Bay 12 Games Forum

Yes, but what I'm getting at is a pump at one end pushing just enough to get 6/7 at the first wheel. The problem then with that is that by the 4th one, there's not enough water. But if you have a few extra points just adding 1/7, between wheels, you can delay the inevitable loss, and thus get more power for one pump (and a bit of space)

A sprinkler system?  Interesting.  All the mist won't be nice to your FPS though, I don't imagine.

What about a square instead of a vertical water return system?

1 pump. 1 tile wide channel. Huge square loop. 50+ waterwheels.

Code: [Select]


WWWW.WWWW
W*.....*W
W.......W
.........
W.......>
W*......p
W.......P
.........
W.......W
W*.....*W
WWWW.WWWW

Obviously you will need to build some walls to prevent spillage from the pump. The pump pumps water clockwise around the loop. This design should be able to be scaled up almost indefinitely. As the water is only traveling along a 1 tile wide channel and only 1 pump is involved there should be minimal FPS death.

That is a larger-scale implementation of the Loop-Reactor, and it should work fine...except for its unexpected failure*.  I urge you to build one to see for yourself.  Working with the thing itself is what made me realize that my design was flawed.  To the extent that the system is pressurized, then the upstream wheels will fail.  To the extent that the water-level is low enough to avoid pressurization entirely, the downstream wheels will be dry.  Trying to create a "perfect" water level that balances between these two is micromanagement hell.

*failure being defined as failing to deliver constant power from all wheels

However, I have been able to make a 3-wheel loop which generates a steady 300 power:
Z=0
......      '.' is ground, 'wWw' is a wheel, '|' is an axle
.wWw..    '*' is a gear assembly, '#' is walls, 'v' is an open channel
.w|w..    'pP' is a pump that pumps from the west
.W*W..
.w|w..
.v*###
.vpP7# <--Upper reservoir (note that it is at 7/7 water, which limits the pumps activity.)
...###    (when water drains from the reservoir into the channel it drops to 6/7, and 1 unit of water is pumped in)

Z=-1
######
#667## <-- Numbers indicate fluid depth (this is a snapshot of it during operation)
#6#6##
#4#7##
#5#7##
#4#7## <-- Pressure-break from reservoir to channel
#3##7# <-- Bottom of reservoir
######    (the size of the reservoir can be used to adjust the water-level in your channel, if you filled the channel to the brim)
I hope that diagram is decipherable.

Channeling into a solid rock wall results in 2 z-levels of stone getting removed at once.  Dangerous if you're dealing with fluids.  Either dig into the wall directly (but make sure your miner has an easy escape route) or try the "smooth and carve fortifications" method.  Both of these involve putting dwarves in harm's way, it's true, but in general dwarves can run off faster than magma can ooze.

Alternately, construct 2 walls on either side of the site you want to dig out, and then build a magma-safe door (alunite, basalt, or obsidian all work) on the spot that your miner would stand on to do the digging.  Order the digging, and your miner will stand in the doorway to do his work and then when done, will head off and slam the door shut, trapping the magma.  You can then hook the door up to a lever and now you have a controllable magma supply.

I've found that to save on FPS you want as few pumps as possible. Also you want the water to move in a narrow path as possible. The typical water reactor complex is hideously expensive in terms of CPU power if you build it to any significant size.

A boring, but FPS friendly design would be to have linear water reactors.

1 pump, a line of 10+ waterwheels, 1 tile wide channel. At the end of the linear reactor have it drops down one level to feed back into the pump. Not all that compact, but it should be FPS friendly.

I was of this opinion until I tested it for myself.  It turns out that that DOESN'T actually work.  I was sorely disappointed to learn this.  Pressure is the enemy of the water-wheel.  All of the wheels that have 7/7 water under them will produce no power, since the water teleports past them without actually traversing the intermediate distance.

Optimally functioning water channels actually require a careful balance between 2 factors: 
-If the water is at 7/7, then the wheel will be "bypassed" and give no power.
-If the water is less than 4/7, then the wheel won't receive any power due to low water levels.

If the flow is too high, the "upstream" wheels get bypassed, and if too little, the downstream ones get nothing.  Experimentation though has given me the solution to this dilemma:  Pump your water into a reservoir which feeds your stream through a diagonal pressure-break.  You can control the size of this reservoir to adjust how much water is in your channel.  I've built one prototype of this based on the design posted by Schmlok on page 2, but with out the sunken pump.

If you don't want a reactor but instead want a force amplifier do not connect the water wheel to the pump so that it will not self power. It will only run when a dwarf is actively pumping the water through the wheels. If you want it shut off you can instantly order the pump to no longer be operated manually. Essentially its just a lever. It allows a single dwarf to power more than one pump at a time.

I never thought of that!  That's a good idea.  You can turn large amounts of power on and off without having to muck around with levers and slow-installing mechanisms.


In unrelated news though, my research efforts have caused me to stumble across a most peculiar form of water reactor.  It consists of a pump, a wheel, and a 2-tile ditch, and creates power without actually moving any water.  Most peculiar indeed.  It will get its own thread when I can reliably reproduce it.  I think water-level might be important, as it is turning out to be in all my designs.

Wow, so a single bolt of cloth is like a lifetime supply for a hospital?  I personally think it would be more consistent with DF's handling of the relationship between materials and finished goods if bandaging used up the entire unit of cloth, given that crafting a rope-reed amulet does.  I mean it's not like cloth is scarce or hard to come by.

How much cloth is used in each bandaging?

Though "Sober Fortress" could certainly be an interesting challenge.

Try building a raised platform for the well to go on.  That might not help, but it would be interesting.  Or if there's a deep part in the middle of the lake, you could construct a platform out to there and put your well at the end of it.  You could also fish off your pier.

There are 3 kinds of water:
-Water which is a fluid.  This has depth and can leave mud.
-Water which is a contaminant.  You can have puddles and smears of water in a full cistern.  Melted ice boulders make these.
-Water which is an item.  This is the stuff that fills buckets and flasks. 1 unit of fluid water will yield a stack of 10 waters in a bucket.

With the exception of buckets, there is no way to convert between these kinds of water, and they all behave completely differently.

Now, for some useful advice:  If you can cause some glacial ice to collapse down into a non-ice level below it, then it will melt, giving you some water without the need for any machinery.