Is it just me or are Liquid Fuel tank masses backwards?

[Fwiffo]

Does anyone else find it backward that several of the smaller liquid fuselage tanks give more fuel per mass than the larger ones? e.g.:

Tank	Mass (t)	Fuel	Fuel per tonne
Mk0 Liquid Fuel Fuselage	0.275	50	181.82
Mk1 Liquid Fuel Fuselage	2.25	400	177.78
Mk3 Liquid Fuel Fuselage Short	14.29	2500	174.95
Mk3 Liquid Fuel Fuselage Long	28.57	5000	175.01

I would think that generally a larger tank (which not to mention costs more research) should yield better mass efficiency, not worse.  From a gameplay perspective, this sequencing seems silly.

Edited December 3, 2015 by Fwiffo

[Fwiffo]

ps. Here are some vacuum Delta-V stats for those interested (care of KER).  The fact that #2 has more dV than #1 is a bit of a rub.

#4 shows I can get eight Mk0 fuel tanks and three decouplers for less than the mass of a Mk1 tank - it gives the same amount of fuel but 118 more Delta-V.

Edited December 3, 2015 by Fwiffo

[Streetwind]

32 minutes ago, Fwiffo said:

I would think that generally a larger tank (which not to mention costs more research) should yield better mass efficiency, not worse.  From a gameplay perspective, this sequencing seems silly.

 

Most mass fractions in KSP stay pretty static and don't actually change with size. For example, rocket fuel tanks are always 8:1, regardless of size. The only ones different are the Oscar-B and ROUND-8 tanks in their (for stock KSP) very nonstandard form factors. Unless of course that changed too and I just haven't realized it yet...

By the way, if you actually want to affect a change in the way things are done in KSP, this is the wrong forum for it. You should try Suggestions & Development Discussion.

 

Holy Saint in a camper van, how do you delete tables from quotes with this editor?!

[steuben]

At larger sizes more structure is needed to keep things from going squish. The square-cube law and all that, ex a doubling of each dimension means a factor of 8 increase in mass, and a halving of each dimension means 1/8th the mass.

there will be a point where structure vs fuel support (pipes, heating, pressure regulation, hidden snacks) dominates the mass. at low masses of  fuel you just have to keep the fuel in place. at low mass plastic wrap and a couple of zip-ties will work. but for larger masses you will need 1/8 aluminium skin and a steel frame.

 

 

[Chaos_Klaus]

The Oscar B now uses the same mass fraction as the other LOX tanks.

 

The other difference is due to the dry mass being rounded to sane numbers.

Mk1 liquid fuel fuselage weighs 0.25t and holds 2t. --> 1:8

MK0 liquid fuel fuselage weighs 0.025t and holds 0.25t.  1:10

It's a little inconvenient. But a dry mass of 0.031t looks kinda silly. The tanks in KSP obviously contain structural elements (since nobody builds cylindrical tanks). Im' fine with that.

MK3 Tanks are shaped in a completely different way. Makes sense that they don't hold that much fuel. All in all, the differences are minor.

 

Also: You picture clearly shows that it's not worth staging with 3900m/s of delta v on a Nerva. As a rule of thumb: Stage when your delta v is about 10 times the ISP of the engine. With an ISP of 800s, that means you can go up to 8000m/s until you should stage.

[Fwiffo]

@Chaos_Klaus Thanks for clarifying the reasoning.  It's interesting they didn't make them work the same way the regular rocket tanks do (i.e. identical fuel/mass ratios down the whole line).

Also, thanks for the rule of thumb (good to know!), but that picture doesn't show my whole rocket.  Incidentally this came up while I was trying to figure out how to squeeze more Delta-V out of my Nerv(s).  What I seem to be arriving at is it makes sense to replace each of my Mk1 tanks with a setup like #4 instead (which weighs in just under the Mk1, has the same amount of fuel, and benefits from being able to drop dead weight more rapidly).  Of course that's assuming you don't care about part count, rocket size, design hassle, etc. and don't mind turning your ship around every few minutes to spew out another cylinder.  Just investigating building blocks for the moment.

@Streetwind Thanks for the tip about the forums; I wasn't sure where would be best.  I wanted to see if maybe there was something I was missing before implying this is broken or needs changing.  It's not a huge deal, but in spite of all the above, I still find it a bit odd.

Edited December 3, 2015 by Fwiffo

[Streetwind]

I know what you mean. Thankfully, KSP is just so incredibly open about everything. All the stats of all the parts are defined in plain text files, easy to edit - and there are even mods that let you make changes without even touching the base files.

I make a hobby out of changing the game to match my own personal ideal (which often changes as time passes). I like it so much that I practically spend more time tweaking the game than playing it

[Hupf]

Argumenting on the KSP game mechanics side: using a number of small tanks to replace a larger one means larger part count, thus more wobble (=> more struts, even more part count+mass), higher System load and in Career mode, reaching the part count Limits on early VAB Levels sooner.

[WuphonsReach]

I used to be against all the tanks having the same mass fraction, but it does make the game less frustrating when you hook up a TL-400, then decide to slot a TL-200 in its place, or use a pair of TL-400 tanks in place of a single TL-800.

If all the tanks were of different mass fractions, it would be more interesting, but also more frustrating because you would have to do more launches to test the design change.

[Arugela]

I think some of that may be calculating the drop in weight from the loss of the fuel tank. AKA it's assuming you are doing asparagus staging by the looks of it. The rest could be aerodynamics or something else(maybe the diameter of the front of the tank at the top). I don't think it's weight. Only the odd tanks are different base mass per weight. Like MK2 and a few others. And those are like .57 vs. of .5  or equivalent.

 

When you measure tanks mass. take out the fuel weight then compare to another tank. They are generally the same. I think there are now only two tank types for liquids weighing 0.005 per unit. They equal the a factor of 0.5 or 0.57 empty, as stated above, when compared to total fuel.

 

fuel weights:

Xen: 0.0001 per unit

MP: 0.004 per unit

LF: 0.005 per unit

OX: 0.005 per unit

Ore: 0.01 per unit

Edit: Actually, in this case the weight is smaller for the lighter tank.

MK0vsMK1:

400/50=8

.275-0.25=0.025x8=0.2 Total weight if it were mk1 size.

MK1:

2.25-2=0.25

 

So it's the empty weight of the tank(Similar to what Chaos_Klaus was stating) for the first 2. And the tanks after are probably calculating for asparagus staging because of the decouplers.

 

NVM, I assume you know most of that. 8) It's getting late!

 

 

 

 

Edited December 4, 2015 by Arugela

[Fwiffo]

@Arugela Yeah I think you basically said the same thing just another way around.  For a given amount of liquid fuel, the dry mass using a bunch of smaller tanks is less than you get using larger ones.

[NathanKell]

If formfactor is kept the same, a pressure vessel (which all tanks are) will have a constant mass:volume ratio. While the FL-T200 is not the same formfactor as the FL-T800, KSP makes the simplifying assumption that that ratio is kept even then.

Square-cube law is for non-pressure-vessel containers, folks.

[Fwiffo]

Hey @NathanKell, great info, and thanks for chiming in!  But I have no qualms with the FLT series tanks.  It's the liquid-fuel-only ones which kinda surprised me, since:

which (I think?) is unlike other series tanks of tanks in the game (such as the combined fuel/oxidizer ones you pointed out), where bigger tends to give equal, or more, bang for your buck / kilogram.

So it encourages you to do crazy things like this:

Spoiler

Now, I realize the fuselages are intended primarily for pointy-end-forward vehicles, which aren't as sensitive to mass as their pointy-end-up cousins (nor as amenable to being built long-and-skinny).  And I get that the bigger tanks offer other benefits (lower part count, smaller envelope, etc) to offset their crummier fuel-to-dry-weight ratio.  But if you want to push an LV-N "Nerv" as far as it will go without it taking ages, I'm guessing a bunch of Mk0's with generous staging to rapidly shed off dead weight is the way to go...

(Side note - interestingly, a Mk1 tank weighs the same as the equivalent volume of Mk0 tanks plus a strut.  So I have a sneaky suspicion the guys at C7 Aerospace just strapped a bunch of the latter together, strutted it on the inside, and painted on some "Mk1" logos)

Edited December 4, 2015 by Fwiffo

[sal_vager]

Just a thought, but a larger tank would need more structure than a smaller one to support not only the extra mass of its contents, but also itself ...

Not to mention plumbing, pumps and associated hardware.

[Stoney3K]

On 4-12-2015 21:20:30, sal_vager said:

Just a thought, but a larger tank would need more structure than a smaller one to support not only the extra mass of its contents, but also itself ...

Not to mention plumbing, pumps and associated hardware.

Not so true for LF or O tanks, since they contain liquified gas that expands. It's not as much a tank that could crumple under its own weight because the top half is empty, a rocket fuel tank is more like a balloon.

The pressure of the fuel inside also has a structural function.

[moogoob]

On 2015-12-04, 2:51:38, Fwiffo said:

Hey @NathanKell, great info, and thanks for chiming in!  But I have no qualms with the FLT series tanks.  It's the liquid-fuel-only ones which kinda surprised me, since:

which (I think?) is unlike other series tanks of tanks in the game (such as the combined fuel/oxidizer ones you pointed out), where bigger tends to give equal, or more, bang for your buck / kilogram.

So it encourages you to do crazy things like this:

  Hide contents

Now, I realize the fuselages are intended primarily for pointy-end-forward vehicles, which aren't as sensitive to mass as their pointy-end-up cousins (nor as amenable to being built long-and-skinny).  And I get that the bigger tanks offer other benefits (lower part count, smaller envelope, etc) to offset their crummier fuel-to-dry-weight ratio.  But if you want to push an LV-N "Nerv" as far as it will go without it taking ages, I'm guessing a bunch of Mk0's with generous staging to rapidly shed off dead weight is the way to go...

(Side note - interestingly, a Mk1 tank weighs the same as the equivalent volume of Mk0 tanks plus a strut.  So I have a sneaky suspicion the guys at C7 Aerospace just strapped a bunch of the latter together, strutted it on the inside, and painted on some "Mk1" logos)

I love the "squid staging" but have one question: how much do all those decouplers weigh? If you recall history, fuel tanks with low dry mass have the benefit of being able to be held onto longer without staging it off. As an example, see the Atlas - it had a single fuel tank in its launch stage and just flung off 2/3rds of its engines when TWR got too high.

Maybe using fewer decouplers (if any) might give you higher delta-v and lower part counts? Just have arrays of static size-0 fuel tanks that all get ejected at the same time, or if decouplers are too heavy then just stick with the craft.

[Fwiffo]

9 hours ago, moogoob said:

I love the "squid staging" but have one question: how much do all those decouplers weigh? If you recall history, fuel tanks with low dry mass have the benefit of being able to be held onto longer without staging it off. As an example, see the Atlas - it had a single fuel tank in its launch stage and just flung off 2/3rds of its engines when TWR got too high.

Maybe using fewer decouplers (if any) might give you higher delta-v and lower part counts? Just have arrays of static size-0 fuel tanks that all get ejected at the same time, or if decouplers are too heavy then just stick with the craft.

Feel free to try it out and post if you get better dV out of it!

[Galinette]

On 4/12/2015 03:56:33, NathanKell said:

If formfactor is kept the same, a pressure vessel (which all tanks are) will have a constant mass:volume ratio. While the FL-T200 is not the same formfactor as the FL-T800, KSP makes the simplifying assumption that that ratio is kept even then.

Square-cube law is for non-pressure-vessel containers, folks.

Square-cube law works well for gas cylinders. It also works well for insulated liquids (hot water, cryogenic liquefied gas). The container structure (surfacic weight)  is governed by pressure in the first case, and temperature difference in the second case.

However hydrostatic pressure in large liquid containers require to quit the square-cube law, as larger containers will have more pressure to withstand at the bottom, when facing gravity and/or acceleration.