atmospheric isp fuel efficiency thrust changes, what?

[quasarrgames]

so i heard somewhere that in 1.0, the isp of engines will affect their thrust as well as their fuel efficiency, like real life.

So, what does this mean for me and my rockets? will they generate more thrust in the atmosphere, but less in vacuum? or the other way around? will i have to strap more engines to my rockets to get them off the ground? or will they be less efficient, meaning i need to strap more tanks to them?

im confusd

[cantab]

More thrust in vacuum.

In a real rocket engine, the fuel flow is constant. Outside air pressure gets in the way of the exhaust and makes it generate less thrust. The following equation links thrust and specific impulse:

Thrust = Specific Impulse * Fuel Flow Rate

So reduced thrust and the same fuel flow rate means a lower specific impulse.

In KSP as it stands, the engines compensate for the thrust drop by increasing fuel flow. This is technically possible in real life, but there's no reason to do it. As per the above equation, if you hold thrust constant and increase fuel flow rate, again that means a lower specific impulse.

[Cirocco]

what will happen is that thrust will vary: the higher the atmosphere, the less thrust. Isp will remain the same throughout the flight path. In real life that is because the thrust of an engine is related to the difference in pressure between the inside of the combustion/reaction chamber and the outside environment. More pressure outside = smaller difference (gradient) = less thrust. Specific impulse (which you can think of as engine efficiency) is related to engine-specific parameters and does not change if the engine itself doesn't change. (note that I think this isn't entirely true, but for simplification purposes it can be assumed to be so)

In 1.0, you will no longer be controlling the thrust with shift and ctrl, but rather you will control the fuel rate, just like in real life. So at 100% throttle, you'll be feeding the maximum amount of fuel into the engines, and they'll operate at peak power. Now, as they rise, they will go into thinner atmosphere and pick up thrust until they hit vacuum, at which point they wil reach peak thrust.

So in short: what will change is that rather than changing the amount of fuel being consumed throughout the flight (engines are really greedy at the start and become more efficient throughout the flight but always push equally hard), engines will now always consume fuel at a steady rate, but start pushing harder and harder as the flight goes on. So what you want to do is strap either more powerful engines to your rocket (to compensate lack of thrust early on) but you'll need less fuel (because you'll have a steady rate of fuel consumption, same as you would now have in vacuum). Other option includes strapping additional engines to the ship in the first stage to help overcome the lack of thrust.

In other words: the basic gist is that you'll need more thrust early on in the flight, so the solution is, in fact, to ADD MOAR BOOSTERS!

[Streetwind]

The most immediate gameplay effect will be that engines with a large Isp gap between sea level and vacuum will be a lot worse as first stage engines.

Prime example: the LV-N. Yes, it has low thrust, but some people still manage to build something that can launch on nothing but LV-N's. 60 kN per engine works out for them. The engine is kinda thirsty at 220 Isp in-atmosphere, but as you rise, Isp quickly climbs towards 800 and things are fine and dandy.

Come v1.0, that will no longer work. On the launchpad, the LV-N will still have the same fuel flow rate as it would in vacuum, and it will no longer be extra thirsty... but, it will only offer 16 kN of thrust* instead of 60 kN. Only as the rocket climbs and Isp increases, the thrust will climb back up towards full. However, with 16 kN, you're not going to move off of the pad. At all.

Instead, you're going to want an engine with very similar Isp values for your first stage, because those a.) actually deliver close to their full thrust on the pad, and b.) don't see their thrust change as much as they climb out of the atmosphere, meaning you can better plan how your acceleration and TWR profiles develop.

* Assuming the LV-N will still feature 220-800 Isp and 60 kN of peak thrust in 1.0. Can't say for sure right now, because Squad stated that they will rebalance ALL THE THINGS.

[nekogod]

Instead, you're going to want an engine with very similar Isp values for your first stage, because those a.) actually deliver close to their full thrust on the pad, and b.) don't see their thrust change as much as they climb out of the atmosphere, meaning you can better plan how your acceleration and TWR profiles develop.

Interesting so something like the aerospike would have almost identical thrust at sea level as in vacuum

[KerikBalm]

Yes, even thought thats not how aerospikes work IRL.

If it was like IRL, you could take the mainsail, KR-2L and poodle as examples.

Engine______Isp@1atm_____Isp@0atm

Mainsail______320s_________360s

KR-2L________280s________380s

Poodle_______270s_________390s

Basically, the higher your vacuum dV, the lower your atmospheric dV (assuming the same fuel, this is related to nozzle design)

An aerospike would have basically the optimal nozzle the whole flight, and would get something like 320-390 Isp.... which is way better than 270-390 Isp, and way better than 320-360 Isp

[OhioBob]

what will happen is that thrust will vary: the higher the atmosphere, the less thrust. Isp will remain the same throughout the flight path. In real life that is because the thrust of an engine is related to the difference in pressure between the inside of the combustion/reaction chamber and the outside environment. More pressure outside = smaller difference (gradient) = less thrust. Specific impulse (which you can think of as engine efficiency) is related to engine-specific parameters and does not change if the engine itself doesn't change. (note that I think this isn't entirely true, but for simplification purposes it can be assumed to be so)

In 1.0, you will no longer be controlling the thrust with shift and ctrl, but rather you will control the fuel rate, just like in real life.

That's not entirely correct, Isp is not constant. In real life the only constant is the fuel flow. As a rocket rises through the atmosphere and the ambient pressure drops, the thrust increases. Since Isp = thrust / (fuel flow * g_o), as the thrust increases, so does the Isp. The exhaust velocity doesn't change (maybe that's what you're thinking of), but the Isp does because of the changing pressure difference at the nozzle exit.

- - - Updated - - -

Interesting so something like the aerospike would have almost identical thrust at sea level as in vacuum

That's not true (aerospikes are not correctly depicted in KSP). The nozzles of engines are optimized for the ambient conditions at which must operate, i.e. low altitude (near sea level pressure) or high altitude (near vacuum). At low altitudes, nozzles have small expansion ratios, and at high altitude they have large expansion ratios. An aerospike effectively acts like a variable expansion ratio nozzle engine, obtaining a sea level performance similar to a small expansion engine, and a high altitude performance similar to a large expansion ratio engine. In a specific environment, however, aerospikes tend to perform a little worse than a fixed nozzle engine that is optimized for that specific condition. Here are some hypothetical Isp numbers to illustrate this point:

Small expansion ratio: 260 s sea level, 300 s vacuum

High expansion ratio: 200 s sea level, 330 s vacuum

Aerospike: 250 s sea level, 320 s vacuum

[Red Iron Crown]

An aerospike would have basically the optimal nozzle the whole flight, and would get something like 320-390 Isp.... which is way better than 270-390 Isp, and way better than 320-360 Isp

Funny, the 320/360 engine will lose less thrust at sea level than the 320/390 engine. This change requires some rethinking of how we evaluate engines.

[Pecan]

so i heard somewhere...

Did you also hear that apart from changing Isp/Fuel Flow:

There will be a completely new aerodynamics model (similar to, but not, FAR).

There will be a rebalance of all the engines and other parts.

And LOTS of other things.

Which means:

The specifications to which you design need to change.

The parts you build with will need to change.

How you fly will need to change.

Squad really are asking for trouble making this the release version without a single beta test.

We are all going to be back at square 1, without even the assumption that it requires ~4,500m/s to reach orbit.

[Renegrade]

We are all going to be back at square 1, without even the assumption that it requires ~4,500m/s to reach orbit.

This part is cool. I'm looking forward to it

Squad really are asking for trouble making this the release version without a single beta test.

This part is not. DO NOT WANT. :C

I'd have definitely gone with the 'release candidate' model. I'd roll 'em betas until I got one that played very well and was quite stable. I'd call that version "long term support" and then move on to 1.1... assuming I ever did a "1.0" thing.

[FlyingPete]

Funny, the 320/360 engine will lose less thrust at sea level than the 320/390 engine. This change requires some rethinking of how we evaluate engines.

It should make launcher design interesting. Currently I can launch most mid-size payloads into orbit using 3 Skipper engines as a first stage (plus a couple of SRBs), and 1 for the second stage. I think there will end up being much more defined 'First stage engines' and 'Second stage engines'

[MarvinKitFox]

Squad really are asking for trouble making this the release version without a single beta test.

No beta?

just what do you think you are playing right now?

[Tank Buddy]

It may have already been answered, but this is how rocket engines (all engines for that matter) are controlled:

The throttle controls fuel flow, which in turn controls thrust. So more throttle = more thrust.

The thing is that because of the expansion of hot gasses in a conventional engine bell, the thrust generated per unit fuel (Isp) varies at differing atmospheric pressures, or lack thereof.

So, IRL, given a constant fuel flow (controlled by throttle) an engine will have a higher thrust in a vacuum than in atmosphere.

SQUAD is now changing to this method, rather than throttle directly controlling thrust and fuel flow differing in atmosphere.

[quasarrgames]

I'd have definitely gone with the 'release candidate' model. I'd roll 'em betas until I got one that played very well and was quite stable. I'd call that version "long term support" and then move on to 1.1... assuming I ever did a "1.0" thing.

Amen!

[gazzareth]

No beta?

just what do you think you are playing right now?

We are all playing the beta now right enough, but surprising that there are going to be massive changes to the core of the game for v1.0 .....

G

[Pecan]

We are all playing the beta now right enough, but surprising that there are going to be massive changes to the core of the game for v1.0 .....

^ This - we are playing THE only beta release, without all those features as they are planned for 1.0. There isn't going to be a (public, as has been pointed-out elsewhere) beta version that includes all those rather massive changes, just straight to "yep, that's the full game, as we intended".

[Cirocco]

That's not entirely correct, Isp is not constant. In real life the only constant is the fuel flow. As a rocket rises through the atmosphere and the ambient pressure drops, the thrust increases. Since Isp = thrust / (fuel flow * g_o), as the thrust increases, so does the Isp. The exhaust velocity doesn't change (maybe that's what you're thinking of), but the Isp does because of the changing pressure difference at the nozzle exit.

This is true! I was indeed thinking of actual exhaust velocity, which is not the same as the specific impulse (also called effective exhaust velocity). Rookie mistake, thanks for the clarification

[KerikBalm]

Funny, the 320/360 engine will lose less thrust at sea level than the 320/390 engine. This change requires some rethinking of how we evaluate engines.

Well, what is really means is that thr 320/360 engine is losing thrust in the vacuum due to underexpansion of the exhaust, while the aerospike is not losing thrust in a vacuum.

http://en.wikipedia.org/wiki/Altitude_compensating_nozzle

Caption:

Nozzles can be (top to bottom):

Underexpanded

Ambient

Overexpanded

Grossly overexpanded

If under or overexpanded then loss of efficiency occurs. Grossly overexpanded nozzles have improved efficiency, but the exhaust jet is unstable. Conventional nozzles become progressively more underexpanded as they gain altitude

[Red Iron Crown]

Well, what is really means is that thr 320/360 engine is losing thrust in the vacuum due to underexpansion of the exhaust, while the aerospike is not losing thrust in a vacuum.

http://en.wikipedia.org/wiki/Altitude_compensating_nozzle

http://upload.wikimedia.org/wikipedia/commons/thumb/4/47/Rocket_nozzle_expansion.svg/599px-Rocket_nozzle_expansion.svg.png

Caption:

Engines will always produce the most thrust in vacuum, though by what degree depends on over/underexpansion as you mention. Engines designed for atmo work are just less bad in atmo and less good in vacuum, they will still have higher thrust in vacuum.

My point was that the amount of thrust variance between atmo and vacuum is proportional to the ratio of atmo Isp and vacuum Isp. So for your example engines:

An engine with 320/360 Isp will have just under 89% of its vacuum thrust at sea level.

An engine with 320/390 Isp will have about 82% of its vacuum thrust at sea level.

With fuel flow scaling as we currently use, the 320/390 engine is better than the 320/360 engine in every situation, all other things being equal. With thrust scaling, the 320/360 engine will have the same efficiency and better TWR at sea level than the 320/390. As I said, this change will make us reevaluate engines for atmospheric work as the ratio between atmo/vac Isp becomes almost as important as the absolute value of Isp.

[cantab]

One more gotcha - you're assuming the rated thrust is in vacuum. I won't be surprised to see the rated thrust actually at sea level, thereby ensuring that (aside from Eve and Jool) the player always gets at least their expected TWR.

[KerikBalm]

With fuel flow scaling as we currently use, the 320/390 engine is better than the 320/360 engine in every situation, all other things being equal. With thrust scaling, the 320/360 engine will have the same efficiency and better TWR at sea level than the 320/390. As I said, this change will make us reevaluate engines for atmospheric work as the ratio between atmo/vac Isp becomes almost as important as the absolute value of Isp.

Well... the "all other things being equal" is the problem...

It could be they are set to have the same(or similar) TWR at sea level, and the aerospike has a superior TWR in a vacuum....

but yes, the ratio between Isp_atm and Isp_vac will determine the ratio of atmo thrust to vacuum thrust.

I think balancing should be done mostly around atmospheric thrust.

There are few bodies where you need good TWRs, yet there is no atmosphere... tylo is the only one that comes to my mind.

[Empiro]

Engines will always produce the most thrust in vacuum, though by what degree depends on over/underexpansion as you mention. Engines designed for atmo work are just less bad in atmo and less good in vacuum, they will still have higher thrust in vacuum.

My point was that the amount of thrust variance between atmo and vacuum is proportional to the ratio of atmo Isp and vacuum Isp. So for your example engines:

An engine with 320/360 Isp will have just under 89% of its vacuum thrust at sea level.

An engine with 320/390 Isp will have about 82% of its vacuum thrust at sea level.

With fuel flow scaling as we currently use, the 320/390 engine is better than the 320/360 engine in every situation, all other things being equal. With thrust scaling, the 320/360 engine will have the same efficiency and better TWR at sea level than the 320/390. As I said, this change will make us reevaluate engines for atmospheric work as the ratio between atmo/vac Isp becomes almost as important as the absolute value of Isp.

It sounds like we'll want the UI to show Thrust (vac) and Thrust (atm) instead of a single static Thrust value like it is now. Other than that, I don't foresee too many players having much trouble adjusting to the new system.