Optimal number of LV-N engines for big multipart ships

[Colmath]

Hi,

I am building an interplanetary mothership (to be exact: refining my old one from 0.19, completely rebuilding it) to serve my expeditions and colonisation at least to Duna and Eve, but possibly even farther. It is built in orbit out of 2 parts, the engine section and the command section, and assembled it should look something like this:

I was wondering what the optimal number of engines would be for this beast. I've read that due to the big weight and low TWR of the LV-N engines sometimes less is more and better. Is there a formula or calculation method to calculate how many engines are needed for a certain tonnage of spacecraft?

BTW the stats with 8 engines for this craft are as follows: 201,9 t ; 480 kN thrust ; 9807 m/s DeltaV, 2,376 m/s2 max acc ; 0,85 max TWR (according to MechJeb).

I haven't calculated how these would change with only 6 or maybe 9 engines (I'll try to do that tomorrow), but I hope you guys can help.

Thanks in advance

Damaris Industries

[stupid_chris]

Optimally, if you plan onto using that thing as a tug, you don't want to have a TWR below 0.3 on Kerbin, else your transfer burns could be so long you'd end up wasting fuel. The less engines you put, the more efficient you are, ultimately you want the lowest amount possible to have a respectable TWR. It depends on your craft, mess around with MJ in the VAB and with test payloads to see what's best.

[RadHazard]

I've made a simple calculator with Java that might help. You input the mass of your payload (without fuel), how much dV you want (You can use one of the interplanetary calculators available on the forums to see how much dV it will take to get to a particular planet), and what TWR you want (As stupid chris mentioned, 0.3 is pretty good for the interplanetary stage). The calculator is not very easy to use right now, but I've been meaning to update it once I can get a handle on my college workload.

[sephirotic]

If you already on orbit, the less, the better. Ideally you would only want a single nerva booster for the whole thing. Since you don´t have to fight gravity to leave the ground, any small push in microgravity will be more efficient (at the same ISP) if the % of mass in fuel is bigger.

Problem is, of course, by having a very massive tug in orbit with very low acceleration (less than 1m/s²) leaving orbit could be complicated. What i usually do is just put a pair of lvt-30 with their own fuel tanks dedicated to boost to escape velocity on such a tug and jettison then later when they are already in interplanetary space.

Other option, that i usually do that is efficient you can see me doing in this video:

Fragment the escape burn in several smaller burns in several different orbits.

At 6:44 of the video you can see the TWR and mass of the ship: 290 tons, 240kn of thrust and a ridiculous 0.09 TWR. Still i managed to leave kerbin's orbit eventually and with less fuel than i would have needed with more mas on engines.

The only reason i used two nerva engines per booster instead of one was by pure laziness since i knew burns would be awfully slow with just 120kn of thrust for such a massive interplanetary ship. Tough i knew by using 4 LV-Ns isntead of two, i would be considering reducing the effective d/v of each booster by more than 500m/s in total.

Edited September 17, 2013 by sephirotic

[Sirine]

For my self experience without any math, mechjeb or anything.

If your ship (already at the Kerbin orbit) 150~200tons = 8 LV-N.

If your ship 100 tons and below = 4 LV-N.

Do apply asparagus. The mass does matter even in vacuum space. The lighter your mass the faster you travel. So, drop your LV-N if you don't use them, they are damn heavy (lot of mass).

Unlike the fuel tank, their dry mass are acceptable.

[Evrion]

LV-N engines are heavy, so the fewer the better.

However, if you don't have that many, then you will have to do you transfer burn over several passes. That means, set up the node to do the full transfer, burn say 3 min. early to 3 min after the node to get 6 min of the transfer done, set up the node again for the next pass and burn more, repeat until you get the full transfer done. This is really annoying, but very fuel efficient.

Personally, I did a transfer using 10 orange tanks and 20 LV-N engines with 5x750 units of mono-prop to Jool in one pass (fuel depot). This made it to Jool with 5 full orange tanks I maybe should have split it, but I'm lazy. Should give you some idea of where to start.

[Colmath]

Thanks for the answers folks!

I've considered using asparagus for the interplanetary booster, and I will use it in case of further missions (like to the Joolian system). However, this ship is intended to be a reusable mothership that only needs refueling and adding payload (landers, cargo, probes, etc) and can do multiple round trips to at least the closest planets. I know I could use something much smaller for first flights to planets, but I like to do complex missions for exploration (a bit of realism), where I complete several objectives at once. My first Duna flight had several firsts for me all in one mission that I flew with the predecessor of the currently planned ship carrying 2 heavy landers and 2 satellites (I'll post a mission report on that one when I have the time):

- First interplanetary travel (with all its mistakes and bad piloting)

- First landing on Duna and first landing on Ike

- First satellites installed into Duna and Ike orbit

Later on I will try boosters for interplanetary stages - if I use docking ports the ship can even be refitted and thus reusable. For now I'll stick to the current design. I've played around with engine configurations, and I got the following data for this design:

6 engines: 197,3 t ; 0,19 TWR ; 360 kN thrust ; 10276 m/s deltaV ; 1,825 m/s2 max acc.

7 engines: 199,56 t ; 0,21 TWR ; 420 kN thrust ; 10035 m/s deltaV ; 2,105 m/s2 max acc.

8 engines: 201,9 t ; 0,24 TWR ; 480 kN thrust ; 9807 m/s deltaV ; 2,376 m/s2 max acc.

9 engines: 204,2 t ; 0,27 TWR ; 540 kN thrust ; 9590 m/s deltaV ; 2,645 m/s2 max acc.

So it would seem that I need some compromise: any way I'll try doing my interplanetary burns in smaller portions (thanks for the advice on that). 6 or 9 engines seem to be unreasonable, so I'll drop that.

7-8: I think I'll stick to 8, considering that I'll want to add some payload to this design as well (landers and probes and cargo), so it will become even more slower, so I need some more thrust even if it means more weight. And thanks to the design of the ship I can have some tanks attached to the ship with fuel that I can drop, so if need be I can add some deltaV to these numbers.

[RoboRay]

The right answer for the number of engines if you are seeking maximum efficiency is always "one."

However, since installing two engines doubles your thrust for only a small increase in dry mass, it's generally worthwhile except on very small ships. Beyond that, considering that higher TWRs offer no value to craft that never land, and that each additional engine produces a proportionally smaller fraction of the total thrust, it's really not worth adding more engines (and the structures to support them).

If your goal really is higher TWR rather than efficiency, you're probably better off just selecting lighter and more powerful engines to start with.

[WafflesToo]

The right answer for the number of engines if you are seeking maximum efficiency is always "one."

Pretty much took the words right out of my mouth. It really comes down to how patient you are and how far in advance you're willing to plan ahead.

Personally, most of my interplanetary has been with ion-powered probes so even the most overburdened LV-N feels like driving a sports car to me

[Alistone]

The main drawback to lower T/W is that you cannot get maximum effect from the oberth effect. Oberth basically says that the faster you are moving when you burn the more efficient you are. It requires 4 times the kinetic energy to get from 100-200mph than from 0-100 mph. But for a rocket 100 delta V cost is the same for either; so it is better to use it when you're already traveling as fast as possible.

If you cannot complete your burn in Low Kerbin Orbit then you are losing oberth efficiency opportunity. The further from Kerbin the slower your orbit and the less efficient your burn.

This is why many calculators will have an "optimal" Kerbin orbit for you to leave from. If you have an extremely low T/W ratio you may as well refuel from near a minmus orbit before setting off since your oberth efficiency will be so low you may as well start with a full tank as far outside the gravity field as possible.

Edited October 9, 2013 by Alistone

[WafflesToo]

Yes, if you do it all in one burn your losses will be tremendous. But you can minimize them by splitting your burns across multiple orbital passes.

[arq]

Even with multiple passes, once you hit Kerbin escape that is your last chance to use Oberth. For places like Duna or Eve, which lie just a sneeze past escape, that's fine. However, for the 2000m/s to the outer planets, only half of that can be performed using periapsis kicks. You will lose efficiency on the second half if your TWR is low because of lost Oberth effect.

[WafflesToo]

So the question becomes which way costs the most dV: multiple engines adding to your non-fuel mass, or lost efficiency from losing the advantage of the oberth-effect? I'm genuinely curious and I'd like to see how the math stacks up.

[Starwaster]

I've made a simple calculator with Java that might help. You input the mass of your payload (without fuel), how much dV you want (You can use one of the interplanetary calculators available on the forums to see how much dV it will take to get to a particular planet), and what TWR you want (As stupid chris mentioned, 0.3 is pretty good for the interplanetary stage). The calculator is not very easy to use right now, but I've been meaning to update it once I can get a handle on my college workload.

That link isn't working, can you double check plz?

So the question becomes which way costs the most dV: multiple engines adding to your non-fuel mass, or lost efficiency from losing the advantage of the oberth-effect? I'm genuinely curious and I'd like to see how the math stacks up.

I think you either need to do some (maybe a bunch) of test burns in orbit for comparison. One of MechJeb's functions will display delta-V loss to gravity. (it's not in any default window, I think you have to add it. It's in the Recorder category when editing windows)

Either that or find a tool to calculate gravity loss during planetary transfer burns.

Just curious, ARE those two side tank/engines boosters? Are they going to be dropped? They're sort of reminiscent of the nuclear tug option von Braun was proposing for the Mars mission in the 70s. Two tugs in identical configuration to the main rockets (I think) would boost it until just before escape velocity then they'd be jettisoned so they could be recovered in orbit and reused. (damn we were ambitious back then)

I'm considering something like that myself.