Tips to optimize Delta V

[tuguley]

Here are a few things I keep in mind when building my ships, They mostly sacrifice aesthetics for efficiency, so be weary.

Please feel free to add your own tips!

1) On large 3m rockets, don't use the grey ASAS. Instead use the smaller 1m ASAS and stick it on the top of your command pod.

2) Instead of using the 3m decoupler, use the smaller, 1m decoupler and reinforce it with 2-3 struts. If you're feeling particularly adventurous, use the probe decoupler.

3) Instead of using the big SRB, use the longest 1m Fuel Tank and slap on a T-30.

4) Do not use the disk shaped battery (the one with 500 charge). It has the lowest charge/weight ratio of all the battery packs.

5) 4 winglets on the main body of your lifter stage can do wonders for craft stability and maneuverability. Try using fins instead of a bunch of gimballing engines. This is especially prominent with engine clusters when you can use T-30s instead of the heavier T-45. I've notice having 3,6,8 winglets can introduce lots of wobble. 2,4 is the sweet spot for winglets.

Put the winglets in-line with 000,090,180,270 headings.

6) Poodle Engines are almost never the answer to your problems. 3-4 LV-909s have a lower weight, identical ISP (in vac) , and almost the same thrust.

7) You never really need as much mono-propellant as you think you need. 100-150 units of monoprop will last a journey to Jool and back while including docking. (For a reasonably sized ship)

Edited May 20, 2013 by tuguley

[kahlzun]

delta-V just boils down to minimising 'dry' weight, and using the most efficient engines possible.

Asparagus staging FTW.

After you pass 1800m in altitude (on Kerbin), the Nuclear Engine exceeds 390 Isp, making it (still) the most efficient engine (apart from ion).

When doing a transfer, especially interplanetary, burn as early as reasonably practical. A small change makes a huge difference during a long trip: a single m/s deltaV perpendicular to travel in a journey of a weeks length makes a difference of over 600km from where you would otherwise be..

[zarakon]

moar boosters!

[Sabastian]

That's an interesting tip about using 3-4 LV-909's but one problem I always seem to encounter is staging. To be specific, with a Poodle, you can easily build a stage under it, but how do you do that with a cluster of 909's? Also, does clustering 3-4 909's on a 3m rocket require enabling clipping?

[tuguley]

That's an interesting tip about using 3-4 LV-909's but one problem I always seem to encounter is staging. To be specific, with a Poodle, you can easily build a stage under it, but how do you do that with a cluster of 909's? Also, does clustering 3-4 909's on a 3m rocket require enabling clipping?

Instead of clustering them, just mount then on the side with small fuel tanks and connect them with fuel lines.

[Hakuryu]

My biggest tip I found by accident - turn off gimballing for your center engine in an asparagus setup. Got a rocket that wobbles? Do this and it will be ultra-smooth, saving fuel. Only problem is when you get rid of all the asparagus stages; gotta turn gimballing back on or you might go out of control.

For stacking 3 engines on a large fuel tank, use the 3-splitter structural -> 3 engines -> 3 decouplers -> 3-splitter structural. NOT 3-splitter->3 engines->3-splitter->decoupler. You will not see 3 fairings doing it the second way (hence issues).

[AndreyATGB]

Just use asparagus staging and nuclear engines. Also replace boosters with more asparagus staging.

[superm18]

This tread should go sticky!

[Rhomphaia]

That's an interesting tip about using 3-4 LV-909's but one problem I always seem to encounter is staging. To be specific, with a Poodle, you can easily build a stage under it, but how do you do that with a cluster of 909's? Also, does clustering 3-4 909's on a 3m rocket require enabling clipping?

You can cluster them without clipping using radial attacment points or cubic octagonals struts, the results are pretty ugly though. To stage attach a modular girder segment to the tanks main attachment point and your decoupler to that.

[numerobis]

An aerospike is the same mass but better than 3 LV-909s in most ways (higher thrust, lower drag, better atmospheric Isp), except for the lack of an attachment underneath. But you'll want to ignite it first since it is mainly efficient in the atmosphere, so that's not usually a problem.

For most purposes you can replace a radial decoupler (25 kg) with a cube strut and a probe decoupler (15 kg). And struts.

Cube struts have no mass. You can replace structural elements with loads of them, at the cost of part count.

Landing struts have mass; landing gear don't.

[CalculusWarrior]

Essentially it boils down to:

• Adding fuel increases delta-v
  
• Adding engines increases TWR, neccessaary to get off the pad, but decreases delta-v
  
• Any other sort of parts (capsules, batteries, etc) decrease delta-v and TWR, but provide an reason for your rocket to be up there in the first place.
  

So you want a high enough TWR to be able to fly, but sufficient delta-v so that you can go places.

KSP Rocket design for dummies, right here

[Rickenbacker]

My biggest tip I found by accident - turn off gimballing for your center engine in an asparagus setup. Got a rocket that wobbles? Do this and it will be ultra-smooth, saving fuel. Only problem is when you get rid of all the asparagus stages; gotta turn gimballing back on or you might go out of control.

I solve that by simply turning off gimballing (with an action group) for ALL engines except the center one. One gimballing Mainsail is usually enough to keep the rocket straight, but not enough to cause much wobble.

[Wait- Was That Important?]

Essentially it boils down to:

• Adding fuel increases delta-v
  
• Adding engines increases TWR, neccessaary to get off the pad, but decreases delta-v
  
• Any other sort of parts (capsules, batteries, etc) decrease delta-v and TWR, but provide an reason for your rocket to be up there in the first place.
  

So you want a high enough TWR to be able to fly, but sufficient delta-v so that you can go places.

KSP Rocket design for dummies, right here

This is an excellent summary. Probably the best presentation of information I've seen in this thread.

[Sovek]

An aerospike is the same mass but better than 3 LV-909s in most ways (higher thrust, lower drag, better atmospheric Isp), except for the lack of an attachment underneath. But you'll want to ignite it first since it is mainly efficient in the atmosphere, so that's not usually a problem.

For most purposes you can replace a radial decoupler (25 kg) with a cube strut and a probe decoupler (15 kg). And struts.

Cube struts have no mass. You can replace structural elements with loads of them, at the cost of part count.

Landing struts have mass; landing gear don't.

Umm, an AS engine IS NOT as good as several LV-909s, due to the lack of gimble,

[stupid_chris]

Adding "MO4R B00STERZ" as I have often seen people recommending before is rarely the case. If you want to get a very large payload into orbit (100t+), your best bet is to make a "three stage" rocket

-First stage is the main engine, usually the most powerful you can get even if it only has 250 of isp ASL. You're gonna need the strength to get off the ground. You surround this by a cluster of 3-4 asparagus pairs. Here again, you want to have some high thrust, but here you want to tune the whole thing to get a TWR of about 2, and that stays about constant with each pair of boosters dropped.

-"Second stage" Is basically just dropping the asparagus boosters.

-Third stage is above the main engine/main tank. You want a high efficiency in vacuum/medium thrust engine to give you the last push into orbit. At this point (you should be around 40-50k by now) you don't need as much of a high thrust. You really just need an efficient engine. If you work with stock only, engine clusters are your friend. Staging them above a tank is harder, but it'll work charms on your payload. If you use mods as KW Rocketry or NovaPunch, they have lots of high efficiency/medium thrust engines. Just pick the one that suits your payload the best. You put the payload you want to send in orbit above this. This stage should not be a part of your payload.

Basically, you should plan the asparagus stage to get you steadily above 10k, your main engine should do most of your gravity turn and give you the most of your velocity to about 40k, your orbital engine should give you this final push into LKO. As always, recommended to turn off the gimbal on the outer engines, to add a few ailerons for better control and to strut your thing correctly.

Of course there are other ways to do it, but this can make very efficient and good looking designs. I really hate seeing people asparagus stage a countless number of mainsails with an orange tank above each. More boosters is never the best solution. Can work, but it's very inefficient.

[Tripod27]

Adding "MO4R B00STERZ" as I have often seen people recommending before is rarely the case. If you want to get a very large payload into orbit (100t+), your best bet is to make a "three stage" rocket

-First stage is the main engine, usually the most powerful you can get even if it only has 250 of isp ASL. You're gonna need the strength to get off the ground. You surround this by a cluster of 3-4 asparagus pairs. Here again, you want to have some high thrust, but here you want to tune the whole thing to get a TWR of about 2, and that stays about constant with each pair of boosters dropped.

-"Second stage" Is basically just dropping the asparagus boosters.

-Third stage is above the main engine/main tank. You want a high efficiency in vacuum/medium thrust engine to give you the last push into orbit. At this point (you should be around 40-50k by now) you don't need as much of a high thrust. You really just need an efficient engine. If you work with stock only, engine clusters are your friend. Staging them above a tank is harder, but it'll work charms on your payload. If you use mods as KW Rocketry or NovaPunch, they have lots of high efficiency/medium thrust engines. Just pick the one that suits your payload the best. You put the payload you want to send in orbit above this. This stage should not be a part of your payload.

Basically, you should plan the asparagus stage to get you steadily above 10k, your main engine should do most of your gravity turn and give you the most of your velocity to about 40k, your orbital engine should give you this final push into LKO. As always, recommended to turn off the gimbal on the outer engines, to add a few ailerons for better control and to strut your thing correctly.

Of course there are other ways to do it, but this can make very efficient and good looking designs. I really hate seeing people asparagus stage a countless number of mainsails with an orange tank above each. More boosters is never the best solution. Can work, but it's very inefficient.

Could have told me that last part earlier... Wasted weeks trying to get anything heavier than an unmanned probe/rover/ manned non-landing craft into orbit by using "all da mayne salez" and it rarely worked out

[Mr Shifty]

Of course there are other ways to do it, but this can make very efficient and good looking designs. I really hate seeing people asparagus stage a countless number of mainsails with an orange tank above each. More boosters is never the best solution. Can work, but it's very inefficient.

My workhorse is 5 Mainsails with 2 orange tanks on each, clustered in two asparagus pairs around a center tank, with locked outside gimbals and fins all around. Then above that I stack my transfer vehicle: two LV-Ns with FL-800's. Payload goes on top. Total delta-V is in the neighborhood of 11km/s. (And I end up adding a bunch of RCS and tanks to the center rocket because the whole thing is so long it won't pitch or yaw in vacuum without help.) The Mainsails get me into orbit with enough left in the center tank for another 1200-1300m/s. Would it be better if I used clustered engines instead of the Mainsails?

[Francesco]

Landing struts have mass; landing gear don't.

I've been trying to understand this, landing gear has mass inside the VAB/SPH but doesn't in the flight scene?

or is it a on the ground vs in the air thing?

[Djhcc26]

I'm no rocket scientist(outside of KSP anyway ) but i find that D-v IS what it IS. i had 14000 Dv on a simple fuel tanker that went from ground to LKO. I'm sure i could have gone all the way to Eeloo with it but it was meant for refueling a station in orbit it was a sinple 6x asparagus staging with only 8 orange tanks used (2 in the center) with mainsails it lifted 80 tons to lko with a full orange tank for refueling. soo i would say less is more in all cases. Its a delicate balance between TWR/efficiency thats gives you your Dv. If a lifter has 1.2 TWR on its first stage and its says it has 4300 Dv, chances are if you do your launch right it will get you into LKO without staging it. As long as your TWR is above 1:1, it will lift you, and the Dv is calculated accordingly. Yes the less weight you have, the less thrust you need, but you dont need anything more than a 1.2 TWR to get launched. The more fuel you carry, the more thrust you'll need to launch. The more thrust, the faster it burns, the more fuel you'll need. Putting your 1st stage TWR above 1.2 is wasteing fuel, and in turn wasting Delta-v.

[PolecatEZ]

I've been trying to understand this, landing gear has mass inside the VAB/SPH but doesn't in the flight scene?

or is it a on the ground vs in the air thing?

Yes.

Landing gear shows in VAB/SPH as having .5 tons each. Once you spawn on the runway, they are weightless or nearly so.

[stupid_chris]

My workhorse is 5 Mainsails with 2 orange tanks on each, clustered in two asparagus pairs around a center tank, with locked outside gimbals and fins all around. Then above that I stack my transfer vehicle: two LV-Ns with FL-800's. Payload goes on top. Total delta-V is in the neighborhood of 11km/s. (And I end up adding a bunch of RCS and tanks to the center rocket because the whole thing is so long it won't pitch or yaw in vacuum without help.) The Mainsails get me into orbit with enough left in the center tank for another 1200-1300m/s. Would it be better if I used clustered engines instead of the Mainsails?

On the outer mainsails, no need. However on the middle one, if you want to keep using it in outer space, I recommend indeed that you would switch to an engine cluster. If you turn part clipping on and add the tail wing parts to the side of your tank, you could cluster up to 17 engines easily on your middle stage. So if you use 5 LV-45s for control and the rest are LV-30s, you can get up to 3580kN on the core stage for the same ISP as the LV-30/45s have. Of course, you don't need to cluster so much of them, but if you cluster enough of them you will get pretty much the same TWR as with a mainsail but a much better Delta V due to a way higher ISP. Of course it's more parts and more weight but it can be worth it.

Personnally, for heavy lifting I use a design similar to this. I stick with mainsails for up to three pairs of double orange tanks in asparagus until it's not enough to get me in orbit. Then, I start to scratch off more delta V by making engine clusters. Or, if my TWR is too low, I switch to more powerful clusters.

Mainsails aren't evil, but stacking mainsail over mainsail is far from being the best thing.

[Nodrak]

Not entirely true.

7x LV-30 = 8.75 mass, 1,505 Thrust, 17.5 TWR

1x MainS = 6 mass, 1,500 Thrust, 25.5 TWR

5x LV-45 + 12x LV-30 = 22.5 mass, 3,580 Thrust, 16.22 TWR

Add on the weight of whatever tail wing piece you are mentioning and its not worth it, from a TWR standpoint.

If you have a large lifter built already and need some extra detla-v, try adding some Areo-spikes on the main rocket that fire whenever the main rocket fires. This will add to the avg ISP of the entire lift stage, and do little to add to the weight. I use as many as will fit given the symmetry of the booster stages. The reason this works is because of the high atmosphere ISP of the spikes. Some people use Jet engines like this as well, but then you need to add intakes as well. Drop the jet engines like you would any other stage, I wouldn't know the best height for it though.

Edited June 7, 2013 by Nodrak

[numerobis]

Clusters are better than mainsails on stages that burn for a long time, due to their improved Isp. A central tower that burns about 2 km/s is better off with a cluster of T30 and T45 than with a mainsail, for equivalent TWR.

That said, on most craft I build, I'm better off making a few asparagus stages of T30/45 than a big cluster of them.

[WhiteWeasel]

6) Poodle Engines are almost never the answer to your problems. 3-4 LV-909s have a lower weight, identical ISP (in vac) , and almost the same thrust.

Even better: Two aerospikes. Only for a marginal increase in fuel consumption, you get 60% more thrust 350 kn vs 220 kn. Same ISP in vac, and FAR better in ATM.