[0.23.5] Spherical and Toroidal Tank Pack (Updated 05/02/14) (New download link)

[Talisar]

Quick question for those of you who downloaded the v3.11pre version I posted: Did that fix the issue where using the tanks as a root part caused them to shrink when reloaded? Just want to make sure I squashed that one before I make the update official. Thanks!

[XbasmanX]

Some good news: I have found and fixed the problem causing the full tanks that were supposed to be empty. The fix will be in the next update, which will hopefully be soon as I am hot on the trail of the borked-tanks bug .

[edit]stupid censorship killing a valid description of the tanks.

Are the storage tanks supposed to be empty and tweakable so you can designate what fuels go in them?

[taniwha]

Certainly tweakable, but for empty, I don't know. Probably.

[XbasmanX]

Certainly tweakable, but for empty, I don't know. Probably.

Okay, the next question, one of your pic's in the OP shows that you can edit the tanks. I tried but when i go to the action groups screen no menu appears to allow me to select and edit fuel types. Am I missing something?

[taniwha]

XbasmanX: um, are you asking the right person? Or on the right thread? Anyway, no menu appears. Instead, just left-click the part and a panel appears over in the action groups panel.

[Talisar]

Are the storage tanks supposed to be empty and tweakable so you can designate what fuels go in them?

The storage tanks are not set up with any contents by default. You have to open the action group menu and select the part as taniwha said in order to bring up the menu which allows you to add the resources you want to that particular container.

Depending on the resource you select, the tank may automatically fill or be empty. The TAC Life suport consumable resources such as oxygen and water should auto-fill, while waste resources should be empty. Additionally, the EL resources will all start out empty with the exception of RocketParts.

[XbasmanX]

The storage tanks are not set up with any contents by default. You have to open the action group menu and select the part as taniwha said in order to bring up the menu which allows you to add the resources you want to that particular container.

Depending on the resource you select, the tank may automatically fill or be empty. The TAC Life suport consumable resources such as oxygen and water should auto-fill, while waste resources should be empty. Additionally, the EL resources will all start out empty with the exception of RocketParts.

Ok I have tried that and the menu does not appear.

just like the pic below

i left click and nothing opens

[NathanKell]

Did you install Modular Fuel tanks?

[XbasmanX]

Did you install Modular Fuel tanks?

That was it TY NathanKell

[katalex-3]

Ok I have tried that and the menu does not appear.

Insert tank as second part.

[legolegs]

Hello. I am pretty much sure those nice looking parts are not properly balanced. Spherical tanks are tiny bit OP and toroidal tanks are TOO MUCH HEAVY.

I suggest to change masses to some more sane values. Here I did some math about existing vanilla tanks and default settings from Talisar:

[table=width: 500, class: grid]

[tr][td]tank[/td][td]mass[/td][td]fuel+oxy mass[/td][td]dry mass[/td][td]q[/td][/tr]

[tr][td]oscar[/td][td]0.079[/td][td]0.064[/td][td]0.015[/td][td]0.235[/td][/tr]

[tr][td]round 8[/td][td]0.136[/td][td]0.111[/td][td]0.025[/td][td]0.225[/td][/tr]

[tr][td]fl-t100[/td][td]0.563[/td][td]0.5[/td][td]0.063[/td][td]0.126[/td][/tr]

[tr][td]fl-t800[/td][td]4.5[/td][td]4[/td][td]0.5[/td][td]0.125[/td][/tr]

[tr][td]rockomax x200[/td][td]9[/td][td]8[/td][td]1[/td][td]0.125[/td][/tr]

[tr][td]TO-S-1[/td][td]1.701[/td][td]0.201[/td][td]1.5[/td][td]7.463[/td][/tr]

[tr][td]TO-M-1[/td][td]6.892[/td][td]3.142[/td][td]3.75[/td][td]1.194[/td][/tr]

[tr][td]HSP-S-1[/td][td]3.856[/td][td]3.556[/td][td]0.3[/td][td]0.084[/td][/tr]

[tr][td]SP-S-1[/td][td]7.613[/td][td]7.113[/td][td]0.5[/td][td]0.07[/td][/tr]

[/table]

The q = dry mass / fuel+oxy mass and means "how bad tank is". Toroidal tanks are awful, the small one is pure dead weight.

My proposal is to change dry mass of toroidal (and maybe spherical) tanks (see the 4th col)

[table=width: 500, class: grid]

[tr][td]tank[/td][td]mass[/td][td]fuel+oxy mass[/td][td]dry mass[/td][td]q[/td][/tr]

[tr][td]TO-S-1[/td][td]0.225[/td][td]0.2[/td][td]0.025[/td][td]0.125[/td][/tr]

[tr][td]TO-M-1[/td][td]3.533[/td][td]3.14[/td][td]0.393[/td][td]0.125[/td][/tr]

[tr][td]HSP-S-1[/td][td]4.005[/td][td]3.56[/td][td]0.445[/td][td]0.125[/td][/tr]

[tr][td]SP-S-1[/td][td]7.999[/td][td]7.11[/td][td]0.889[/td][td]0.125[/td][/tr]

[/table]

The spherical tanks may be left as is. They are lighter than vanilla tanks but it is not a huge difference.

Hubs also need some tweaking. I believe the small hub needs to be 0.01 and medium 0.03.

[Monkeh]

Wow. That is just epic. I think you've gotten more use out of that pack in that one ship that I have altogether

I especially like the way you used the structural adapters as platforms. I never thought of anything like that, but I am totally going to steal that idea. I didn't plan on including those in the pack originally (I didn't think they would be very popular, just made them for my own use), but I forgot to move them out before zipping it up. I'm glad you get some use out of them.

Ha, only just seen your reply. Me bad at forums...

If you have any interest then please do have a look at my mission report, it's still my favourite ship and most interesting mission I think. There's a disaster included for good measure as well. Good times

http://forum.kerbalspaceprogram.com/threads/60009-The-harsh-realities-of-engineering-Home-Sweet-Home-Mk2-To-Duna-and-back

[NathanKell]

Hello. I am pretty much sure those nice looking parts are not properly balanced. Spherical tanks are tiny bit OP and toroidal tanks are TOO MUCH HEAVY.

Congratulations! You've just discovered the well-known fact that toroids are bad for fuel storage.

Consider this example. First, consider a perfect sphere of volume 1m^3. It has a radius of about 0.5m. Now imagine pressurizing it with fuel.

Now, consider a small cuboid, 10m x 10m x 0.1m. Same external volume. Fill it with fuel to the same pressure. Do you *really* think they'll both have the same structural mass? Heck, do you even think, given how thick the sides must be on the cuboid, they'll have the same internal volume?

[CleverLikeMe]

Congratulations! You've just discovered the well-known fact that toroids are bad for fuel storage.

Consider this example. First, consider a perfect sphere of volume 1m^3. It has a radius of about 0.5m. Now imagine pressurizing it with fuel.

Now, consider a small cuboid, 10m x 10m x 0.1m. Same external volume. Fill it with fuel to the same pressure. Do you *really* think they'll both have the same structural mass? Heck, do you even think, given how thick the sides must be on the cuboid, they'll have the same internal volume?

And therefore the reasons for using toroids would most likely be aesthetic or, perhaps, due to some kind of structural limitation. Otherwise you would use something else.

[Surefoot]

They do make sense from a structural point of view, but in-game this is quite limited indeed.

[taniwha]

torus: useful for when you need to get your volume away from the center for some reason (eg, artificial gravity). Otherwise, eat the abysmal mass ratio.

sphere: fantastic for maximizing usable volume while minimizing dead mass. Lousy for packing.

cylinder: Not too shabby for mass ratio, only the ends are a worry for strength and rounding helps. Pack moderately well and can be streamlined.

box: Fantastic for packing (doesn't get better). Not so good on the mass ratio, but that depends on structural needs.

cone: just a cylinder with size issues. However, doesn't pack well.

[legolegs]

Congratulations! You've just discovered the well-known fact that toroids are bad for fuel storage.

You have a point. However it is hard to believe someone makes a tank which have SEVEN TIMES MORE METAL THAN FUEL. Even 1:1 ratio seems silly.

[RoboRay]

You have a point. However it is hard to believe someone makes a tank which have SEVEN TIMES MORE METAL THAN FUEL. Even 1:1 ratio seems silly.

How many toroidal fuel tanks do you encounter in real-life?

They do exist, for niche purposes where their form-factor is more important than their efficiency in materials required. But they are quite rare.

[legolegs]

How many toroidal fuel tanks do you encounter in real-life?

There are propane tanks for automobiles:first link from search engine Yes, they are heavy (dry mass / contents = 0.86) for reasons: 1) liquid propane has low density, 2) the contents is pressurized, 3) the tank must sustain car accidents and 4) actually nobody cares about weight because cars do not fly into space. And super advanced space tanks made from superkerbal metal are heavier than cheap ones humans use on the ground? Really?

Also: when toroidal tanks are better than spherical ones? When height is restricted.

[RoboRay]

Right. That's one of the niche examples I mentioned.

[taniwha]

Legolegs: You might be surprised.

The sphere and the torus shown below both have the same exterior volume (115.56m³). The sphere's interior is 115.09m³ and the torus's interior volume is 114.74m³ (I gave each a skin 0.5cm thick).

Now we get to the important part: the volumes of the skins are 0.573m³ for the sphere and 0.923m³. If we use iron (density 7.8t/m³), that gives the tanks masses of 4.47t for the sphere and 7.20t for the torus. Since the density of LFO is 1t/m³, this gives mass ratios of 115.08/4.47 = 25.74 (sphere) vs 115.75/7.20 = 15.94 (torus).

Note that a torus is inherently weaker to internal pressure than is a sphere and thus to hold the same contents must have a thicker skin.

The tanks in this post have little intended relation to the tanks in this mod, though the torus has a major radius of 3.75m and a minor radius of 1.25m. The sphere's radius was calculated to hold the same volume as the torus.

[legolegs]

taniwha, thanks for supporting my point Your tanks have q = dry mass / fuel = 0.04 and 0.06 and thus both are way better than all of KSP tanks. Imagine if your toroidal tanks had dry mass 863 tons which would happened if it had dry mass / fuel ratio of 7.46, just like TO-S-1 does!

[taniwha]

Ok, here are some real stats, taken from EL's (not yet published) display panel and fed through a simple python script:

Name    Wet      Dry   Ratio
TO-XL-1 52.4478, 9.0,  5.8275333
TO-L-1  18.873,  6.0,  3.1455
TO-M-1   6.8926, 3.75, 1.8380267
TO-S-1   1.7009, 1.5,  1.1339333

• don't forget that those arms from the torus ring to the center have mass too.
  
• (this might be what gets you upset): it looks like the skin thickness does not get progressively thinner for the smaller tanks. I believe we decided the smaller tanks might be too flimsy, taking into account pressurization.
  

That said, looking at the four tanks set against each other, the balancing may need some more work.

Edited January 11, 2014 by taniwha

[legolegs]

I still do not understand why you're using wet/dry ratio instead of more illustrative contents/dry ratio, but whatever, let it be.

Do not forget the old good toroidal tank from vanilla KSP:

Name    Wet      Dry   Ratio
round 8  0.136, 0.025,  5.44

Of course larger tanks must be more efficient because they have better surface/volume ratio while material thickness remains the same. Thats why I expected toroidal tanks to be not worse than Round 8.

Do not forget that most tanks in game also carry all weight of upper stages of rocket even under acceleration while toroidal tanks only carry up themselves.

As a summary:

1) real life tanks have better dry/contents ratio than TO-M-1 and TO-S-1

2) theoretical calculation performed by yourself predicts better dry/contents ratio

3) Existing tank in KSP has better dry/contents ratio than all tanks except TO-XL-1

taniwha, why are you still disagree with me? Did I spoke unclear (which may be case because English is not my native language) or did I some miscalculations?

[taniwha]

taniwha, why are you still disagree with me? Did I spoke unclear (which may be case because English is not my native language) or did I some miscalculations?

Actually, after doing that post, I decided I really need to take another look at the toruses. Various things just did not make much sense to me.

As for wet/dry vs dry/contents: delta-v. The rocket equation is all about the wet/dry mass ratio, so that's what we care about. Neither is more illustrative of the other, they're just useful in different contexts.

3) Existing tank in KSP has better dry/contents ratio than all tanks except TO-XL-1

Not quite true. Even the TO-XL-1 is abysmal compared to even stock tanks: 5.44 (TO-XL-1) vs 9 (most stock LFO tanks). The spherical tanks in this mod are 15.222. (note: all wet/dry)

2) theoretical calculation performed by yourself predicts better dry/contents ratio

No argument there. I'm wondering what went wrong. They probably should be similar to cylinders except any dead-weight. I am still poking at this.

1) real life tanks have better dry/contents ratio than TO-M-1 and TO-S-1

is there a real-life tank that doesn't have a better ratio than any KSP tank (except in RF)?

Of course larger tanks must be more efficient because they have better surface/volume ratio while material thickness remains the same. Thats why I expected toroidal tanks to be not worse than Round 8.

This, actually, is dead wrong. Here's how it works for a spherical tank...

You have a relative internal pressure (unavoidable) of P and a material tensile strength of ÃÆ’, both of which are in Pascals. To work out the skin thickness required to prevent the tank from bursting, work out the force required to keep two hemisphere shells together. Integrating the force resulting from the internal pressure that is perpendicular to the plane between the two halves produces a net force F_p=P2Àr_i² (r_i is the internal radius, 2 because we have two halves, not just one). The force our material can provide comes from the tensile strength and the area of the intersection of the sphere with the plane cutting it: F_ÃÆ’=ÃÀ(r_o²-r_i²) (r_o is the outer radius of the sphere).The skin thickness is Ä=r_o-r_i. Doing a bunch of cancellation and rearranging, we get P/ÃÆ’=2(Ä/r_i)+(IJ/r_i²). For large ratios between r_i and Ä, the squared term can be ignored.

The equations again for visiblity:

F_p=P2Àr_i²

F_ÃÆ’=ÃÀ(r_o²-r_i²)

Ä=r_o-r_i

P/ÃÆ’=2(Ä/r_i)+(IJ/r_i²)

I won't show all of my work because typing all that in is a RPITA .

So in the end, the skin thickness is proportional to the radius, and with that, the mass of the tank is proportional to its volume. Thus, the mass ratio of the tank and the size of the tank have no relation.

Also, I just remembered that I worked out that the mass ratio of a toroidal tank of fixed minor radius is independent of its major radius (I was rather surprised by that). Add to that that the mass ratio of a small segment of a toroidal tank will be the same as for the full tank (minus end caps, but I'll neglect those for the moment). Then consider that a cylinder is just a small segment of a torus with a finite minor radius and an infinite major radius. This means that, ignoring end caps (thus we're dealing with pipes), a cylinder and a torus segment of the same (minor) radius and same "volume" will have the same mass ratio. Putting on end caps, they're still the same because the radii are the same. Until you go to a full torus. Then the torus wins because it had no end caps.

However, Talisar's toroidal tanks have extra support material: the arms and the inner ring. They will have more mass than the end caps if you don't want them to be as flimsy as paper. Their mass is about 1/4 that of the torus itself (assuming the same material thickness all round).

Also, it turns out that the jumbo-64 is about right at 4t assuming a skin thickness of 9mm and being made of iron (and a few other not too unreasonable assumptions).

I will now go back to the drawing board and rework the torus masses and volumes (even that looks a little iffy to me).