Are cluster engines effective in 1.0.4?

[maceemiller]

Hi all. So, im trying to get enough fuel together for a trip to Moho and after watching a video about cluster engines it appeared that this was the way to go. In a test that was conducted 1 engine (I forget the name but its stock and has 1500 thrust) was strapped to a single orange tank and launched straight up and the apoapsis height was recorded, lets say 2 million meters as an example.

The same test was done but this time with 3 LV engines in a cluster formation and the apoapsis was way over 4 million meters. Great I thought, but when I replicated the test .I found the single engine reached a far higher apoapsis than the cluster by over 2 million meters? Why is this? Is it because the test in the video was on an older version of KSP and not 1.0.4?

Also, should I just use cluster engines in space as apposed to using them on launch?

Im just having a hard time getting the required dv to Moho so trying different things.

Any help would be much appreciated

Edited August 12, 2015 by maceemiller

[Slam_Jones]

Well, each additional engine adds a lot of extra weight, effectively lowering the dv. You really only need to cluster them when you don't have enough TWR at that particular stage. If the TWR is already 1.0+, then you shouldn't need any additional engines, as they will end up (basically) as dead weight. Let's say an engine weighs 4 tons. if it's able to lift the payload by itself, then it's fine. If you now add two extra engines, that's basically 8 extra tons that aren't helping it do anything it couldn't before.

I recommend installing a mod like Kerbal Engineer or MechJeb that will give you a dv/TWR readout. Very helpful

[Taki117]

You don't need as much dV as you think to get to Moho, check out this thread for more information

[Starwhip]

If I read this correctly, you're saying that in an old version, the Mainsail engine was less efficient than three of the LV series, but in the newest version, the Mainsail was more efficient? Did you repeat the test with the cluster engines again?

If you did, I can say that the Mainsail is a very good atmospheric engine, better than the LV series.

The problem as a whole is a very complicated one, and highly dependent on your scenario. If you've got a big and heavy ship, loaded with fuel, and you don't mind waiting a while, a single LV-N will do the trick. If you want to get somewhere faster, the KR-2L is a great vacuum engine, though it is big and heavy. A small craft is sometimes more efficient with a smaller, albeit less efficient, engine, simply because of the mass difference.

[mhoram]

Temstar wrote a nice guide about cluster configuration here:

http://forum.kerbalspaceprogram.com/threads/28248?p=346702&viewfull=1#post346702

Here are also two threads that help with finding good engine configurations:

http://forum.kerbalspaceprogram.com/threads/65011

http://forum.kerbalspaceprogram.com/threads/127691

[maceemiller]

If I read this correctly, you're saying that in an old version, the Mainsail engine was less efficient than three of the LV series, but in the newest version, the Mainsail was more efficient? Did you repeat the test with the cluster engines again?

If you did, I can say that the Mainsail is a very good atmospheric engine, better than the LV series.

The problem as a whole is a very complicated one, and highly dependent on your scenario. If you've got a big and heavy ship, loaded with fuel, and you don't mind waiting a while, a single LV-N will do the trick. If you want to get somewhere faster, the KR-2L is a great vacuum engine, though it is big and heavy. A small craft is sometimes more efficient with a smaller, albeit less efficient, engine, simply because of the mass difference.

Yes, you read it correctly and yes I did repeat the test with the same results. It may be the new atmosphere, etc in 1.0.4 thats made the differece from the old video test. Thankyou for your reply though as its given me things to think about

- - - Updated - - -

You don't need as much dV as you think to get to Moho, check out this thread for more information

Thankyou, very informative

[cantab]

From some quick experiments with the engine charts app it looks like engine clusters are now rarely better than fewer more powerful engines. The most notable exception I found is Aerospike clusters sometimes doing better than KR-2Ls because you can't have half a KR-2L. Even then KR-2L(s) plus aerospikes is probably better than aerospikes alone.

[Randazzo]

If you're going for efficiency then you'll want to find the best single engine that suits the task at hand.

Clusters are purely for lifting power.

[pincushionman]

The real question I think is: how are you attaching the clusters? Are they all mounted to the same tank somehow, or their own tanks? Part attachment to anything but a node is assumed to be exposed to freestream air.

[maceemiller]

The real question I think is: how are you attaching the clusters? Are they all mounted to the same tank somehow, or their own tanks? Part attachment to anything but a node is assumed to be exposed to freestream air.

I attatch them for example to a single orange tank using the small cube strut thingys

[Fearless Son]

My observation has been that clusters of smaller engines are no longer necessarily more effective than a single more powerful engine. Most of the engines performance has been tweaked to remain balanced now. If you can get the same thrust-to-weight ratios and ISP from using one big engine instead of several smaller ones, then there is no reason to opt for the cluster other that aesthetic considerations.

[pincushionman]

I attatch them for example to a single orange tank using the small cube strut thingys

Then each strut/engine "sees" full drag at the front and the back. Plus the tank will have additional suction drag at the back from the open node. I don't know how big an effect this will be, but I expect it won't be negligible. Try again with drag disabled in the debug menu.

[GoSlash27]

"Moar boosters" <> more DV except where you are staging unless you're underpowered to begin with. DV is all about how much propellant mass you expend, what velocity your engine(s) accelerates it to, and how much mass you are accelerating.

Adding engines increases mass that is not propellant, and so reduces DV. Adding tanks to feed those engines will get you close to the original figure, but you'll lose some efficiency due to increased drag.

It's only when you stage (dropping dead weight as you go) that you will begin to see increases in DV.

Best,

-Slashy

Edited August 13, 2015 by GoSlash27

[Pecan]

...Also, should I just use cluster engines in space as apposed to using them on launch?

Im just having a hard time getting the required dv to Moho so trying different things.

Aye, there's the rub.

You want "a" stage or vehicle to go through space to Moho. Design it to go to Moho, through space.

You then want a launch vehicle that will deliver it to orbit in the first place. Design it to go to orbit, through atmosphere.

Those things are very different and require different engines. Some are great for working in a vacuum but they usually lack thrust (LV-N or Ion) so are useless for lifting all the mass of a large vehicle into orbit. The KR-2L has both vacuum efficiency and thrust, but uses huge amounts of fuel at sea level. Other engines, however, are very good at launching things but not efficient in space.

Separate stages. Separate vehicles. Separate engines. Once you've chosen your engine you might find you need more thrust than one can provide so, then, it might be worth clustering them.

Do yourself a favour too - install the KER or MJ mods so you can see deltaV and TWR stats for your vehicles. Squad won't give them to us in stock because 'some beginners might be overwhelmed by the numbers' but as soon as you start designing vehicles, instead of just throwing stuff at the sky to see what stays up, you need the figures. KER is probably the most popular but MJ can also double as an autopilot for various functions.

Edited August 12, 2015 by Pecan

[GoSlash27]

^ The Pecan is good, the Pecan is wise.

While I'm personally an advocate of learning the math and applying it instead of using KER or MJ, I also know that not everyone can realistically just jump in and do that. So if that's not your thing, then do yourself a favor and make use of these tools.

Best,

-Slashy

[Red Iron Crown]

In sandbox, I think only the non-chemical orbital engines are worth clustering (the Ion and LV-N Pecan mentioned). For lifter engines the new drag model and the nerfing of the most OP engines (48-7S and, much earlier, the aerospike) has made clusters less desirable. Choose the appropriate engine size for the stack, if it's not powerful enough then build a bigger stack.

In career, it can be a bit different as you might have a largish payload to loft into orbit but only have smaller engines available.

[Edax]

Hi all. So, im trying to get enough fuel together for a trip to Moho and after watching a video about cluster engines it appeared that this was the way to go. In a test that was conducted 1 engine (I forget the name but its stock and has 1500 thrust) was strapped to a single orange tank and launched straight up and the apoapsis height was recorded, lets say 2 million meters as an example.

The same test was done but this time with 3 LV engines in a cluster formation and the apoapsis was way over 4 million meters. Great I thought, but when I replicated the test .I found the single engine reached a far higher apoapsis than the cluster by over 2 million meters? Why is this? Is it because the test in the video was on an older version of KSP and not 1.0.4?

Also, should I just use cluster engines in space as apposed to using them on launch?

Im just having a hard time getting the required dv to Moho so trying different things.

Any help would be much appreciated

It sounds like you aren't conducting these tests with the intent to orbit Kerbin. More thrust allows you to more effectively use delta V to attain orbit, but every engine you add reduces delta v. A single reliant engine powering a 400t ship will produce a lot of delta v (if your in space), but it certainly wont lift that spacecraft off the ground. The only reason to cluster engines is that you haven't unlocked the bigger engine yet, or your using a LF- only spaceship powered by LV-N's that require more thrust then the 60 units of thrust that it provides (such as for landing or escaping sub-orbital trajectory). *fingers crossed that squad makes a jumbo LV-N*.

[maceemiller]

Well thankyou all for your replies. I have learnt a lot by reading them. As soon as I return home from my day job, imma try out KSP again. Thankyou all so much

[maceemiller]

Well, after an hours redesigning without using cluster engines and using KER I am proud to say I have built my largest rocket to date which is now happily in LKO at 115,508m.......and I have managed to take with me all required dv (and a bit of spare) to get to Moho and back. Thankyou to all who have posted.....it really has helped. As for clustering in the future, I need to experiment more......great fun

[Pecan]

Congratulations. Once you get your head around how to design it does add a lot :-)

...great fun

That's the important thing. And remember, when you need a break you always have the option of just slapping some outrageous construction together and watching what happens.

[jondugger123]

I don't really use clusters in my upper stages, just because I'm a fuel hoarder. The best method I've found, and I'm sure there's better because I'm not genius at this game, is when I strap on one 2.5m fuel tank with a nuke engine at the bottom, and add 4 1.25m tanks to the sides connected by fuel lines so they all feed into the engine. The TWR is really bad, and you don't go anywhere fast, but I've noticed I get a hell of alot of delta-V. I've gotten burned escape velocity, Duna interception, and a low orbit setup with a mentionable amount of fuel left when it's all said and done.

TL;DR:

Clusters = Lifting Power

1 engine and lots of fuel = delta-V

Also somebody correct me if I'm wrong or there's a better way to go about efficiency.

Fly safe.

Added note: the 4 connected tanks are on decouplers, so when they're drained they just shoot off into deep space.

Edited August 14, 2015 by jondugger123
Added note

[Sharpy]

I strap on one 2.5m fuel tank with a nuke engine at the bottom, and add 4 1.25m tanks to the sides connected by fuel lines so they all feed into the engine.

Even more delta-V and increasing TWR: Attach these 1.25m tanks (six of them in total) through decouplers (place them as if in 6x symmetry but actually use 2x symmetry three times). Connect them not directly to the engine but 1st pair -> 2nd pair -> 3rd pair -> main tank. This is "asparagus staging".

Dump any empty tanks.