Fuel Tank Mass Ratio - Something seems off

[CorBlimey]

For comparison:

[TABLE=width: 466]

[TR]

[TD]Family[/TD]

[TD]Wet[/TD]

[TD]Dry[/TD]

[TD]Mass Ratio[/TD]

[TD]Cost per fuel tonne[/TD]

[/TR]

[TR]

[TD]Mk3[/TD]

[TD=align: right]28.0[/TD]

[TD=align: right]3.0[/TD]

[TD=align: right]9.3[/TD]

[TD=align: right]600.0[/TD]

[/TR]

[TR]

[TD]Kerbodyne extra large[/TD]

[TD=align: right]82.0[/TD]

[TD=align: right]10.0[/TD]

[TD=align: right]8.2[/TD]

[TD=align: right]316.7[/TD]

[/TR]

[TR]

[TD]Rockomax (aka Jumbo Oranges)[/TD]

[TD=align: right]36.0[/TD]

[TD=align: right]4.0[/TD]

[TD=align: right]9.0[/TD]

[TD=align: right]400.0[/TD]

[/TR]

[/TABLE]

Cylindrical tank volume = height * pi * r^2

Thus, with the largest cross-sectional area, the Kerbodyne extra larges have the lowest surface area per volume and so should (assuming equal skin thickness or material density - and we shouldn't because the Mk3's have a higher crash tolerance) have the least massive structure to contain a given volume of fuel.

The only benefit (and I admit it is a big one for very large rockets), is that the Kerbodynes need less height for the same fuel.

Anyone else think that the Mk3's should be bumped down to perhaps 8.8 to reflect crash tolerance, and the Kerbodyne's bumped up to 9 or 9.2ish?

edit: ok, added in cost per fuel tonne. Kerbodynes are substantially cheaper and the Mk3's are substantially more expensive (perhaps to reflect the superior tank materials and the premium SSTO's place on mass ratio over upfront cost). Perhaps I withdraw my argument, or repitch it to a slight increase to 8.5-9 for Kerbodynes and the others left the same. What a pointless post I made!

Mods, feel free to call me an idiot and lock this should you agree

Edited February 22, 2015 by CorBlimey

[Renegrade]

Well, the mass ratios are all over the place in stock right now. Given that a big rebalance is coming in 1.0, I wouldn't read too much into the weirdness. (Don't forget that the mk2 parts are significantly worse than the rest, if I recall correctly..)

I'm hoping that Squad is paying attention though and carefully considering the tank ratios in the big rebalance..

[Jouni]

I like the idea that the 3.75 fuel tanks have worse mass ratios than the standard 1.25 m and 2.5 m fuel tanks. Maybe the kerbals had to switch to stronger and heavier materials or to a more robust design, because the standard design couldn't bear the weight of tall rocket stacks.

[Pecan]

It's worse than you think if you look at some of the smaller tanks (I'll leave you to find out which). Subtracting empty from full mass and cost it turns out that sometimes the fuel itself has a different mass and cost, it's not just the tanks.

[Renegrade]

Maybe the kerbals had to switch to stronger and heavier materials or to a more robust design, because the standard design couldn't bear the weight of tall rocket stacks.

This is a good point..

It's worse than you think if you look at some of the smaller tanks (I'll leave you to find out which). Subtracting empty from full mass and cost it turns out that sometimes the fuel itself has a different mass and cost, it's not just the tanks.

What? I thought fuel had a defined cost?

RESOURCE_DEFINITION
{
  name = LiquidFuel
  density = 0.005
  unitCost = 0.8
  flowMode = STACK_PRIORITY_SEARCH
  transfer = PUMP
  isTweakable = true
}
RESOURCE_DEFINITION
{
  name = Oxidizer
  density = 0.005
  unitCost = 0.18
  flowMode = STACK_PRIORITY_SEARCH
  transfer = PUMP
  isTweakable = true
}

the unitCost is the price I believe. Oh wait - if I subtract the fuel from the tank, the tank's cost is reduced by that much .. but if I look at the config files, the cost of the tank you see on the screen is actually listed in the part itself (ie an FL-T100 is 250 space bucks, and if you take the fuel out, it's 204. . but it's .cfg file says it's 250... so Squad must be subtracting the price of the fuel on the fly when loading the parts or somesuch)... I guess some tanks hold more fuel value than they're worth?

[cantab]

Considering the performance problems the game has, anything that encourages the player to use more smaller parts is bad. With that in mind, there's a case for the rebalance making the mass ratios at least equal, if not better for the wider tanks. Though the differences should be fairly subtle, for example an 8.5-9.5 range, not a 5-25 range!

[NathanKell]

Fuel tanks are pressure vessels so the cube-square law does not apply. I would expect the NASA parts to have their dry mass decreased and the Mk3s' increased.

[Bill Phil]

I would say that the Kerbodyne tanks should have a better ratio than the Rockomax tanks, as per the norm.

[NathanKell]

What norm is that? Dry mass scales pretty linearly with volume, for pressure vessels.

[Bill Phil]

Typically, larger tanks require not much more mass in terms of support structure. The effect is more noticeable at larger scales. So, a larger tank is a little better than a smaller one. By like 5%, or something like that.

[Whackjob]

You are not a fool, OP. Data is always data. The important thing is to base your view of the world on the data. Now, if you'd said, "Ok, the numbers don't bear out what I'm saying, but something IS STILL OFF!" then I'd call you a fool.

No. You went with the data. Not a fool.

[Renegrade]

Fuel tanks are pressure vessels so the cube-square law does not apply. I would expect the NASA parts to have their dry mass decreased and the Mk3s' increased.

Noooo~ the Square-Cube law is love, the Square-Cube law is life

Actually it looks like you're right there - it basically boils down to M = 3/2P*V for a given material for a spherical tank and uhhhh..something very close for a cylinder with rounded ends. Although I think an extra R is sneaking in. Possibly. I'm tired...

[PB666]

I like the idea that the 3.75 fuel tanks have worse mass ratios than the standard 1.25 m and 2.5 m fuel tanks. Maybe the kerbals had to switch to stronger and heavier materials or to a more robust design, because the standard design couldn't bear the weight of tall rocket stacks.

The mass ratios on the larger tanks don't really have as much impact as the smaller tanks because they are only inflight during boost phase. In addition during post orbital phase there are small engines on bigger tanks which means the tanks mass is more of a critical factor as the tank empties.

I use the MK1 fuel tank and use the 400L tank ratios on everything.

I have scale "400", "200" for:

0.5, .75, 1.0, 1.5, 2, 3, 4, 6, 8 (0.625, 0.875, 1.25, 1.75, 2.5, 3.5, 5.0, 7.0, 10.0)

I don't use the stock tanks at all. (ugly)

[Pecan]

What? I thought fuel had a defined cost?

Since you ask then ...

The difference between an empty and full tank must be the cost and mass of the fuel, since that's all that changes.

In nearly all cases the fuel masses 0.005t/unit. The cost per unit varies a small amount between different tanks - perhaps some are harder to fill than others?!

In an Oscar B, however, each unit of fuel masses 0.00503t, in an FL-T100 0.00501t and in an X200-32 only 0.00497t. Hard to justify that, unless there's some handwavium-leakage or something.

[Renegrade]

Since you ask then ...

The difference between an empty and full tank must be the cost and mass of the fuel, since that's all that changes.

In nearly all cases the fuel masses 0.005t/unit. The cost per unit varies a small amount between different tanks - perhaps some are harder to fill than others?!

In an Oscar B, however, each unit of fuel masses 0.00503t, in an FL-T100 0.00501t and in an X200-32 only 0.00497t. Hard to justify that, unless there's some handwavium-leakage or something.

Bur...how can you tell the numbers so precisely? The difference I'm getting for LF value in an Oscar is about 5, or 6 with the oxidizer..

[Red Iron Crown]

Since you ask then ...

The difference between an empty and full tank must be the cost and mass of the fuel, since that's all that changes.

In nearly all cases the fuel masses 0.005t/unit. The cost per unit varies a small amount between different tanks - perhaps some are harder to fill than others?!

In an Oscar B, however, each unit of fuel masses 0.00503t, in an FL-T100 0.00501t and in an X200-32 only 0.00497t. Hard to justify that, unless there's some handwavium-leakage or something.

Really? I never noticed that before, how did you determine it? None of the tools I use for calculation have that sort of precision for mass.

Edit: Reneninja'd.

[cantab]

In nearly all cases the fuel masses 0.005t/unit. The cost per unit varies a small amount between different tanks - perhaps some are harder to fill than others?!

In an Oscar B, however, each unit of fuel masses 0.00503t, in an FL-T100 0.00501t and in an X200-32 only 0.00497t. Hard to justify that, unless there's some handwavium-leakage or something.

That's less than one percent difference. It seems blindingly obvious to me that it's just a rounding error. Quite possibly not even in the physical values but merely in the numbers shown in the user interface.

[Renegrade]

Oh wait, that was mass, not cost. Still, those are still very precise numbers. even KER just says 64kg fuel total (29+35kg of fuel/oxy).

The actual numbers for the Oscar should be (according to it's cfg):

Dry mass: 15kg (0.015)

Wet mass: 78.675kg (0.078675)

KER says dry 15kg, wet 79kg. KSP says dry 0.0t, wet 0.1t

(I find it interesting that the fuel value is subtracted from the tank value at runtime though in terms of money though. Means that the karb SRBs have a negative value )

[Pecan]

Just (full mass - empty mass) / fuel units. Yes, it's a small difference and probably just rounding, but it's another reason apart from part-count not to use the small tanks if you can avoid them.

[NecroBones]

No, the units are meant to be exactly 0.005 tons/unit. Anything else is rounding error. You can find the definitions yourself:

GameData/Squad/Resources/ResourcesGeneric.cfg:

RESOURCE_DEFINITION
{
  name = LiquidFuel
  density = 0.005
  unitCost = 0.8
  flowMode = STACK_PRIORITY_SEARCH
  transfer = PUMP
  isTweakable = true
}
RESOURCE_DEFINITION
{
  name = Oxidizer
  density = 0.005
  unitCost = 0.18
  flowMode = STACK_PRIORITY_SEARCH
  transfer = PUMP
  isTweakable = true
}

[Renegrade]

Just (full mass - empty mass) / fuel units. Yes, it's a small difference and probably just rounding, but it's another reason apart from part-count not to use the small tanks if you can avoid them.

But..where are you getting such precise values? I ran some myself, but the values were very far off, rather than bit, but the tools I have are a little coarse-numbered... Oh, I have an idea - I could build a probe or such and run the delta-v numbers backwards!

[Gambi]

It is not really breaking news that most of the parts are unbalanced.

And I'm not even talking about physicless parts....

[WuphonsReach]

I'm hoping that Squad is paying attention though and carefully considering the tank ratios in the big rebalance..

This is one of those things where it would be nice if you could upgrade tanks through further research. For instance, a 10t wet tank might have 0.5t unavoidable dry-mass and 1.0t dry mass that can be reduced through further research. So initially, you have a dry:wet mass ratio of 6.67. Research the part again and you cut that 1.0 in half to get a 1:10 ratio tank. Research it again and you have a 0.75t:10t ratio (13.33).

After the rebalance, my hope is that:

0.625m parts end up with a 1:8 dry:wet ratio, 1.25m parts end up with a 1:9 ratio, 2.5m parts end up with a 1:10 ratio, and 3.75m parts have a 1:11 ratio.

Unpainted parts should definitely have better dry:wet ratios, but not drastically better. So a painted 2.5m tank would have 1:10, but the unpainted version might be 1:11 or 1:12 ratio.

[magnemoe]

This is one of those things where it would be nice if you could upgrade tanks through further research. For instance, a 10t wet tank might have 0.5t unavoidable dry-mass and 1.0t dry mass that can be reduced through further research. So initially, you have a dry:wet mass ratio of 6.67. Research the part again and you cut that 1.0 in half to get a 1:10 ratio tank. Research it again and you have a 0.75t:10t ratio (13.33).

After the rebalance, my hope is that:

0.625m parts end up with a 1:8 dry:wet ratio, 1.25m parts end up with a 1:9 ratio, 2.5m parts end up with a 1:10 ratio, and 3.75m parts have a 1:11 ratio.

Unpainted parts should definitely have better dry:wet ratios, but not drastically better. So a painted 2.5m tank would have 1:10, but the unpainted version might be 1:11 or 1:12 ratio.

I think they saved 500 kg on the shuttle fuel tank then they did not paint it. it was far heavier than 50 ton.

[WuphonsReach]

I think they saved 500 kg on the shuttle fuel tank then they did not paint it. it was far heavier than 50 ton.

It was 272kg on a dry mass of 26-35t (depending on the tank version), that's about a 1% change. So on a 10t wet tank, with 1t of dry mass - painted version would be 1:10 ratio, unpainted would be 1:10.1 ratio.

I'd still like to see a slight improvement of +0.2 to +0.5 to the dry:wet mass ratio for unpainted.