Xenon Tank Weights

[Wcmille]

Why is the wet/dry ratio for Xe tanks so much lower than the ratio for liquid fuel/oxider rockets? Is this based on reality or designed for game balance?

[Starman4308]

I strongly suspect the answer is going to be "yes"; xenon is stored as a pressurized gas instead of a liquid, thus bulky, heavy pressure tanks, and for game balance, you still get vastly better delta-V out of ion engines due to their specific impulse.

[Sandworm]

"... carbon composite pressure vessel was qualified to launch with 992 lbs of xenon at 1,250 psig. The full assembly weight of the tank and skirt is under 49 pounds."

Call that 1000lbs of xenon in a 50lb tank.  That's a ratio of 1/20 for a flown tank on a real spacecraft.

http://www.cobham.com/mission-systems/composite-pressure-solutions/space-systems/xenon-propellant-tank-datasheet/

[Starman4308]

3 minutes ago, Sandworm said:

"... carbon composite pressure vessel was qualified to launch with 992 lbs of xenon at 1,250 psig. The full assembly weight of the tank and skirt is under 49 pounds."

Call that 1000lbs of xenon in a 50lb tank.  That's a ratio of 1/20 for a flown tank on a real spacecraft.

http://www.cobham.com/mission-systems/composite-pressure-solutions/space-systems/xenon-propellant-tank-datasheet/

By comparison, I think the Space Shuttle main tank had a ratio of 1/50, so pressurized gas tanks are still significantly behind conventional liquid propellant tanks. Plus, I'm pretty sure the ratios get worse for lighter gases*, as each mol of gas inside your tank has less overall mass.

*Xenon is expensive.

[Sandworm]

2 minutes ago, Starman4308 said:

By comparison, I think the Space Shuttle main tank had a ratio of 1/50, so pressurized gas tanks are still significantly behind conventional liquid propellant tanks. Plus, I'm pretty sure the ratios get worse for lighter gases*, as each mol of gas inside your tank has less overall mass.

*Xenon is expensive.

Well, the link was to a xenon tank.  So the ratio won't be any worse than 1/20, and certainly nowhere near KSP's 1/2 ratio.  Xenon is a heavy gas.  As for cost, it's roughly 1$ per gram, or 1000$ per kilo.  That isn't cheap, but the tank will still likely cost more to build and install than the gas it carries.  The extended service life permitted by Xenon means that sat owners realy couldn't care less about its cost per kilo.

[Starman4308]

51 minutes ago, Sandworm said:

Well, the link was to a xenon tank.  So the ratio won't be any worse than 1/20, and certainly nowhere near KSP's 1/2 ratio.  Xenon is a heavy gas.  As for cost, it's roughly 1$ per gram, or 1000$ per kilo.  That isn't cheap, but the tank will still likely cost more to build and install than the gas it carries.  The extended service life permitted by Xenon means that sat owners realy couldn't care less about its cost per kilo.

My point is that for any gas lighter than argon, the ratio should (in theory) be worse, because given fixed pressure/temperature/volume, you can carry a fixed number of mols of gas, and xenon has an advantage in being heavy per mol of gas.

And yes, 1/2 is absurd, but so is the 1/9 ratio for stock LF+O tanks.

[Wcmille]

6 hours ago, Starman4308 said:

By comparison, I think the Space Shuttle main tank had a ratio of 1/50, so pressurized gas tanks are still significantly behind conventional liquid propellant tanks. Plus, I'm pretty sure the ratios get worse for lighter gases*, as each mol of gas inside your tank has less overall mass.

*Xenon is expensive.

If this is true, then I think KSP xenon ratios should be closer to 1:9 x (20/50) = 1:3.6

[Wcmille]

6 hours ago, Starman4308 said:

I strongly suspect the answer is going to be "yes"; xenon is stored as a pressurized gas instead of a liquid, thus bulky, heavy pressure tanks, and for game balance, you still get vastly better delta-V out of ion engines due to their specific impulse.

Compared to NERV, I get 4200/800*LN(2.27)/LN(9) = 1.96 times better, which doesn't feel that great, given the poor thrust, no mining refueling, power requirements, and higher part counts.

[Streetwind]

9 hours ago, Starman4308 said:

By comparison, I think the Space Shuttle main tank had a ratio of 1/50, so pressurized gas tanks are still significantly behind conventional liquid propellant tanks. Plus, I'm pretty sure the ratios get worse for lighter gases*, as each mol of gas inside your tank has less overall mass.

*Xenon is expensive.

Not correct, I'm afraid. The mass ratio was slightly less than 30.   ( 760,000 wet / 26,500 dry = ~28.679 )   I believe that SpaceX today claims a mass ratio of about 30 for the Falcon 9 as a whole, but I don't have any numbers to back it up.

That only makes the xenon tanks worse by comparison, of course I did bring this up before in the suggestion forum, and was unofficially told that it would get earmarked for a future patch, likely along with the general rocketry part revamp... unfortunately, 1.2 has come and gone without said revamp, and the person who told me this no longer works at Squad. Soooo I'm not sure if and when any xenon tank changes might happen.

Last devnotes said that we would get more info on the way forward to 1.3 in the coming devnotes. Today is Friday, where devnotes are released. Perhaps we will find out later today, then. Perhaps not.

[KerikBalm]

13 hours ago, Starman4308 said:

I strongly suspect the answer is going to be "yes"; xenon is stored as a pressurized gas instead of a liquid, thus bulky, heavy pressure tanks, and for game balance, you still get vastly better delta-V out of ion engines due to their specific impulse.

Why would they store it as a pressurized gas rather than a liquid in KSP though? KSP keeps LF and Ox ambiguous, but if the Oxidizer is liquid O2, then we should have liquid xenon. Some will claim the Oxidizer is more properly di-Nitrogen tetraOxide. (Ie two main options: Kerosene and liquid O2, or Aerozine and N2O4)

If KSP is not dealing with the problems of cryogenic storage of propellant, then we should have liquid Xe.

 

13 hours ago, Starman4308 said:

By comparison, I think the Space Shuttle main tank had a ratio of 1/50, so pressurized gas tanks are still significantly behind conventional liquid propellant tanks. Plus, I'm pretty sure the ratios get worse for lighter gases*, as each mol of gas inside your tank has less overall mass.

Yea, but Xenon is far from a lighter gas. You're also comparing a large tank (small surface to volume ratio) meant to store a cryogenic liquid for hours(and only minutes without the support infrastructure of the launchpad), compared to a small tank (high surface area to volume) meant to store the propellant for several years.

Liquid O2 boiling temperature: 90.2 K

Liquid Xenon boiling temperature: 165k

-> Xenon is the clear winner, being able to be stored at a temperature 75K warmer

Liquid O2 density: 1.14 gm/cm^3

Liquid Xenon density: 2.94 gm/cm^3

-> Xenon is the clear winner, being over 2.57x as dense as O2

Clearly, liquid Xenon is much better suited to long term storage than liquid Oxygen, and lets not even talk about liquid hydrogen. Liquid Xenon tank mass ratios would be much higher than those of O2 or H2.

Even argon looks like it could better than O2:

Boiling point: 87.3 K  (ok, this is 2.8K worse than O2)

Density: 1.395 gm/cm^3 (this is over 22% better)

Argon is really cheap when compared to Xenon (not sure about O2)... about 1000x cheaper than xenon

As a compromise for availability and expense of propellant: Krypton, 10x cheaper than Xenon, but about 100x more expensive than argon

119.93k boiling point (~30k better than O2)

2.414 gm/cm^3 (2.1x that of O2... more than twice as good).

The lighter gasses give more Isp, they aren't used though because of power restrictions, much more energy is "wasted" to ionize them, leaving less energy available to actually accelerate them. So if one doesn't want to sacrifice Isp, you get much much less thrust for a given power input - and if you reduce exhuast velocity to get the same thrust as Xe, you end up with a much worse Isp.

Put a nuclear reactor on a ship, or have it fly close to the sun (like venus/mercury close) rather than farther out (like ceres), and Krypton will be more appealing than Xenon.

Xenon's main appeal is the low ionization energy per unit mass. As energy becomes easier to come by, the other noble gasses will look more attractive.

It will always have the advantage of easier storage, but that's not the reason its used today.

[MaxL_1023]

The tank weights are a game balance decision. So is the fact that ion engines produce more than a thousand times as much thrust as they should. With realistic tank weights and "balanced" ion engines it would be too easy to get to other planets. The same would be true to some degree with realistic LFO tank weights - SSTO rockets would be trivial (not spaceplanes) due to the lower delta-v requirements. 

The tank weights are a compromise with the simplified engine mechanics (unlimited ignitions + throttling) and the scaled down solar system. Changing one would require change to the others.

You can try RO when it is updated for 1.2.1 -> it has realistic tank weights. I had it on accidentally without putting in RSS, and the minuteman solid fuel booster (basically a thumper in stock) launched my Mk.I pod onto an escape trajectory. 

[Gaarst]

OK, let's make ion more realistic and have better fuel fractions on Xe tanks. Good luck with your months-long milliNewtons burns though.

Edited November 25, 2016 by Gaarst

[GoSlash27]

19 hours ago, Starman4308 said:

I strongly suspect the answer is going to be "yes"; xenon is stored as a pressurized gas instead of a liquid, thus bulky, heavy pressure tanks, and for game balance, you still get vastly better delta-V out of ion engines due to their specific impulse.

^ This.

 Part realism, mostly game balance. The KSP solar system is extremely tiny in comparison to the real thing, so there's simply no call for the absurd DV that RL ions can produce. As @Gaarst points out, if ions were *truly* realistic in KSP, nobody would ever use them.

Best,
-Slashy

 

[panzer1b]

Given the smaller solar system and lower dV requirements its a matter of balancing and keeping any engine type from being so good there is no point to another.  Despite being unrealistic, the extremely high dry masses of tanks and such is to balance the parts and make it an actual engineering challenge to get somewhere and not a simple matter of slapping a probe core onto a SRB and sending those 2 parts into the sun.  That, and ions right now are kinda perfect (at least imo).  They offer the best range for any engine, but come with the energy requirement (you kinda need to either spam solar panels or spam batteries, or if you are like me, edit the ion engines to have built in batteries so that i dont need to have 500 parts just for a ion drive), and the downside of fairly low thrust even though it is over 1000 times that of real life ion engines which can take years to perform a burn.

Now if you do actually install RSS then you will need to have somewhat realistic mass fractions as you wont get anywhere using the stock parts, it just cant be done without utter insanity (ive actually tried it, got a satellite into LEO with a rocket that could have circumnavigated the entire stock solar system in the stock game).  Still fun to try though...

[Xyphos]

Argon would have been a better choice than Xenon, but you'll need to use a mod for Argon engines.

[MaxL_1023]

Generally, the lighter the molecular weight of the exhaust gas the better the ISP, as it is proportional to the exhaust velocity. However, the thrust decreases assuming the same number of propellant molecules being accelerated. Argon is harder to store in quantity due to its lower molecular mass and lower boiling point. I also am pretty sure that Argon is harder to ionize than Xenon, so that extra energy might offset the efficiency gain from a lighter propellant. 

The ion thrust is as much a limitation of the physics engine as anything - KSP can't calculate a continuous thrust trajectory. The patched conic approximation gives you trouble when you are dealing with realistic ion propulsion - a constant thrust trajectory will have a major discontinuity when crossing SOI boundaries and will end up putting you in a completely different spot than real physics would lead you. I would love to use Ions in RSS if I could just do the burns on autopilot at 1000x warp - it makes a lot of sense for mercury probes, multiple-asteroid missions in RSS Extended or for anything which wants to sit at a Lagrangian point (which we sadly don't have - I want to put the Kerbal Space Telescope at L2 DAMMIT!). 

Either way, Ions are fine as implemented. They are not really overpowered (good luck landing on anything larger than Minmus with them) but have a substantial niche. If anything, I would want to see a slightly better physics engine to smooth out SOI transitions. 

[Archgeek]

I have to wonder, just how ionized do ion engines make their remass gas?  Do they just rip off a shell or two, or do they strip the atoms bare to the nucleus?  If the latter, things grow interesting.

Keep in mind that Isp comes from exhaust velocity, not directly from molecular -- or in the case of noble gases, atomic -- mass.  If we rip an argon atom's electrons clean off, we've got 18 protons worth of electropositvity in nearly 40 AMU, we yank that towards the the electronegative exhaust grid and out it goes, nice and fast.  'Same for xenon, but with a whopping 54 protons worth of charge in 131.3 AMU.  That does admittedly give argon a better electromotive force to mass ratio of .45 to xenon's .41, but not by very much at all.  For some reason argon's a bit neutron-heavy on average.  So, we wind up with Argon ions accelerating about 10% faster, thus for the same engine we get that much more Isp over Xenon, but a lot less thrust, around a third (30% the mass, but with the benefit of 10% more velocity, so probably around 34%).  Or, we can slightly lengthen the cavity for the xenon engine and get the same Isp by just giving the ions more time to speed up.

To make matters weirder, there's a practical limit to how much either particle can be accelerated in a given space, so for long engine cavities, the Isp approaches equivalence and Xenon maintains the thrust advantage.

Of course, that's all if ion engines tear every last electron (or at least most of them) off the poor atoms.

[MaxL_1023]

I don't think that they completely ionize the atoms - that would take a LOT of energy, more than is practical. The most important factor is first ionization energy (or the first few electrons) and molecular weight. 

Of all the noble gases which are non-radioactive and naturally occurring, Xenon is the easiest to ionize.