Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Sanctume]

I am theorycrafting here.


z+2      z+1      z+0      z-1
#######  #######  ###x###  #.....#
#######  ##+.+##  ##x.x##  #.....#
7x¢#¢x7  ##+.+##  ##x.x##  #.....#
#######  ##+.+##  ##x.x##  #.....#
#######  #######  ###x###  #.....#

#= wall  +=floor  ¢=hatch  x=floodgate  .=open space  7=liquids

On z+2 are the liquid source for water and magma.
Precise 7/7 liquid controlled by flood gate. 
A hatch drops this 7/7 down to z+1. 

On z+1 is a 1x3 tile floor.  The theory here is that liquid will spread to at least 2/7 + 3+7 + 2/7, then those spill over the open space down to z+0. 

From old test where 1-tile tube of magma and water that are 1 tile apart will generate obsidian z-1 below and causes cave-in the obsidian wall to fall.

Now, the theory here is that if floodgates will support and hold forming obsidian wall(s), then it would be possible to open the floodgates and releasing supports of the obsidian wall(s).

[carldong]

Single piece of 3x3 obsidian. Main question is if you want to preserve all buildings used in construction, and if you (probably) do , how do you make the center tile.

Seems like one would best cast it it at higher z-level, and have it collect together in cave-in at the bottom. So for map-wide cavein it'd make pyramid or several pyramids.

@bloop_bleep: The upside-down J was an arbitrary example of a hanging structure more than an actual goal
Ultimately, it seems like it'd be simplest to remove scaffolding afterwards.

Beyond a bottomless pit, there's also the magma sea. Ice-based system with air biomes might be better, though ice has some tendency to not notice it should cave in, unlike obsidian, and leave behind unwalkable empty space once it melts.

As for creating the center tile, I am extending from one center edge, pouring a waterfall from an aqueduct, casting a piece precisely outside the waterfall. Now the waterfall should be extended on the three sides of the newly created obsidian tile. If I can get a dwarven state machine (or just simple levers) to pour magma precisely at those points, the obsidian should be extruded. Like this:

Code: [Select]

#: Wall
O: Obsidian
.: Open space
....  ....  .O..
#...  #O..  #OO.
....  ....  .O..

Well, let me see how long I need to breach that pesky 3-level aquifer with ring cave-in in my current embark.

[carldong]

Well, let me see how long I need to breach that pesky 3-level aquifer with ring cave-in in my current embark.

OK, miners keep channeling from underneath water and one decided to dig the ring in panic. Now 3 hours of effort has been lost.

Even with pumps set up to reduce water it is still a problem. Miners keep getting pushed around and have a hard time getting back to shore.

Not even mentioning that somehow pump operators like swimming, too.

And that no stone = no mechanism = no gear assembly = no mechanical pumping.

[Fleeting Frames]

Pumps can carry power too, yknow, so if you have logs both waterwheels and windmills could be an option. Unless, you have no axes either and visitors haven't brought any.

@Sanctume: Well, floodgates will not provide support. however I'll note that if you have water and magma meet in the center at the same time on Z+1 they'll be supported from Z+2 unless you replace the walls above with, say, floating windows.

[carldong]

I thought of using a pump stack, but then discovered a hole in the non-aquifer ring which rendered it useless. Next time I will do a 2-wide ring, but for this test I abandoned and found another 1x1 small site. With goblins around, too. And no aquifer. And DFHack reveal so my miners don't fall into cavern as they find it (sometimes they just keep.falling, I had to save scum a few times).

[Sanctume]

Pumps can carry power too, yknow, so if you have logs both waterwheels and windmills could be an option. Unless, you have no axes either and visitors haven't brought any.

@Sanctume: Well, floodgates will not provide support. however I'll note that if you have water and magma meet in the center at the same time on Z+1 they'll be supported from Z+2 unless you replace the walls above with, say, floating windows.

Ok, what if the z-1 is a retracting bridge?

z+2      z+1      z+0      z-1
#######  #######  ###x###  #rrrrr#
#######  ##+.1##  ##x.x##  #rr1rr#
7x¢#.x7  ##+.1##  ##x.x##  #rr2rr#
#######  ##+.1##  ##x.x##  #rr1rr#
#######  #######  ###x###  #rrrrr#

#= wall  +=floor  ¢=hatch  x=floodgate  .=open space  7=liquids  r=retracting bridge


So my theory here is that I drop 7/7 magma first, hopefully it spread like this. 
Then drop 7/7 water to obsidianize 1x3 on the retracting bridge. 

Question: Will un-retracting make the formed obsidian fall?

[Fleeting Frames]

What I meant is that you want the Z+2 look like this:

Code: [Select]

#######
#######
7x¢#.x7
#######
######## = not wall, but a floating window, raised bridge, etc. that blocks magma/water without providing support.

Alternatively, just have water/magma fall for 2z-levels instead of 1. Hm, that's vastly simpler actually.

In that situation you describe, when you pour water it'll form unsupported obsidian, which will then cave-in. This cavein will go straight through your bridge, destroying it like all other buildings, regardless of it being retracted or not (Yes, I know this makes no sense.)

[wierd]

It could be doable I suppose, this way:

Complex set of overlapping floors overhead, with dump zones leading to the open space at the edge for minecart dumpers. Each set of dump zones is pushed furhter and further into the volume of the room. (hard to explain, so here is cross section picture. This is a cross section over several Z levels.)

Code: [Select]

|DFFFFFFFF|
|*DFFFFFFF|
|**DFFFFFF|
|***DFFFFF|
|****DFFFF|  <--Overlapping cascade of floors with dump zones that fall into open space
|*****DFFF|
|******DFF|
|*******DF|
|********D|
|*********|<--Floodgates to inject water from calibrated reservior
|*********|
|*********| <--Build area
|*********|

....

|*********|
FFFFFFFFF <-Bottom of build area/ Floodgate leading to drain.

Large print compartment below:  Top of area has flood gates to spill in water in calculated amounts, sufficient to raise water level up one z level per "flush."

Then it is just a matter of using minecart logic to send a minecart to a specific dump zone on a specific floor above, and drop a "bomb" of magma into the build area below.  Overhangs will require the construction of "support" hardware to support the magma floor when it gets created.

[taptap]

What should work is two layers of floating floodgates and a bridge on top. It is a pain to designate because you have no maintenance access whatsoever for later changes in case you make any mistakes (you have to start with the upper layer before you remove the constructed walls on top of which it was built).

1) everything closed, magma above
2) enter pattern into 1st floodgate layer
3) open bridge above / fill 1st layer with magma
4) close bridge above
5) open complete 2nd layer -> drop magma on wet floor, entered pattern now obsidianises

if you can raise waterlevel at will (not hard) you can do some limited multi-storey printing. but when all is said and done, it remains a vast vanity project (i tested the floating floodgate layers thing, it works)

[bloop_bleep]

I'm getting a bit confused about which designs are for what the OP asked and which designs are for the general 3D printer...
What I really want to know is if magma always falls at the same speed -- if several tiles of magma start falling at exactly the same tick, will they all reach a certain level at exactly the same tick? If so, my design would allow overhangs, because then all the obsidian on a certain level gets created at the same time; thus, a cave-in would not happen.

[Fleeting Frames]

Nah. Though they do tend to fall in same pattern. Object testing arena can test this question in a matter of seconds (though it took many hours for me to read your post):

||

||

||
(That's mist, not obsidian btw)

All images taken in same step, and letting it run out generated no caveins, merely bunch of obsidian walls at the bottom and few at second level from bottom and above on that pyramid shape.

I wouldn't know if it keeps doing this in a busier fortress, too, though.

[PatrikLundell]

I'm not 100% certain "naturally" generated magma can't flow sideways, as I've had a couple of weird cases where magma appeared on ledges beside open tiles where magma forms (to fall down into the magma sea below). The ledges were in non magma pile mid level tiles (48 * 48 2 * 2* meter in game tiles), just beside the magma forming air tile belonging to a magma pipe mid level tile.

[bloop_bleep]

Ah, well. Perhaps that means that fully automated 3d-printing is not possible -- the user has to manually enter the positions of obsidian for each layer in the correct order to prevent cave-ins. Though, maybe, some sort of dwarfputing contraption can be made to make sure that the "supporting" (those that aren't a part of an overhang) tiles of obsidian are the first that are printed... I'll look into it.

EDIT: I think I've found something. In the dwarfputer, there will be a bridge and two layers of hanging floodgates like taptap described, on top of which will be a water source, and beneath which will be an equally-sized square of pressure plates. To print layer x (assuming x!=1), the following happens:

1. Load the obsidian positions for layer x into the top row of floodgates.
2. Load the obsidian positions for layer x-1 into the bottom row of floodgates.
3. Open the bridge.
4. Print the tiles corresponding to the pressure plates that were activated.
5. Close the bridge.
6. Open all bottom floodgates to flush out remaining water in the top layer.
8. Load the inverse of the obsidian positions for layer x-1 into the bottom row of floodgates (that is, close the floodgates corresponding to obsidian, and open the others.)
9. Open the bridge.
10. Again, print the tiles corresponding to the pressure plates that were activated.
11. Close the bridge.
12. Flush water out again.

Note that during the first print session, the activated pressure plates correspond to tiles that are supposed to be filled with obsidian for both layer x-1 and layer x, which therefore are tiles that are not part of an overhang in layer x. The second print session prints the rest of the tiles (inverting the second layer of floodgates was simply to make sure the same tiles aren't printed twice.)

The dwarfputer will be very complex to design completely and build, and will probably also be bigger than the construction area itself, but I think it's doable. We're dwarves, aren't we?

[Fleeting Frames]

You'll ultimately have to decide and input what you're printing in any case.

You could make the dwarfputer, but...It might be simpler to do it manually, because you're not likely to use the machine twice, considering why you'd want one (and for 1-tile adjustments later, 1 minecart holding watercart+magmacart is simpler).


At least, until boat arc, after which you could print 3D obsidian art pieces onto ships and send them off to do an art exhibition in human town.

[bloop_bleep]

It is much simpler to do it manually, if you're an elf.
I agree this whole dwarfputer thing isn't that practical. But ultimately, it is the dwarfiest thing to do. So, I'm going to go build it anyway.