"Dawn" Ion Engine Suggestion

[SpaceplaneAddict]

I propose a tweak to the Ion engine, as to me currently it seems highly wacked out. So much electricity, and so little Xenon.

 

I ask of this:

Reduce electrical usage, and slightly lower ISP. This will make it less efficient, but also easier to power.

Alternatively, make it so that you can tweak the usage. You could balance usage between the two, maybe, if possible.

 

-SPAddict
Suggestions to add to this, welcome!

Edited February 14, 2016 by SpaceplaneAddict

[UmbralRaptor]

I'd lean towards double checking that the number/size of solar panels needed isn't inconveniently huge. Since 1.0, one gigantor should be very roughly enough for two Dawns at Kerbin, or one at Duna. Dropping to 2/3 thrust (blasphemy, I know), should push that out to Dres. I suppose an argument could be made for intermediate sized panels between the OX-4 L/W and Gigantor (or uprating the OX-4) when less power is needed, though.

 

14 minutes ago, SpaceplaneAddict said:

-Keep ISP values same,

-Increase Xenon usage,
-And decrease electricity usage.

This set of changes is impossible without significant resource or engine system tweaks. Within KSP, electriccharge is a massless resource, so all the propellant usage implied by that 4200 s Isp comes from the xenon. Reducing Xenon usage in this environment would mean raising the Isp, or lowering the thrust.

To get a bit into the history of the engine, The Dawn supposed to be semi-realistic (high Isp, high power requirements, low thrust), though admittedly with a TWR several orders of magnitude higher than any actual ion engine. It suffered from a mismatch between electricity usage and available panels for several versions, though that should have been fixed with changes circa 0.25 and 1.0. OTOH, if you were using ion engines out at Jool before 1.0, keep in mind that solar panels in that era did not follow the inverse square law.

 

[Nich]

hell no IONs need love nukes are better in just about every way.  The only time I consider ION is when 1 nuke is too big.  Double there ISP and reduce their TWR by half

Edited February 14, 2016 by Nich

[SpaceplaneAddict]

11 minutes ago, Nich said:

hell no IONs need love nukes are better in just about every way.  The only time I consider ION is when 1 nuke is too big.  Double there ISP and reduce their TWR by half

Not getting what your saying.

What I understand is that you want to: make the thing EVENMORE wacked out (Even for KSP), and reduce the thrust still.

That would make interplanetary travel quite rediculously taxing (part-wise, engine spam to maintain decent TWR), no?

 

Also, updated OP with revised suggestion.

Edited February 14, 2016 by SpaceplaneAddict

[H2O.]

I disagree with the OP. I think the ion engine is quite well balanced actually.

On the other hand, what we need is a big ion engine (let's say x10 the thrust) plus a nuclear electricity generator and a bigger xenon tank. This combo would be the solution of choice for everything past Jool.

[soulsource]

Well, given the constraints of KSP - namely that it's impossible to do a Low Thrust Transfer - the ion engine is done quite well, imho.
Edit: Of course I'd really like to see low thrust transfers in KSP, even if they only allow prograde burns, or if they are limited to thrust along the current craft orientation like the sadly no longer up to date Orbit Manipulator mod did. If the game would allow such transfers, I'd be perfectly fine with a significant reduction of ion engine thrust, along with reduced xenon, and possibly electric charge usage.

What I'd second would be a VASIMIR analogon in the game, an ion engine that allows to sacrifice some ISP to get higher thrust. It's not like such things would be impossible to implement within KSP, the afterburner of jet engines already goes in that direction.

Edited February 14, 2016 by soulsource

[Pds314]

Realistic = the ion engine uses LUDICROUS amounts of electricity and about the same amount of propellant. Remember. Every kilo of ion propellant has (4200*9.8)^2/2=847,072,800 Joules. The engine uses a bit under 50 grams per second, or about 41,000,000 watts at full thrust. That means that at 20% efficiency, it needs like 150,000 m^2 of solar panels, that's comparable to the floor space of a thousand houses, and is actually more powerful than some jet engines, including the CF-6.

Edited February 16, 2016 by Pds314

[KerikBalm]

what we need is "on rails" thrusting, and a much lower thrust.

Everything ellse will just be running in circles as we chase balance.

A nuclear reactor as an alternative to RTGs or fuel cells would also be nice for deep space operations

[Pds314]

4 minutes ago, KerikBalm said:

what we need is "on rails" thrusting, and a much lower thrust.

Everything ellse will just be running in circles as we chase balance.

A nuclear reactor as an alternative to RTGs or fuel cells would also be nice for deep space operations

A reactor is about the only thing that could practically power the current Dawn engine for sustained periods realistically.

[_Astra_]

A BIG reactor, you mean?

 

The game actually lacks a powerful reactor.

[wizzlebippi]

Dawn is fine as is.  Use a bunch of batteries, 12-15k EC, and you'll still be able to achieve a few minutes of run time at Eeloo with 7 Dawn thrusters.  I only use 4 gigantor solar panels, ~3 EC/s at Eeloo, and one RTG incase I end up with no solar panels facing the sun during time warp.  You still end up with 1.5'ish hour burn times and need to start from a very deep orbit to make the burn work.  Bigger Xenon tanks and ISRU support would be nice though.

[The Optimist]

Try having a large store of batteries, enough for any conceivable transfer burn, and a few gigantors. That way, you can have a long burn and timewarp to rapidly recharge your stores. 

 

What I just said was probably incoherent.

15 hours ago, Pds314 said:

Realistic = the ion engine uses LUDICROUS amounts of electricity and about the same amount of propellant. Remember. Every kilo of ion propellant has (4200*9.8)^2/2=847,072,800 Joules. The engine uses a bit under 50 grams per second, or about 41,000,000 watts at full thrust. That means that at 20% efficiency, it needs like 150,000 m^2 of solar panels, that's comparable to the floor space of a thousand houses, and is actually more powerful than some jet engines, including the CF-6.

No. It is not "more powerful than some jet engines". Ion thrusters have a maximum amount of energy you can apply to each particle, and after that limit the only way for more thrust is to have a larger engine area.

Edited February 17, 2016 by The Optimist

[Pds314]

4 hours ago, The Optimist said:

Try having a large store of batteries, enough for any conceivable transfer burn, and a few gigantors. That way, you can have a long burn and timewarp to rapidly recharge your stores. 

 

What I just said was probably incoherent.

No. It is not "more powerful than some jet engines". Ion thrusters have a maximum amount of energy you can apply to each particle, and after that limit the only way for more thrust is to have a larger engine area.

Here:
Our Ion engine:
2000 Newtons * 4200 s * 9.8 m/s^2 / 2 = 41160000 watts

By comparison:
CFM56-7B18 Turbofan engine:
(86700 N) ^ 2 / (307 kg/s) /2 = 12242492 watts.

That's nearly 3.5 times the thrust power, so yeah, a whopping lot more powerful than some jet engines. In fact, more powerful than some entire airliners. That's why real ion engines don't run at Kilonewton-level thrusts, and the most anyone has ever used is about 0.5% that powerful.

Edited February 17, 2016 by Pds314

[MatterBeam]

35 minutes ago, Pds314 said:

Here:
Our Ion engine:
2000 Newtons * 4200 s * 9.8 m/s^2 / 2 = 41160000 watts

By comparison:
CFM56-7B18 Turbofan engine:
(86700 N) ^ 2 / (307 kg/s) /2 = 12242492 watts.

That's nearly 3.5 times the thrust power, so yeah, a whopping lot more powerful than some jet engines. In fact, more powerful than some entire airliners. That's why real ion engines don't run at Kilonewton-level thrusts, and the most anyone has ever used is about 0.5% that powerful.

That's a very interesting formula for calculating jet engine output.

Where did you find/learn it?

[Pds314]

20 minutes ago, MatterBeam said:

That's a very interesting formula for calculating jet engine output.

Where did you find/learn it?

I forget where I learned it, but the reason it works for jets/rockets/reaction engines is because of the fact that with fuel flow held constant, the only variable that can alter thrust is exhaust velocity. Doubling the exhuast velocity doubles the thrust, but requires 4x the power.

So for example, say we have a rocket with 10 kg / s fuel flow? We don't know the exhaust velocity, but we know it has 30 kN of thrust?

We COULD determine the thrust power as follows.

V_Exhaust=30000 N / (10 kg/s) = 30000 kg*m/s^2 / (10 kg/s)=3000 m/s.

And then use the equation for kinetic energy to find the energy density of the exhaust:

Energy_Density_Exhaust=(3000 m/s)^2 / 2 = 4500000 J/kg.

And then multiply by the fuel flow to get power:

P_Exhaust= 10 kg/s * 4500000 J/kg = 45 MW.


But this method is tedious.
Rather than:
P_Exhaust=Propellant_flow*Energy_density
Energy_density=V_Exhaust^2/2
V_Exhaust=Thrust/Propellant_flow.

We can combine these formulas.

P_Exhaust=Propellant_flow*(Thrust/Propellant_flow)^2/2

And simplify:

P_Exhaust=Thrust^2/Propellant_flow/2.

And indeed, in the above example, (30000 N)^2 / (10 kg/s) / 2 = 45000000 W.

Edited February 17, 2016 by Pds314

[SpaceplaneAddict]

Yeah, the Dawn is OP compared to real life. But so are the rest of the rockets in game

All I really want is to make the Ion more useable in the name of interplanetary motherships etc

[Pds314]

24 minutes ago, SpaceplaneAddict said:

Yeah, the Dawn is OP compared to real life. But so are the rest of the rockets in game

All I really want is to make the Ion more useable in the name of interplanetary motherships etc

Actually, most of the rockets are anything-but.

For ex, the LV-T45 Swivel is comically obese. It weighs 1.5 tonnes and has 200 kN of thrust, for miserable TWR of 13.

The KS-25 Vector is 4 tonnes and 1000 kN. Better, but a TWR of 25 isn't by any means excellent.

All the chemical rockets are substantially underpowered. Even the long SRBs don't have realistically high thrust for their diameter, and the real ones on the shuttle had thrust-vectoring!

Meanwhile, jets still have twice the specific impulse they should, ions are magically energy-efficient, and the LV-N is about realistic other than that it uses kerosene instead of H2 and radiation isn't modelled. KSP really undersells chemical rocketry in most respects. The one thing KSP gives us that most real rockets don't have is unlimited restarts and throttle. IRL rockets, except for pressure-fed hypergolics, can't usually restart infinitely or throttle under 70% or so. Most of the time, the way launch vehicles reduce thrust during their ascent profile is to permanently shut down the engine that's generating too much thrust.

The reasonable thing to do is to nerf the ion's thrust, but then use equations of motion that assume a constant force pushing the craft, unless it runs low on electricity, and allow on-rails warp with ions.

And also to allow switching between absolute and rotational reference frames so the engine can be pointing constantly prograde or constantly retrograde.

 

Edited February 17, 2016 by Pds314

[Laie]

6 hours ago, SpaceplaneAddict said:

All I really want is to make the Ion more useable in the name of interplanetary motherships etc

With all due respect... would you be open to the notion that you're doing it wrong?

The domain of ion drives are relatively small satellites. If by "mothership" you mean a rather large vessel, you should be looking towards nukes (which are still underpowered for the job, requiring one to accept large part counts -- but that's another topic for another day).

It's entirely possible to build a useful RTG-powered ion vessel. It's been a while, but IIRC with something on the order of one large 4k battery and three RTGs I could do ~400m/s in one go and would need only 30-40 minutes to recharge the battery between burns. That was entirely sufficient for a scansat to map the whole Joolian system, and Eeloo as well. A chemical rocket was used to propel the satellite to Jool, though (not hard, the whole thing was about 1t).

[chrisb2e9]

I just do what i always do and edit the part file. 400 thrust sounds about right...

[KerbonautInTraining]

On Wednesday, February 17, 2016 at 11:21 PM, Pds314 said:

I forget where I learned it, but the reason it works for jets/rockets/reaction engines is because of the fact that with fuel flow held constant, the only variable that can alter thrust is exhaust velocity. Doubling the exhuast velocity doubles the thrust, but requires 4x the power.

Does this take into account the fact that jets use air as reaction mass? Their actual exhaust velocity is usually <350m/s (36s Isp) unless it's afterburning.

Edited February 19, 2016 by KerbonautInTraining

[MatterBeam]

1 hour ago, KerbonautInTraining said:

Does this take into account the fact that jets use air as reaction mass? Their actual exhaust velocity is usually <350m/s (36s Isp) unless it's afterburning.

Jets have a different formula. They're accelerating a mass of slow moving air,unlike a rocket where the entirety of its energy comes from accelerated propellants.

[KerbonautInTraining]

3 minutes ago, MatterBeam said:

Jets have a different formula. They're accelerating a mass of slow moving air,unlike a rocket where the entirety of its energy comes from accelerated propellants.

That's what I was saying. Looking at it again, I see the formula assumes the only reaction mass is the fuel even for jet engines. Although I don't want to get too far off topic.

Edited February 19, 2016 by KerbonautInTraining

[Pds314]

On 2/19/2016 at 8:20 AM, KerbonautInTraining said:

Does this take into account the fact that jets use air as reaction mass? Their actual exhaust velocity is usually <350m/s (36s Isp) unless it's afterburning.

Yes. It is looking at the exhaust velocity vs. airflow. I ignored fuel because the aforementioned HBTF jets have maybe 1% the fuel flow of their airflow, if that. In this particular case, the ISP is about 28.

Edited February 21, 2016 by Pds314

[Pds314]

On 2/19/2016 at 9:52 AM, KerbonautInTraining said:

That's what I was saying. Looking at it again, I see the formula assumes the only reaction mass is the fuel even for jet engines. Although I don't want to get too far off topic.

No. The formula is based on propellant flow, not fuel flow. If you want to use fuel flow, you need to assume an effective ISP of hundreds of times times that, and you can't meaningfully compute power output without knowing the bypass ratio and some other data.

For reference, a typical low-bypass fighter engine without afterburner might have an effective ISP of 4000 and an actual ISP of like 80 or 90. That is to say, even fighter engines match ions for fuel consumption vs. thrust generated, but of course have much lower ISPs. Computations based on effective exhuast velocity and fuel flow are not useful. Clearly, a GE90 with 12000 seconds of ISP is not generating 9 times the energy per unit fuel of an ion thruster. It is in fact generating considerably less energy per unit fuel than an ion thruster.

Edited February 21, 2016 by Pds314

[Pds314]

As a point of comparison:

A 1 kN ion engine at 4000 ISP has more-or-less the same thrust power as:

A 9 kN Hydrolox rocket.

An 11 kN Metholox rocket.

A 13 kN Kerolox rocket.

A 15 kN SRB or Alcolox rocket

A 35 kN Afterburning LBTF.

A 50 kN LBTF

A 140 kN HBTF

5 Fully loaded Sikorsky CH-53E Super Stallion Helicopters producing 1650 kN of thrust between them.

600 Fully loaded Robinson R22 Light helicopters producing 3700 kN of thrust between them.

78000 100 kg bikers + bikes and gear at 5 m/s accelerating at 1 m/s^2! These are what are known as Very Very Very high bypass hybrid geared low-temperature combustion engines.

Or...

A 131 MilliNewton Flashlight (40 Megawatt laser of doom).

Edited February 21, 2016 by Pds314