Measure of a rocket's "goodness"

[ShunterAlhena]

Hi guys,

I'm trying to wrap my head around ISP, delta-v, TWR etc. I think I have a pretty good grasp of these concepts, but still there are many questions I'm struggling with.

1. Do you agree that delta-v is the single most important characteristic of a rocket? If it has a high enough TWR to achieve 100% atmospheric efficiency during takeoff, then basically all that counts is its delta-v budget. A launch system with a larger delta-v will be able to carry a bigger payload farther, and that's what matters.

2. I've been reading on the forum that many think that the new ARM engines are very powerful. If it is so, then how is it possible that I get about the same delta-v budget (around 10000) for an asparagus-staged rocket with 2 rings of stages (18+1 engines altogether) made of old LV-N parts, and for an asparagus-staged rocket with the same topology built entirely of the new parts? Why isn't the new rocket "better" (higher delta-v), when it should be vastly more powerful?

Thanks for helping me out - this stuff ain't trivial.

[regex]

1. TWR and delta-V are both very important to me. I sent a lander to Moho last night and added two additional nuclear engines and some extra fuel (on top of the two it already had) because sub 0.25 TWR is terrible for interplanetary burns; way too much error introduced for my tastes. Others prefer to be more efficient bored out of their skulls with single-engine designs and that's cool.

E: Suppose I should point out here that delta-V is quite important because without enough you won't be reaching your destination. Most people, me included, would say that is the most important thing to consider.

2. The ISP of the new parts doesn't really mesh with the old parts in terms of balance. It is generally preferable, from an efficiency standpoint, to use the best parts for the job, and those parts have the most thrust as well as having excellent fuel efficiency compared to the older parts.

[vidboi]

Hi guys,

I'm trying to wrap my head around ISP, delta-v, TWR etc. I think I have a pretty good grasp of these concepts, but still there are many questions I'm struggling with.

1. Do you agree that delta-v is the single most important characteristic of a rocket? If it has a high enough TWR to achieve 100% atmospheric efficiency during takeoff, then basically all that counts is its delta-v budget. A launch system with a larger delta-v will be able to carry a bigger payload farther, and that's what matters.

2. I've been reading on the forum that many think that the new ARM engines are very powerful. If it is so, then how is it possible that I get about the same delta-v budget (around 10000) for an asparagus-staged rocket with 2 rings of stages (18+1 engines altogether) made of old LV-N parts, and for an asparagus-staged rocket with the same topology built entirely of the new parts? Why isn't the new rocket "better" (higher delta-v), when it should be vastly more powerful?

Thanks for helping me out - this stuff ain't trivial.

Delta-v is the single most important thing when you're in space.

For a lift vehicle, that's not the case, as all launchers should have a set amount of delta-v (enough to get into orbit). Other things like the rocket's TWR and aerodynamics (although not in the stock game) will determine what this delta-v requirement is, but it's roughly the same across all lift vehicles. Your real measure of a lifters efficiency is it's payload proportion - how much the payload weighs in proportion to the lift vehicle. A more efficient launcher will be able to heft a heavier payload for its weight.

On the topic of SLS parts it's only the engines that are more powerful due to both their highish ISP and very good TWR. The tanks themselves have the same wet/dry mass ratio as the older tanks, and so your rocket will have approximately the same wet/dry mass fraction for each stage and therefore similar delta-v.

Hope this helps your understanding!

[maccollo]

For a launch vehicle, two things:

Mass fraction to low orbit.

How safe it is during flight, which is why I prefer vertical staging. Radial boosters often have the tendency to smack into the core stage.

[Felsmak]

Its physical unity at the end of its mission.

[Red Iron Crown]

I wrote a little primer on comparing engines for another poster, you might find it helpful.

It is the balance between TWR and Isp on the new engines that is drawing complaints from some.

Your example in #2, comparing the LV-N asparagus and a new engine asparagus, illustrates the imbalance: The new engines are nearly as efficient as the LV-N while having a far higher TWR. Have a look at the TWR of the two rockets, you'll see what I mean.

[Jouni]

1. In addition to TWR and delta-v, maneuverability is also important. For orbital maneuvers, you want a ship that doesn't wobble, turns fast enough, allows fine control, and stops turning when you tell it so. For landings and takeoffs, you want that even more.

2. Some of the new engines don't feel reasonable, considering their roles:

• The LFB is not supposed to be a Mainsail replacement but a booster that provides a lot of thrust somewhat inefficiently at a low price. Lowering its Isp figures a bit below Mainsail's would make it more in line with the real F-1 and F-1A engines.
  
• The KR-2L is a bigger version of something between the Poodle and the Skipper. It's Isp is fine, but the thrust would be more reasonable at around 1200 kN.
  
• The KS-25x4 engine cluster is mostly fine, as it's supposed to be an efficient engine that can't lift heavy rockets on its own. A more faithful imitation of the real SLS engine cluster (according to some scaling) would have 2400 kN of thrust and 310/380 seconds of Isp. That doesn't work too well in KSP, because the scaling assumes that each of the boosters is about as heavy as the first stage of the SLS, while providing twice the amount of thrust.
  

[Whackjob]

I usually go by, "How many of the largest engine does it have?"

[sal_vager]

For me a "good" rocket is one that can do the job without much excess, no redundant parts, not much spare fuel, not even excessive batteries/panels.

I enjoy it when I almost fail due to inefficient flying, it's fun to keep within the limitations of the craft and my ability, that risk of running out of fuel or power adds an edge to the flight that you can't get with a monster rocket.

[Nemoricus]

There are many measures to judge what a 'good rocket is. In no particular order...

1. Part count

2. Mass efficiency

3. Fuel efficiency

4. TWR

5. Delta-v

6. Cost (eventually)

7. Aesthetics

Which of these is most important depends on the purpose of the rocket and one's personal preferences. Delta-v and TWR are the most important in determining where a rocket can go, but all of the others have their own value. Fuel efficiency is important for reasonable rockets, for example, since it means that they need to be refueled less often, while mass efficiency means your boosters can be smaller.

[Zourin]

1. Structural rigidity. Massive amounts of energy is wasted to wobble and badly balanced designs.

2. TWR. How easily can the assembly reach Kerbin Orbit. Are you overburdening your launch sequence with too much fuel?

3. Payload size. How much total mass reaches Kerbin orbit? Do you have enough parts to complete the mission (eg, enough fuel capacity)

4. delta-V. How far can that payload go? Does it require refueling?

5. Mission Support and Mission Independence. Is there support infrastructure available to run multiple missions without having to land back on Kerbin? Can it interface with that infrastructure? Does it enable personnel to return without a 'retrieval' mission?

If you go down this list and reach a point of failure, you need to redesign and go back to #1. You cannot discuss dV until you have addressed whether it is structurally sound, can get into orbit, and what you have to work with once you're in orbit. #5 is a point of failure if it hinges on Kerbonauts returning to Kerbin, but cannot do so because the design (or pilot) isn't capable of making the return trip, with or without orbital support.

Edited April 9, 2014 by Zourin

[samiamthelaw]

Its physical unity at the end of its mission.

Physical unity is overrated. Besides, I'm sure a part or two survived.

[Red Iron Crown]

For me a "good" rocket is one that can do the job without much excess, no redundant parts, not much spare fuel, not even excessive batteries/panels.

I enjoy it when I almost fail due to inefficient flying, it's fun to keep within the limitations of the craft and my ability, that risk of running out of fuel or power adds an edge to the flight that you can't get with a monster rocket.

Nicely said, playing just at the edge of my abilities is the best fun for me, too.

[ayana]

"Can I make the gravity turn without losing control?"

"Can I make the transfer burn in one attempt?"

If a rocket has enough thrust and holds together under it, that's the important part for me.

[Kasuha]

Rocket is good if it can get the job done and is not hard to control. That last part also means that it does not take eternity to turn it around.

I don't care if I'm taking too much fuel or RTGs, as long as the rocket behaves I'm happy with it.

[Ralathon]

The 3 most important things for a engine's goodness is:

Mass. You want it as low as possible.

Thrust. You want it as high as possible.

Isp (efficiency). You want it as high as possible.

In general engines in KSP are good at one of these stats and crap at the others. The NERVA has good ISP, Bad thrust, Bad weight. The mainsail has good thrust, bad Isp, bad weight. And the ant engine has good weight and sucks at everything else.

So you try to pick the right engine for the right job. During liftoff you don't care about efficiency because you have to fight gravity, so you want a mainsail. For interplanetary burns you only care about efficiency so you use a NERVA etc.

The reason the ARM engines (And the Rockomax 48-7S) are overpowered is because they're good in 2 or even 3 of those categories. For example, look at the 3m upper stage engine:

It only weighs 0.5t more than a mainsail. So it is pretty light for its size.

It has a thrust nearly twice as high as a mainsail. So it is powerful.

It has a vacuum Isp of 380, almost as good as the best non-NERVA engines.

If the engine is almost as good as the best in any field that combines to make it overpowered.

[cantab]

1. In addition to TWR and delta-v, maneuverability is also important. For orbital maneuvers, you want a ship that doesn't wobble, turns fast enough, allows fine control, and stops turning when you tell it so. For landings and takeoffs, you want that even more.

Absolutely agreed.

As for general rocket performance

If we're talking about interplanetary ships, delta-V is king. More dV means you can go more places. Not enough means you'll get stuck.

If we're talking about launchers and landers, enough delta-V and enough TWR are givens for the job. It either has it and can succeed, or doesn't and will fail.

For launchers in theory payload fraction - how much of the launch mass is the payload and how much is the launcher - is the main measure of performance. If you strive to make efficient rockets, that's what you should be going for.

in practice though in KSP mass on the launchpad is free. (Even in the real world fuel on the pad is cheap). What is arguably more important for playability and enjoyment of the game is payload per part. A launcher with fewer parts will let me send up a bigger, more complicated, and more satisfying ship with less lag and thus less real time taken. That's why my next launcher is going to use the ARM tanks, even though they have worse tankage than the Rockomax and FL-T ones.

[MKI]

The new rocket parts are the most efficient in almost every aspect. In sandbox mode there really isn't a reason why you shouldn't use them. Obviously they are overkill for simpler launches, and it might not be very appealing to use a large launch system to get a simple satellite up. Use what you want, generally if you can't get something into orbit with older tech, use the SLS nasa parts and its almost guaranteed

I rate my rockets or launch systems on 3 things.

1. can it get payload X into orbit? - mainly Delta-v and TWR.

2. Is it reliable and simplistic? - not overwhelming parts, doesn't clutter my orbit with upper stages and not confusing to set up on different launches. So yea i don't use asparagus, and generally keep my missions small enough where i don't need it.

3. Stable - before the update most of my launchers were designed with stability in mind. I wanted a launcher that could handle almost any payload of a certain weight. So there were a lot of structual elements that went into my launcher. Even my 4x4 engine cluster that fueled my payloads was designed to handle an engine failure. (has happened on a couple of cases!)

[TerLoki]

Versatility and reliability are the most important factors to me. If I can use the same upper stage and first stage in different combinations for different payloads to different destinations without trouble, then I say I've made a good rocket. I like a design that can mix and match parts like a Falcon 9/Falcon Heavy or Delta series rocket and handle everything from crew transports to orbit, to deploying station components to Minmus, to launching probes out toward Jool or Eeloo with little modification of the core components.

[RocketBlam]

Get the furthest with the mostest, the lightest.

How far it can go, how much it can lift, and how light it is.

At some point though, the part count becomes a drag on your computer, so part count becomes important unless you want to watch a slide show.

[Corw]

Tool that fits the task.

Sufficient delta v (always bit more for reserve, for emergencies), low part count, low stage count. Function is primary goal, looking good secondary, but mist look reasonably realistic (no pancakes). Must be controllable and reliable. It is preferable not to leave debris in orbit.

Personally I prefer not to abuse asparagus staging, it looks wrong. Love the Soyuz-like configurations, but I never go beyond single ring of side-boosters (no side-boosters on side-boosters). Really love to reuse lifers to orbit, but above that is almost always custom made.

[Streetwind]

2. I've been reading on the forum that many think that the new ARM engines are very powerful. If it is so, then how is it possible that I get about the same delta-v budget (around 10000) for an asparagus-staged rocket with 2 rings of stages (18+1 engines altogether) made of old LV-N parts, and for an asparagus-staged rocket with the same topology built entirely of the new parts? Why isn't the new rocket "better" (higher delta-v), when it should be vastly more powerful?

Wanted to address this because nobody else really has. This is a concept that goes very much into the advanced construction school, and I ran afoul of it when trying to build a really large rocket once. No matter what I tried, I couldn't get a meaningful amount of extra dV compared to much smaller rockets. Here is the thread I made on the subject, it contains many good answers and links to similar discussions or related resources.

Basically, any given single rocket stage has a practical limit on how much delta-V it can provide. I say "practical" because, while the theoretical delta-V is probably infinite, there is a point at which returns begin to diminish at a rapid pace. This results in having to add gigantic amounts of mass for any sort of meaningful dV increase, which in turn results in you running out of thrust to lift the rocket. For a ship assembled and fueled in orbit, liftoff thrust is obviously not important, but you still have to deal with the greatly diminishing returns. This starts to kick in at a certain payload fraction - in other words, the larger your current stage is in comparison to what is on top of it, the less efficient it becomes. This is independent of engine technology (but obviously better engines will achieve a higher total dV).

This means that for example SSTO rockets are a terrible idea from an efficiency standpoint. Rather than trying to maximize the dV of a single stage, it is far better to have a multi-staged rocket. In the real world, Earth orbit is traditionally achieved with two stages; there were three-stage designs that did afford additional efficiency gains, but these gains were small enough that opting for the simpler two-stage design with less points of failure is generally referred. The Saturn V on the other hand was a three-stage rocket, because anything less than three stages would have been inefficient for the dV required to perform a translunar injection. And then it also had the Apollo spacecraft itself which was technically a fourth stage.

The distilled relevance for KSP gameplay is this: your rocket is not getting that much more dV out of the new, giant NASA parts than your old rocket did because you've in all likelyhood run headfirst into that diminishing returns scenario. You probably have one stage with a really large amount of dV, and other stages which have comparatively less. For example, if you have one stage with 6000 dV, one with 2000 dV and one with 1000 dV, that's a really inefficient setup. If you instead had three stages with 3000 dV each, the whole rocket still has exactly the same total dV - but it would be a whole lot smaller and lighter. Conversely, sticking with the 6k-2k-1k example, if you wanted to increase the total dV of that rocket, it would be catastrophically inefficient to try and do it in the first stage. You would get a much, much higher benefit if you increased the second and third stages. Only small amounts of extra mass there would result in much more total dV. Of course, then your first stage may run out of TWR if it is already at the limit. In that scenario, removing 1000 dV from the first stage might allow you to add 2000 dV to the second, and another 2000 to the third stage. In my case, I was able to turn a 22,500 dV rocket into a 26,500 dV rocket with identical TWR purely by evening out the dV amounts between the individual stages - despite being limited in that regard by individual stage TWR and couldn't get it fully even.

If you now think that the highest efficiency is achieved with a huge amount of stages, that's only partially true. That works so long as your rocket mass is dominated by fuel, but in KSP, rocket parts - especially engines - have unrealistically high amounts of dry mass. If you make hundreds of tiny stages with tiny dV budgets, your rocket's total fuel/mass ratio drops off and you get less dV again because you're simply not carrying as much fuel as you could.

Ultimately it depends on what your mission profile is. Most missions in stock KSP require only moderate amounts of dV, so you hardly ever run into these construction limits, and you'll be fine with a rocket that only has 2 or 3 stages in a fairly inefficient mass distribution. But if you ever have a scenario where you need more, and somehow can't manage to get it - then you might want to even out your stages and add another one or two to the mix.

[Nemoricus]

The [new] tanks themselves have the same wet/dry mass ratio as the older tank

This part isn't true. The new tanks have a worse wet/dry mass ratio than the old tanks at 8.2. The old engines have a wet/dry mass ratio of 9.

[bonyetty]

It's gota look (Kerbal) cool (seriously all previous posters how did you overlook the most critical factor for rocket goodness) and then hopefully do what it's designed to do.

[helaeon]

The great rockets that tend to end up in my sub-assembly kit (as in I use those to launch things other than what I originally built it for) have the following characteristics generally:

They're all going to meet basic requirements of TWR and Delta-V for a given payload, and usually exceed it to make it more versatile.

I like to add :

Flies nice. No rotation and enough control authority to not have to fight it, but not so much that any tap of the stick gets it moving a lot.

Clean separations, I don't like explosions that are a mystery if they're damaging the actual rocket or just parts being jettisoned.

The top decoupler or fairing can be easily changed to fit payloads with different release requirements and there isn't stuff in the way.

I'm also not a fan of dropping engines when I can help it. I like to do drop tanks whenever possible to get more delta-v but not drop engines if I can help it. Will drop engines if transitioning from atmo-launch engine to upper stage space engine though in order to get the better ISP.