Efficient Launch Stage

[Rusty6899]

I decided to try to design an efficient rocket for an unmanned Mun return trip, now obviously getting into LKO is the major challenge ("half way to anywhere" and all that), but I managed to keep my rocket mass down to 3.3 tonnes for the rest of the trip (about 1.4t fuel).

I really had no idea how to design an efficient launch stage. All I knew was that Delta-V needed to be > 4550m/s and TWR needed to be > 2.

I ended up going for an asparagus staging using 5 FL-T800 Fuel tanks (normally I use 6 x symmetry for Asparagus, but I calculated that 4 would do the job) and Toroidal Aerospike Engines. I managed to get into LKO, with 20.84 litres(?) of fuel left, but my launch stage had a mass of over 30 tonnes.

Is there a better way to design a launch stage? I considered splitting the stage up into maybe a stage to ascent to 10km then another to achieve orbit, but I had no idea how much Delta-V each stage would need.

Edited November 20, 2013 by Rusty6899

[Tank Buddy]

I always use SRBs without igniting the liquid engines for smaller payloads. I don't know if it makes too big of a difference but it's a little of a boost.

[Rusty6899]

I looked into using 2 of the larger SRB's mounted radially and then a single FL-T800 with the Aerospike, but it fell short of the required Delta-V. More than 2 SRB's wouldn't have really improved things as there are diminishing returns for every additional engine you add, and it wouldn't have been saving on weight either.

[arq]

Design TWR to start around 1.7, as fuel burns it will increase. You actually want a TWR of ~2 (not much higher or lower) for the first 10km, and 1.7 usually will split that relatively well. After that, it matters less (often times my last 1000m/s to orbit happen with TWR<1), but higher is slightly better.

[Taki117]

all a higher TWR means is that any maneuvers done are done faster. Are you doing a gravity turn and compensating for terminal velocity? both will greatly help reduce the ammount of dV needed for orbit. (I average about 4.4 km/s) also, the dV map uses the vaccum dV values, if you are using the atmo values (with MechJeb or KER) you might actually make it to orbit with a sub par rocket.

[Rusty6899]

There's also the fact that I don't know how to perform a proper gravity turn. I just go up 10k and then pitch to 40-45 degrees from horizontal. I think you are supposed to just pitch very slightly and then follow the prograde marker as it moves to the horizontal, but I have no idea when to carry that out as it seems like your thrust/weight ratio, would greatly effect your final orbit.

[el_coyoto]

all a higher TWR means is that any maneuvers done are done faster. Are you doing a gravity turn and compensating for terminal velocity? both will greatly help reduce the ammount of dV needed for orbit. (I average about 4.4 km/s) also, the dV map uses the vaccum dV values, if you are using the atmo values (with MechJeb or KER) you might actually make it to orbit with a sub par rocket.

This.

You do it "manually" by starting to pitch a little (5°) at ~8 km, then at 10 km, you can start pulling the velocity marker towars the hozrizon. I do this by pointing the rocket on the "horizon" side of the velocity marker and follow it as it drops. With a TWR of ~2, this often leaves me pitched at 45° when at 20 km altitude.

A more "numbery" solution is checking KER "Atmospheric efficiency" : during my gravity turn, I make sure that it stays close to 100%. If it is significantly lower, it means that could have turned earlier and if it is more than 100%, it means that I am wasting fuel fighting drag losses because . In that case, I correct the situation by pointing the rocket up a bit to climb faster until my speed matches terminal velocity more reasonably.

You will notice that this part of the flight is rather short (10km --> ~30km) because at some point terminal velocity will increase exponentially while your ship velocity does not.

At this point, the problem is finding a balance between accelerating as close to prograde as possible (to minimize gravity losses) or continue climbing to limit drag losses but losing to gravity : this is where I'm not sure I found the optimal method to determine the best ascent path.

With a good TWR rocket, I usually burn @45° pitch until ~35km, where I usually pitch down towards prograde (pointing prograde @ ~45km) and burn until I am happy with the apoapsis. During that phase, I check KER values and make sure that I still have a reasonable vertical speed.

I must confess that I also use the shape of the orbit in map view to check this, which is a highly scientific method...

Also, if you are going the asparagus way, I'm not sure you should add another stage. It's possible to setup your rocket so that the central core does most of the circularization burn, thus avoiding the penalty of another stage and a "dead weight" engine that is not firing during ascent. This is how I understand one of the aspect of asparagus staging efficiency : as many engines as possible have to "pay" for the weight they add by firing continuously until they are discarded...

Cheers!

[panzerknoef]

You need aspergus staging for sure

[capi3101]

Asparagus is your best bet for an efficient launch - and Temstar's got formulas for that here. You can usually get about a 15% payload fraction with asparagus.

Now, recently I've had some discussion on a different thread regarding serial staging, and one of the designs we came up with for a five tonne payload would up with a 10% payload fraction, if that's something you might be interested in (simplicity of the design and all that). Asparagus is more efficient, of course.

SSTO (a single-stage booster) is also possible, though those aren't efficient at all - they usually run around 4% payload fraction. For a 3.3 tonne payload, though, they're viable and you can't beat the part count (a consideration for folks with slow machines) - try a Mainsail, Rockomax Decoupler, two orange tanks and an X200-16. 9 struts to stabilize = 14 parts, 4550.664 m/s and 1.398 TWR. Those require you to keep a hand on the throttle.

Gravity Turn: If you don't have KER, start by turning to 45 degrees elevation on course 090 when you get to 10,000 m. Watch your ascent in the map view. When you get to the point where you are at least 35 seconds from Apoapsis, turn your ship to face your prograde marker and follow it down. If you go below 30 seconds to apoapsis, return to 45 degrees elevation and stay there until it goes back up past 35 again. Once you're a minute to apoapsis, turn to burn along the horizon and keep doing so until your apoapsis is about ten kilometers past where you want it (you'll probably be still in atmosphere at this point, the extra margin will help with losses due to drag). While you're doing this, watch your gee meter - if it climbs out of the green zone, throttle back until it's back in there. Preferably you want the gee meter needle to be right at the top the entire way up. Those are good general guidelines if you don't have mods to handle your ascent.

Edited November 20, 2013 by capi3101

[Rusty6899]

Thanks for all the tips. I wasn't really sure what constituted a good payload fraction. Mine is about 9%, which isn't that bad, and the whole rocket only takes up 50 parts (although I could probably remove 10 or so of them if necessary. I don't have any mods to help me with the ascent but I reckon that with a few attempts with this advice in mind I'll get better.

I'm guessing that it might be easier to get a higher payload fraction with larger rockets. Temstar's formulas look pretty interesting, I hadn't really looked at using differing engines for Asparagus staging (which should give me some added fun calculating Isp for each stage).

[capi3101]

I'm guessing that it might be easier to get a higher payload fraction with larger rockets. Temstar's formulas look pretty interesting, I hadn't really looked at using differing engines for Asparagus staging (which should give me some added fun calculating Isp for each stage).

It's not necessarily the case that larger rocket = higher payload fraction. Check out some of the submissions in the Payload Fraction Challenge; some of those designs are pretty small. Light payload = 48-7S as an asparagus booster; no joke.

LV-T30 and LV-T45 have the exact same Isp ratings; just thought I'd mention that...

[Rusty6899]

In general, though, if I am going to use different engines, then I can expect to have to calculate Isp's.

Seems that I was wrong about the payload fraction being lower for smaller rockets, there doesn't seem to be much difference between the categories (aside from the heaviest, but only 3 people are on the board). I can't really get close to any of the top values there for any size of rocket.

[capi3101]

In general, though, if I am going to use different engines, then I can expect to have to calculate Isp's.

True - and the easiest way to do that is with the formula found here:

For example, say you've got a Skipper with 4 LV-T30s. Skipper outputs 650 and has an atmo Isp of 300; ratio of output to Isp is 650/300 = 2.167. LV-T30 outputs 215 and has an atmo Isp of 320; it's ratio is 215/320 = .671875. The total thrust for a Skipper and 4 LV-T30s is (4*215)+650 = 1510 kN, and you'd divide that by (4*.671875)+2.167 = 4.854167. 1510 / 4.854167 = 311.073, which is the Isp you would use in that case (you could round it down to 311 if you'd like).

Been a while since I looked at that challenge - looks like I got knocked off the leaderboard and didn't realize it...well, that sucks.