Efficiency (or not): ISP vs Trust: Vaccum stats

[Kar]

Hi all.

I always have this problem in my career game. I usually tend to build big landers. The reason is, when I can I like to take most science tools, a crew of 3 and sometimes I take drills converters,etc. 

I play with mods, like life support, MKS, KAS+KIS, RT2, and many others to add a bit more difficulty to my game and to make it a bit more realistic. 

My problem is in the earlier tech tree, with no probes, no scans for resources to convert to fuel, I have a hard time to land my monsters landders and return them to kerbin from places like the Mun!

Most of my problems is solved after I get the nuclear engines. My lander have 2 engines on the side, each with is own fuel tank that are feed from the main tank (a rockmax) with fuel lines.

Some mods I use also add several type of engines, in the beginning I start using the 909's 2 of them on each side. But I have a problem regarding fuel, I endup with no fuel after trying to leave the SOI of the mun on the return to Kerbin.

What should I consider? ISP on vacuum or Trust on Vacuum? The 909 have about (cant remember now) 345 isp on vacuum but trust of 60kn. I have other engine, with about 500 ISP but only around 6 trust Kn. All this on vacuum.

Should I consider isp or trust in a engine regarding travel time, burn time, and enough force to take off from the mun (in this case) and return home?

The Nerva Nuclear engine have 60 Kn and 800 Vacuum ISP if a remember correctly, but I need another option better then the 909 considering that I have to use 2 engines. 1 engine, for example is the bottom of my lander is not a option, since I play with stage recover mod, and my lander has 3 stages that deploy on Kerbin entry to recover everything.

Any tips on this would be appreciated.

Thanks!

[Veskenapper]

In case you're not familiar with it, I_SP tells you how much power (delta-V, if you're familiar with the extremely helpful MechJeb panel for that) you'll get for each unit of fuel the engine burns. That means you can go further while using less fuel, which also saves the weight of the extra fuel, meaning you don't need as much thrust. Thrust usually comes at the expense of I_SP. An easy real-world example is cars – a Ferrari with a powerful V8 engine guzzles gasoline, while you can drive forever in a tiny little Ford Fiesta.

As for your landers, thrust is not that important. As long as you have enough thrust to get the lander moving, delta-V is a much more important stat, so I'd suggest you try out the various engines you have access to with the DV-panel open, to see how far you'll get with each setup.

Edited March 6, 2016 by Veskenapper

[Kar]

1 minute ago, Veskenapper said:

First off, I'd suggest you try building smaller landers early in the game

Lol.

Yea, I know.

I do small landers when I come up with the contracts to do fly by on the mun or mimnus, I also manage to land them with no problems. I already play ksp for some time, but I always have difficulty when I start building big landers. I know there other options but I am a "one man, on ship do all" thing lol.

Serious now, the main reason I do big landers, is when I have several contracts to get temp scans or have to collect samples on the mun or mimnus. Since I use kerbonite mod and all the stuff, I need to do a lot of flying in the moons of kerbin. Land, take off, take different course corrections and so on. They get big cause I need to drill kerbonite, convert to LFO to keep going. I guess I should say that my lander is not a really a lander, but a mobile refinery with crew, life support (those resources have some weight) batteries plus science stuff to do many biomes also in one go. So I need to refill my tanks.

The main thing in the earlier / mid game, is the lack of nuclear power. I am almost there, but I usually take my research for other areas before.

Im going to try different ISP engines to see if I get better results, at least untill I research nuclear engines, then the problem is gone.

Thanks a lot for your help.

 

[ForScience6686]

I use a small Lander with a fuel tank and transfer vehicle in orbit.  Hit a couple biomes and return to refuel.  That makes for much less weight to land with.  I also selected thus method to negate the time for mining.  I have a separate setup and can rendezvous with my fuel station to top off if needed.  It still took hours to mine all the science, so I don't want to add to that waiting to mine resources.

[Snark]

+1 to @Veskenapper's advice.  As long as you have "enough" thrust (e.g. a local TWR of, say, 3 or so on the Mun-- i.e. about 5 kN of thrust per each ton of ship mass), you'll do fine.  Isp is much more important, as you've discovered (with the nukes).

One way you can help fuel woes is to use more specialized equipment.  Lugging around extra mass just kills you.  For example:  Suppose you put a fuel depot in orbit around the Mun.  Doesn't have to be anything big or fancy, basically just a big empty fuel tank (plus a tank for monopropellant, Karbonite, and/or whatever other resources are important to you).  Equip it with a fair number of docking ports of various sizes.

Then you could have a specialized Karbonite-miner that has nothing but Karbonite drills & refining equipment.  It doesn't waste mass on other equipment it doesn't need.

And a specialized science lander that could be very small and light-- basically just a Mk1 lander can, two tons or so of fuel, a single Terrier or couple of Sparks, all the science instruments, nothing else.  You could build the whole thing for around 3 tons including fuel.  It can go hop around quite efficiently with very low fuel usage.

By having a family of specialized vehicles for different purposes, you're a lot more efficient because you don't lug around mass you don't need on a particular sortie.  For example, a single refueling run by your Karbonite miner might provide enough fuel for your science lander to hop to half a dozen different biomes, with refueling stops at the depot.

Anyway, just a thought. 

[Norcalplanner]

What @Snark said, except that I'd bring at least 4.5 tons or more worth of fuel in the lander so you can do more biome hopping before heading back up.

[numerobis]

"Should I consider isp or trust" -- both, and then some! Three things matter:

1. Efficiency (Isp).
2. Thrust, but really only when you're landing.
3. Mass.

The amount of mass you need depends on engines you use, plus how much fuel you need to bring. How much fuel you need to bring depends on Isp. How much thrust you need depends on mass. How many engines you need depends on how much thrust you need. How much mass depends on the number of engines. Oh snap, we're back to the start!

Really the best way to design is to try a few designs and do the math with each, then pick your favorite (for whatever design criteria you have -- mass, cost, beauty, reliability, ...)

Contra @Snark, I tend to build with TWR about 1.5 when I'm about to touch down, not 3. I prefer lower TWR because it's more tolerant to error: I don't accidentally overshoot as much. It's also cheaper, due to reduced engine mass -- example, a single LV-909 can lift about 20t of payload and fuel on Mun; a single LV-N can lift about 22.5t (plus itself). A trick for landing particularly heavy loads but still get good Isp is to pack two engine types: an efficient engine that doesn't have enough thrust on its own, and a lighter but less efficient engine to give a boost on the final stages of landing and the start of liftoff.

[Kar]

You guys are awesome !!! 

I guess I am doing all wrong lol.

My science / miner / refiner / lander (lol) weights more or less 50t ! So yea, it is a pain in the a** to get the thing move. I never built a station at any of Kerbins moons, but I will consider it.

By making my career game more difficult due to the type of mods I use, the main problem for making several infrastructures orbiting is the tech tree. I do not transmit science for example, I can't. I use remote tech so I always have to consider that. Probes also is a bit of a problem in earlier game. Life support is also another, since I cannot leave kerbals forever in space. I need to build habitats for them.

Considering also, that I must select my tech tree very carefully due to the mods and  want I need most (usually I try to unlock the most science stuff I can to get me more science per trip).

I almost hitting the 300 mark nodes, I still have to make the very expensive upgrade, but due to contracts mods I have, money is getting in. Also I use stage recover mod, I try to recover almost everything I launch into space, with an occasional tank+engine I left in orbit to save on acceleration when I am doing a transfer to save the fuel of the main lander.

I guess in all my time playing ksp, I always got playing like this, I never consider making small stuff to orbit the moons of Kerbin to help me.

I guess thats why I never visited Duna or other planets since I play KSP...

  Again, thank you all for the great tips and ideas, I go back to ksp now and going to change my strategy.

36 minutes ago, numerobis said:

The amount of mass you need depends on engines you use, plus how much fuel you need to bring. How much fuel you need to bring depends on Isp. How much thrust you need depends on mass. How many engines you need depends on how much thrust you need. How much mass depends on the number of engines. Oh snap, we're back to the start!

That one just made me a huge LMFAO !

p.s.

Hope I could vote best answer to all of you. You deserve it.

Edited March 6, 2016 by Kar

[Snark]

41 minutes ago, numerobis said:

Contra @Snark, I tend to build with TWR about 1.5 when I'm about to touch down, not 3. I prefer lower TWR because it's more tolerant to error: I don't accidentally overshoot as much. It's also cheaper, due to reduced engine mass -- example, a single LV-909 can lift about 20t of payload and fuel on Mun; a single LV-N can lift about 22.5t (plus itself). A trick for landing particularly heavy loads but still get good Isp is to pack two engine types: an efficient engine that doesn't have enough thrust on its own, and a lighter but less efficient engine to give a boost on the final stages of landing and the start of liftoff.

Yah, TWR is one of those things where it's really a matter of taste.

On vacuum worlds, I prefer high TWR because it's more efficient (less gravity loss).  Also, for cases where I need to do a pinpoint landing in a particular spot (e.g. land my fuel hauler right next to my refinery), higher TWR makes it a bit easier to plot my approach, since the point of ground intercept moves less when I'm retro-thrusting (the suicide burn is a lot shorter).

The notable exception is with LV-N's, since their Isp is so high that it offsets gravity losses, as long as the TWR doesn't go too low.  I've also used the trick you mention about packing some small, light, high-thrust engines for that last bit right before landing, or that first bit right after takeoff.  A pair of LV-909's will nicely complement an LV-N that way, for example-- just an extra ton of mass for the engines, and you can briefly triple your thrust.  I'm a big fan of Porkjet's Atomic Age mod, and the LANTERN engine there is great for this sort of thing-- a switchable-mode engine that can toggle between high-thrust, low-Isp LFO mode and low-thrust, high-Isp LF mode.

16 minutes ago, Kar said:

I do not transmit science for example, I can't. I use remote tech so I always have to consider that.

I don't follow... why can't you transmit science?  RemoteTech is fine for transmitting science, you just have to have a communications link.  Or are you running some other mod that specifically disables science transmission?

[Kar]

2 hours ago, Snark said:

I don't follow... why can't you transmit science?  RemoteTech is fine for transmitting science, you just have to have a communications link.  Or are you running some other mod that specifically disables science transmission?

I can transmit science, I guess I did not explain it right. What I mean is, since it is a pain getting most of the science without losing points by transmitting (for example I will never transmit goo and materials), I need to recover every bit of science during my career until I´m comfortable with the current tech I have in a latter moment to just do not care.

EVA reports and Crew reports, yea, I do it, not the rest.

Everything is much more harder, since I have to make money, I have to minimize the losses of money due to not recover every bit of a ship, I need to grind a lot of science, but then sometimes the money is not enough when I am starting. Even getting a RT network in orbit is a pain because of the proper rockets to launch satellites. My RT network usually is up a bit before mid science grind, between the 300 and 500 node mark, that is because before I do that, I need to get better energy batteries, solar panels, and start putting probes in orbit to scan planets.

This is my mod list in my game at the moment:

The laaaaggggg oh the laaaggg

[Snark]

13 minutes ago, Kar said:

I can transmit science, I guess I did not explain it right. What I mean is, since it is a pain getting most of the science without losing points by transmitting (for example I will never transmit goo and materials), I need to recover every bit of science during my career until I´m comfortable with the current tech I have in a latter moment to just do not care.

Ah, okay.  Yes, that makes sense.

One thing to bear in mind is that you can do transmit-then-recover.  For example, let's say you're at a kind of "lull" where you're starved for science points and need to unlock another tech node or two to help you take it to the next level.  So you can launch some low-tech probes to, say, Eve.  If you're not returning, going to Eve is really simple and needs only very low-tech, and Eve is worth a bushel of science points.  So you can send some cheap, light, low-tech, one-way probes to Eve, and transmit back lots of science.  A huge percentage gets lost in transmission, sure... but Eve science is so valuable relative to what you get around Kerbin that it's worth the investment, even just to get a percentage.  And you can then use that (partial) science return to boost your tech, which you can then use to build more sophisticated ships that can come back.

For example, in my own career games:  one thing I like to do is shoot for the Mun early.  I mean, super early, before patched conics, before upgrading any buildings.  By making the probe unmanned, just something super light with a few science instruments, I can transmit back some science that gives me a huge jump up from the sort of science I get in near-Kerbin space.  And with that, I can unlock several more nodes that make it much easier for me to send a manned mission to follow, and get the real paydirt.

Edited March 7, 2016 by Snark

[Kar]

9 hours ago, Snark said:

Ah, okay.  Yes, that makes sense.

One thing to bear in mind is that you can do transmit-then-recover.  For example, let's say you're at a kind of "lull" where you're starved for science points and need to unlock another tech node or two to help you take it to the next level.  So you can launch some low-tech probes to, say, Eve.  If you're not returning, going to Eve is really simple and needs only very low-tech, and Eve is worth a bushel of science points.  So you can send some cheap, light, low-tech, one-way probes to Eve, and transmit back lots of science.  A huge percentage gets lost in transmission, sure... but Eve science is so valuable relative to what you get around Kerbin that it's worth the investment, even just to get a percentage.  And you can then use that (partial) science return to boost your tech, which you can then use to build more sophisticated ships that can come back.

Never thought on that possibility.

Well I guess you gave me an excuse to see a outer planet for the first time in my entire KSP game time lol.

Ty.

 

Oh, now I have to launch a satellite that has enough range to Eve and can transmit to  Kerbin.....

Back to the drawing board.

Edited March 7, 2016 by Kar

[Warzouz]

As many said, it's a matter of taste.

The higher the local-TWR, the more easy and quick it is to manoeuvre but you get extra weight. On atmospheric bodies, high TWR may not be very comfortable, you can end  burning your ship while ascending. On airless bodies, there is no such issue.

For my part, I prefer to have generic science landers which behaves always the same and which can land mostly anywhere. I favour having much more d-V than needed (7T lander with 3000 m/s and Terrier - variant with chutes for Duna), that way I can do various missions (biome hop, rescue kerbals, even travel to other SOI).

As I stick a space station near every planetary body I use a heavy lander (40 to 60 tons) to refuel from mostly every bodies. One trip of the refueller can refill 4 times a 3000m/s science lander.

[Snark]

2 hours ago, Kar said:

Oh, now I have to launch a satellite that has enough range to Eve and can transmit to  Kerbin.....

That's actually pretty easy-- the Communotron 88-88 has easily enough range to cover the entire inner Kerbol system, out to Duna.  Small, light, doesn't require super-high tech.

The main thing is to make sure you're comfortable with the "delayed command" functionality of RemoteTech's flight computer, so that you can handle Eve reentry without getting the antenna ripped off.  You close the antenna (thus going "dead" and having no comms) just before hitting atmosphere, but before you close it you set it up so that it'll open itself again a few minutes later after your probe is safely drifting down on its parachute. 

(Note, to be able to do this, you'll need some finagling, since there's no way to send a delayed "open" command to an antenna while it's already open.  A couple ways to do this:  one is to have the antenna's "toggle" action in an action group, and send a delayed "toggle" action.  The other is to have two antennas, and send a delayed "open" command to the closed one.)

Apologies if you already know all this stuff, just thought I'd mention it because I've noticed that a fair number of RemoteTech players are unaware of this technique, which can really make your life easier.

[RizzoTheRat]

ISP/Thrust/Weight are all linked obviously as the lower the ISP the more fuel/weight you need to carry, the heavier the engine the lower the TWR and dV for a given amount of fuel, etc.

For a given payload and TWR there is a optimum engine configuration that will give you the minimum overall weight, but cost also comes in to it.

http://forum.kerbalspaceprogram.com/index.php?/topic/130427-best-engine-for-mun-landing/&page=1

Personally I favour reusability so like to use Nukes where I can as in the long run I have to lift a lot less fuel.  I have a 2 seater science lander that was able to all science on Minmus with only 2 refuellings, and can do multiple hops on the Mun.  Not tried it further out in the solar system yet.

 

I also have a twin NERV lander than I built to be able to land base station modules (up to a orange tank size) on the Mun, I'm thinking it wouldn't take a lot of modification to make it a multi purpose lander with different payloads depending on the mission, ie science payload of a couple of scientist and all the experiments, a drilling/refining payload for a refueller, etc.

 

 

Edited March 7, 2016 by RizzoTheRat

[Kar]

37 minutes ago, Snark said:

That's actually pretty easy-- the Communotron 88-88 has easily enough range to cover the entire inner Kerbol system, out to Duna.  Small, light, doesn't require super-high tech.

The main thing is to make sure you're comfortable with the "delayed command" functionality of RemoteTech's flight computer, so that you can handle Eve reentry without getting the antenna ripped off.  You close the antenna (thus going "dead" and having no comms) just before hitting atmosphere, but before you close it you set it up so that it'll open itself again a few minutes later after your probe is safely drifting down on its parachute. 

(Note, to be able to do this, you'll need some finagling, since there's no way to send a delayed "open" command to an antenna while it's already open.  A couple ways to do this:  one is to have the antenna's "toggle" action in an action group, and send a delayed "toggle" action.  The other is to have two antennas, and send a delayed "open" command to the closed one.)

Apologies if you already know all this stuff, just thought I'd mention it because I've noticed that a fair number of RemoteTech players are unaware of this technique, which can really make your life easier.

No Apologies needed. I know a bit how to work with RT but never did anything outside of Kerbin. A few hours ago, I forgot that I did not have Protractor Mod, I installed and lol.... Eve is over 400 days interception, closest in Mohoo with 15 days. I made a probe but I mess it up since I forgot to add a Satellite on Kerbin with enough range to reach the probe out of Kerbin SOI.

We learn with the mistakes. Also, since I never made transfers to another planet, I used Mechjeb for that. It is not very accurate. Also I made something wrong when I gave the command to the probe to execute the node, it dind´t . I guess I have much more study to do.

Back to Scott Manley space school

In any case, thanks alot for the help Snark and all of you. You guys gave me many new cool ideas to work on in my career game. I just have to be more carefully planing. I tend to get a bit lost and try to do everything at the same time lol. I need to get more organized in my space program.