Moonwards

« first day (398 days earlier) ← previous day next day → last day (828 days later) »

kim holder

22:00

so that is 10 m in diameter and 20 deep in a tank of 30 m diameter and 30 deep

after one round of day and night

with continued cycling, it will change

the thing is just to keep the losses down to much less than if it just went straight into the ground

see, my mind goes all over from here. if you are drawing off heat from the outside in, then you'd need to switch to different flow loops as the heat store is drained by that, moving in towards the center

Hohmannfan

I get a loss of 220 W/m² on the cylinder wall on the 5m radius setup.

kim holder

which changes as it cools, right?

and some can be reflected back in?

Hohmannfan

The code is now in the repo under /tools/. It is CLI, and basically just asks you to input numbers for various questions. Then it prints a list of the temperature in all the layers, plus the loss.

@kimholder yes, it changes with temperature

kim holder

oh, very cool :)

Hohmannfan

@kimholder Not sure. (about reflecting)

g++ -o gradient temperatureGradient.cpp && ./gradient

To compile and run it.

kim holder

22:10

that might be a big help to Mike. he has worked before with microturbines.

so, hopefully he can add his practical experience and say whether this idea works and how it should really be

i mean - i'll compile it. but i really don't know what i'd do with it : /

i'm still not sure whether all this means we should even try to make a tank, or just put in extra capacity that makes up for the heat that drains away

Hohmannfan

At the same setup, the loss is just 138 W/m² at 1200K. It is indeed lower at lower temperatures.

78 W/m² at 1000K

kim holder

that seems to argue for draining the heat for the turbines from the perimeters of the heat store

Hohmannfan

Remember that heat energy is less useful at smaller gradients though.

kim holder

even if that means a series of loops have to be created so the working fluid remains in an area of the proper temperature

Hohmannfan

Ah, preheating.

kim holder

22:17

i had thought maybe it would help to have a working fluid that uses as low a temperature as possible

Hohmannfan

Forget what I said, this is all the hot part of the loop.

It should stay hot.

kim holder

so move the loop as the volume cools, inwards to where it is still hot

and make the loop longer so it has enough time to pick up the heat it needs

and use something that will still get enough heat at the end of the night, when the store is depleted

so, a working fluid (or gas) that uses a rather low temperature

Hohmannfan

I think the loop should start where it is hot and move outwards. To make use of the countercurrent flow principle for heat exchangers:

kim holder

and maybe there is something wrong with that, but it is the idea that comes to me

Hohmannfan

kim holder

22:21

i don't understand that...

Countercurrent exchange

Countercurrent exchange is a mechanism occurring in nature and mimicked in industry and engineering, in which there is a crossover of some property, usually heat or some component, between two flowing bodies flowing in opposite directions to each other. The flowing bodies can be liquids, gases, or even solid powders, or any combination of those. For example, in a distillation column, the vapors bubble up through the downward flowing liquid while exchanging both heat and mass. The maximum amount of heat or mass transfer that can be obtained is higher with countercurrent than co-current (parallel...

let's see...

Hohmannfan

"The maximum amount of heat or mass transfer that can be obtained is higher with countercurrent than co-current (parallel) exchange because countercurrent maintains a slowly declining difference or gradient (usually temperature or concentration difference)"

kim holder

so... the heat pipe would move the heat outward in the thermal store as it works... but the heat surrounding it would remain fairly steady along the total route?

Hohmannfan

Correct.

kim holder

and this wouldn't make much difference to the amount of heat that reaches the outer walls? or does it not matter because it is more efficient?

i guess if it is taking heat from the central portion, then the 'pressure' moving the heat outwards is less

Hohmannfan

Wait, you did actually describe a countercurrent flow to start with...

kim holder

22:27

i did? :D

Hohmannfan

So yeah, inwards movement.

kim holder

but i wasn't thinking in terms of running the pipes through the middle of the mass, where it is hottest

Hohmannfan

That would probably melt them anyway. So close to the core, but not completely.

kim holder

this sounds better still. i thought we'd have to switch loops as the store loses energy.

Hohmannfan

Now the loop is actually combating the loss! Great.

kim holder

22:30

and maybe we still need to switch loops, but only to make it a little longer, not to switch to a completely different loop that is further in where the store is still hot enough. so less piping.

because of the countercurrent thing. i really hadn't thought about that, i just put the pipes far out.

i really should make a sketch, it is hard to describe in words.

Hohmannfan

BaseTemp + (PoolTemp - BaseTemp) / ℯ^(Distance / 3.6) gives an ok fit for the temperature gradient if you are in a hurry.

kim holder

i'll let Mike know :P

Hohmannfan

Another thing: A small radius of the cylinder means a larger surface area per volume, right? But there is a secondary effect as well. The loss per surface area shrinks with radius too as the heat must diffuse more slowly, making the gradient less steep.

So a large pool is good for many reasons

The bonus shrinks slowly though:

Loss per area of different lava pools by radius:
5m: 220 W/m²
10m 187 W/m²
100m 150 W/m²
1000m 147 W/m²

kim holder

22:48

the size of the second TESS on the hilltop would be the place to think about this...

by the time that is made, Cernan's Promise has plenty of infrastructure to build big things, if it wants to

Hohmannfan

even a 2km wide lava pool?

kim holder

no, the whole colony at that point is only about 4 km across

the city could try something like that

my reaction is that there would be diminishing returns for other reasons

Hohmannfan

Combining the diminishing return with the square-cube law to get the loss per volume of stored lava: (wall component only)

5m 88 W/m³
10m 37 W/m³
100m 3 W/m³

That is, if the lava has an uniform temperature. In practice, it would be hotter in the middle, giving another small boost.

kim holder

that does seem pretty significant...

Hohmannfan

tl;dr, it gets better with size in many different ways.

kim holder

22:59

go big. that's what i've been saying for months :D

oh, there is this thing i can't find right now in blender's controls that would make this look so much better

but anyways, this is sort of how i was thinking of the loops

you only run the fluid through a few of them at first, and as the night wears on, you run it through the ones in between too

but maybe they should go out farther towards the edges of the reservoir

right now this is the modest tank 30 m across and 30 deep, and the loops are 16 m wide

this is from above

then the heating pipes from the mirrors run down the very center

they wouldn't get any closer than about 4 m from the pipes for the turbines

if you loop out nearer to the edges you can make an access tube and have a decent chance to have valves that can be serviced, at least by a robot.

@Hohmannfan so, that is a lava pool that is 10m in radius?

« first day (398 days earlier) ← previous day next day → last day (828 days later) »

all rooms

Transcript for

Apr17

Apr '1719

Apr20

about moonwards.com and the Virtual Moon Colony project

colony nuclear-engines settlement space tethers the-moon

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC