No Suitable Second-Stage Engines

[ShermanPVK]

The Issue:

Currently, in KSP, we mostly just have launch engines and lander engines. We really lack dedicated second-stage engines for orbital insertion and transfer burns. This pretty much leaves us with the choice between putting another launch engine on the second stage and making the staging itself almost useless if not counterproductive, adding so many boosters that the launch core reaches orbit, or using one of the lander engines for orbital insertion. The latter is the most similar to how real launches work and is likely to be what new players will do. In KSP, it works well enough for low-efficiency orbital insertion at high altitude, but is generally impractical for realistic, high efficiency LKO insertions due to the low thrust. Dozens of my launch attempts have failed due to not having enough second stage thrust. I also don't think it makes much sense that we should use lander engines for orbital insertion and transfer burns, or even that they should be good for it. Lander engines are designed to fill their role (landing small craft on low-gravity bodies) by sacrificing efficiency and thrust to have low mass and be short enough for the landing legs and ladder to touch the ground.

Comparison to Real Rocketry:

Stock KSP and Realism:

I understand and support Squad's decision for stock Kerbal Space Program not to closely follow realistic physics and engineering, but this is still a rocketry game and it is important to consider how rocketry really works when making design decisions. Considering the relatively recent decision to implement re-entry heating and sensible aerodynamics, I'm assuming that there is motivation within Squad to ensure that KSP is not too dissimilar in its mechanics to reality.

Scale:

KSP is, of course, at a much smaller scale than reality, making it much less difficult in general to play. It does, however, introduce a problem. While there is a much greater margin for error on Delta-v, there is not any more time or space to get things done. High rates of acceleration are needed to get from a low sub-orbital trajectory to low orbit regardless of how much extra Delta-v the stage has. Lander engines simply don't have the thrust to propel decently heavy craft to low orbits.

Lander Engines:

We have three main lander engines in KSP: the LV-909, the "Poodle," and the 48-7S. The 48-7S is a tiny, low efficiency engine capable of powering most Mun landers, ascent stages, or combined lander-ascenders, but is not suitable for other uses in craft of significant mass. The LV-909 and "Poodle" are not only far beyond suitable for almost any Mun landing and ascent, but are also the most efficient chemical engines in KSP at an iSP of 390s despite having comparatively tiny engine bells. They are second only to the LV-N in their usefulness for high energy transfers. In reality, the LEM descent engine had an iSP of only 310, lower even than the old configuration for the Mainsail. It was required to give the lander more Delta-v than is typically used for the entire Munar descent, ascent, and return transfer. The ascent stage engine had the same low iSP and was used for a similarly demanding task. The LV-909 and the "Poodle" have much greater efficiency than is sensible for landing and ascent purposes.

Second-Stage Engines:

Real orbital launch systems generally use engines specifically designed to be used for the second stage. These engines are characterized by very high specific impulse, low-moderate thrust, and long engine bells. Notable examples include the RL-10 and the J-2, both having iSPs of over 450s. The RL-10 has a mass of just under 280 kg and a thrust of about 110 kilonewtons. The J-2 has a mass of just under 1800 kg and a thrust of over 1 meganewton. The closest things to engines of this kind in KSP are the launch engines: "Skipper" and KR-2L. Both have much higher thrust for their size and lower specific impulse than the LV-909. The use of the latter as a second stage engine is problematic as it produces thrust-to-weight ratios that often destroy payloads with g-forces. In addition, neither engine is suitable for 1.25m craft. This effectively leaves Kerbal Space Program without a major class of engine.

Suggestion:

New Engines:

I think that the addition of dedicated second stage engines would be a major improvement to stock Kerbal Space Program. Basing their performance too closely on real life analogues, however, would break the stock pattern without being better at their jobs than the existing lander engines. Realistic thrust would not be enough to solve the problem of LKO insertion caused by the game's scale. What I am proposing is a set of long, moderately heavy engines with higher thrust and somewhat higher efficiency than the existing lander engines but not enough thrust to be used as launch engines. To me, it seems like a great way of both improving gameplay and making KSP a bit more realistic.

Implementation:

I have noted potential issues with this idea. There may not be enough of a contrast between the power of the launch engines and the power required for good Kerbal second-stage engines. This may be solved by slightly increasing the power of the launch engines. I always thought that they were not quite powerful enough. Particularly, the LVT-30 & 45, the most powerful engines of the 1.25m size, often seem to have trouble achieving an acceptable thrust-to-weight ratio on relatively small rockets and often cannot lift a second stage and payload enough to achieve orbit without boosters. The 2.5m second-stage engine may be very similar in thrust to the "Skipper", which is just barely powerful enough to use as a launch engine for small 2.5m craft. Adding a 3.75m second-stage engine makes the usefulness of the KR-2L much more limited, so a rebalance of it may be a good idea. Finally, as I mentioned before, the landing engines are significantly overpowered for the job of landing. With new dedicated engines filling the role the lander engines were used for, reducing their iSPs to more reasonable levels for their job would not impede gameplay.

Public Release:

This is the subject that I am most unsure about. Personally, I won't feel that the game is complete until we have second-stage engines, so I'd like to see them in version 1.0, but I don't know if it would be best to add a whole new class of engine along with everything else that will change in version 1.0. It may be better to know what the community thinks of the major rebalance before deciding whether or not to implement it in the retail release, so I wonder if it would be best to release it as an official mod or something like that to find out if it would work.

[Starwhip]

Do you really need second-stage engines? I generally use Mainsail-Mainsail-Poodle combos for simple craft. Or Mainsail-Skipper-Poodle.

The NASA ARM Parts are extremely overpowered: one stack can reach orbit with no staging, and fuel to spare.

The 1.25 meter LV-T45 Liquid Fuel Engine is also a decent second-stage for smaller craft, it has good thrust and vectoring. As such it is also useful as a core stage, in tandem with the LV-T30.

We will see what the new aerodynamics and realistic Isp-Thrust relationship brings to engines, hopefully it is for the better.

[Scoundrel]

A well thought out post, though I do find it amusing that I touched on most of what you are saying in another thread.

We really lack dedicated second-stage engines for orbital insertion and transfer burns...

Actually we do have second stage engines: the Skipper, the KR-2L, and (arguably) the LV-T45. They mostly have a lower mass than their lift counterparts (the Mainsail and the KS 25x4) and a lower thrust, but marginally higher Isp. Also the post Starwhip made pretty much sums up what I was going to say anyways.

Lander Engines...

Okay, firstly there is an assumption that smaller bells = lower Isp. That's not quite how it works. What you are thinking of is over-expansion/under-expansion and its relation to the area of the nozzle exit, and how all that affects Isp. The thing is that we don't know if these engines use small plain ol' bell nozzles or if they use something like expansion-deflection nozzles, or dual expander nozzles, or dual throat nozzles, or plug nozzles, or <insert one of the other myriad advanced nozzle/throat systems>, which would explain why they have high vacuum Isp and small nozzles. Secondly, you are confusing an aesthetic choice with an actual physical working engine. I get people getting caught up in cognitive rationalizations, but the engines aren't reflective of actual engine parameters. They're made by artists who go "ooh, pretty!" and push polygons around until it looks cool, so don't get too caught up in them as engineering reference points. And thirdly, you can't compare a LFO engine to a hypergolic engine for Isp because they use two separate fuel types which have different Ve ranges.

Yeah, rocket science is tricky stuff!

Second-Stage Engines...

Um... if your rockets are shaking themselves apart due to high Gs, then you need to be throttling back or use that handy thrust limiter function. FYI, the Skipper and the KR-2L are in fact second stage engines... for the bigger rockets. Unfortunately nobody can really see that since the actual problem is that Isp is treated as being linear and there is no Pe attribute for them from which under-expansion and over-expansion can be calculated, which results in our second stage engines making for perfect launch engines. It also doesn't help that virtually all of the engines with a specific power greater than 300 are SSTO capable.

New Engines...

I too would love to see more engines!

Implementation... This may be solved by slightly increasing the power of the launch engines... <and other tweaks>

Gah! No! The engines in this game are already overpowered for a universe that is essentially 1/10 scale, and one of the reasons they are so overpowered is to ensure a high margin for error so new players don't get turned off right away. As I've stated above, all of the engines with a specific power >300 are SSTO capable, and cranking up the power will just make your suggestion of second stage engines moot.

Public Release... Personally, I won't feel that the game is complete until we have second-stage engines...

Heh, careful. People might take a statement like that as a claim of entitlement and nail you with it.

But you are quite correct. We need to wait and see what 1.0 holds (and yes, the engines are being rebalanced!) before we can come to any conclusions.

[Sean Mirrsen]

The 909 and the Poodle are the quintessential pusher stage engines in my book. Lightweight and efficient. Second stage, identified as "the thing that brings you to apoapsis", is usually a Skipper, or a mix of 30s and 45s, though a cluster of 909s could work just as well with a good enough first stage.

I echo the above statement that we'll need to see what changes with the aero update, but I see no need for a dedicated "second stage" engine series as of yet. I expect most would consider it to be needless increase of part count - as you've already said, there is very little difference between lifter and second stage engines in regards to KSP in particular. It's just too narrow a performance band.

[Tw1]

I think we have a fairly good set of engines, but then I have no qualms about clustering them, and tend to avoid stack staging.

[NoMrBond]

It will be interesting to see what the mass flow ISP correction and subsequent balance pass by Ted do to all the stats

Lots of changes coming

[cicatrix]

I had no idea I needed a special 2nd stage engine. There are plenty, even in stock. You just need to pick one that has acceptable characteristics. I know some people are against mods, but I've never had such a problem. I also use KW Rocketry with a wide range of engines suitable for everything.

[Superfluous J]

Just use the 48-7S for most everything, and the LV-N for everything else.

...until the balance comes, of course, and then (hopefully) we'll actually have choices for which engines to use where.

[ShermanPVK]

I get what all of you are saying, especially about the "Skipper.," I have also used the LVT-45 as a second-stage engine for heavy 1.25m payloads, but I still think that engines specifically designed to be balanced for second-stage propulsion would be a very good addition. Really, the 1.25m rockets are the ones that I most feel need this kind of engine. When I mentioned increasing the power of launch engines, I was really just talking about the LVT-30/45 which really struggle to lift another LVT-45 powered stage without boosters. Certainly not the Mainsail or the 3.75m engines. I am quite content with the parts we have for heavy payloads.

About my comments on lander engines, I did not intend to say that iSP was a direct result of the size of the rocket bell, merely that the rocket bell size was something that lander and second stage engines do not have in common. I do also understand that the "Poodle" and 909 use a different kind of fuel, but what I was really saying was that the engines are extremely and unnecessarily overpowered as landing engines, but don't really have the thrust for insertion to LKO with payloads heavier than capsules and there's really quite a jump between the "Poodle" and "Skipper".

To clarify, I think it's weird that one of the lowest iSP classes of engine in the real Apollo program has the highest iSP available for chemical rockets in KSP (even though I understand that they use different resources). I think that while the parts we have now are most certainly adequate at least, there is an area where two or three new engines based on a kind of real engine would fit quite nicely and not having them has been my biggest pet-peeve about designing launch vehicles in KSP for a long time.

[NoMrBond]

Just use the 48-7S for most everything

We should start playing the death knell for that one right now

[Jouni]

I do also understand that the "Poodle" and 909 use a different kind of fuel, but what I was really saying was that the engines are extremely and unnecessarily overpowered as landing engines, but don't really have the thrust for insertion to LKO with payloads heavier than capsules and there's really quite a jump between the "Poodle" and "Skipper".

The Skipper may be 3x more powerful than the Poodle, but the gap between the two is only 1.5x: 3 tonnes vs. 2 tonnes. You can always use an engine that's more powerful than necessary, as long as it's not too heavy.

There is a real gap between the LV-909 (0.5 tonnes) and the LV-T45 (1.5 tonnes) / the Poodle (2 tonnes). An intermediate 1.25 m engine could fit in there.

[Sean Mirrsen]

The Skipper may be 3x more powerful than the Poodle, but the gap between the two is only 1.5x: 3 tonnes vs. 2 tonnes. You can always use an engine that's more powerful than necessary, as long as it's not too heavy.

There is a real gap between the LV-909 (0.5 tonnes) and the LV-T45 (1.5 tonnes) / the Poodle (2 tonnes). An intermediate 1.25 m engine could fit in there.

The 30 is 1.25t. Yes I know it's more a lifter engine. You could also cluster two 909s.

[Jouni]

The 30 is 1.25t. Yes I know it's more a lifter engine. You could also cluster two 909s.

Neither of those is a single engine with thrust vectoring. There are always workarounds, but they usually don't let you build what you want.

[Sean Mirrsen]

I think the same thing happens IRL often enough.

I mean, how many different rocket engines are in active service nowadays anyway?

[Jouni]

I mean, how many different rocket engines are in active service nowadays anyway?

Around 30, according to the Wikipedia. A bit more, if you consider different variants of the same engine (e.g. launch/vacuum versions of Merlin 1D). A bit less, if you discount ion thrusters and similar stuff.

[rkman]

I agree with the OP's points, but i think a bigger (related) problem is that there is generally not enough choice of engines - especially because clustering of engines is barely supported in stock KSP. There's only one radially mounted engine that can be used for clustering on ascent stages, and it will be outside of any engine fairing.

The issue is that the amount of thrust needed for various stages varies greatly depending on the payload mass - and the available engines simply don't cover that range of thrust, not without some trickery anyway.

An engine that can serve in a 2nd- or upper stage on one rocket, will be greatly under- or overpowered (and thereby be useless) for that same task on another rocket. The cleanest way of solving that in stock is to build the stage as a bundle of parallel stacks so that the stage has enough thrust, but that comes at the cost of making that stage (or the entire rocket) overly wide/flat/pancake.

[Wabbit]

48-7s for tiny size

LV-909 for 1.25m

Poodle for 2.5m

And you can stack them

3.75m parts are just massive, but I would use the KL-R2 as basically my only choice

But I use KW Rocketry as a good solution, because it adds more balanced engines (and fairings, yay!) that are dedicated to a specific rule. Also adds different capacity fuel tanks and battery packs.

[Jouni]

The issue is that the amount of thrust needed for various stages varies greatly depending on the payload mass - and the available engines simply don't cover that range of thrust, not without some trickery anyway.

An engine that can serve in a 2nd- or upper stage on one rocket, will be greatly under- or overpowered (and thereby be useless) for that same task on another rocket.

This only becomes a problem, if you always try to build the same rocket in different sizes. The problem disappears, if you build different rockets for payloads of different sizes.

Look at the Delta IV, for example. The upper stage always uses the same engine, even though there is 3x variation in payload capacity, depending on which upper stage and which boosters (if any) are used. You can easily do something similar in KSP.

Start with a Poodle/X200-16 upper stage and a Skipper/Jumbo-64 lower stage, and you have a rocket that can lift 8 tonnes to orbit. Add two BACC boosters and the payload capacity increases to 10 tonnes. Switch to bigger SRBs and throttle the core down when they're attached, and the payload capacity goes up to 15 tonnes. Use two additional lower stages as boosters, and throttle the core down a little less, and the payload capacity becomes at least 20 tonnes.

[sumghai]

I personally think the current range of second / upper stage engines is adequate - I'm particularly partial to the Poodle and the LV-909.

[TMS]

It'd be useful if the new fairings had some way of supporting engine clusters.

We can hope.

[panzer1b]

Im all for more balanced engines, which HOPEFULLY will be the case in 1.0. Right now, you more or less pick the engine that has teh thrust you need to oush whatever you need to push at a reasonable rate.

Also, the KR-2L is overpowered in too many ways right now. Im pretty sure most of us know this. that engine needs to have at a maximum 2000 thrust (right now better isp then the quad, mass almost as low as a mainsail, and almost twice the thrust of a mainsail). The 48-7s is overpowered, but i think we all know that super low mass, very good thrust (for that mass), and respectable ISP is a bad combination right now.

Anyways, until 1.0 comes out, im not even gonna bother complaining about unbalanced engines. Right now im sure most of the community is aware that there are very few engines that make sense from an efficiency perspective, and many engines are made rather redundant by them, aesthetics and or part count aside.

[tsotha]

I don't understand how you can talk about "second stage" engines in a vacuum (heh) like this. What constitutes a reasonable size for a second stage engine depends entirely on your payload and mission profile. I've used everything from the 48-7S to the KR-2L for a second stage and thought to myself "Self, this engine is the perfect size for my second stage".

Same goes with landers. The only thing that makes the 48-7S, LV-909 and Poodle "lander" engines is their low profile. But you're making a whole lot of assumptions about what the lander looks like and its intended use. I've used at least a half dozen other engines on my landers as a result of this or that constraint. And I can't count high enough to number all the LV-N lander designs I've used myself or seen on forum/reddit threads.

Now, if what you're really getting at here is the fact that there are thrust/weight gaps in the engine lineup, then I agree. I'd like to see something between 50 and 220 besides the aerospike, which you don't get until the very end of the tech tree. Preferably something that produces power, unlike it's smaller bretheren. But that's really the only gap I see. Here the mk 55 doesn't count because, paraphrasing the old Lay's tag line, you can't use just one.

This pretty much leaves us with the choice between putting another launch engine on the second stage and making the staging itself almost useless if not counterproductive, adding so many boosters that the launch core reaches orbit, or using one of the lander engines for orbital insertion.

You've left out the third and most common option here, which is to use multiple smaller engines. You can use radial engines, engines on radially attached tanks, or with girders and COS. I mentioned I think there's a gap between 50 and 220 kN, but I can get 150 kN with three 909s or five 48-7Ss (for 1/3 the weight. Sigh).