48-7S fan club.

[Giggleplex777]

I really wan the to 48-7S to be heavier (0.1 to 0.18 tons). This would be the trade off for its efficiency.

[magnemoe]

Its a good small engine for small payloads. For larger (>10-15t) payloads there are better engines.

I wouldn't go as far as to say that it is obsolete. There is probably a large area of overlap where the LV-909 is within 5% of optimal efficiency of the 48-7s. The problem with your graphs is that they don't really show you how close is the next runner up. Since a large vehicle will require a lot of 48-7s engines, the LV-909 will win out on part count in many cases which leads to better fps and performance for larger rockets, while only loosing a marginal amount of dV.

The situation is probably more inline with the LV-T30 and the aerospike, both of which are really close to one another in allround performance but occasionally outperform the other at a given task.

Personally I think the 48-7s just fills a gap in the stock parts for smaller in-line engines, which is a good thing.

The main factor is cargo weight, if you use the two man lander can the 909 win, one man can or lighter and the 48-7S win in most settings.

If your cargo weight is very low like 100 kg the 48-7S is pretty much unbeatable, it give better performance than even the LV-N.

[Tiron]

Hmm, I didn't know that about lower ISP engines, learn something new every day. I would be interested in seeing just how much the 48-7s wins over the LV-909, if it isn't too much trouble to plot. I suspect it won't be that big of a difference, but I could be wrong.

The problem there is he's assuming the same mass, Delta-V, and TWR. It's basically impossible for that to happen on an actual rocket, because the engines vary in mass, thrust, and TWR. Generally, a craft with a lower Isp engine is going to have to carry extra fuel to end up at the same delta-V, which is PROBABLY going to increase the weight and negate the bonus.

The thing you DO run into on practical designs is that less efficient engines sometimes weigh less than more efficient ones: The LV-N is particularly bad this way, because it weighs 2.25 tons. What can happen in this case is that the delta-V gain from the decreased weight can more than make up for the loss in efficiency.

My mapping probe's upper stage for example, consisted of a FL-T800 with the scanners, probe core, batteries, solar panels all just stuck onto it. It turns out it actually has MORE Delta-V with a 48-7S or a LV-909 than the LV-N, simply because they weigh so much less and it was such a big portion of the thing's weight.

[tavert]

The problem there is he's assuming the same mass, Delta-V, and TWR. It's basically impossible for that to happen on an actual rocket, because the engines vary in mass, thrust, and TWR. Generally, a craft with a lower Isp engine is going to have to carry extra fuel to end up at the same delta-V, which is PROBABLY going to increase the weight and negate the bonus.

You got it. I said "all other things being equal," lower Isp wins. This can happen, but only for certain combinations of payload, delta-V, and TWR. Basically along the boundaries of the different regions in my mass-optimal engine plots, where two different engines are effectively tied.

Again assuming everything else is exactly equal, you would want your TWR to increase faster for landing or takeoff, since it'll take less delta-V to do the same landing/takeoff as your TWR increases. See http://forum.kerbalspaceprogram.com/showthread.php/39812-Landing-and-Takeoff-Delta-V-vs-TWR-and-specific-impulse

[Captain Sierra]

For the record, Several of the lower end Javelin series lifters I am working on rely on these buggers cluster staged.

[lammatt]

LV-909 is still good on landers and small planes (one small liquid tank + 2 smallest 1.25 liquid+oxidizer tanks)

[regex]

I used four to bring a 16 ton rover down to Duna's surface (most of the weight was fuel that was consumed, but anyway...). My Grand Tour lander is 10 tons and uses four for an excellent combination of TWR and delta-V; it'll do everything from Moho on down.

Like others, I have noticed the tipping point where you switch over to a 909, but it is an excellent little engine. I even made a half-scale radial attachment point for it.

[astropapi1]

SSTO, 0.2 tons to orbit.

Thread: http://forum.kerbalspaceprogram.com/showthread.php/47380-Pico-SSTO-Lifter-A-K-A-The-R-48-7S-needs-a-nerf

[leax256]

that engine seriously needs to be nerfed. I say it should have the same or similar ISP to the little radial red engines

the difficulty of making something into a radial engine is bound to have things that lower the isp of the engine, that said remember that all values are really only placeholders till the final release of the game. not everything will be balanced.

[tavert]

Hmm, I didn't know that about lower ISP engines, learn something new every day. I would be interested in seeing just how much the 48-7s wins over the LV-909, if it isn't too much trouble to plot. I suspect it won't be that big of a difference, but I could be wrong.

It took more refactoring of the code than I had hoped, but here you go.

Second-best engine: http://imgur.com/a/Hz4XV

and margin of victory (difference in craft mass between best and second best engine types, divided by craft mass with best engine type): http://imgur.com/a/3D3ux

The LV-909 does make a pretty good showing for second place in a lot of cases, and around TWR of 0.3-0.4 there are some specific areas where it's only behind by a few percent. Go ahead and use it wherever you want, if you'd rather save a few parts than a few fractions of a ton.

[Captain Sierra]

Which is why we use it!

Clustering those suckers makes for some really cool looking rockets. Though I wish we had more sizes of 0.625m tanks to use as boosters.

[M5000]

Okay okay okay, for those who are saying the LV-909 is outdated and irrelevant, I'm sorry, but I have to disagree with you. There is still a niche where the LV-909 excels over all other stock engines. And that is when it is used as a main engine on OTVs based on a 2.5m platform. It offers as much thrust as the NERVA, obviously with significantly lower specific impulse, but with also significantly lower mass. Really, for an OTV that will remain in orbit of a single planet, the NERVA is far overkill. However, using the 48-7S will give about 500m/s less delta-V than the LV-909, in this situation. I was building an OTV the other day and MechJeb indicated that the 909 was undoubtedly the most effective engine for the job. In my opinion, the LV-909 is by far the best orbital engine for non-interplanetary transfers between SOI's.

Also it's my favorite stock engine. It's light, has decent thrust (great thrust for its mass), looks pretty good, and isn't a very "big" engine. It's compact and refined.

Also, also, it makes a better sound than the 48-7S. Sorry, but I'm not sitting through a rendezvous with the sound of that tiny engine.

However, I'm not saying that the 48-7S is a bad engine by any means. I'm just trying to make a point that the LV-909 is, in fact, still relevant and that the 48-7S has its own applications.

Long live the LV-909!

-M5K

[wilt57]

It took more refactoring of the code than I had hoped, but here you go.

Second-best engine: http://imgur.com/a/Hz4XV

and margin of victory (difference in craft mass between best and second best engine types, divided by craft mass with best engine type): http://imgur.com/a/3D3ux

So let me know if I'm reading these graphs correctly. In the TWR range of 1.5-1.7, after a payload reaches about 1 ton, the advantage that the 48-7S has over the LV-T30 is very small to virtually none. For payloads under 1 ton, the 48-7S becomes better than the LV-T30 by enough to where it's worth it to cluster them.

[leax256]

...has its own applications...

this describes every engine in the game, of only some people would learn this.

[Nao]

The numbers back up liking the 48-7S for many many uses: http://forum.kerbalspaceprogram.com/showthread.php/45155-Mass-optimal-engine-type-vs-delta-V-payload-and-min-TWR

The LV-909 is almost completely obsolete now. Handful of uses for it on the border between 48-7S and LV-N, that's it.

These are pretty graphs .

Did something similar after 0.21 launch, less pretty but since i've already done it, might as well post it (its all in one too)

If only our new midget engine overlord didn't have such silly sound i would have considered accepting his rule... but alas LONG LIVE THE LV-909!

[M5000]

These are pretty graphs .

Did something similar after 0.21 launch, less pretty but since i've already done it, might as well post it (its all in one too)

If only our new midget engine overlord didn't have such silly sound i would have considered accepting his rule... but alas LONG LIVE THE LV-909!

Wow, it looks by your graph that the LV-909 is only marginally better than the 48-7S in a very thin line of applications..

But I still assert that it's the best engine for 2.5m OTVs. It may be the only thing it's the best at anymore, but there are a LOT of uses for OTVs.

[Javster]

Might be but think that the red radial is underpowered,

ant isp: 290

24-77 isp : 300

48-7S isp : 350

909 isp : 390

The 20 kN engines is 13 times stronger than the ant, the 909 is 2.5 times stronger than them so they should be closer to the 909 than the ant and the 48-7S lays pretty much in the middle.

Only thing might be to increase the weight to 0.15 that would put it more evenly between the ant and the 909.

Might also be that radial engines should be weaker, look at the terrible mark 55. its very easy to just group 4-8 of an radial engine around an tank.

I don't know any red radial engine

[tavert]

Okay okay okay, for those who are saying the LV-909 is outdated and irrelevant, I'm sorry, but I have to disagree with you. There is still a niche where the LV-909 excels over all other stock engines.

It's a rather small region where the 909 is superior. http://imgur.com/a/x4dIn

For a fair comparison between LV-909 and 48-7S, add as many of each engine as necessary to satisfy your TWR requirement. Then the 48-7S should be provided with some additional fuel such that the total craft mass is equal. Only then should you compare delta-V, when total craft mass is equal and either setup satisfies your TWR requirement.

So let me know if I'm reading these graphs correctly. In the TWR range of 1.5-1.7, after a payload reaches about 1 ton, the advantage that the 48-7S has over the LV-T30 is very small to virtually none. For payloads under 1 ton, the 48-7S becomes better than the LV-T30 by enough to where it's worth it to cluster them.

The banding in the region you're looking at for TWR 1.5-1.7 where 48-7S and LV-T30 alternate is due to the TWR requirement. When you need to add another LV-T30 to satisfy the TWR requirement, you're bumping up the engine mass too much and using the 48-7S allows you to meet your TWR requirement more precisely, with less engine mass. One limitation of my charts is that they only show single-engine-type combinations. I've done some optimization experiments looking at mixing multiple engine types, and I suspect the actual optimum in these situations is to use one fewer LV-T30 than necessary to meet your TWR requirement, then a small number of 48-7S's to increase TWR by the small additional amount you need.

These are pretty graphs .

Did something similar after 0.21 launch, less pretty but since i've already done it, might as well post it (its all in one too)

If only our new midget engine overlord didn't have such silly sound i would have considered accepting his rule... but alas LONG LIVE THE LV-909!

Thanks! Yours is nice too, though the downside of your data is you're assuming engines and tanks are infinitely divisible, which is a bad assumption for small craft.

I don't know any red radial engine

Probably referring to the Rockomax 24-77 that we've had for a while. It looks orange to me, some people might think it looks red though.

Edited September 12, 2013 by tavert

[Temstar]

Let's not forget, there are plenty of landers out there that are not use once and throw away. Even with worlds as big as Duna it's simple enough to build single stage to orbit landers that can do a round trip down to the surface and then back up to orbit all in one stage. These landers are reusable and are usually left in orbit for the next mission to come along to refuel and go again. For such craft that are fuelled over and over again high Isp is a big advantage since it reduces fuel usage each time. The total saving spread out over multiple missions will soon exceed the extra engine weight.

Obviously for the big reusable landers you want to use LV-N even if it involves some engineering in working around it's unusual form factor. But for smaller landers LV-909's 40s Isp advantage will still outperform 48-7S once the vehicle dry mass + refuel weight is added together over a few trips.

[tavert]

True. I've never bothered intentionally making a reusable design for anywhere other than Kerbin, but when career mode comes around that kind of mission architecture will probably make sense.

If you're using a craft multiple times, then you don't necessarily want to optimize for total craft mass, you want to optimize for fuel use and/or payload fraction for landing + ascent for the body in question. This is tough to do in closed-form for planets with atmospheres, but for the airless bodies you can derive payload fraction from my numbers here: http://forum.kerbalspaceprogram.com/showthread.php/39812-Landing-and-Takeoff-Delta-V-vs-TWR-and-specific-impulse

Here's an example for Tylo, which shows how tough single-stage landing plus takeoff is on Tylo: http://i.imgur.com/TSKN6Ia.png

For the other bodies the peak payload fractions will of course be much higher. Like Nao's plots from earlier, this makes the assumption that engines and tanks are infinitely divisible, whereas in reality the discrete possible designs won't always get close to the continuous theoretical peaks. So absolute fuel use might be minimized with a smaller craft at a lower payload fraction than optimum...

Edited September 12, 2013 by tavert

[magnemoe]

I don't know any red radial engine

The 24-77 is an radial engine, the only other is an low performance mark 55. You can make any engine radial with struts or an small side mounted fuel tank with the engine below and a fuel line to the small tank.

Obviously for the big reusable landers you want to use LV-N even if it involves some engineering in working around it's unusual form factor. But for smaller landers LV-909's 40s Isp advantage will still outperform 48-7S once the vehicle dry mass + refuel weight is added together over a few trips.

No, if one lander has 2300 m/s dV with the 48-7S and 2000 m/s with the 909 the 909 design will use more fuel on each trip.

Same is true with the LV-N, an lander with an LV-N who has lower dV than an lander with another engine will use more fuel.

However if you increase the dV demands for the lander, say you do not enter low Duna orbit but stay in an eccentric orbit to make the return burn cheaper the LV-N will win and you will probably save fuel.

This little lander can land and take off from any body in except kerbin, Eve, Laythe and Tylo, it has landed on all of them except Moho. Has an junior docking port on the side so it can be reused, its only weakness is that the TWR then returning is so high its hard to do precise burns, accuracy is 0.4 m/s

[Nao]

True. I've never bothered intentionally making a reusable design for anywhere other than Kerbin, but when career mode comes around that kind of mission architecture will probably make sense.

If you're using a craft multiple times, then you don't necessarily want to optimize for total craft mass, you want to optimize for fuel use and/or payload fraction for landing + ascent for the body in question. This is tough to do in closed-form for planets with atmospheres, but for the airless bodies you can derive payload fraction from my numbers here: http://forum.kerbalspaceprogram.com/showthread.php/39812-Landing-and-Takeoff-Delta-V-vs-TWR-and-specific-impulse

Here's an example for Tylo, which shows how tough single-stage landing plus takeoff is on Tylo: http://i.imgur.com/TSKN6Ia.png

For the other bodies the peak payload fractions will of course be much higher. Like Nao's plots from earlier, this makes the assumption that engines and tanks are infinitely divisible, whereas in reality the discrete possible designs won't always get close to the continuous theoretical peaks. So absolute fuel use might be minimized with a smaller craft at a lower payload fraction than optimum...

Fully agree on the infinite division but the ideal charts are so... ideal in showing trends and stuff .

It kind of sucks that for many small applications that don't require high mass (using only 1 or 2 engines) 48-7S will be almost always better LV-909 because of lower mass.

And even considering same fuel load for comparison over multiple missions, one 48-7S will still have almost equal Dv to a ship with one LV-909 (assuming TWR is irrelephant).

@Tylo chart

Ohhh i've never got around to making a chart for LV-N on Tylo landers but it seems that starting TWR<1 is actually best. Bookmarking this one! So i can mess with peoples heads

[Pbhead]

Fully agree on the infinite division but the ideal charts are so... ideal in showing trends and stuff .

It kind of sucks that for many small applications that don't require high mass (using only 1 or 2 engines) 48-7S will be almost always better LV-909 because of lower mass.

And even considering same fuel load for comparison over multiple missions, one 48-7S will still have almost equal Dv to a ship with one LV-909 (assuming TWR is irrelephant).

@Tylo chart

Ohhh i've never got around to making a chart for LV-N on Tylo landers but it seems that starting TWR<1 is actually best. Bookmarking this one! So i can mess with peoples heads

Well, think about it.

That chart is taking off AND landing.

When you are doing your landing burn, you can get away with starting with TWR<1.

I think I am reading that right, right?

[magnemoe]

Well, think about it.

That chart is taking off AND landing.

When you are doing your landing burn, you can get away with starting with TWR<1.

I think I am reading that right, right?

You can, however outside of Tylo this is pretty irrelevant, other bodies has atmosphere to slow your down or an escape speed of 1400 m/s or less and you uses less than that landing from low orbit so you will not burn much fuel landing resulting in an low TWR during takeoff. You can land an 1.1 TWR lander but it require piloting skills and planing

[Nao]

@Pbhead Yep, it's pretty obvious, I just never got to do the maths myself. It's kind of funny how it's the best way too.

You can, however outside of Tylo this is pretty irrelevant, other bodies has atmosphere to slow your down or an escape speed of 1400 m/s or less and you uses less than that landing from low orbit so you will not burn much fuel landing resulting in an low TWR during takeoff. You can land an 1.1 TWR lander but it require piloting skills and planing

Actually from what i tested, optimal ascent TWR for one stage LV-N is only around 1,25 (Mun TWR). It would probably be around 1,1 starting TWR to land from low Mun orbit on close to optimal profile. So maybe both Moho and Vall qualify for the "less than 1 TWR" one stage land and ascent.

Also I find low TWR landings quite easy and fun. Hitting the right spot is a pain thou .