Help me improve efficiency

[Enkiel]

Hey all.

I've reach the point where i now need to start looking at other planet, an to efficiently build my craft...

here's roughly what i'm flying with right now ;

Don't mind the lander on top, its a returning drop pod for a science vessel, Right now i'm using 9 mainsail in asparagus staging, which usually get me to either moon with enough fuel left in the last tank to start slowing down on the moon surface.

I sadly dont have picture of my "true" lander, which has all the research gear i currently own, a MK 1-2 command pod, a hitchhiker (for tourist!), 3 thud and roughly 1500-2000 unit of fuel. (and obviously, enough landing gear/parachute to land on mun and minmus, and get back)

I have roughly 500 research stored (and more, once i finally succeed planting a research center on the mun), and what i currently show is pretty much the best i can get (no nuclear engine for example).

The craft currently cost roughly 150k. I'm looking at either decreasing its cost, or making it more efficient (so i can reach further). What would be the next step?

I'm tempted to add more stage to the asparagus, but i think it goes against what i'm trying to do.

Edited June 1, 2015 by Enkiel

[capi3101]

Hmm...a screenie of your actual lander would be nice. Perhaps one from the VAB with KER active. I'll have to guesstimate based on the screenie provided.

From the description, it sounds like you're running with a lander of around twenty tonnes and its delta-V is about 1500 m/s or so, right? And then you've got a Mainsail engine underneath that with a J64 tank which acts as a transfer stage providing the remaining 1500 m/s or so of delta-V, right?

I'd start by changing out the booster rockets with Skippers - you don't need quite as much oomph as you did pre-1.0. That's assuming I'm anywhere in the ballpark of where your current mass is.

The other thing I'd think about is "do you really need to send to send seven Kerbals to the Mun all at once, along with all the science gear"? If so, more power to you; I might suggest combining your lander and transfer stage into one big package, so that you could get away with using a lower-thrust/higher efficiency rocket (say, a quad of Terriers or Sparks) once you're up there.

Edited May 29, 2015 by capi3101

[Feradose]

You need more fuel on your booster stages

[Prepper-Jack]

I would imagine a poodle would be more of use than thuds in vacuum, giving you more total dV with a bit less thrust. The thuds also produce a pretty heavy amount of drag on the way up - so streamlining can save you even more dV on your main stages.

Aerodynamically, I believe those large nosecones are pretty bugged. They offer no benefit over a flat top. You might want to consider the Rockomax adapter and a smaller nosecone that actually works. The long plane tails make excellent noses.

As for weight - the 1-2 command pod at 4.12 tons is really damned heavy for what it offers. You're better off using a Mk1 lander can and another hitchhiker can. 3.16 tons for 5 crew instead of 3.

What's the dV of each of your stages?

[Enkiel]

this lander has a 30 t mass (pretty sure my standard one is close to that, maybe 32-33t), and 3000 delta-v (i'm not very good at leaving a planet yet, and i rarely have that much left when i land back on kerbin).

all 9 engine are mainsail yes. And each one is attached to an orange tank.

Skipper in the middle? Should he start with the rest, or i just start him once in or near orbit?

I'll try to snap some picture tonight to show you the complete stats.

As of right now, i reach stable orbit with just the middle tank left (usually run out of fuel on the other 2 while pushing to get my orbit right).

- - - Updated - - -

You need more fuel on your booster stages

I'm not sure. Maybe, but i tried adding 4 S1 SRB-KD25k when launching, and it did nothing to improve my fuel. I still ended up running out of fuel on the 2nd and 3rd last tank before finishing my orbit...

I would imagine a poodle would be more of use than thuds in vacuum, giving you more total dV with a bit less thrust. The thuds also produce a pretty heavy amount of drag on the way up - so streamlining can save you even more dV on your main stages.

Aerodynamically, I believe those large nosecones are pretty bugged. They offer no benefit over a flat top. You might want to consider the Rockomax adapter and a smaller nosecone that actually works. The long plane tails make excellent noses.

As for weight - the 1-2 command pod at 4.12 tons is really damned heavy for what it offers. You're better off using a Mk1 lander can and another hitchhiker can. 3.16 tons for 5 crew instead of 3.

What's the dV of each of your stages?

i kind of got used to use the thud... they provide enough oomph to slow down my decent on mun very late. My first version had poodle tho, i could always revert to that (especially since my normal lander have large tank on the side, to cut on the height of the lander).

Large nosecone bugged? Damn, i guess i'll try tonight and see how it goes.

Not sure why i used that command pod on this craft. Since its my science one, i wanted the smallest one possible, just to be able to bring a pilote on board. I'm pretty sure my standard one has the conic one.

[Red Iron Crown]

I would start by paring down your payload as much as possible. That will ripple down through the rest of your stages.

[Enkiel]

I would start by paring down your payload as much as possible. That will ripple down through the rest of your stages.

in what way?

For the lab lander, i tried many different way to shorten it, but i always had the trouble that i want the driver to return home before the lab, so nothing can connect to the lab.

[Red Iron Crown]

It's kinda hard to be more specific without seeing your payload a bit closer, to be honest.

[FancyMouse]

My upper stage for Mun/Minmus lander is 10t with 360 unit of LF and LV-909 giving 1570m/s dV. It also has 5 crew capacity (lander+hitchhiker), all sciences instruments included (including from some mods), only without the lab. It has return from Mun/Minmus lots of times, and it only has 1 Mainsail during launch. Cost is 224k though - but the lander 184k. I guess instruments are expensive.

Without fuel and chute the lander is 5t, so given the mass of the lab, I guess I just need to double all my boost/fuel and should be able to handle it. 9 Mainsails does sound a little bit overkilling to me.

One thing that might be useful is that you can drop one stage on the surface of the target body, so the lander only needs the rest of dV to land and the dV to return.

[Enkiel]

So here it is ;

lander is 24t, with 2000 delta v.

Whole thing is 378t with 8550 delta v.

My lander is the shortest i could make it, with all equipment and a hitchhiker. And considering the following quest, i doubt i'll ever take the hitchhiker out....

- - - Updated - - -

Just noticed there's some stage missing...

S0 1.52 2,065

S1 1.04 2,201

S2 1.40 1,073

S3 1.50 2,408

S4 1.55 1,335

S5 1.58 0,581

[jarmund]

Since nobody else has, I will be the one to say it: MOAR BOOSTERS

Now that we've gotten that detestable expression out of the way, I'll add some pointers from a serious perspective:

- Those radial engines you have on the lander are somewhat inefficient. If you can, swap to an undermounted engine with higher Isp instead. My Duna lander worked great despite only having ~1.1 TWR on launch. I recommend the poodle for you, as a single one of those is similar to two of the radial ones, but with better efficiency and lower mass. It's also fairly short, so it's good for landers. My landers are normally built on Poodle, LV-909, or Aerospike, depending on the thrust needed.

- Consider scaling down. That pod plus the hitchhiker can should be a total of 7 kerbals. Do you need that many? An MK2 lander can is both cheaper and lighter, with only one seat less.

[DeMatt]

What-all's in the Service Bay? Materials, Goo, Thermometer, Gravioli, Barometer, and Seismometer? Anything else? 'Cuz if that's all, you could try moving the Materials Bay to between the capsule and the parachute, then switching out the 2.5 metre Service Bay for a 1.25 metre Service Bay and moving it there too. That'll knock 0.2 tons off right there.

You ARE stripping out nonessential Monopropellant, right?

If you only need to move five tourists, keep the Hitchhiker, but replace the 3-man pod with a 1-man, and add a probe core for control. They're not paying for a GUIDED tour.

Maybe try rotating the outermost boosters inwards, so from above it looks more like this:

1 2 1
3 4 3
1 2 1

[cephalo]

No middle stage eh? If that center tank is supposed to take you to Duna, maybe you should put a Poodle on there.

[Empiro]

To start with, switch the Mk55 Thuds with the Terrier or maybe a single Poodle (with a 2.5m fuel tank at the bottom to make things more aerodynamic). It will give your craft a bit more more crash tolerance too -- a single parachute might be too fast for the Hitchhiker can.

You also have far more total delta-V than you need. Around 7000 should be plenty to get to the Mun and back with some room for mistakes. This means that you can remove your first two stages and still have more than enough for a Mun trip and return.

Finally -- you're actually using Skippers all throughout by the look of things. I love those engines, but a pair of mainsails might be appropriate here, and it would let you reduce the width of your rocket

Edited May 29, 2015 by Empiro

[capi3101]

Okay. So at 24 tonnes, you need 39.12 kN of thrust for a 1.0 Munar TWR, 58.68 kN for a 1.5 Munar TWR. That's really all the more thrust you need this thing to have. At the moment, you've got a Mun TWR of 9.202...far above what you really need. You'd benefit in many, many, MANY different ways by swapping out those Thuds for something else. Hell, a Terrier would give you the thrust you need easily. Scale it down from 24 tonnes with 3 Thuds to a single Terrier; your new mass becomes 21.8 tonnes, your Isp increases to 345, your Mun TWR is 1.689 - and you suddenly have 2700 m/s of delta-V. Your cross section decreases in the process - and you're a lot less draggy as a result. 2700 is a bit low for a Mun transfer, landing, launch and return but it is doable.

Did the math on a trio of Sparks; the mass savings isn't worth it in this case on account of the lower Isp of Sparks.

So then you go through Temstar's asparagus design process for 21.8 tonnes; 15% payload fraction, so assume your rocket will have a final mass of around 145 tonnes. 1.5 launch TWR these days - you want in the ballpark of 2131.5 kN of thrust at launch. 22% of that in the core = 468.93 kN; you want a Skipper in the core. The rest distributed to six boosters = 277.095 kN - Skippers, or to eight boosters = 207.82 - Reliants. Let's say you go with eight Reliants and a Skipper - you have 13 tonnes of engines and 21.8 tonnes of payload... leaving 110.2 tonnes, which works out to a little over 12 tonnes per stack. In the center, use a X200-16 and X200-8 tank with the Skipper. For the boosters, try two FL-T800s and one FL-T400 per stack. Adjust the thrust of the boosters as necessary to give you your ~1.5 with each stage. Take any mass saved by my rounding errors here and use it for nosecones, fins, adapters and decouplers.

Yes, Temstar's method was for KSP 0.20, and produced a rocket with 4500 m/s of delta-V as a rule. I've found that the addition of the nosecones and fins usually lowers things down to the 3500 regime though, so it's still a valid method of building good asparagus. The only problems with 1.0.x asparagus is A) asparagus staging is generally more costly unless you play with a mod like StageRecovery (which is awesome, BTW) and you have to be careful to keep things to where most of the mass stays closer to the nose, or you're going to tumble out of control. I usually achieve this these days by keeping the core stage up higher than the boosters as much as I possibly can; that way by the time you're flying on the core alone the drag is too low to flip you around.

Start with those design changes and see what happens to your craft. Out of curiosity, what all do you have in your Service Bay?

Edited May 29, 2015 by capi3101

[Jouni]

Here is a quick version of the ship. The price is a bit under 100,000 funds.

The design principle was "don't touch the payload".

You were apparently using the Thuds for landing at Kerbin, as a single parachute isn't enough for a safe landing. It's more efficient to use Terriers instead of Thuds, and replace the nosecones of the engine nacelles with parachutes. It also turned out that you get more delta-v from the entire rocket with two engine nacelles instead of three, so I removed one of them. Finally I removed the unnecessary monopropellant from the pod.

The launch vehicle is a simple 2.5-stage rocket. The upper/transfer stage uses a Poodle, the lower stage uses a Skipper, and the boosters use Mainsails. The adapter/nosecone combinations look ugly, but they generate less drag than 2.5 m nosecones.

[Enkiel]

Wow, so many answer, thanks guys.

Since nobody else has, I will be the one to say it: MOAR BOOSTERS

Now that we've gotten that detestable expression out of the way, I'll add some pointers from a serious perspective:

- Those radial engines you have on the lander are somewhat inefficient. If you can, swap to an undermounted engine with higher Isp instead. My Duna lander worked great despite only having ~1.1 TWR on launch. I recommend the poodle for you, as a single one of those is similar to two of the radial ones, but with better efficiency and lower mass. It's also fairly short, so it's good for landers. My landers are normally built on Poodle, LV-909, or Aerospike, depending on the thrust needed.

- Consider scaling down. That pod plus the hitchhiker can should be a total of 7 kerbals. Do you need that many? An MK2 lander can is both cheaper and lighter, with only one seat less.

I had the poodle at first, but the legs were too short, so i had to move to side engine, that's pretty much the only reason i got the thud. Now that i have the longer legs, maybe it could work.....

about the hitchhiker, the problem i had was that 2 weren't enough (most mission i had, at least 3 or 4 kerbal were in tow....), and yeah, 7 is too much. And it is much cheaper (i think) to do one big trip instead of 2 "medium". Since the travelers mission seems to be more rare now, i guess i could remove it!

What-all's in the Service Bay? Materials, Goo, Thermometer, Gravioli, Barometer, and Seismometer? Anything else? 'Cuz if that's all, you could try moving the Materials Bay to between the capsule and the parachute, then switching out the 2.5 metre Service Bay for a 1.25 metre Service Bay and moving it there too. That'll knock 0.2 tons off right there.

You ARE stripping out nonessential Monopropellant, right?

If you only need to move five tourists, keep the Hitchhiker, but replace the 3-man pod with a 1-man, and add a probe core for control. They're not paying for a GUIDED tour.

Maybe try rotating the outermost boosters inwards, so from above it looks more like this:

1 2 1
3 4 3
1 2 1

Service bay have everything. I use the big one because i found the small one between the pod and the cone, it was f*ugly.... yeah, i make decision choice on look sometimes....

Uhm, wait.... the engine i have right now need oxidizer, no?

right now its something like this ;

    1
    3 
2 4 5 4 2
    3 
    1

No middle stage eh? If that center tank is supposed to take you to Duna, maybe you should put a Poodle on there.

No real middle stage, except like you said, the center tank which is usually almost full when i'm in Kerbin orbit.

To start with, switch the Mk55 Thuds with the Terrier or maybe a single Poodle (with a 2.5m fuel tank at the bottom to make things more aerodynamic). It will give your craft a bit more more crash tolerance too -- a single parachute might be too fast for the Hitchhiker can.

You also have far more total delta-V than you need. Around 7000 should be plenty to get to the Mun and back with some room for mistakes. This means that you can remove your first two stages and still have more than enough for a Mun trip and return.

Finally -- you're actually using Skippers all throughout by the look of things. I love those engines, but a pair of mainsails might be appropriate here, and it would let you reduce the width of your rocket

I actually have 5 parachute, there's 4 around the pod.

you mean, i should move from 9 tank to 5 ? that seems pretty extreme...

the engine are all mainsails underneath (unless i took a wrong picture...)

Okay. So at 24 tonnes, you need 39.12 kN of thrust for a 1.0 Munar TWR, 58.68 kN for a 1.5 Munar TWR. That's really all the more thrust you need this thing to have. At the moment, you've got a Mun TWR of 9.202...far above what you really need. You'd benefit in many, many, MANY different ways by swapping out those Thuds for something else. Hell, a Terrier would give you the thrust you need easily. Scale it down from 24 tonnes with 3 Thuds to a single Terrier; your new mass becomes 21.8 tonnes, your Isp increases to 345, your Mun TWR is 1.689 - and you suddenly have 2700 m/s of delta-V. Your cross section decreases in the process - and you're a lot less draggy as a result. 2700 is a bit low for a Mun transfer, landing, launch and return but it is doable.

Did the math on a trio of Sparks; the mass savings isn't worth it in this case on account of the lower Isp of Sparks.

So then you go through Temstar's asparagus design process for 21.8 tonnes; 15% payload fraction, so assume your rocket will have a final mass of around 145 tonnes. 1.5 launch TWR these days - you want in the ballpark of 2131.5 kN of thrust at launch. 22% of that in the core = 468.93 kN; you want a Skipper in the core. The rest distributed to six boosters = 277.095 kN - Skippers, or to eight boosters = 207.82 - Reliants. Let's say you go with eight Reliants and a Skipper - you have 13 tonnes of engines and 21.8 tonnes of payload... leaving 110.2 tonnes, which works out to a little over 12 tonnes per stack. In the center, use a X200-16 and X200-8 tank with the Skipper. For the boosters, try two FL-T800s and one FL-T400 per stack. Adjust the thrust of the boosters as necessary to give you your ~1.5 with each stage. Take any mass saved by my rounding errors here and use it for nosecones, fins, adapters and decouplers.

Yes, Temstar's method was for KSP 0.20, and produced a rocket with 4500 m/s of delta-V as a rule. I've found that the addition of the nosecones and fins usually lowers things down to the 3500 regime though, so it's still a valid method of building good asparagus. The only problems with 1.0.x asparagus is A) asparagus staging is generally more costly unless you play with a mod like StageRecovery (which is awesome, BTW) and you have to be careful to keep things to where most of the mass stays closer to the nose, or you're going to tumble out of control. I usually achieve this these days by keeping the core stage up higher than the boosters as much as I possibly can; that way by the time you're flying on the core alone the drag is too low to flip you around.

Start with those design changes and see what happens to your craft. Out of curiosity, what all do you have in your Service Bay?

That's alot of changes... The first thing, i'm not too fond of booster, especially 12 of them (if i count correctly?) I know that real one had it, but i dunno, seems like it was good at first, now, i'm less inclined to use them. I'll try the Skipper in the middle to see what it does with my middle tank... should that one be running during the aspartagus stages?

Here is a quick version of the ship. The price is a bit under 100,000 funds.

http://jltsiren.kapsi.fi/ksp/1.0/example_ship.jpeg

The design principle was "don't touch the payload".

You were apparently using the Thuds for landing at Kerbin, as a single parachute isn't enough for a safe landing. It's more efficient to use Terriers instead of Thuds, and replace the nosecones of the engine nacelles with parachutes. It also turned out that you get more delta-v from the entire rocket with two engine nacelles instead of three, so I removed one of them. Finally I removed the unnecessary monopropellant from the pod.

The launch vehicle is a simple 2.5-stage rocket. The upper/transfer stage uses a Poodle, the lower stage uses a Skipper, and the boosters use Mainsails. The adapter/nosecone combinations look ugly, but they generate less drag than 2.5 m nosecones.

Wow, that's kind of what i'm thinking... maybe i'm overdoing this and just end up weighting more and adding more power to compensate? I'm trying to find the middle between single craft that just goes in the air, and big thing that is too big for its own good.

I really like the simple design of this ship, but its a case of "i'm scared to flip it"...

- - - Updated - - -

EDIT :

i just tested, and the poodle is too long for the current legs i have....

- - - Updated - - -

EDIT 2 :

with the poodle tho, i'm down to 16 t, after removing the hitchhiker and the side tank replaced with a x-200 that gives me the same delta V.

[Empiro]

You're definitely using Skippers in the lower stages -- your craft size is 380T, 9 skippers provide 5850 kN of thrust for a TWR of about 1.5 or so.

And yes, you can drop the first two stages (for 5 orange tanks) and still have enough delta-V. In fact, you can drop the first two stages, make it to the Mun, land, get back into orbit, transfer to Minmus, land there, and then get back to Kerbin.

Use this delta-V map: http://i.imgur.com/UUU8yCk.png (the only difference is that now, you only need ~3500 m/s to get into orbit instead of 4500 m/s).

Edited May 30, 2015 by Empiro

[Enkiel]

oh yeah, i always have it in hands... i think its a bit on the "too close" range, but gives a good idea....

for the record, i tried the same setup with skipper, and didnt improve anything. If anything, it was worst.

[FancyMouse]

My suggestion would be -

Don't ever bother with Mk1-2 pod - it's heavy as hell. Mk1 lander can+hitchhiker is less than 3.2t with 5 crew capacity. Put another lander can or probe for control. Put another reaction wheel if you need it. This is already saving you ~3t payload.

And forget about the XL chute - the most useful chute is the radial one. When fully deployed, Mk2-R gives 30.4 effective diameter for 0.1t, while XL gives 37.8 (about 1.5x area) for 0.3t (3x mass)

And most importantly, as I said before, lander doesn't need to carry that much dV, at least to Mun/Minmus (not considering Interplanetary yet - that needs some different consideration).

All these payload mass reduction has cascading effect, as you might figure that it might be possible to use just a LV-909 as landing engine, and then you can save engine/fuel mass, causing an even lighter lander.

[Enkiel]

My suggestion would be -

Don't ever bother with Mk1-2 pod - it's heavy as hell. Mk1 lander can+hitchhiker is less than 3.2t with 5 crew capacity. Put another lander can or probe for control. Put another reaction wheel if you need it. This is already saving you ~3t payload.

And forget about the XL chute - the most useful chute is the radial one. When fully deployed, Mk2-R gives 30.4 effective diameter for 0.1t, while XL gives 37.8 (about 1.5x area) for 0.3t (3x mass)

And most importantly, as I said before, lander doesn't need to carry that much dV, at least to Mun/Minmus (not considering Interplanetary yet - that needs some different consideration).

All these payload mass reduction has cascading effect, as you might figure that it might be possible to use just a LV-909 as landing engine, and then you can save engine/fuel mass, causing an even lighter lander.

To test your theory, i used my lab lander without a return home (decreasing the landing pod to 16t). When i reached Mun, i still had my last orange tank over half full...

[FancyMouse]

Not sure what you mean by over half full because I don't know your baseline number - but it sounds like you're seeing a better result. If that's the case, that's great.

Let me show you my lander:

DOrbit science mod has a micro science bay to save the science bay volume/mass but there are more new instruments + TAC life support stuff so a similar stock lander shouldn't be much heavier/lighter. dV is distributed so that lower 2 stages drop on Kerbin and middle stage drop on Mun/Minmus. It has earned me thousands of science points by surface samples etc. with no trouble. When returning from Minmus I can even retro-burn for ~1k m/s upon aerobraking.

Lower stages are simple - Mainsail*1, Skipper*1, Poddle*1 (Rockomax has a great product line!), finally 909*1 (the lander, tanks in 4-way symmetry)

[Enkiel]

I think i can say i've reach much better efficiency.

Here is my new one ;

8000 delta V, 213t and 79k $.

compared to my old one ;

8550 delta V, 378t and 133k $

So far i landed on the mun without any issue, and did a pass by on minmus. Both time, i had plenty of fuel left (thank god, because when i tested with minmus, i only had one parachute).

Next target ; Duna !

Thanks guys.

Oh, for the record. I'm using mainsail. I did try skipper, but i was having trouble leaving the pad....

[FancyMouse]

Yeah depending on what you mean by efficiency. My stuff cost a lot because of cramming all science instruments there (see lander itself $~184k, engines and fuel below are just worth 40k), and I can fit 5 kerbals in there. Usually I just do a batch of rescues with it, then shoot to Mun, land, do science, plant flags, then return.

8000 seems overkill - mine is just enough for Mun landing + returning (~300 remaining) and a lot remaining if the same ship is used for a Minmus mission.

[Enkiel]

Yeah depending on what you mean by efficiency. My stuff cost a lot because of cramming all science instruments there (see lander itself $~184k, engines and fuel below are just worth 40k), and I can fit 5 kerbals in there. Usually I just do a batch of rescues with it, then shoot to Mun, land, do science, plant flags, then return.

8000 seems overkill - mine is just enough for Mun landing + returning (~300 remaining) and a lot remaining if the same ship is used for a Minmus mission.

Well, efficient compared to what i have

My "Driving" is the main reason i need so much Delta V. I'm practicing getting better at getting into orbit of Kerbin ; i think its where i waste the most power, ending up with an Ap of 130 km and Pe of 80 km most of the time...

To test, i took picture of each stage...

I start with ; 8055

kerbin orbit ; 3915

Mun intercept ; 3057

Mun Orbit ; 2675

Mun Landed ; 1499

Kerbin intercept ; 358

I did this with 3 tourist, and a probe (to free up an extra seat for the tourist)