Forum | Lander fuel tank size/dimension discussions

Brmj and I chatted a little about lander fuel tank size and dimension issues on IRC tonight.  Pasting the relevant parts of the discussion here because they include some hard numbers that it would be nice to have handy for future reference.

<brmj> Yeah, that would be great.
<brmj> Do we even have an oficial lander design at the moment, though?
<brmj> The latest tow both have issues.
<brmj> two*
<lmaurits> I'm not sure anything is really "official" yet.  I know the latest render is suboptimal from a fuel tank mass perspective, but if it is under our maximum mass then I'm inclined to just accept that, on account of the great flexibility that design offers.
<brmj> Fair enough.
<brmj> I do definitely like the flexibility.
<lmaurits> Me too.  I would really like to see a more principled estimate of the empty mass of the current render, though, just to make sure it is a worth-while trade off.
<brmj> Right now I'm looking at a set of four spherical tanks like the Langely Lightest. It might just compare favorably to your design with a flat tank, and it could make it less tall.
<lmaurits> Sounds great, I have no objection to us investigating alternatives like that.
<lmaurits> The thing with a set of 4 spherical tanks, though, is that because we will need different volumes of LH2 and LOX, the only way to get things into 4 identical spheres is to have each sphere have an internal bulkhead and hold both.
<lmaurits> Which is perhaps not that big a deal.
<brmj> They don't need to be identical.
<lmaurits> But if they are not, the whole lander will be unbalanced, unless they are stacked vertically.
<brmj> Just two sets of two, one for oxidiver, one for fuel.
<brmj> Opposite each other, on the axis.
<lmaurits> Like:
<lmaurits> o  O
<brmj> It will ballence, if in both directions the center is at the lander's center as a whole.
<lmaurits> O  o
<brmj> Yes.
<lmaurits> Hmm.
<lmaurits> I suppose that works just fine.
<brmj> If we want to make it behave the same for manuvering in either direction, we can even tweak that by adjusting how far from the center they are.
<brmj> Or handle it in software. Either way.
<lmaurits> True.
<lmaurits> I am quite a fan in general of pushing as much complexity as we can into software.
<lmaurits> Since software is much more amenable to parallel internet magic than hard.
<brmj> Now, any chance we have estimates for how much fuel and oxidizer we need so I can do calculations specific to our lander?
<brmj> I really suck at making engineering estimates, unfortunately.
<lmaurits> Should be able to find that, one moment…
<lmaurits> Hmm, stupidly, that figure is not in the overview document.
<lmaurits> Still, the details to reproduce it are.
<brmj> That's unfortunate.
<lmaurits> 4700 m/s delta-v for descent and ascent.
<brmj> But better than nothing.
<lmaurits> And 300 kg empty lander.
<lmaurits> 450 s Isp for LH2/LOX.
<lmaurits> Let's see…
<lmaurits> Hmm.
<lmaurits> Also need astronaut and suit mass.
<lmaurits> We've been using 75 kg astronaut and 100 kg suit.
<lmaurits> (suit estimate is pessimistic, assuming non-Webbian design)
<lmaurits> So that's 475 actual unfuelled total mass.
<lmaurits> Which gives 904 kg of propellant.
<lmaurits> Now…LOX:LH2 is 6:1?
<brmj> Depends.
<brmj> The chemistry isn't the deciding factor.
<lmaurits> I think that's optimal for vacuum?  It's what I've used before anyway.
<brmj> Not sure what's right for vacuum off the top of my head.
<lmaurits> It gives 130 kg of LH2 and 774 kg of LOX.
<lmaurits> Well, 904 kg is the total mass, anyway.
<lmaurits> You can experiment with a range of ratios.
<lmaurits> Or just find a definitive source on vacuum optimal ratios.
<brmj> Wait… Double check if the ratio was by mass, by volume, by mols or what.
<lmaurits> I think we decided it was by mass?
<brmj> If you have that readilly avlaible.
<brmj> Sanity checking it with the space shuttle right now.
<lmaurits> Ah!
<lmaurits> I have done that before, actually.
<lmaurits> Well…
<lmaurits> Assuming 6:1 by mass…
<lmaurits> And getting volume figures using Air Liquide's online converter app.
<lmaurits> I got a volume ratio that almost perfectly matched the shuttle's external tank.
<brmj> okay, cool.
<lmaurits> The volume ratios differed in the second decimal point or something like that.
<lmaurits> I was quite pleased.
<brmj> I bet.
<lmaurits> Mainly because it gave me some confidence in Air Liquide's figures.

…

<brmj> Now, to get off this topic for a moment, I did the calculations on the tank area.
<lmaurits> Sure, go for it.
<brmj> A pure cylindrical tank, with a bulkhead and a diameter of 4 meters (perhaps a bit large) is .19646 m tall and has a total wall area of 40.17879 m^2
<brmj> 4 spheres have a total area of 13.688 m^2, assuming 6 to one fuel ratio
<lmaurits> Yikes, that's a very significant difference.
<lmaurits> 4 m is a very large diameter.  I would be more inclined to try 2 m?
<brmj> For comparison, a falcon 9 payload shroud is 4.6 meters wide on the inside.
<brmj> Maybe I'll recalculate it with three meters of diameter.
<lmaurits> Our CM will only be about 2 m in diameter, so I can't really see the LL getting much larger than that.
<lmaurits> It is a shame that we are wasting so much space inside the shroud.
<brmj> 3 meters gives 24.5781 m^2
<brmj> trying two next
<lmaurits> Could you also paste the heights for 3 and 2 m?
<brmj> k
<brmj> just a sec.
<brmj> height for 3m is 0.347817 m
<brmj> height for 2m is .786066 m
<lmaurits> Hmm.
<lmaurits> .78 m is much taller than I had imagined.
<lmaurits> What was the wall area for 2m?
<brmj> Same thought here.
<brmj> 14.3638 m^2
<lmaurits> Hmm, so negligibly more than the spheres.
<lmaurits> What were the dimensions for the spheres?
<brmj> Just a sec.
<brmj> Two are 0.8643 meters in diameter, the other two are 1.196 meters in diameter
<lmaurits> Hmm.
<lmaurits> So if they were packed quite close together the overall diameter would be about 2.5 m?
<brmj> Yeah.
<brmj> Something like that.
<brmj> Depending on if they were packed around the engine or not, and how big it is if so.
<lmaurits> Hmm, the fuel section would be 1.196m high, too.  That's more than 0.78m.
<brmj> Yes, but spheres can be packed around an engine and a cylender can't
<brmj> So the total height could still end up being better.
<lmaurits> Hmm.
<brmj> hmm is right.
<lmaurits> Feels like it is actually quite a close call.
<brmj> Quite.
<lmaurits> I would probably advise holding off on making too hard of a decision until we had an estimate on the engine size.
<brmj> Makes sense.
<lmaurits> It's excellent to have all these figures finally there to look at, though.
<lmaurits> A log of the relevant sections of this chat should probably be posted to the forum for future reference.
<brmj> Note: I used the physical properties Wolfram Alpha gives for "liquid hydrogen" and "liquid oxygen". It occasionally makes boneheded assumptions when asked such things, so the density numbers I used may not match those for the states we store things in.
<brmj> Ought to be close, though.
<brmj> I hope.
<brmj> Agreed about this needing to go in the forum.
<brmj> You mind doing it? It's almost 6AM for me, and I would like to get to bed eventually, if only for a few hours.
<lmaurits> No problem.
<brmj> Thanks.