the landing (broken? glitch?) but the ship was stable and took off fine.  Went into a polar orbit since that was how it came in--wanted to land on, well, land and not in the sea.

Started at relatively low altitude too, 654m ASL (heh, the previous test landed in the Highlands ).  Ended up with over 1150 m/s DV left after the circularization burn.

The first version of the lander included a top-docked rover, my latest "mega mini" type, the plan being to transfer the scientist Kerbal from the ship to the rover after then pop off the rover before landing and have it drive back to the ship, visiting several biomes on the way  That proved to be successful, so I stopped bothering to include it on later versions.  I might even change the plan to land the rover separately, or even several rovers...but that would require a 2 or 3-Kerbal version of the ship...guess what will be next?

 

 

[Foxster]

Well done!

A few suggestions:

Use a mk1 capsule rather than a lander can, it weighs a little more but has much less drag and that is more important. Drag is the major factor for Eve. 

Also lose  the docking port at the front for streamlining too. 

That many engines is overkill. You are adding a lot of weight. You will be more efficient with one Vector and a couple of aerospikes. 

I don't see any airbrakes. They will help a lot in getting it down safely. 

Put radial 'chutes and the landing struts on decouplers and lose them before the return launch to save weight.

Streamline the last stages that have multiple tanks. Will be causing a lot of drag. 

 

[AlphaMensae]

Yup, this proved to not be the "final" version, as I quickly started thinking of a redesign, with the goal being to reduce the size and weight.

Here's the newest slimmed-down version:

It uses four stages, three in the central stack and the fourth being the 6 radial Ratites.  Top stage uses 2 Terriers, middle stage uses 2 aerospikes, bottom one still uses a Mainsail.

My first test entered Eve's atmosphere just fine, but I ended up landing on way too steep of a slope, and it started sliding and breaking stuff, so no launch from the surface first time out.

There's a docking port in front because the ascent vehicle will have enough DV to meet up with whatever ship brings this to Eve.  I'm also considering bringing a largish rover down with the lander, it will separate before touchdown and the scientist Kerbal will be able to visit (hopefully) several biomes while driving to his/her ride off the purple rock.  That's why there's no science package on the lander (not yet at least  ).

[Rune]

A few useful tips to (further) reduce the mass of the ascent vehicle:

-Given Eve's great atmospheric pressure, you are better off using engines with the flattest Isp curve possible. The Aerospike is awesome in that regard, and it also has a high-ish Isp that will greatly reduce your launch weight.

-Put your Eve entry stuff so that you can ditch it before ascent, to save weight. Those parachutes and legs will weight (and drag!) you down on ascent, and you could use radial decouplers to put them in.

-You can use a "cruise stage" on the bottom to put the docking port, then ditch it just as you meet Eve's atmosphere. That way you can dock it to a transfer vehicle without taking any RCS down to the surface (or back up), and you can use a Sr. Docking port without screwing with the aerodynamics of your ascent vehicle. "But no docking on ascent", I hear you say? Well, you are going to make orbit with a tiny expendable capsule, just park it close to the return vehicle and EVA the kerbal and the scientific data over. Or Klaw it.

Using all those tips, I have an engineering prototype to use when the time comes, but I am afraid it is completely untested. I imagine you would prefer to come up with your own thing, but if I'm wrong, or you can get some inspiration from it... it uses a few fancy building tricks, so feel free to ask me if you can't replicate it or something.

Note how it fits the 3.75m aeroshell, which will protect it during Eve entry, after it ditches the "cruise stage" (which doubles as upper stage for the launcher) which has a Sr.docking port, RCS, and some solar panels. Once things are merely supersonic, the plan is to pop the fairing, extend the airbrakes to go subsonic, pop chutes, and proceed to landing. When the time comes to go back home, you eject the chutes and ladder, use the heatshield and landing gear as a launchpad, and lift up on a very aerodynamic ultra-asparagus tiny ascent vehicle with almost 8km/s in the tanks and a more than respectable 1.78 TWR at liftoff.

 

Rune. I think 8km/s are enough from Eve's sea level, these days.

Edited January 20, 2016 by Rune

[Redhaze]

Question, Rune, how do you keep your return vehicle stable on ascent? Mine keeps wanting to do handbrake turns.

[Rune]

13 minutes ago, Redhaze said:

Question, Rune, how do you keep your return vehicle stable on ascent? Mine keeps wanting to do handbrake turns.

As I said, 100% untested, and you have probably brought up a good cause for a redesign. But. There is some thought put into that. The basic thing is that the rocket doesn't empty any tank before the one below it is empty (if you look at your design, you empty the small grey tanks on top of the boosters first). That helps a lot, because as long as your center of mass is forwards of the center of drag, the rocket will be stable, or in other words, empty tanks on the back have a stabilizing effect on the rocket. Empty tanks on the nose of the rocket is a bad thing, of course. Another important thing is having a very drag-less nose, of course (parachutes are very draggy, BTW), and if all else fails I guess that I could always chuck a couple of winglets in there on the boosters to bring CoL further back.

But also, good piloting can make a very unstable rocket get to orbit. The trick is to do a perfect gravity turn, starting at just the right time, so you are never more than a couple degrees from prograde, where all the aerodynamic momentums are very small. Keep on trying to find the perfect altitude to start it on your own design!

 

Rune. Note that you are tackling one of the toughest jobs in this game, too!

[AlphaMensae]

Thanks, Rune, some good ideas there.  I had already started thinking about putting the landing legs on the bottom-mount lander can (acts as an EVA elevator instead of putting a bunch of ladders on), which will be the first thing ejected after launch.

I might just have to switch to 1.05 so I can use the bottom-node aerospike, that would simplify matters.  Or go back to the 4 1.25m Poodles for the upper stage

I'm definitely not going for a minimal or low-mass design, as the ship that will take this to Eve will be using Porkjet's Nuclear Lightbulb, so it can be a bit heavy.  Plus I like the challenge of launching an all-up multi-functional crew ship from Eve.  I also want it keep the capability of bringing my planned rover down with it:

That's just the base version for low-grav moons, I made various slightly alternate models, only differing in number, type and placement of docking ports.  The Eve version though uses aerospikes instead of Sparks and has 4 radial chutes.

[Redhaze]

3 hours ago, Rune said:

(if you look at your design, you empty the small grey tanks on top of the boosters first).

Whoops! Don't confuse me with the OP! I just butted in for my own purposes.

3 hours ago, Rune said:

As I said, 100% untested, and you have probably brought up a good cause for a redesign. But. There is some thought put into that. The basic thing is that the rocket doesn't empty any tank before the one below it is empty (if you look at your design, you empty the small grey tanks on top of the boosters first). That helps a lot, because as long as your center of mass is forwards of the center of drag, the rocket will be stable, or in other words, empty tanks on the back have a stabilizing effect on the rocket. Empty tanks on the nose of the rocket is a bad thing, of course. Another important thing is having a very drag-less nose, of course (parachutes are very draggy, BTW), and if all else fails I guess that I could always chuck a couple of winglets in there on the boosters to bring CoL further back.

But also, good piloting can make a very unstable rocket get to orbit. The trick is to do a perfect gravity turn, starting at just the right time, so you are never more than a couple degrees from prograde, where all the aerodynamic momentums are very small. Keep on trying to find the perfect altitude to start it on your own design!

 

Rune. Note that you are tackling one of the toughest jobs in this game, too!

I kinda cloned your design and got issues with stabilities. I'm a little confused, 'cause yeah, the fuel flow SHOULD keep it stable, but it flips out at any fuel load regardless. I did try (small) winglets, but they didn't help squat. Didn't know parachutes are that draggy, though. I'll try and see if that makes a difference, but I fear not.

Of course, shooting straight up till the bottom core stage is gone works, but that's not exactly efficient.

[AlphaMensae]

Made more revisions:  Reduced the radial boosters to four, replaced the Mainsail stage with another Ratite, replaced the Mk16 XL chutes and Rockomax 16 tanks with the FTP nose tanks, put 3 radial chutes on the boosters, added another set of legs and reduced the size to 300%.  Almost made it to orbit...more tinkering.

 

[AlphaMensae]

Well, after many revisions and scrapped redesigns, I got my lander down to 244t (minus the heat shield and deorbit engine for testing purposes).  One of the intermediate designs that worked was a revised version of my previous slimmed-down lander as noted above.  That came to 310t or so, but I thought I could do better.  So this was the result:

The first version of this lander worked great, had enough fuel to leave 1400+ m/s dV in the crew stage after the circularization burn, but that seemed a bit much, plus it was a bit too wide with too many parts, so I did some rearranging, deleting and adding to come up with the version in the picture.

The legs are on decouplers, and if released just after liftoff, make a surprisingly large fireworks show when they hit the ground.

The radial launch boosters use the Moa engine from Space Y, with the base toggled off so they'll "fit" in a 1.25m tank, in this case the T1200 tank from Fuel Tanks Plus, plus a T800 and a 1.25m nose tank from FTP.

Main (lower) stage uses a Mainsail, the tanks being the 3/4 Jumbo and and adapter tank from FTP.

The middle stage uses Rhinos Tweakscaled to 1.25m, both in the core and radials.  It's my alternate to the Aerospike, which are rather heavy and have no gimballing.

The upper crew stage is simply a T400 and Terrier.  I added a nose cone and separator on top of the docking port for the launch, it's ditched when the Terrier is lit.

[Blaarkies]

On 1/27/2016 at 2:47 AM, Laguna said:

Well, after many revisions and scrapped redesigns, I got my lander down to 244t...

It feels awesome to actually get it working the first time! How mass does your final stage have?(the very last stage when dry)

I figured an all or nothing approach: 1 command seat, on a ockto probe(thin), a 200ec battery, oscar, ant engines. 

This all weighs 0.42t dry as the payload/final stage. I built a 4stage rocket onto this, with fairings, ladders, legs, heatshields and chutes all detachable. It all weighs 52t when fully fueled.

Now back to topic: Do you use a ratio to determine how much mass your stages should be? or do you just wing it?

[AlphaMensae]

I actually revised it again, in order to see how much smaller I could make it and to provide direct access to the surface from the crew stage so I could actually take back the science collected from the surface.  I decided to drop RCS and docking ability, will use a small retrieval craft to bring in the Kerbal tasked with this mission.

This is still in 1.0.4, so no Vector and non-stackable Aerospikes.  I started up 1.0.5 to see what I could do there, and it's true, the Vector makes an Eve lander all too easy.

I started with the final stage, and used KER to figure out the boost stages, in this case optimizing for dV and mass.

Final result came to 168.5 t without the heastshield+retro engine, 210t with them.  

The final stage is essentially just the Mk1 capsule, a mostly-empty FL-T400 tank and a Terrier, so a bit under 2t.  Lower stage of the core is a Rhino with radial-decoupled Aerospike boosters. Main engines are four Mainsails.  The adapter tanks and slanted nose tanks are from Fuel Tanks Plus.  Those big radial decouplers are from Space Y and include sepratons. Landing gear are Tweakscaled 200% LT-2s on radial decouplers.

[RA3236]

When I saw this thread looking for a good rocket for Duna I passed it and then went 'Wait, WHAAAT?' Hurriedly scrolling back towards it. It's not every day you see an eve lander.

[Foxster]

Some of these craft show just how the addition of a few or wrong items at the tip of a rocket leads to over-engineering of the rest of it. 

You have to be very focused on the essentials for Eve and pare them to the bone. Don't add extra probe cores, batteries, mono-propellant, ladders, solar panels, decouplers, etc to the bit that will be left by orbit. Also make sure that all the landing stuff like struts, landing engines, 'chutes, air brakes, etc are detached before lifting.  

And make very sure to reduce drag to the minimum. That is slightly more important than mass for Eve because the atmosphere is so soupy. For instance, a Mk1 capsule will take less dV to loft than a lander can despite it's extra weight.  Make sure you use skinny stacks (tanks and engines) rather than fat ones too. 

You can get an Eve return craft down to about 45t before de-orbit and 35-40t before lifting again and it will make a 100km orbit with 500dV to spare.  Sure, your crewman will need to hitch a ride home but that Kerbin-return craft will be much smaller than some of the lifters above. 

KISS. 

(PS - Any craft that you can get down from orbit and back to space at Eve is a very impressive achievement! It shows an engineering prowess and a lot of work). 

Edited February 3, 2016 by Foxster

[Starlionblue]

Looking at your vehicles, they seem way too wide and short. Given the thickness of the atmosphere, the trick to Eve ascent is a narrow, streamlined profile. Fairings are key. I did some tests and adding the fairing shown below to the middle and upper stages increased first stage performance by over 30%. Also, wide and short vehicles easily become unstable and start tumbling in a thick atmosphere. I had to add a reaction wheel thingy to the middle stage of my vehicle or it would easily tumble while staging (no gimbal on the Aerospike.)

Interstage fairings are important if you have gaps. Basically anything that gets rid of gaps and sharp transitions along the stack will be worth the extra weight. Even using the slanted nose cones instead of the centered ones on the side boosters helps. On the mid stage, I used Oscar-B tanks for the side boosters in order to keep the cross-section narrow.

Note on fairings: The do add drag to the front, which makes your vehicle want to back flip. A judicious use of fins/canards and gimbal authority will solve this.

You must limit Q so you don't just push against massive drag. My rudimentary testing shows that somewhere around 80-100k seems to be the max Q sweet spot for Eve launches. If you set max Q lower, you won't accelerate enough through the lower atmosphere and your second stage will be left without enough oomph to carry on. You need to get your lower stage well up, at least to around 15-18km altitude, and don't turn before 15-18km, and keep the turn very gradual. You shouldn't be horizontal before at least 65-70km.

My vehicle has around 8400 delta-V, which is more than enough for a 100km orbit from sea level. However this is only true if you have a streamlined vehicle, good thrust to weight ratios and a good ascent profile.

I started by building the ascent vehicle and testing it with HyperEdit. Once I knew it worked, I added the descent vehicle, a "basket" with a heatshield at the bottom. The tops of the "arms" have fins and RCS to keep the thing pointed in the right direction during entry. Once the chutes are out, the heatshield is deflated and jettisoned. After landing, the "arms" are jettisoned, leaving the ascent vehicle on a "launch pad" made of the bottom of the "basket".

Ascent vehicle specs below. Craft file is here: Eve descent and ascent vehicle.

• Lower stage.
  • One central Vector . One FL T-800 and one FL T-200 tank.
  • 6 Dart Aerospikes in asparagus boosters shedding in pairs. Four boosters with one FL T-800 and one FL T-200. Last two boosters one FL T-800.
• Mid stage.
  • One central Dart Aerospike. One FL T-400
  • 8 Sparks in asparagus boosters shedding in pairs. 5 Oscar-B tanks per booster.
• Upper stage. One Spark. One FL T-200.

Initial climb from Eve. Large fairing to keep airflow smooth around the messy mid stage. Slanted nose cones. Narrow and reasonably smooth cross section:

In the VAB. You can see the 8 mid stage boosters under the fairing.

Entering the Evemosphere. The fins and RCS on the "arms" keep the right end pointing forwards. The "basket" has two FL T-200 tanks at the bottom for RCS fuel.

After landing. "Arms" jettisoned. I had to add little hardpoints to my "platform" to keep it from sliding down hills.

Mid stage Two pairs of boosters gone. Two to go.

 

On 2/4/2016 at 1:11 AM, Foxster said:

Some of these craft show just how the addition of a few or wrong items at the tip of a rocket leads to over-engineering of the rest of it. 

You have to be very focused on the essentials for Eve and pare them to the bone. Don't add extra probe cores, batteries, mono-propellant, ladders, solar panels, decouplers, etc to the bit that will be left by orbit. Also make sure that all the landing stuff like struts, landing engines, 'chutes, air brakes, etc are detached before lifting.  

And make very sure to reduce drag to the minimum. That is slightly more important than mass for Eve because the atmosphere is so soupy. For instance, a Mk1 capsule will take less dV to loft than a lander can despite it's extra weight.  Make sure you use skinny stacks (tanks and engines) rather than fat ones too. 

You can get an Eve return craft down to about 45t before de-orbit and 35-40t before lifting again and it will make a 100km orbit with 500dV to spare.  Sure, your crewman will need to hitch a ride home but that Kerbin-return craft will be much smaller than some of the lifters above. 

KISS. 

(PS - Any craft that you can get down from orbit and back to space at Eve is a very impressive achievement! It shows an engineering prowess and a lot of work). 

You speak truth. Any extra 100kg of you can leave on the ground will dramatically improve performance. And streamline, streamline, streamline. The extra weight of a fairing will be repaid multiple times in lower drag.

35-40 tons? Colour me impressed. I know it is possible but the best I have managed is about 60 tons for the ascent vehicle. Typically if I add 5-10 tons on top of that performance improves dramatically. The one in the pics is 67 tons at blastoff.

I went back and checked an older design I made. 119 tons, but it had less delta-V than this one! That's what you get for using a Mammoth.

Edited June 23, 2017 by Starlionblue

[Foxster]

8 hours ago, Starlionblue said:

35-40 tons? Colour me impressed. 

Actually, my best weight to date was for a challenge here:  ~18t for a sea-level to orbit manned craft. About 25t is pretty comfortable and requires no special flying or dodgy offsetting of parts. 

[Starlionblue]

9 hours ago, Foxster said:

Actually, my best weight to date was for a challenge here:  ~18t for a sea-level to orbit manned craft. About 25t is pretty comfortable and requires no special flying or dodgy offsetting of parts. 

Time for some pondering. And testing. And more pondering. And more testing...

[Starlionblue]

44 tons from Eve sea level to a 150km orbit. Boo-ya! Pondering apparently helps. It's not 25 tons yet but I'm getting there. 

I do love the Vector engine but it weighs 4 tons. Replaced the central stack engine with a Spike. Cut tankage all over, added two extra Sparks to the upper stage for a total of three, exchanged the fins for canards (the Spike doesn't gimbal) and my ISP is still higher than before. Not as fast in the first couple thousand meters due to lower TWR, and that is an efficiency loss, but once she hits the max Q wall any excess thrust is wasted weight. Using a Vector at one third to a quarter thrust is a lot of wasted weight.

A key aspect seems to be high TWR until you're almost at the circularization burn. And not burning up. Thing get really hot near the top.

Craft file for the ascent vehicle here.

Edited June 24, 2017 by Starlionblue

[Starlionblue]

Further tweaking got me down to just under 40 tons. I'll have to ponder a lot more to get down to 25 tons....

Tweaked the tankage and replaced the three sparks in the upper stage with a Terrier. It adds 0.2 tons but the vacum ISP goes from 320 to 345, which makes a big difference above 50km.

Craft file here.

[Rune]

3 hours ago, Starlionblue said:

Further tweaking got me down to just under 40 tons. I'll have to ponder a lot more to get down to 25 tons....

Tweaked the tankage and replaced the three sparks in the upper stage with a Terrier. It adds 0.2 tons but the vacum ISP goes from 320 to 345, which makes a big difference above 50km.

Craft file here.

With a Terrier for an upper stage, I'd move more fuel to it, moving the mass ratio closer to the ideal e. That would give you more dV at the cost of less TWR (when you no longer need it), boosting the core to lower velocity and saving the energy for the upper stage. That little engine sure packs a punch. You'd get more total dV, and who knows, maybe that means you can drop the first onion stage, dramatically lowering total liftoff mass, and most importantly, drag (by about 30%). BTW, can't that fairing have a smaller, flush cross-section?

For other drastic weight reductions, you could fit a whole other stage inside the fairing: Chair, tiny fuel tank, ant engine. More dV that you'll know what to do with, tough control becomes complicated. That way you can certainly drop two spikes, since final mass on orbit is truly ridiculous(-ly small).

 

Rune. The eficientest stage always has the golden mas ratio.

[Starlionblue]

Thanks for the ideas.

I don't really want to go chair. It doesn't feel "realistic".

The fairing is that wide due to the MechJeb box on the side. 

[Foxster]

Here's a simple 26t craft that can make sea-level to orbit...

[Starlionblue]

1 minute ago, Foxster said:

Here's a simple 26t craft that can make sea-level to orbit...

Cool. That's inspiring thanks. I had a feeling Vector was the way to go somehow... 

Are those intakes? Any particular reason?

Also, why hardpoints instead of decouplers?

[Foxster]

Circular intakes because they are lighter and have less drag than nose cones. 

Hardpoints have a lot less drag than decouplers. Even though they weigh a little more, the reduced drag is more important.   

The other thing to check out, if you haven't already looked at it, is the GravityTurn mod. 

You can use it to show you how to optimise your flight to orbit through throttle control and minimising AoA. You'd be surprised how much dV is usually wasted. 

[Starlionblue]

Thanks for that.

Unfortunately that add-on seems no longer supported. Guess I'll have to do the old trial and error in MechJeb instead.