Poodle vs Lv30/45

[architeuthis]

By my calculations even with a fairly small rocket (consisting of a single Rockomax X200-16 propellant tank), you get superior delta-v with a Poodle than with a LV-T45. The LV-T45 is usually better than the LV-T30 because the thrust vectoring gives you a lot of steering torque, and the mass difference is less than the mass of a single reaction wheel. The exception is for a first stages using fins for control, in that case you will have a better TWR using LV-T30s. For in-space applications though the Poodle's thrust is overkill, so the LV-909 is usually a better deal. The LV-909s don't give as much steering torque, and they don't have an alternator, but those aren't really very big issues IMO.

EDIT: My calculation was wrong actually the break even point for a bare bones rocket is roughly 57 tonnes. Which is about what UmbraRaptor suggested in an earlier post.

Edited November 24, 2013 by architeuthis

[mmeinild]

You're welcome

Note how the poodle isn't even included with the charts. The 909 outperforms it on every single aspect with the only downsides being that they are not 2.5m and part count.

The LV-909 is hardly there either.

For those who wants the quick run down of the charts:

Vacuum

For stages with target TWR ratio below 1 use multiples of the LV-N (atomic engine) or multiples of the 48-7S if you wish to lower the weight of the stage. For very heavy payloads use only 48-7S's

For stages with target TWR ratio above 1 use multiples of the 48-7S

Atmospheric

Use multiples of the 48-7S, unless your target TWR is less than 1.8 - in that case you can use aerospikes if you have a very heavy payload with high DeltaV requirements (IE heavy stuff with lots of fuel in the stage).

TL;DR: Use multiples of 48-7S and attach the engines in symmetry by using the tiny "Cubic Octagonal Strut" as adapter. It will attach pretty much anywhere on most parts.

[Dispatcher]

Off topic: for rockets, my preferences include the Aerospike, LV909 and 487S for general work and the LVN for interplanetary travel.

Back on topic: the Poodle's TWR is 9, while the LVT30 TWR is 17.5 and the LVT45 has 13.6 TWR. So I tend not to use the Poodle, except under wide body landers on low Gee places; for stability and assembly purposes.

[manskal]

Back on topic: the Poodle's TWR is 9, while the LVT30 TWR is 17.5 and the LVT45 has 13.6 TWR. So I tend not to use the Poodle, except under wide body landers on low Gee places; for stability and assembly purposes.

Quick question: does the superior TWR of the LVT30 make it more efficient thant the LVT45? Or is it the other way 'round?

[phunk]

Quick question: does the superior TWR of the LVT30 make it more efficient thant the LVT45? Or is it the other way 'round?

TWR alone isn't enough to judge efficiency, but in this case, the 30 is more efficient, as it has the same ISP as the 45 but has higher thrust and lower mass. The only advantage the 45 has is thrust vectoring.

[Space Viking]

I think the OP could have elaborated the subject more. It's like apple and oranges – the Poodle is never being used as a first stage booster in place of an LV-T30/45. Respectively, you'd be better off with a Poodle in space rather than using an LV-T30/45 as the main engine.

[phunk]

Respectively, you'd be better off with a Poodle in space rather than using an LV-T30/45 as the main engine.

On small ships, the difference in mass can make the poodle worse than the lv-t30/45.

[numerobis]

You need to be doing a long burn for the Poodle to be better than the T30 on fuel efficiency grounds; and by then, you'd have been even better off with a cluster of LV-N.

Let's compare two setups: the Poodle with m tonnes of fuel, versus the T30 with (m+1) tonnes of fuel. The two setups start with almost the same mass (Poodle is 2.5t, whereas T30 + FL-T200 is 2.375t). I ask the question: how big must m be before it imparts the same momentum?

The answer: you must burn at least 18.5t of fuel with the Poodle, or else you'd have gotten more momentum from the T30. In other words, if your total burn time with the Poodle is less than 6 minutes, use a T30. Conversely, if burn time with the T30 is more than 6 minutes, consider a more efficient engine (the OP limited it to Poodle vs T30, ignoring the many other choices).

Momentum is thrust multiplied by time. Time is the amount of propellant divided by the mass flow out of the rocket. Mass flow is thrust divided by exhaust velocity.

p = F t
t = m / mdot
mdot = F / (Isp g0)
thus p = m Isp g0

Let's set the momentum equal for the Poodle with m tonnes of fuel, and the T30 with (m+1) tonnes of fuel:

p = m 390 g0 = (m + 1) 370 g0
m = 370 / (390 - 370) = 18.5

Momentum isn't exactly what we want; we only care about accelerating the payload. But then the math gets much more complicated, and you don't end up with an analytical solution.

Doing that more complicated math via a python script (you need KSP-scripts for this to work):

import engine

def evaluate(Isp, mpay, meng, mprop):
    m1 = mpay + meng + 9 * mprop / 8.0
    m0 = mpay + meng + mprop / 8.0
    return engine.deltaV(m1, m0, Isp)

def compare(mpay, mprop0, engines):
    (meng0, Isp0) = engines[0]
    print ("%g: %g\n" % (Isp0, evaluate(Isp0, mpay, meng0, mprop0)))
    for (mengI, IspI) in engines[1:]:
        diff = meng0 - mengI
        mpropI = mprop0 + diff * 8 / 9.0
        print ("%g: %g\n" % (IspI, evaluate(IspI, mpay, mengI, mpropI)))

compare(100, 18.5, (
            (2.5, 390), # poodle
            (1.25, 370), # T30
            (0.8, 350), # cluster of 8 48-7S
            (2.25 * 4, 800))) # cluster of 4 LV-N

Shows that for all payloads, if you burn 18.5t of fuel with the Poodle and you compensate the fuel when you switch to a different engine, the T30 is better for all payload sizes. The 48-7S is better for very small payloads. The LV-N is better for all but the smallest payload.

[Space Viking]

On small ships, the difference in mass can make the poodle worse than the lv-t30/45.

Yeah, if the ship weights around 5 tons, then it's possible to save up to 25% of Delta-V by running on an LV-T30/45 instead of a Poodle. However, when a typical spacecraft starts to exceed the weight of 15 tons, then the weight saving advantage even of the LV-T30 is starting to fade out. Although, it might still be more value because that engine still only costs about half the price.

Of what I understand is despite the Poodle's weight, it is still actually relatively cheap given its rather decent thrust of 220 kN along with its wide vectoring, and also having a high ISP of 390. I think the main issue is most things has been taken for granted because all the parts has been provided for free in KSP so far. For example, an LV-909 supposedly costs 750 and a Poodle 1600, so a system of multiple LV-909 might not even be viable the day an actual budget is implemented, since a Poodle is basically five LV-909 with 30 kN less thrust (which is practically meaningless in space anyway), but with wider thrust vectoring and the ability of generating electricity.

[Edit]

Oh, I just saw your post after I posted mine, numerobis. I was actually looking for information like that, so thanks for the enlightenment!

Edited November 27, 2013 by Space Viking

[numerobis]

I should have linked to tavert's post about this question: which engine should you use for a given payload mass, required deltaV, and required TWR.

The Poodle isn't in that table because the LV-909 is strictly better in the ways relevant for the table: for the same mass, five LV-909 provide more thrust and the same Isp. But the LV-909 almost never wins either.

The conclusion is that you can't win a case for using the Poodle on efficiency grounds. You can quite reasonably argue that it's prettier or that it is fewer parts or whatever.

Edited November 27, 2013 by numerobis

[Superfluous J]

Very astute analysis Numerobis, I'd give you a star but the forum tells me I'm too much of a fan already

You actually made me feel a bit better about using the Poodle on my distant probes. I didn't burn them for 6 minutes but the inefficiency was less than I'd feared. Especially because I'd opted for flybys instead of trying for orbits. Also, I hadn't yet unloced LV-Ns so they were not an option at the time.

[WafflesToo]

In most cases, you really should use the 48-7S; it even outperforms the Mainsail in almost every single way... I more-or-less have been trying to not use it because it feels like a cheat and WHEN it gets nerfed I don't wanna feel too reliant on its godlike performance.

LV-30 sees a ton of love from me for a variety of uses from launchers to spaceplanes (90% of the benefit of an aerospike doesn't even get used because you're on jet engines until you're well over 20km). I don't really use the LV-45 much outside of a second-stage engine on my launch vehicles where I get the best use from its vectoring thrust. I use the LV-909 for almost all of my orbital maneuvering and lander needs beyond that (even sent a 40t mission package to Duna on one once).

I do wish there were an engine that was in-between the two though... just because sometimes you need about 100kN of thrust but only have space for one engine.

Ohhh, the Poodle? Never use it.

Edited November 27, 2013 by WafflesToo
actually answered the OPs question.

[Pixel of Life]

Javascript is disabled. View full album

This might be useful. Not made by me. Includes charts for vacuum and Kerbin atmosphere.

First open the chart (vacuum/atmospheric, whichever you need) with your desired TWR (indicated by the text at the top), go along the Y axis to your payload mass, then along the X axis to however much delta-V you need. The colour indicates which engine is the best for the job.

[ammonia_ocean]

I actually never thought of clustered smaller engines and am going to go try that out immediately. When it comes to low-grav vacuum landers, I usually use one LV-909 or a 48-7S (if I'm doing really low-grav stuff), but using more of them should greatly increase my versatility. Larger landers, I usually use LV-T30s/45s or just two or three LV-Ns.

I rarely use the Mainsail except for the center engine on large asparagus designs, and tend to use clustered LV-T30s instead in most of my lower stages (admittedly, this is mostly because I think it looks so cool). For circularization burns and some transfers/insertions, I go with the Skipper (which actually looks pretty mediocre, now that I look at the stats. Maybe I'll switch to another LV-T30 cluster). I think I've used the Poodle maybe two or three times in my KSP career, and never in 0.22.

[Wallace]

Poodle may be less efficient than the T45 and a whole bunch of other smaller engines, but if you put any of those diddy engines in a 2.5m stack, then it will look ridiculous, and it will snap in two. You can mount some radially, but then you have struts and tanks all over, I personally prefer a nice straight up stack without a load of radial crap stuck all over it, it's more structurally sound and a lower parts count, so the launcher may be less efficient, but it is much more reliable. Now once I get out of the atmosphere, I'd consider the efficency a bit a more closely, but when it comes to launching, I've had too many launchers that snap in two, lag like a donkey or otherwise crap out every 3rd launch, I'd rather just slap a big ol poodle in the middle and a tank on top for my final stage, no fannying about with clusters of dinky engines.

[thereaverofdarkness]

I use the poodle any time I want a large sized LV-909. I don't mind that its TWR is a bit less than the LV-909, that mostly only matters on takeoff and I don't use those on takeoff. It matters a bit on landing but I find the poodle is plenty powerful enough to land large landers on the Mün.

[Psycix]

What I do instead, is run an I beam or girder down from the pod and radially attach 4 1.25m fuel tanks and rockomaxes (or LV-909's) to that.

This way I can have an engine cluster inside a 2.5m stack for both better performance - and looking pretty!

You can achieve the same level of decouple-ability, while using medium/tiny decouplers instead of the big ugly heavy large ones.

This is my standard approach to Poodle-workaround, greatly enhancing performance due to losing a few tons of weight.

An implementation of it is featured on my Kerbal Xi rocket:

Javascript is disabled. View full album

Edited December 1, 2013 by Psycix

[Aqua667]

I actually kinda like Poodles. They a lot shorter than the LV series rockets, so I can fit them under my landers or my space planes. (MY first SSTO used a Poodle as a orbital insertion stage.) Besides, it can BBQ your snacks, your silly LVs can do that! LVs are nice for launching stuff though.