"Best" engine for Mun landing...

[RizzoTheRat]

Has there been an ISP/thrust/etc thread recently?  Most of the calculations I can find seem to be a year or two old so presumably out of date as I believe engines have been rebalanced a bit over the last year.

Anyway after having started a new career game I'm interested in making my early game Mun/Minimus landers as efficient as possible, so hunting around the net I found this post http://imgur.com/a/SYKWj and set to work in excel (slow day in the office)

By my calculations, if I wanted to make a Mun lander with say 2000dv, then to get the lightest vessel (adding extra engines if needed to maintain a TWR>2 on the Mun)  for a given payload (anything that's not engine/fuel/tank) I should use: 

• < 1.5 tonnes = Spark
• 1.5-2.25 tonnes = Terrier
• >2.25 tonnes = NERV

Can't seem to post dropbox link as an image... https://www.dropbox.com/s/m01f6mbcxx400fp/engines.JPG?preview=engines.JPG

I was quite surprised that the Aerospike doesn't get a look in, and also surprised how low a payload you need to make the NERV worth it, that's a payload less than the weight of the engine.

So a single Mk1 lander can, materials lab, goo pod, landing gear etc is nearing the top of the Spark zone, while my currently preferred configuration of a pair of Mk1 cans, various science experiments, wide landing gear, RCS and a docking port is firmly in NERV territory.

Have I got that right?

 

 

 

[Choctofliatrio2.0]

That sounds about right. I'm not sure, though. I never thought to use Nervas that much on Mun landings, mostly because they are unlocked pretty late in career.

[Gaarst]

Aerospikes are not excellent in vacuum, this is why they are not there. The advantage of the Aerospikes is a small Isp gradient with atmospheric pressure; outside a space shuttle/plane (good luck with the gimbal though) or Eve, they are just like any other engine.

Surprised at how light the NERVs become the best, that's less than their own mass, are you sure you did not make a mistake somewhere ?

EDIT: in your Imgur album, the value at which the NERV overcomes the 909 for 2 km/s dV is closer to 4.5t

Edited January 21, 2016 by Gaarst

[RizzoTheRat]

I was trying to see if my last strategy of unlocking the NERV after a couple of Mun landings, and then building a NERV powered lander to do all Minimus and the rest of the Mun was worthwhile.  I'd doubted it when I realised how light the lander was compared to the engine, but looks like I was right originally.  For minimus the lower gravity measn you don't have to add more Sparks so they're best up to about 2 tonnes, but the NERV's still worth it for my bigger lander

I've seen a  few posts praising aerospikes recently, mainly from a TWR perspective, I guess that's for launch from atmospheric bodies rather than airless then.

[MagicFireCaster]

Aero spike is pretty good, those 340 of Isp are pretty great for 1.25 transfer stages, but no so much for landers

[Choctofliatrio2.0]

I generally use Terriers, because they're my favorite engine and I'm bad at the game.

[Miravlix]

For Mun the primary issue for me is that nothing is flat, so you don't want a tall narrow lander, but a wide one that don't risk falling over on slopes.

The NERV seems ill suited for this, as you would have to put it in the center and then build the ship down around the NERV, making it hard to make a solid ship.

[wumpus]

I'd strongly recommend higher TWR.  The Terrier has lots of advantages in limiting gravity losses, especially if you aren't coming down at a completely sideways angle.  Also if you are using lander cans, are you using a NERV to get a larger "command module" to orbit around the Mun?  Maybe a command module around Kerbin (keep it just for reentry).

Remember you are wasting half your thrust while thrusting vertical at TWR=2.0  Anywhere close to that and a higher engine is better.

Other issues:

mk1 capsule: best for Kerbin re-entry.  Does not need any docking gear (both docking ports, RCS modules, RCS tanks, lights on both ships...).  Allows science dance when landed.

landercans: tricky for Kerbin re-entry (I'm not sure how to keep them hot side down).  Allows science dance.

command chair: not for Kerbin re-entry.  Does not allow science dance.  At .1 tons makes landing trivial (but not always docking afterwards).

Nerva engine: can haul from kerbin circularization (and probably a bit of ascent as well, just don't push it) to landing and back with minimal fuel (and thus mass for first stage).  Note that if you use side tanks for stability and landing on the Mun, you might be able to pull off a single drop tank above your capsule (connected to the side tanks by fuel pipe) and drop that after circularization and/or munar insertion burn.  You will likely want to build a landing structure (legs won't reach past the nerva engine) unless your balance is really good (last I looked finding a flat zone on the Mun was *hard*).

There is a lot to say for the Nerva engine, but mostly it is getting to the Mun.  If you can get the mass low enough for a decent TWR, it is ideal.  If you have airbrakes, consider using a landercan and returning with the NERVA (don't think its possible without airbrakes, maybe behind a heat shield and large tailfin structure).  Otherwise I would recommend just using the mk1 capsule and terrier (note the tricks with side+top tanks still work with the terrier, and are probably needed even more).  The spark is for the last bit of landing, while the terrier can get you all the way to the Mun (although probably isn't for much of Kerbin circularization, much less ascent).

Edited January 21, 2016 by wumpus

[Nich]

Why has no one mentioned the ion?

[RizzoTheRat]

The TWR=2 was just as a minimum for the calc, I only expected it make a difference to the spark, but upping the minimum TWR to 4 is interesting

• <1.25 = Spark
• 1.25-2.25 = Terrier
• 2.25-2.75 = NERV
• 2.75-4.5 = Terrier very closely followed by Aerospike
• >4.5 = NERV

 

As for lander height, this was the lander I used for Minus and some of the Mun in my last game, it coped nicely with slopes I think the payload works out at somewhere around 2.5 tonnes so a good guess to use the NERV.  Pushed it's own fuel dump to Minimus and netted nearly 1200 science in one mission

Edited January 21, 2016 by RizzoTheRat

[Jetski]

Use the offset tool to push the NERVA up into the lander can until just the bell sticks out, just make very sure you toggle to disable Shroud in VAB. Strap on 4 radial LF tanks with fuel lines to the engine, 4 legs, and 4 struts to keep it from wobbling and you have a lander with 6k dV for mun, minmus, whatever. Put a docking port on top and use it as transfer engine to push labs and fuel dumps interplanetary, then land when you get there.  Cant do tylo or atmosphere planets, but moho eeloo moons etc easily.

[Norcalplanner]

LV-Ns are really expensive in career. Putting a terrier under a Rockomax 16 works great as a Mun lander with around 3.5 to 4 km/s of delta v.

Randazzo's VX series engines are also great - there's a 40 kN 1.25m engine, and a 120 kN 2.5m engine, both optimized for vacuum, which I use all the time.

[Rune]

29 minutes ago, Norcalplanner said:

LV-Ns are really expensive in career. Putting a terrier under a Rockomax 16 works great as a Mun lander with around 3.5 to 4 km/s of delta v.

Basically, that. Who needs a light lander if you can lob a cheap one to orbit on some SRBs. Nukes are not really worth it unless you plan on using them for a looong time, or recover them. Plus, you get big landers out of them. And don't discount the Poodle, it actually has a higher Isp than the Terrier these days.

 

Rune. The LKO-Munar surface-LKO trip is very doable in a single stage.

[Agost]

Squad should seriously improve a bit the Aerospike, it makes almost no sense right now related to its cost and position in the tech tree...

I'd say 310 sea level ISP would be a better choiche, while keeping other stats the same (well, I'd like to have thrust vectoring too since AFAIK it's feasible with real toroidal aerospikes, although not as much as classic gimbals)

[Waxing_Kibbous]

For fun I occasionally use Thuds, their high TWR cuts landing time considerably

[Tourist]

3 hours ago, Rune said:

Basically, that. Who needs a light lander if you can lob a cheap one to orbit on some SRBs. Nukes are not really worth it unless you plan on using them for a looong time, or recover them. Plus, you get big landers out of them. And don't discount the Poodle, it actually has a higher Isp than the Terrier these days.

 

Rune. The LKO-Munar surface-LKO trip is very doable in a single stage.

I don't think I've ever used anything other than a poodle for my Mun Landers and never had any problem with them.... I'm not a number crunching Kerbalnaut though, more of a "strap-it on, see if it works" kind of Kerbalnaut, so its possible my crafts may be quite inefficient. 

[Violent Jeb]

I put my hat in for the Terrier. I AM a number crunching Kerbalnaut.

I've build several lunar landers, and i typically design them able to land on the mun and return to kerbin in a single stage. no clip or offset.

NERVA is too much engine for the task at hand. By the time you put liquid fuel in there you'll be up over 10t. Landing legs will also be an issue.

Put a rockomax X200 on top of a spark (and your pod), and you get over 3000 dv for 4.6T & 1190 Funds. That's a very competent lander with a minimum of 3 parts.
Something like 5-10 TWR.
Strap on SAS, science and RCS to suit your needs. It's so consistent (part count and weight) I usually don't bother with much else.
FWIW, the COM is so low on this lander that I can consistently land without any landing gear.
Want to bring a friend? Simply throw on a lander can. This design is a huge workhorse for me because of its flexibility.

I find landing takes apx 600-800 dV depending on your orbital height (i'm usually coming in from 50km orbit).Return trip to kerbin can be anwhere from 800-1500 depending on how efficient you are and where you are on the surface. This usually leaves ~1k Dv to play around with, mun "hops", science experiments, (even lunar orbital rendezvous if that's your thing) etc. This method is so efficient I can usually design a launch stage which can accommodate 2 or 3 of these landers without too much trouble, and get 3x individual kerbals and landers on the muns' biomes per mission.

.

Edited January 21, 2016 by Violent Jeb
added some stuff

[Red Dwarf]

7 hours ago, Gaarst said:

Aerospikes are not excellent in vacuum, this is why they are not there. The advantage of the Aerospikes is a small Isp gradient with atmospheric pressure; outside a space shuttle/plane (good luck with the gimbal though) or Eve, they are just like any other engine.

Surprised at how light the NERVs become the best, that's less than their own mass, are you sure you did not make a mistake somewhere ?

EDIT: in your Imgur album, the value at which the NERV overcomes the 909 for 2 km/s dV is closer to 4.5t

I've found the aerospike to be incredibly helpful in vacuum. Both my Tylo and Slate landers ran on them, and they got better efficiency than my vacuum-rated transfer stage

[Guest]

If you can get it there, the poodle is my personal favorite for Mün landings.  It's really good in vacuum.

[Temstar]

I'll like to add: the advantage of nuke increases dramatically when the lander is intended to be reused. Despite making the lander heavier, nuke engines potentially allow the lander to use much less propellant per trip to the surface than chemical rockets and that means much lower running cost as you need to launch fewer tankers to feed the lander.

[RizzoTheRat]

Anyone know if the 8:1 wet:dry mass is still correct for fuel tanks, and is it also true for liquid fuel tanks?  (re comments on that Imgur album, that was some time ago and I believe several engines have been tweaked bit since)

If so then I think I need to expand this spreadsheet and plot lots of nice graphs as it becomes quite interesting.  I've now added in the poodle, and the Mun with a minimum TWR of 4:

Payload	Engine
< 1.5	Spark
1.5-2.25	Terrier
2.25-2.75	Nerv
2.75-4.5	Terrier
4.5-5.75	Nerv
5.75-7.5	Aerospike
7.5-8.25	Nerv
8.25-10	Poodle
>10	Nerv

I'm surprised by the result for Minimus too

Payload	Engine
<1.75	spark
1.75-2.25	Terrier
>2.25	Nerv

However partly why I got thinking about this is because I'm planning on using USI Kolonisation and building a big base on either the Mun or Minimus, so my number of engines to get the right TWR goes out the window a bit as if I go for a skycrane approach I'll need at least 4 engines, and would then add them in pairs, in which case for the Mun at a TWR>4

Payload	Engine
<5	Spark
5-9	Terrier
9-11.5	Nerv
11.5-14	Terrier
>14	Nerv

But that's assuming 2000dV still, I'd still need to look at whether I reorbit and reuse the skycrane, or build disposable one's that I then recycle at the base.

There's a hell of a lot more variables here than I originally thought

 

 

Yes I'm sat at work with Excel when I'd rather be sat at home with KSP

Edited January 22, 2016 by RizzoTheRat

[Streetwind]

Yes, 8:1 fuel mass ratio is correct for all stock tanks except for the ROUND-8, if I recall correctly.

I'm firmly on Team Terrier, by the way. By the time I have the LV-N, I'm not going to the Moon anymore - even for colonies there are more interesting destinations. And for the price of a single Nerv I could probably build a complete lander plus return stage...

[Temstar]

There's no reason why a LV-N lander originally intended for the Mun can't be reused for say, Ike or the smaller Jool moons. In fact a lander like this doesn't even need an interplanetary transfer stage. With the relatively high TWR needed for landing compared to interplanetary burn it's a simple case of putting an extra LF tank and heat shield on top and you can fire it to Jool all on its own.

[RizzoTheRat]

Completely agree that cost isn't something factored in to the above, and that the Nerv makes more sense for reusable craft.  That's my plan with the pictured Minimus lander, it's already done about 20 landings on Minimus and the Mun, with only the original launch and one refuelling mission (with crew return vehicle) launched from Kerbin, so hopefully a lot cheaper than 20 individual missions, and I'm getting 2 sets of science data everywhere (pilot and scientist in separate cans).  The intention is to use KAS/KIS to add additional experiments to it as they become available, and push an extra fuel and life support module that it undocks before landing.  The problem with that is it needs enough fuel to re-enter Kerbin orbit.  I hadn't thought of putting a heatshield on top though.  A heat shield attached to a docking port than be fitted on top for the aerocapture sounds a great idea, the wings and landing gear should withstand a fair bit of heat, I'll just need to swap the solar panels for retractable ones or ox-stats.  Although I've now added Infernal Robotics too so I wonder if I could design something with a retractable heat shield?  I love this game!

As for base building I was thinking of putting a refuel/resupply base on Minimus (or the Mun but probably Minimus) and a relatively cheap crew transport craft between there and Kerbin, so a lot of fairly heavy payloads going in and out of at least one moon.

 

[MrOsterman]

What's odd for me is that I'm trying to engineer a multitask mobile base and lander so for me, the Thud is percolating out as the engine to use to finalize the landing.  Main engines to slow and then Thuds along the mainline of the ship to flip it belly down and land.  I had a design that layered the bottom of the ship with Aerospikes but I couldn't keep it level, and I'm not sure I can do the same with the required number of Terriers to keep the ISRU, drills, and passenger compartments from turning into a crater.

I may have to give up on the belly landing though.. it's proving to be really hard to manage.