|
|
Post by wittyusername on Mar 8, 2017 5:53:16 GMT
I'm beginning to experiment with the module designer and decided to start with propellant tanks. It seems that the best possible mass ratio for any propellant can be obtained with a tank with the highest possible length/radius ratio made of UHMWPE. Also, scaling the amount of propellant does not affect the mass ratio. This discovery inspired a number of questions, such as
1. Why does a long, thin tank give a better mass ratio than a close to spherical one?
2. Shouldn't bigger tanks have better mass ratios than smaller tanks, due to the square-cube law?
3. What are some reasons not to make all propellant tanks have 20/1 length/radius and he made
of UHMWPE?
Hopefully these aren't dumb questions and someone will have some answers.
|
|
|
|
Post by The Astronomer on Mar 8, 2017 5:59:18 GMT
I'm beginning to experiment with the module designer and decided to start with propellant tanks. It seems that the best possible mass ratio for any propellant can be obtained with a tank with the highest possible length/radius ratio made of UHMWPE. Also, scaling the amount of propellant does not affect the mass ratio. This discovery inspired a number of questions, such as 1. Why does a long, thin tank give a better mass ratio than a close to spherical one? 2. Shouldn't bigger tanks have better mass ratios than smaller tanks, due to the square-cube law? 3. What are some reasons not to make all propellant tanks have 20/1 length/radius and he made of UHMWPE? Hopefully these aren't dumb questions and someone will have some answers. 1. Sadly, I have no idea about this. 2. The interior pressure increased as the tank gets bigger, I guess? 3. UHMWPE is expensive, and if your military budget is running low because you spent them on death stars, you might have to use other materials. |
|
|
|
Post by David367th on Mar 8, 2017 6:01:38 GMT
Another point on 3. Boron tends to be cheap, and 20:1 costs you weight and money if you're using armor on your ship.
|
|
|
|
Post by n2maniac on Mar 8, 2017 6:38:23 GMT
2: Pressure vessel mass (eg. fuel tank) scales proportionally to its volume * pressure (assuming other strengths don't end up needier and the walls are thin).
|
|
|
|
Post by bigbombr on Mar 8, 2017 7:26:52 GMT
Tanks try to be cylinders. Narrow cylinders have proportionally smaller flat surfaces (the end caps) which are the part of the cylinder which needs to be the thickest (though thickness is probably homogeneous). Having purposefully bulbous end caps (domes) would lighten tanks, as the end caps wouldn't need to be so thick. Or allowing a wire/wires connecting the end caps to each other would similarly help.
|
|
|
|
Post by AtomHeartDragon on Feb 11, 2018 14:52:39 GMT
Tanks try to be cylinders. Narrow cylinders have proportionally smaller flat surfaces (the end caps) which are the part of the cylinder which needs to be the thickest (though thickness is probably homogeneous). Having purposefully bulbous end caps (domes) would lighten tanks, as the end caps wouldn't need to be so thick. Or allowing a wire/wires connecting the end caps to each other would similarly help. I am still not seeing it. In-game tanks are pill shaped and from geometry it would be obvious that the closer the tank to a sphere the more efficient it is at storing its contents (barring further considerations such as wanting to then pack those tanks in a big armoured sodacan, but those don't affect tank's own mass ratio), so the closer to the spherical, the better mass ration. Unless it's something to do with slosh baffles and water hammers, elongated=effective is plain wrong. Either way warrants some detailed explanations and investigation of how things are calculated in game and why. |
|
|
|
Post by apophys on Feb 11, 2018 16:00:10 GMT
I am still not seeing it. In-game tanks are pill shaped and from geometry it would be obvious that the closer the tank to a sphere the more efficient it is at storing its contents (barring further considerations such as wanting to then pack those tanks in a big armoured sodacan, but those don't affect tank's own mass ratio), so the closer to the spherical, the better mass ration. Unless it's something to do with slosh baffles and water hammers, elongated=effective is plain wrong. It has to do with the pressure of the fluid and hoop stress of the tank. Smaller radius (i.e. more curvature) is better at withstanding internal pressure. A sphere has the most volume to surface area; this leads to it having less curvature than a cylinder of the same volume, and so it needs to be thicker to withstand the pressure. The increased thickness overwhelms the raw surface area savings and causes the sphere to have a worse mass ratio. It would be nice to have toroidal tanks; those should be right by the ultimate limit for an extended cylinder, since the end caps disappear into each other. |
|
|
|
Post by newageofpower on Feb 26, 2018 5:46:52 GMT
Tanks should be rated for a certain amount of G loading and stress; I know my MPD-only ship tanks should be fairly thin but a 50g terminal stage missile shouldn't have paper thin fuel tank walls...
|
|
|
|
Post by AtomHeartDragon on Feb 26, 2018 20:05:02 GMT
I am still not seeing it. In-game tanks are pill shaped and from geometry it would be obvious that the closer the tank to a sphere the more efficient it is at storing its contents (barring further considerations such as wanting to then pack those tanks in a big armoured sodacan, but those don't affect tank's own mass ratio), so the closer to the spherical, the better mass ration. Unless it's something to do with slosh baffles and water hammers, elongated=effective is plain wrong. It has to do with the pressure of the fluid and hoop stress of the tank. Smaller radius (i.e. more curvature) is better at withstanding internal pressure. A sphere has the most volume to surface area; this leads to it having less curvature than a cylinder of the same volume, and so it needs to be thicker to withstand the pressure. The increased thickness overwhelms the raw surface area savings and causes the sphere to have a worse mass ratio. It would be nice to have toroidal tanks; those should be right by the ultimate limit for an extended cylinder, since the end caps disappear into each other. Ok, but I would still like to see some sources - the claim is extraordinary and requires corresponding standard of evidence. Also, there has to be some cutoff point where losses due to volume to surface ratio exceed gains from reduced stresses. |
|
|
|
Post by boersgard on Aug 17, 2018 5:07:36 GMT
This is an issue I noticed quickly as well. Problem: I want a thick but short section on my spacecraft filled with fuel. First solution: Basically a spherical fuel tank. Turns out the mass of the fuel tank + fuel canceled out any delta-v gains entirely. Maybe even lost me some delta-v. Second solution: Look up the maximum circles I can pack into another circle - it turns out this is a big thing math nerds like to play with for some reason! Thanks math nerds. Success. Now I have 19 long but small radius fuel tanks where I once had 1. HUGE mass savings. I now do this for all fuel tanks. 19 is optimal (can't go higher in-game) but 7 is also good.
You will also notice that this circle packing works for anything you're trying to pack into a circle, including guns in the nose or engines, however you don't necessarily get mass savings it just lets you pack them into a smaller space.
To calculate how big the radius of your fuel tank should be, first find the radius of the larger circle you are going to pack your fuel tanks into. Next, multiply that large radius by the optimized circle packing ratio algorithm value thing (if you don't know, wolfram alpha has a circle packing algorithm which will tell you - how useful!) For example, if you intend to pack 7 fuel tanks into the bigger circle, you want each tank to be 0.333333x the radius of the radius of the larger circle. So if the big circle is 18m diameter (9r), you want 6m diameter (3r) fuel tanks. Maximize their height vs radius vs fuel ratio for the size constraints you are working with and then pack 7 fuel tanks in and voila. I honestly can't see any reason NOT to do this. Not only are the fuel tanks more mass efficient, but you have massive redundancies on account of 7 or 19 fuel tanks, making your craft much more resilient. It is, as far as I can tell, all upsides and no downsides - except perhaps in the case where you max out the ratio with fewer tanks, e.g. given a choice between 19 fuel tanks at a maximum fuel:mass ratio and 7 tanks ALSO at max fuel:mass ratio, you should save mass with the 7 tanks vs the 19. |
|
|
|
Post by gedzilla on Aug 17, 2018 6:24:41 GMT
This is an issue I noticed quickly as well. Problem: I want a thick but short section on my spacecraft filled with fuel. First solution: Basically a spherical fuel tank. Turns out the mass of the fuel tank + fuel canceled out any delta-v gains entirely. Maybe even lost me some delta-v. Second solution: Look up the maximum circles I can pack into another circle - it turns out this is a big thing math nerds like to play with for some reason! Thanks math nerds. Success. Now I have 19 long but small radius fuel tanks where I once had 1. HUGE mass savings. I now do this for all fuel tanks. 19 is optimal (can't go higher in-game) but 7 is also good.
You will also notice that this circle packing works for anything you're trying to pack into a circle, including guns in the nose or engines, however you don't necessarily get mass savings it just lets you pack them into a smaller space.
To calculate how big the radius of your fuel tank should be, first find the radius of the larger circle you are going to pack your fuel tanks into. Next, multiply that large radius by the optimized circle packing ratio algorithm value thing (if you don't know, wolfram alpha has a circle packing algorithm which will tell you - how useful!) For example, if you intend to pack 7 fuel tanks into the bigger circle, you want each tank to be 0.333333x the radius of the radius of the larger circle. So if the big circle is 18m diameter (9r), you want 6m diameter (3r) fuel tanks. Maximize their height vs radius vs fuel ratio for the size constraints you are working with and then pack 7 fuel tanks in and voila. I honestly can't see any reason NOT to do this. Not only are the fuel tanks more mass efficient, but you have massive redundancies on account of 7 or 19 fuel tanks, making your craft much more resilient. It is, as far as I can tell, all upsides and no downsides - except perhaps in the case where you max out the ratio with fewer tanks, e.g. given a choice between 19 fuel tanks at a maximum fuel:mass ratio and 7 tanks ALSO at max fuel:mass ratio, you should save mass with the 7 tanks vs the 19. does more fuel tanks require more crew ? |
|
|
|
Post by AdmiralObvious on Aug 17, 2018 7:05:02 GMT
does more fuel tanks require more crew ? Nope, not unless it pushes your ship over one of the magic cross section thresholds I forgot the specifics for. |
|
|
|
Post by bigbombr on Aug 17, 2018 8:30:58 GMT
does more fuel tanks require more crew ? Nope, not unless it pushes your ship over one of the magic cross section thresholds I forgot the specifics for. Are the thresholds based on cross section or mass? I always assumed the latter. |
|
|
|
Post by anotherfirefox on Aug 17, 2018 14:24:20 GMT
Tanks should be rated for a certain amount of G loading and stress; I know my MPD-only ship tanks should be fairly thin but a 50g terminal stage missile shouldn't have paper thin fuel tank walls... Considering real life orbital rocket boosters sometimes come in "balloon" type, which can't even stand themselves without the fuel inside, that would be feasible I guess. |
|
|
|
Post by AdmiralObvious on Aug 17, 2018 20:15:15 GMT
Nope, not unless it pushes your ship over one of the magic cross section thresholds I forgot the specifics for. Are the thresholds based on cross section or mass? I always assumed the latter. I could've sworn it was based on cross section. I've made some incredibly "dense" rockets, using modded liquid lead as a fuel for a NTR before. The crew requirement was notably small. A ship with 500 um spaced aluminium at 25 meters outside the ship does seem to need more command crew. |
|
