Calculations show that the "Dawn" Electric Propulsion System defies law of physics

[Alpaca Z]

Supposedly most ksp players already know that the “Dawn” ion engine in the game is way overpowered. But probably few people have wondered how unrealistic it is.

So I decided to do some simple calculation to show that the ion engine in the game actually defies laws of nature.

 

The Dawn ion engine has 4200 s Isp with a max thrust of 2 kN.

The max power required is P=½*F*c=41.2 MW. 

This means it consumes at least 41.2 MJ of energy per second (not counting efficiency). 

In the latest version of ksp (1.12.3), each Dawn engine consumes 8.74 units of charge per second when at max thrust and Isp. 

This means each unit of charge should be at least 4.71 MJ of energy.

 

Let’s then look at the batteries in game to figure out if this is possible.

Battery: 5 kg equals 100 units of charge. 

Hydrogen fuel cells have the highest specific energy among all practical energy storage devices.

Assuming the batteries are hydrogen fuel cells with 100% mass of reactants, the standard Gibbs free energy of formation for liquid water at RT is 13.17 MJ/kg. Then the total energy that can be stored in 5 kg of H2 and O2 is 65.87 MJ. 

1 unit of charge = 0.6587 MJ.

(If the batteries are Li-ion instead, specific energy ~ 200Wh/kg = 0.72 MJ/kg. In this case, the total energy in 5 kg is 3.6 MJ. 1 unit of charge = 0.036 MJ.)

In summary, real life physics and chemistry indicates that 1 unit of charge < 0.6587 MJ.

In conclusion, there is no way that the in-game ion engine can produce such a high thrust in real life. 

To be even remotely realistic, either the thrust should be reduced by 10 fold, or the electricity consumption should be increased by 10 times. This will make the already-unpleasant experience of using ion engines even more frustrating, so it is understandable that the game creator chose to make the ion engine unrealistic but fun to play.

 

[KerikBalm]

Or batteries are matter anti-matter converters.

It is interesting that you make this argument based on battery mass.

Mostly I see it based on presumed solar irradiance at Kerbin (I think the cfg file value is the same, assuming units of kw/m^2 or something like that), or the upper limit on solar irradiance given Kerbin's atmosphere and temperature, and the sizes of the solar panels.

I remember when the ion engines produced 0.5 kN, not 2.

That was OP Eve then, but without thrust during warp, they aren't usable at lower thrust

[Snark]

On 4/10/2022 at 7:29 PM, Alpaca Z said:

To be even remotely realistic, either the thrust should be reduced by 10 fold, or the electricity consumption should be increased by 10 times. This will make the already-unpleasant experience of using ion engines even more frustrating, so it is understandable that the game creator chose to make the ion engine unrealistic but fun to play.

Well, yes, this is fairly well known. After all, real-life ion engines are a  tiny, almost microscopic, fraction of the thrust of a Dawn.  Thrust that's so low they need to keep it up for weeks at a time to adjust their orbits.  If they modeled the Dawn on real-life ion engines with realistic stats, it would be utterly unplayable in KSP 1.  (Could work in KSP 2, which will allow for continuous thrust while on rails, but not in current KSP.)

Per the Wikipedia article on ion engines:

Quote

Ion thrusters in operation typically consume 1–7 kW of power, have exhaust velocities around 20–50 km/s (I_sp 2000–5000 s), and possess thrusts of 25–250 mN and a propulsive efficiency 65–80%^[3][4] though experimental versions have achieved 100 kW (130 hp), 5 N (1.1 lb_f).

In other words, typical ion engines have a thrust that's 8,000 to 80,000 times smaller than the KSP Dawn's... with experimental units topping out at 1/400th the thrust.

Yes, physically they make no sense in terms of real-world physics, the way they're currently implemented.  It's clearly a game-balance decision, so it would at least give players the idea of "here's what it's like to use a very low thrust but very high Isp engine" and how that differs from the experience of a traditional rocket engine.  It's a step in that direction, even if it had to be unrealistically powerful, by a few orders of magnitude, in order to be reasonably "playable".

[king of nowhere]

the isru mechanics break physics a lot worse than that. you mine ore and convert it to fuel, and burn that fuel to get energy to mine more ore. unless you can mine oil on the other planets, it breaks conservation of energy forever.

[Snark]

KSP has plenty of places where it cuts corners in the name of playability.  For example, there are the magical reaction wheels that can supply torque forever without saturating... or planets 1000x denser than real ones, and so forth.

All of which I, for one, am totally fine with.  It accurately models certain mathematical principles, provides good enough accuracy to get the feel of rocket science across, teaches the "right" lessons IMO, and so forth.  Most importantly, it manages to do all that and still be fun, and I'm absolutely fine with a few unrealistic shortcuts if they help to make it so.

[darthgently]

56 minutes ago, king of nowhere said:

the isru mechanics break physics a lot worse than that. you mine ore and convert it to fuel, and burn that fuel to get energy to mine more ore. unless you can mine oil on the other planets, it breaks conservation of energy forever.

Isn't this kind of like saying that you couldn't run a steam powered chain saw by burning the trees you were cutting down?  Of course you could.  Because the trees contain more energy than the amount required to cut them down.  Same with burning fuel from ore to mine ore.  It isn't oil that is gotten from the ore, or oil that the converters are converting.  It is some kind of hydrogen/oxygen game thing.  Anyway, the point is that you can absolutely power a "harvesting" process using what you are harvesting to power your harvester.  You do this every time you eat, really.

[Snark]

42 minutes ago, darthgently said:

Anyway, the point is that you can absolutely power a "harvesting" process using what you are harvesting to power your harvester. 

Yes... but only if the thing you're harvesting has a net positive energy content. Which, in general, typically only happens with biological processes that store energy (e.g. from sunlight).

Naturally occurring minerals do not typically have that property. If they could release energy by burning them, they would have already done so.

Which means that if you're collecting material on a dead world that has never had a biosphere, you're generally not going to be able to generate net energy from what you mine. You may be able to put energy into it (for example, cracking water to make hydrogen and oxygen), but this requires a net input of energy to the mining equipment.

So, for example, it may be somewhat realistic to be able to have a powered miner that makes LFO from ore. But it's very unrealistic if the power requirements are such that you can burn the LFO produced in a fuel cell to run the refining process-- which KSP allows.

[darthgently]

4 minutes ago, Snark said:

So, for example, it may be somewhat realistic to be able to have a powered miner that makes LFO from ore. But it's very unrealistic if the power requirements are such that you can burn the LFO produced in a fuel cell to run the refining process-- which KSP allows.

Ok, that makes sense.  Order, entropy, thermodynamics, etc.   With just one caveat:  if, as you imply, entropy can only be reduced via biological processes, and life requires a resource that offers a net energy, then which came first?  The resource with an available energy gradient or the biological process?  

So sure, it isn't a given that any regolith or mineral deposit will already be a maximum entropy as processes can get halted, or very slowed down, for a variety of reasons.  It takes time to reach maximum entropy,  and a harvester just has to get to the resource before that time has has gone beyond some cross-over point.

But I get what you are saying.  Unless the oxygen and hydrogen are already separate, and not combined in the same molecule, like water, in the resources available, then outside energy would be needed to split them.

[Alpaca Z]

4 hours ago, Snark said:

Well, yes, this is fairly well known. After all, real-life ion engines are a  tiny, almost microscopic, fraction of the thrust of a Dawn.  Thrust that's so low they need to keep it up for weeks at a time to adjust their orbits.  If they modeled the Dawn on real-life ion engines with realistic stats, it would be utterly unplayable in KSP 1.  (Could work in KSP 2, which will allow for continuous thrust while on rails, but not in current KSP.)

Per the Wikipedia article on ion engines:

In other words, typical ion engines have a thrust that's 8,000 to 80,000 times smaller than the KSP Dawn's... with experimental units topping out at 1/400th the thrust.

Yes, physically they make no sense in terms of real-world physics, the way they're currently implemented.  It's clearly a game-balance decision, so it would at least give players the idea of "here's what it's like to use a very low thrust but very high Isp engine" and how that differs from the experience of a traditional rocket engine.  It's a step in that direction, even if it had to be unrealistically powerful, by a few orders of magnitude, in order to be reasonably "playable".

I agree that it is fairly well known that ion engines are over-powered in the game. Just wanted some math practice for myself ...

Actually, if the thrust of Dawn was reduced by 500 to 1000 times, then everything about it makes sense. The batteries can be Li-ion ones and one electricity unit would be 0.036 MJ.

This compromise is fine for ksp, however I really hope KSP2 will amend this flaw. Because when travelling interstellar using non-chemical propulsion, I believe it is extremely important to have the correct energy balance.

[Alpaca Z]

1 hour ago, darthgently said:

Ok, that makes sense.  Order, entropy, thermodynamics, etc.   With just one caveat:  if, as you imply, entropy can only be reduced via biological processes, and life requires a resource that offers a net energy, then which came first?  The resource with an available energy gradient or the biological process?  

So sure, it isn't a given that any regolith or mineral deposit will already be a maximum entropy as processes can get halted, or very slowed down, for a variety of reasons.  It takes time to reach maximum entropy,  and a harvester just has to get to the resource before that time has has gone beyond some cross-over point.

But I get what you are saying.  Unless the oxygen and hydrogen are already separate, and not combined in the same molecule, like water, in the resources available, then outside energy would be needed to split them.

There are certainly non-biological natural minerals or resources that contain a lot of usable energy (in chemistry, it is often computed in terms of Gibbs or Helmholtz free energy).

For example, there are liquid ethane lakes on Titan, and Titan's atmosphere contains a high concentration of methane. These are all highly reductive substances, ie, fuels.

Ammonia is another abundant chemical in the outer solar system that can be used as fuels.

However, naturally occurring oxidizers are rare.

Hydrogen peroxide and molecular O2 have been found on Europa, but are at low quantities. A study by NASA in 2013 found "a peroxide abundance of 0.12 percent relative to water on the side of Europa that always leads in its orbit around Jupiter". Mapping the Chemistry Needed for Life at Europa | NASA

They are created by intense particle bombardment from Jupiter's strong magnetic field.

So there is definitely a chance to mine liquid fuels, but not so much for oxidizers. With liquid fuel only, you will not be able to run a fuel cell.

Things become different when you consider mining nuclear fuels. Nuclear energy (fission, fusion or radioactive decay) are independent of bio-processes, and there is no difference mining them on Earth or other planets.

Edited April 13, 2022 by Alpaca Z

[Snark]

26 minutes ago, Alpaca Z said:

For example, there are liquid ethane lakes on Titan, and Titan's atmosphere contains a high concentration of methane. These are all highly reductive substances, ie, fuels.

Except that they aren't-- because they don't have oxidizers.  Ethane is a fuel on Earth, sure, because we have all this free oxygen hanging around that biological processes helpfully produced for us.   But on Titan, oxygen would be a "fuel" that you burn.

For that matter, Jupiter and Saturn are mostly made out of hydrogen, which is a "fuel", right?  Uranus is mostly made out of methane-- same deal.  But it's not as though you could go to one of those places and harvest them for fuel that you can get energy out of that can run the fuel-gathering process. Heck, for that matter, there's a lot of water in the cosmos, and you can get plenty of energy out of water-- just add sodium!     (to be clear, I'm being facetious, there-- the point is that any body that has water is not going to have free sodium lying around)

Lots of naturally occurring compounds exist that could react, given a substance that they can react with.  The problem is, if they can react, then they do, and what's left over is the reaction product.  Which is then something you can't get energy out of, because it's already been got.

26 minutes ago, Alpaca Z said:

So there is definitely a chance to mine liquid fuels, but not so much for oxidizers. With liquid fuel only, you will not be able to run a fuel cell.

Yes, that was my point.  Realistically, you're not going to be able to land on a dead planet that has never had a biosphere, and mine some kind of material that you can get energy out of (unless you bring your own reagents to react it with).  It's absolutely realistic to mine ore and turn it into burnable LFO-- but only with a net energy input.  For it to be realistic, the amount of electricity needed to produce the LFO needs to exceed the energy that you could get by burning it in a fuel cell.  (By a lot, actually, given that clearly a lot of energy is being thermally wasted, since a KSP mining/refining rig has to have a lot of radiators.)

[Catto]

It does guzzle EC. I have to put 182842 solar panels on my craft to power 1.

[Snark]

9 hours ago, siklidkid said:

It does guzzle EC. I have to put 182842 solar panels on my craft to power 1.

Sure, but from a realism perspective, it doesn't guzzle enough.

In KSP, it's easily possible to run a mining/refining rig on fuel cells alone, with no solar input at all, with the fuel cells powering the mining and refining using only a portion of the LFO produced. As long as there's a reasonable ore percentage at the location, and you've got a decent engineer aboard, it works like a charm.

And conservation of energy doesn't work that way. That's the part that's unrealistic. To be realistic, it needs to be flat-out impossible to run it on fuel cells.

(To be clear, I'm not advocating that the devs should have done that, or that the current design is "bad". From a gameplay perspective, it makes reasonable sense. It just happens not to be physically realistic, is all, like a few other mechanics in the game.)

[king of nowhere]

15 hours ago, darthgently said:

Isn't this kind of like saying that you couldn't run a steam powered chain saw by burning the trees you were cutting down?  Of course you could.  Because the trees contain more energy than the amount required to cut them down.  Same with burning fuel from ore to mine ore.  It isn't oil that is gotten from the ore, or oil that the converters are converting.  It is some kind of hydrogen/oxygen game thing.  Anyway, the point is that you can absolutely power a "harvesting" process using what you are harvesting to power your harvester.  You do this every time you eat, really.

no, those are very different things.

when you cut down wood with a chain saw, wood is a high-energy material. you can burn it and it releases lots of energy. wood, and other combustibles we have on earth, are very special and they are not found elsewhere.

the thing is, nature tends towards low energy. wood releases energy because some plants took low energy material and used sun power to push them to higher energy.

in other planets, with no life, you don't get wood, oil, or anything like that. you don't get high energy substances. in the billions of years since the formation of those planets, everything had time to react. So, for example, spacex wants to mine methane and oxigen on mars; but you don't get those stuff on mars. on mars you get carbon dioxide and water, which are low energy substances. to convert them to hydrogen and methane, which are high energy, you need to give them energy. as much energy as they release when burning, according to the law of energy conservation. more, because in practice you always have some inefficiencies and losses.

So, mining other planets for fuel is not like cutting down wood for energy. it's like carrying water uphill, to power a hydroelectric dam. and then hoping that the water falling down will get you more energy than you spent to carry it up.

[darthgently]

40 minutes ago, king of nowhere said:

So, mining other planets for fuel is not like cutting down wood for energy. it's like carrying water uphill, to power a hydroelectric dam. and then hoping that the water falling down will get you more energy than you spent to carry it up.

I was just going by the fact that all bodies being at full entropy being extremely unlikely. but yeah, as Snark pointed out, the non-entropy, or free energy, is unlikely to be useful

Edited April 13, 2022 by darthgently

[darthgently]

17 hours ago, Snark said:

If they could release energy by burning them, they would have already done so.

I still disagree with this though.  Many entropy increasing processes require barriers to be surmounted, like water overflowing and washing out a dam.

[Snark]

1 hour ago, darthgently said:

I still disagree with this though.  Many entropy increasing processes require barriers to be surmounted, like water overflowing and washing out a dam.

Yes, it's technically possible; but the assertion here is that it's unrealistic.

Planet-sized bodies  generally form in conditions that involve lots of excess heat (the heat of collisions, the heat of gravitational condensation).  The rocks, minerals, etc. making up a body have typically been molten at some point.  It's highly unlikely that there would be some sort of ore material sitting around that consists of a net energy-positive mixture of reactive materials that can react with each other but haven't done so over the aeons since formation.

Water reacts with sodium.  It's not physically impossible to have free metallic sodium somewhere on a planet, I suppose (how that could happen I can't imagine, but for the sake of argument let's posit that it's possible.)  But you know where you'll never find metallic sodium?  Sitting there mixed in with water ice so that all you have to do is mine it and thaw it out and get useful energy.

Other than nuclear fuel, and the results  of biological processes (i.e. fossil fuels), can you name a single material on Earth that can be harvested to extract net chemical energy from?  (Without using atmospheric oxygen, which itself is a biological product.)

Water can overflow and wash out a dam because water releases energy when it flows downhill.  It can do this because it has the energy in the first place.  It's releasing stored energy.  To release stored energy, you have to get that energy from somewhere in the first place-- in this case, it ultimately comes from the sun.

If you're mining ore and getting a net exothermic reaction out of it, that means you're releasing chemical energy-- energy that never got released before, including through the hot, tumultuous formation of the planet in the first place.  So the question becomes, 1. where did that energy come from, and 2. what stopped it from getting released before now?

I'm having trouble thinking of any realistic situation where that could happen-- can you?

Note that even the term "ISRU" used on the KSP gizmo that does this is taken from real-world space engineering.  The idea of using ISRU for, say, refueling the return leg of a Mars journey is an interesting one.  But nobody's talking about "we're going to go to Mars and get energy out of stuff that we mine there."  It's all about putting energy in (by nuclear or solar power) to manufacture fuel from local materials.  Mars will supply the reaction mass and the necessary chemical elements, yes-- but it will be nuclear or solar power that provides the necessary energy to make it into usable rocket fuel.

[darthgently]

15 minutes ago, Snark said:

Yes, it's technically possible; but the assertion here is that it's unrealistic.

Yes, I understand that.  But the bottom line is, for life to have evolved at all, it can't be completely unrealistic in all circumstances

[Snark]

1 hour ago, darthgently said:

Yes, I understand that.  But the bottom line is, for life to have evolved at all, it can't be completely unrealistic in all circumstances

Can microscopic flecks end up with a net positive energy, due to a supply of something like solar radiation, heat from ocean thermal vents, etc.?  Sure.  In infinitesimal quantities that would be utterly negligible on a planetary scale... unless those flecks end up developing the knack to replicate themselves, in which case they take over everything, are no longer negligible, and form a biosphere.

But in the sense of substantial percentage of a body's surface crust materials, spread everywhere, in bulk, without benefit of a biosphere?  Generally not-- I'm not aware of any realistic model of planetary formation that would allow for such a thing.

[darthgently]

16 minutes ago, Snark said:

But in the sense of substantial percentage of a body's surface crust materials, spread everywhere, in bulk, without benefit of a biosphere?  Generally not-- I'm not aware of any realistic model of planetary formation that would allow for such a thing.

Yes, wasn't disagreeing with that at all given what you've written.

It wasn't until I started messing with kOS that a lot of the KSP stock gui made more sense, ha ha.  Once you start to see it all in terms of vectors and orbital mechanics the GUI makes a lot more sense.  And I've only really just begun to understand most of the math aspects beyond the basic orbital stuff

Edited April 13, 2022 by darthgently
wrong thread