I don't understand why bigger isn't better

[Catthedragon]

Let's say you have a command pod, an orange tank, and a mainsail. That'll get you halfway to the Mun on its own, because its lightweight.

But if you tacked 3 solid fuel boosters onto that, you could have that same setup start at 8000 feet rather than on the ground.

By that logic, bigger rockets should be directly better than smaller rockets with absolutely no downside.

Why is this wrong? I know it is.

[Samniss Arandeen]

In stock sandbox, there's no stopping you building a BFR (work out what that stands for on your own).

But sometimes too big can mean inefficient compared to several smaller in sequence or at the same time. Thus the concept of "staging" to lower vehicle weights during flight. As sizes increase, the amount of force to maneuver them increases exponentially, resulting in big rockets being more of a pain to launch than the lightweights. Hell, too big and the rocket will tear itself apart if you didn't build it correctly. I've spent untold hours trying to get the strengths and weights exactly perfect putting things on top of a KW Rocketry Griffon Century. So many factors to consider including weight, drag, thrust to weight, how long each stage will last and what speed and altitude you'll be firing it at, space junk, etcetera...if you use FAR/NEAR you'll have to consider the shape and flight profile as well as the balls-out thrust and weight. And then of course you have to contend with high part counts as the rockets get massive.

"Bigger is better" depends on your specific application. Really, it's better to tailor not just the size, but the characteristics of rockets and payloads for their mission. And if you're playing Career mode, it's more cost effective too.

In my sandbox and career modes, bigger is almost never better. That doesn't stop things from being massive when they need to be, it just means I have to take the practical considerations into account.

[Wanderfound]

If your sole criteria is "how much can it lift", then yes, bigger is generally better.

But that isn't the only criteria. Cost, reliability, controllability, efficiency, utility...etc.

Rockets run into a very steep wall of diminishing returns once they get big. Yes, another booster will give you more thrust, but most of that thrust is spent on lifting the booster itself. Once you're getting really chunky, it can reach the point where another fifty tons of booster only gets you an extra 10kg of payload.

If you're looking to improve "how far can I go?" rather than "how much can I lift?", shaving a tiny amount of mass from the payload will let you dispense with a huge amount of mass in boosters.

It's inevitably a bit subjective, but if you want a "good" rocket: payload as small as you can manage, boosters as big as you need. No more, no less.

[EdFred]

There's a reason motorcycles get better mileage than semi trucks. Same thing with rockets, or people. Can a 30 stone (420lb man) run further than a 10 stone (140lb man)?

[Geschosskopf]

Everything in KSP ultimately depends on the payload you're trying to get to space because the payload is what will ultimately performs the mission. The rockets are just a means to this end, getting the payload to where it needs to be to perform its mission. The bigger the payload and the longer the trip (including return if desired), the bigger the rocket (EDIT or the MUCH greater time of flight) you need. So basically, rocket size is solely a function of mission parameters as embodied by the payload. Sometimes you need a big rocket, sometimes you don't. A small rocket won't work where you need a big rocket, so in that case, bigger isn't just better, it's required. But OTOH, if you don't need a big rocket, it's at the very least unnecessarily CPU-intensive to use a big rocket (sandbox) and might also waste money (career).

Edited September 10, 2014 by Geschosskopf

[Sirine]

I don't understand why bigger isn't better.

It actually depends on what do you refer to.

- Bigger (Stronger) Thrust. Is always better.

- Bigger (Higher) ISP. Is always better. ISP is refer to how long you can burn with the same given amount of fuel compare with other engines.

[Kasuha]

The meaning of "bigger is not better" is in the fact that fuel tank with engine gives you exactly the same performance as two fuel tanks with two engines, assuming you run them at once.

Apart of that, there's of course many ways how to build rocket bigger to make it better.

[Zylark]

Size is not a quality in and by itself - in KSP at least. Unless you're Whackjob or just want to test your PC to the breaking-point.

It is about having the right size booster for the payload you wish to get into orbit. I tend to make a selection of boosters to kick stuff into orbit and save them as sub-assemblies. That way, when done making some payload, and knowing its weight, I can select the correct booster from the sub-assemblies and just plug it in under the payload. Maybe add some strutting if the payload is some multiple-module ungainly wobbly thing - if not then just hit launch knowing I got enough oomph to get it to orbit without trouble.

Added benefits are keeping payloads reasonable, as there are weight concerns that if exceeded will cost a bit more for the booster - or if exceeding the max of my heaviest booster, would require designing a new heavier booster plus the near endless loop of "simulating" launches before getting it right.

Currently I got four booster designs, ranging from <06 Tons to <60 Tons - can be exceeded only if the payload got its own propulsion to finalize getting to orbit. So far no need to go bigger.

[Moonfrog]

Could you explain what you mean please? I read that you mentioned using solid rocket boosters to get to 8 KM, you haven't mentioned if you are using single stage rocket, multiple staged rocket, serial staging. Or bad asparagus stage method or efficient asparagus stage method.

Although I know a lot of people have said, including yourself that bigger is worse. I actually believe bigger is better in principle. The most important technology in rocket design, in both kerbal and real life technology, is how thin can you build a fuel tank before it the pressure of the fuel bursts it like a balloon. In kerbal majority of fuel tanks share the same mass ratio. This mass ratio can never be exceeded, why we use staging. Since we don't get new more efficient fuel tanks, we rely on staging techniques to get better results. In real life they build bigger and better fuel tanks with greater volume. In the real world it is actually true that bigger is better, the fuel tanks in real life are custom fit to one size as needed.

You may have noticed that after a certain point in serial staging the delta V u get for your fuel mass decreases substantially. This is to do with that fuel tank mass ratio. if you already have a mass ratio up to around 2, including the Ln segment, you could add 2000 more tanks, still get no more delta V. I generally do not build any serial stage rocket beyond a mass ratio including Ln of 1.1. That's just for efficiency. An asparagus staged rocket works differently. It actually does use the principle bigger is better, serial does not. In asparagus because of the fuel lines, u can think of the craft as being a Single stage rocket that is dismantling itself during flight, rather than serial stage which is two or more small rockets lifting eachother.

What you described is actually how I do my rockets as well, I use a serial bottom stage to lift to 8km or above, to give my liquid engines higher isp count.Then Do aspragus for those engines to push it into the atmosphere. I have lately been experimenting with doing solid booster rockets up to around 20km. Reason for this is solid booster rockets have a great mass ratio, perfect for lifting. ALso really cheap.

Although this is different to what others say, bigger is better as long as you know how to build it efficiently. Same with human body and fitness. If you think about it logically, all those 100m sprint runners who add muscle to there upper body are adding weight creating drag, they also run straight up creating more drag than running aerodynamically. Technically speaking we run like idiots. makes sense that our ancestors could run so much faster than us. Also waste energy by moving our arms while running, don't believe me try standing still moving your arms , see if u get anywhere. Useless waste of energy.

Edited September 10, 2014 by Moonfrog

[lammatt]

Because rocket equation tells you your mileage varies as a log function of mass dry / mass wet

It basically tells you there's a marginal diminishing return in action.

[Pecan]

lammatt made the logarithmic diminishing-returns point.

Essentially, a bigger rocket (with staging) will be able to do more.

A rocket that is twice as big won't do twice as much.

Two rockets, each half the size, will be able to do better.

The major problem though is "absolutely no downside" - a bigger rocket costs more and is less robust so you soon get to the point where instead of getting bigger it's better in almost every sense to launch two separate modules and assemble them in space.

[Samniss Arandeen]

a bigger rocket costs more and is less robust so you soon get to the point where instead of getting bigger it's better in almost every sense to launch two separate modules and assemble them in space.

^ This. I've built launchers for 80t payloads and even strutted like a strut spider spun its strut web all over the goddamn place those were the flimsiest pieces of $h!t I have ever flown. One wrong input and all that $$$ goes kaboom.

[O-Doc]

If you're firing the mainsail on the launchpad then you are effectively wasting more energy in drag than you gain from the solids. Try launching with just the solids and stage the mainsail on when the solids run out.

[Red Iron Crown]

I always interpreted "bigger isn't better" as referring to the payload. If you have a 10-ton payload to complete your mission, the lifter for it will be 50 tons or more. If you can ruthlessly pare your payload down to 5 tons, the lifter mass drops proportionally to about 25 tons. All that fuel and engine mass saved is funds in your pocket (not to mention the payload itself is likely cheaper, too). Same for high dV orbital craft, the lighter the payload, the less fuel and fewer engines are needed to meet the dV requirements.

[Samniss Arandeen]

I always interpreted "bigger isn't better" as referring to the payload.

Same here. But think of it this way: Each stage, and all the stages following, are collectively the payload of the stage immediately before. So even if you pare back your actual payload, not putting its insertion stage on a diet means extra weight for your middle stage to bear, ad nauseam. It's a snowball effect that escalates out of control (see: my earlier statement about 80t payload launchers).

If you can ruthlessly pare your payload down to 5 tons, the lifter mass drops proportionally to about 25 tons.

You lucky sonofagun. Launchers in my experience get exponentially (not proportionally) heavier in relation to their payloads for the same total dV.

[Sean Mirrsen]

"Bigger isn't better" is a little misleading, as it's "not better" in one specific area, cost-efficiency. Between equally well-built rockets of different size, the larger one will always be able to do "more", but the catch is that unless you specifically aim to test the limits of what's possible, "more" can end up being more than what you need - or more specifically, more than you really want having to pay for.

If you want to do a bigger thing, you will need a bigger rocket. A bigger rocket for a smaller thing is simply a waste, hence the saying.

[Red Iron Crown]

You lucky sonofagun. Launchers in my experience get exponentially (not proportionally) heavier in relation to their payloads for the same total dV.

My payload fractions stay about the same (about 1/6th for single ring asparagus-staged), assuming the size doesn't limit engine choices to something inefficient. It is increasing the total delta-V where the exponential growth appears for me.

[KerikBalm]

IRL, due to the square cubed law, I suspect the increase is not linear.

Make a bigger tank, and the thing will have to be made sturdier just to not collapse under its own weight on the launchpad.

As things get bigger, the weight of the structure will tend to occupy a greater % of the mass.

In game, I don't think this matters, due to different size tanks mostly having the same fuel ratio, and massless struts.

[Wanderfound]

80 ton payload can be done with a tolerable level of stability and part count, though:

Onion-staged, everything above the white core is payload and full of fuel, hits 70x70 shortly after the last ring of boosters fall away, with the core booster still mostly fuelled. Instant orbital fuel depot, around any planet that isn't too hard to get to.

Not cheap, though.

Edited September 10, 2014 by Wanderfound

[KerikBalm]

Wouldn't you do better using some higher ISP engines? especially in the core stage, especially after the nerf to the LFBs so that mainsails are in some way competetive with them?

[Red Iron Crown]

What is the payload fraction on that, Wanderfound?

[Wanderfound]

What is the payload fraction on that, Wanderfound?

Haven't done the math, it's just something I threw together in a couple of minutes when I needed a fuelling station in my sandbox save for the Minmus races.

Craft file at https://www.dropbox.com/s/xu7b5nt56qwfwst/Monstro.craft?dl=0

Designed for FAR, of course.

[Wanderfound]

Wouldn't you do better using some higher ISP engines? especially in the core stage, especially after the nerf to the LFBs so that mainsails are in some way competetive with them?

See the post above; it's just something I threw together in a hurry when I needed it in sandbox. I didn't actually need it any higher than LKO.

I don't really do serious rockets in my career game, apart from low-tech Muntrippers while I'm science grinding for RAPIERs.

[Red Iron Crown]

Haven't done the math, it's just something I threw together in a couple of minutes when I needed a fuelling station in my sandbox save for the Minmus races.

Craft file at https://www.dropbox.com/s/xu7b5nt56qwfwst/Monstro.craft?dl=0

Designed for FAR, of course.

Oh, FAR makes bigger payload fractions much easier due to the lower dV requirement. Still seems like a lot of engine for the payload.

[Wanderfound]

Oh, FAR makes bigger payload fractions much easier due to the lower dV requirement. Still seems like a lot of engine for the payload.

Like I said, I was in a hurry.

Cost and efficiency wasn't really a consideration; sandbox.