Experiments in Efficiency

[hittingsnooze]

Hello all!

I started playing with KSP back when the only other celestial body you could land on was the Mun. Due to work, my time playing the game has been relegated to 30 minute sessions here and there. So far I've been able to set foot on the Mun and Duna several times; and orbit Jool and three of its moons. My most successful designs have been lightweight probes attached to very heavy lifting vehicles. Because of this, I'd like to work on my design efficiency.

My most recent project has been experimenting with asparagus staging for deep space exploration. I'd attach the craft file or photos, but my account unfortunately can't do that yet.

Here are the stats on my latest craft:

Total Mass: 276.01 Tons

Surface TWR: 2.22

Delta V and time in vacuum by stage:

13: 772 m/s; 31 s - Eight RT 10 boosters, eight Rockomax boosters, and One Mainsail

11: 574 m/s; 19 s - The remaining fuel from the Rockomax boosers, the single Mainsail, and the addition of eight LV - T30 liquid engines

9: 234 m/s; 12 s - The eight boosters and first two LV - T30s fall away. Now flying with one Mainsail and 6 LV - T30s

8: 571 m/s; 29 s - One Mainsail and 4 LV - T30s

7: 715 m/s; 34 s - One Mainsail and 2 LV - T30s

6: 2325 m/s; 86 s - One Mainsail engine

4: 4156 m/s; 5:13 - One LV 909 Liquid engine

2: One LV - N atomic motor

So the questions are: What is considered a good surface thrust to weight ratio? What is a good delta V for the atmospheric stages during a launch? Is there anything glaring that sticks out in these stats? I'd welcome any constructive advice on how to create a more efficient lifter. Thank you!

Edited June 30, 2013 by hittingsnooze

[Sirrobert]

For surface TWR you need to be above 1, or you won't be able to get off the ground.

You need about 4500m/s deltaV to get into orbit.

While already in space, go for as efficient as possible. So atomic motors

Ofcourse if you want to land somewhere with highish gravity, you need bigger engines

[HoY]

I think you could lose the LV909 and just place the LVN there instead and save a stage.

Stage TWR's for optimal efficiency should be around 2.0-2.2 for the first 1500m, then you don't want to get much higher than 2.3 until your past 10,000m otherwise you'll be traveling faster than terminal velocity and will lost a lot more to drag

With less than 1.5 you will list a significant amount to gravity and end up costing you closer to 4800m/s to get to an73,000 circular orbit, where as keeping it around 2.2-2.5 you can get the same orbit with just 4300m/s

Your numbers don't look bad, except you Might have a little extra fuel on the center stack compared to what you need. As I said, if you take the LVN down a stage, lose the 909 and install the fuel tanks it used to use as drop tanks Beside the LVN stage instead of under it you'll gain another 6000m/s instead. A picture would help too, but you have to use an external hosting site as the ksp forums don't allow direct uploading

[hittingsnooze]

Here's some photos. Thanks for the responses so far! Excellent information.

http://imgur.com/a/2JsWy#0

[UmbralRaptor]

For minimum ÃŽâ€V expenditure you want a TWR of 2.2ish. However, engine mass significantly affects mass ratios. As such, you're better off with a stage starting with a lower TWR. I aim for ~1.8. Later stages (once you're mostly pitched over) can go down to 1.2-1.4. maybe even lower if you're careful.

edit: Are you planning on sticking some landing legs on the upper stages and going to the Mun/Minmus? (No, seriously. It seems doable.)

Edited June 30, 2013 by UmbralRaptor

[sgt_flyer]

mmh, i'd say drop the mainsail much sooner it's so inefficient and it drinks your fuel like mad (my Saturn V's first stage x5 mainsails burns around 4 tons of propellant per second... ) also, what is your payload weight ? (to compare to the total weight - i'd say for non asparagus designs, a good payload / total mass ratio would be around 10% - max 12% for high efficiency designs - asparagus being more in the 14%/16% range.

Edited June 30, 2013 by sgt_flyer

[UmbralRaptor]

Eh, the Mainsail's high TWR offsets much of the low Isp. It really only suffers in comparison with the LV-T30.

[hittingsnooze]

So it looks like my TWR is a little high then, especially when the LV - T30s kick in. Once I get the lifting stages tweaked, I'm planning on changing around the upper stages for different missions (landers, probes, etc.) while keeping the lifter the same.

[sgt_flyer]

yup, but the mainsail needs a huge sustainer stage - which is ultra heavy

[UmbralRaptor]

yup, but the mainsail needs a huge sustainer stage - which is ultra heavy

The large payload makes up for it. Certainly you can run LV-T30/45 clusters to 16% payload fraction, while Mainsails have trouble breaking 14%, but mind the part count.

edit: Admittedly most people fail to appreciate how much fuel a Mainsail needs to work properly. (~1.5-2+ orange tanks)

[hittingsnooze]

Just to be certain, do you compare the dry mass of the engine to the total mass of the fuel tank(s) to calculate the the payload fraction?

[UmbralRaptor]

I go with Payload_mass / Mass_on_pad_including_payload

[hittingsnooze]

Excellent! Thanks! Do you have any experiences with asparagus staging you could share, or perhaps an alternative to consider?

[SRV Ron]

Last night's experimenting with engines and fuel tanks found a sweet spot in launch speed through the atmosphere that greatly improves efficiency. Find that, along with using the ideal gravity turn to get into low orbit, and you can greatly increase the tonnage inserted into orbit. The design, weighing in at 451 tons on launch, placed 51.5 tons into Mun orbit which included the second stage with some fuel left. Once in orbit, use low powered higher efficiency engines.

A note on using the nuclear engine. While much more efficient, you will need to weight how much range you will get with its engine compared to a far lighter but less efficient Poodle or similar engine. You have to use extra fuel to get that heavy thing into orbit as well as deal with the extra dead weight of that engine when doing inter planetary flying.

Edited June 30, 2013 by SRV Ron

[K^2]

Fuel-optimal TWR for vertical ascent is exactly 2. That follows from the fact that optimal ascent rate is equal to terminal velocity^* and except for actual liftoff, acceleration is negligible until you reach thinner atmosphere. Of course, as fuel is used up, TWR at maximum thrust will increase. So you should start at either a little lower than 2, as UmbralRaptor suggests, or you have to cut the throttle a bit as you go up.

Of course, once you begin gravity turn, the turn itself will determine optimal thrust, and I haven't been able to compute the optimal solution even for the simple case of Kerbin's atmosphere. A lot of people are recommending different things, and I have been able to confirm at least a few as not true optimums, but for the most part, if you follow general advice, it's pretty close.

^* Yes, even for the case when atmospheric density decreases with altitude. At least, for exponential case used in the game. So your velocity-altitude profile is going to be v(h) = v₀ exp(h/(2H)), where v₀ is terminal velocity near Kerbin's surface, which is about 100m/s, and H is the scale height, which is 5000m for Kerbin. In other words, you should be traveling at 100m/s very soon after liftoff and reach 270m/s when 5km up. Again, once you begin the gravity turn, this formula breaks down.

[CaptainKipard]

Of course, as fuel is used up, TWR at maximum thrust will increase.

And if you drop a stage with powerful engines it'll decrease, forcing you to throttle up. I notice my atmospheric efficiency drop drastically like that sometimes.

[K^2]

And if you drop a stage with powerful engines it'll decrease, forcing you to throttle up. I notice my atmospheric efficiency drop drastically like that sometimes.

On asparagus with auxiliary engines? Yes, definitely. But it still only really matters the first 20km or so. If your upper stages have the "wrong" TWR it will be a relatively minor setback in terms of fuel, so long as your turn corresponds with TWR your rocket is capable of.

[numerobis]

I too would recommend ditching the mainsail earlier. If you're using it for the entire climb (almost) out of Kerbin's atmosphere, you're better off replacing it with a cluster of T-30s and a small number of T-45s for vectoring thrust.

Even better, rather than replacing the mainsail, build asparagus spears around your LV-N. Use the mainsail just on the first stage.

[Sirrobert]

You can probably ditch the Poodle engine, and instead put some more fuel over the atomic motor. If you are already in space, just go for highest possible efficiency.

If you want highish thrust with them, put for atomic motors ridially from 1 big central tank (the one that now fuels your poddle tank for example). Don't forget fuel lines to each of the engines/tanks supporting them

[Gargamel]

Just eyeballing your pic, it looks like (and this is just a guesstimate) that 75% of the ship is waste. The main point of most of those boosters seems to be lifting other boosters. Given the weight/size of your payload, you should be able to get away with a simple 2-3 stage single coloumn rocket.

[SRV Ron]

Just eyeballing your pic, it looks like (and this is just a guesstimate) that 75% of the ship is waste. The main point of most of those boosters seems to be lifting other boosters. Given the weight/size of your payload, you should be able to get away with a simple 2-3 stage single coloumn rocket.

Seeing the use of the solid stage boosters to lift more of them, I agree. My Onion style liquid fuel first stage gets the rocket high enough to allow second stage insertion to orbit with enough fuel to reach an orbit around Mun with some fuel left. That's 51.5 tons orbiting Mun from a launch weight of 451 tons. That's leaves a fully fueled third stage with a three Kerbal capsule and a thruster pack in Mun orbit to play with. Now, I need lander legs.

[sgt_flyer]

yup - non asparagus boosters primary utility should be to increase your TWR at liftoff to an acceptable point - until your main stage have lost enough weight by burning their fuel to have a sufficient TWR by itself.

[SRV Ron]

Played with a recommended lander from the KSP forum. It had a bit of spin issues during launch so the orbital plane was a bit off for Mun insertion. I did get it in orbit with the first two stages. Then, got it into a capture from ahead Mun insertion with sufficient fuel left in the insertion stage to bring down the lander so as little fuel is used from it as possible for landing. That fuel is needed for return to Kerbal. Had I missed the burn, the lander would have been slingshot back to Kerbal. A capture from behind would slingshot the lander into solar orbit.

Certainly, how a manuver is planned can make a great deal of difference in the amount of fuel burned.

[hittingsnooze]

Using the advice from this thread, I've shaved about 137 tons off of my total launch mass, and I'm still able to reliably get to Jool and orbit a moon of my choosing. Thanks for the help!