Atomic vs LV-909

[BgDestroy]

What is better LV-909 "Terrier" Liquid Fuel Engine or LV-N "Nerv" Atomic Rocket Motor ??

	LV-909	Atomic
Mass	500kg	3000kg (6time more)
Trust	60kN	60kN
ISP	345s	800s (only 2.32 time more)

so: Atomic egines is no effective becouse  for mass ! Why ?

[Kyrt Malthorn]

If you have a craft that's very heavy, the LV-N will be more fuel efficient.  But they will be long burn times, unless you have a lot of LV-N's, which often defeats the purpose.  The LV-N is a niche engine:  heavy craft that need the efficiency and can afford to make all their maneuvers at low thrust.

[BgDestroy]

1 minute ago, Kyrt Malthorn said:

If you have a craft that's very heavy, the LV-N will be more fuel efficient.  But they will be long burn times, unless you have a lot of LV-N's, which often defeats the purpose.  The LV-N is a niche engine:  heavy craft that need the efficiency and can afford to make all their maneuvers at low thrust.

How is minimal mass (dry and full), for atomic engines ?

[Luke Skycrawler]

You also have to factor in that Nukes use only liquid fuel so you can omit half of the fuel load if you use Procedural Tanks. 

[MedwedianPresident]

Generally, nukes are more efficient for large, bulky interplanetary vessels that are adapted to long and slow burns while LV-909's and comparable modded chemical engines provide good acceleration to smaller vessels, being suitable for landers or as OMS engines.

[MiniMatt]

The nuke engine masses 2.5t more. So the nuke becomes more efficient once you're hauling an extra 2.5t of fuel for your more fuel hungry terrier craft for a given delta-v requirement.

So low total mass craft, or low delta-v requirements, will often see the terrier/poodle/etc a better choice.

[stenole]

http://meithan.net/KSP/engines/

This webtool makes graphs that show which engine is better depending on payload, TWR and deltaV requirement. It can give you a good idea of when to use the one and when to use the other. Something else to take into account is also that LV909 is much shorter which is useful for landers and has gimbaling which is also especially useful for landers. In general the LV909 works better for small things and the nuke is better for bigger things.

[Mastikator]

Going from low Kerbin orbit to low Moho orbit and then back again would take over 8km/s. At that huge delta-v it's actually cheaper to use a the Nerv.

Another example is if you have an already heavy payload. If for example you're moving a research station to another planet's orbit then the Nerv becomes cheaper. Since you'll need orders of magnitude fewer fuel tanks.

Edit- if however you're just sending a tiny probe then the LV-1 Ant can give you more than enough delta-v even if you're going very far. The ant is super light so your final weight will be very light too.

Edited August 13, 2016 by Mastikator

[BgDestroy]

17 minutes ago, Luke Skycrawler said:

You also have to factor in that Nukes use only liquid fuel so you can omit half of the fuel load if you use Procedural Tanks. 

wow, this is very very good info for me  

FL-T400 = L180 + O220 = 2 250kg

• LV909 (L-1.596/s) = 112sec = 60 * 112 = 6 720kN
• Atomic (L-1.53/s) = 117sec = 60 * 117 = 7 020kN , !!! but 3 600kg more mass (engine and oxider) 

 MK1 Liquid Fuel = L400 = 2250kg

• Atomic = 261sec = 60 * 261 = 15 660kN, !!! but 2 500kg more (engine), 1100kg more fuel

 

I tested for 1 000kg sattelite and with atomic fuel is deltaV better  and with example 100kg sattelite is better LV909

[rkman]

1 hour ago, BgDestroy said:

so: Atomic egines is no effective becouse  for mass ! Why ?

They are effective depending on how much fuel and payload you bring along: 

http://forum.kerbalspaceprogram.com/index.php?/topic/41594-mass-optimal-engine-type-vs-delta-v-payload-and-min-twr/&page=2  

http://forum.kerbalspaceprogram.com/index.php?/topic/46824-atomic-motors-not-worth-it/&page=3#comment-660859  

"above 10 tons payload and 2000 m/s delta-v, the LV-N is the king"

[magnemoe]

1 hour ago, Mastikator said:

Going from low Kerbin orbit to low Moho orbit and then back again would take over 8km/s. At that huge delta-v it's actually cheaper to use a the Nerv.

Another example is if you have an already heavy payload. If for example you're moving a research station to another planet's orbit then the Nerv becomes cheaper. Since you'll need orders of magnitude fewer fuel tanks.

Edit- if however you're just sending a tiny probe then the LV-1 Ant can give you more than enough delta-v even if you're going very far. The ant is super light so your final weight will be very light too.

This, tips 1, LV-N is for heavy stuff interplanetary, 2 Its nice for going to Mun or Minmus if you reuse it. 
Its main downside is TWR, it would be perfect for an Tylo SSTO but the low TWR kills it, on the other hand I have an Moho science and  utility craft who works very well and can reach orbit and land twice as  the TWR requirements is lower, i would not use it on Minmus as I don't need so much dV, On Moho it was perfect for sampling all the biomes, shuttling crew up and down and future missions. 

[Kyrt Malthorn]

2 hours ago, stenole said:

http://meithan.net/KSP/engines/

This webtool makes graphs that show which engine is better depending on payload, TWR and deltaV requirement. It can give you a good idea of when to use the one and when to use the other. Something else to take into account is also that LV909 is much shorter which is useful for landers and has gimbaling which is also especially useful for landers. In general the LV909 works better for small things and the nuke is better for bigger things.

I had never seen this before.  I like it!

[Bill Phil]

2+ times the Isp gives you 2+ times as much deltaV for a given mass ratio. If you don't care about thrust, go LV-N.

[SchwinnTropius]

4 hours ago, rkman said:

 

"above 10 tons payload and 2000 m/s delta-v, the LV-N is the king"

Good thing I like to read the forums, I've been wondering about this for a while and was not far from asking a question that this answers.

[GoSlash27]

It depends on what you're doing.
 The LV-909 is superior for orbital insertion from Kerbin and interplanetary travel under 2 km/sec DV (which is a surprising amount of travel). The LV-N is superior for lower thrust applications and interplanetary travel above 2 km/sec *but* is still disproportionately expensive.

Best,
-Slashy

[The Optimist]

I do not believe the OP has a fundamental grasp on the relationship between specific impulse and mass.

[GoSlash27]

3 minutes ago, The Optimist said:

I do not believe the OP has a fundamental grasp on the relationship between specific impulse and mass.

Probably not... but I suspect most players are in the same boat.

Best,
-Slashy

[Red Iron Crown]

On 8/13/2016 at 9:32 AM, BgDestroy said:

What is better LV-909 "Terrier" Liquid Fuel Engine or LV-N "Nerv" Atomic Rocket Motor ??

	LV-909	Atomic
Mass	500kg	3000kg (6time more)
Trust	60kN	60kN
ISP	345s	800s (only 2.32 time more)

so: Atomic egines is no effective becouse  for mass ! Why ?

It can be more effective, once the fuel mass savings from the higher specific impulse exceed the additional dry mass. 

A simple example:

Say you have a design that uses a single LV-909 with a Rockomax X200-16 tank (9 tons full): 
Tank + engine mass = 9.5t
Burn time = 451s @ 60kN

Replace it with a single LV-N with two Mk1 Liquid Fuel Fuselages (2.25t each)
Tank + engine mass = 7.5t
Burn time = 522s @ 60kN

You can see that the LV-N version weighs less while delivering the same thrust for a longer time. The advantage gets greater and greater as more fuel per engine is added.

A simple rule of thumb for choosing between LV-909 and LV-N: If you need more than the longest 1.25 tank for each LV-909 then it is better mass-wise to replace them with nuclear engines and liquid fuel tanks.

[wumpus]

On 8/13/2016 at 2:25 PM, Bill Phil said:

2+ times the Isp gives you 2+ times as much deltaV for a given mass ratio. If you don't care about thrust, go LV-N.

While this is technically true, remember that the thing that matters is the *logarithm* of the mass ratios.  To make up the deltaV with the terrier, you will need exponentially more fuel to match the nuke.  If you don't want to be bothered with long burns/Mangallayan maneuvers, don't be afraid to add more nukes.

[GoSlash27]

A chart that illustrates the relative masses:

As we already know, the chemical engines are superior below 2km/sec DV, but what is surprising is how little the advantage is either way for DV budgets typically encountered in KSP. Total chemical stage mass doesn't reach twice that of nuclear stages until 5 km/sec, which is a heck of a lot of DV in interplanetary terms.

 In career play, the added mass of chemical rockets to Jool and beyond may be worth the relative bargain, even at launch.

Best,
-Slashy

 

[Bill Phil]

1 hour ago, wumpus said:

While this is technically true, remember that the thing that matters is the *logarithm* of the mass ratios.  To make up the deltaV with the terrier, you will need exponentially more fuel to match the nuke.  If you don't want to be bothered with long burns/Mangallayan maneuvers, don't be afraid to add more nukes.

Yeah. You have to square the mass ratio to get an equivalent delta v from the 909.

[Sharpy]

Now just add Ion.

In another discussion we came to a conclusion that for best TWR on Ion engines you should use fuel cell batteries.

  "Dawn".

Mass: 0.25t

0.485 /s

3.666 ⚡/s

----------

Fuel cell array:

Mass: 0.24t

18 ⚡/s

0.02025 f/s

0.02475 o/s

-------

18/3.666=4.9

So almost 5 engines per one array. Let's round up to 5 and consider such a block as one 'engine' for our purposes.

1.49 ton

0.225kg of lf+ox per second (at 5kg per unit of LF and Ox)

0.2425kg of xenon per second (at 0.1kg per unit)

10kN vacuum thrust

4200 ISp.for xenon alone, but if we consider almost double the mass flow due to fuel, we get ISp almost halved. Somebody calculate it better, but still over 2000s.

10kN is 1020 kilogram-force. vs 1490kg mass - 0.68 of TWR at 1.49 ton of own mass, and 0.4675 kg/s of mass flow.

--------

Meanwhile, LV-N:

60.00 kN  at 3 tons. So, about three times better TWR. But

1.53 /s at 5kg per unit, that's 7.65 kg/s - versus our 0.4675!

Anybody care to continue? How many Dawns to lift a Mammoth to Eve orbit?

 

[Red Iron Crown]

29 minutes ago, Sharpy said:

Somebody calculate it better, but still over 2000s.

It is ~1290s for fuel cell-driven ions, IIRC. Hardly worth it when you consider how much lower TWR will be as well as the poorer mass fraction of xenon tanks.

[foamyesque]

1 hour ago, Red Iron Crown said:

It is ~1290s for fuel cell-driven ions, IIRC. Hardly worth it when you consider how much lower TWR will be as well as the poorer mass fraction of xenon tanks.

 

I actually ran the numbers pretty recently; it's bang on 1200s, to the point where I suspect it was chosen intentionally.

 

On 2016-08-13 at 8:54 PM, foamyesque said:

I wonder what the Isp of a fuel-cell powered ion actually works out to. I should run the math sometime.

 

EDIT:

Mass flow of xenon: 0.0486kg/s

Flow of EC: 8.74 EC/s

Conversion ratio of fuel cells: 72 EC/kg

Mass flow of LFO: 0.121kg/s

Total mass flow: 0.17kg/s

Thrust: 2000N

Exhaust velocity: 11765m/s

Isp: 1200 seconds.

 

I'm actually not sure it's worth it. That's better'n a nuke engine, but xenon tanks have crappy mass ratios, and a nuke gives much better TWR, are cheaper once you account for the cost of fuel cells/solar panels, and can be refuelled by ISRU.

 

[Sharpy]

11 minutes ago, foamyesque said:

 

I actually ran the numbers pretty recently; it's bang on 1200s, to the point where I suspect it was chosen intentionally.

 

 

Did you use 10m/s for g or such?

I just tried an experiment:

380m/s of delta-V (starting in a rather high Gilly orbit, burning prograde to make gravity drag negligible)
4202 wet mass
4078 dry mass

dv = ISp/g ln(m0/m1)

Wolfram Alpha says 1294