The single tile dip loader is news to me, it is NS-track on the ramp and a ramp before to ensure proper dipping? How does it get slow enough to load or is it bouncing one minecart out of the dip as well? I had a straight ramp two tile dip that worked fine with pushes, but i did not get the speed right for automation.
The main trick is that it's always pushing a filled cart out of the moat. Track on the ramp inside the trench is NS, i.e. it's a ramp that's normally traversible up/down but doesn't provide acceleration. So what happens is:
1. full cart sits in the trench, empty cart arrives, goes over the impulse ramp just north of the trench, pushes cart in the trench.
2. cart inside the trench leaves south, at ~24% of the pushing cart's speed, due to momentum-conservation rules. it probably still gets slowed down by magma upon leaving or somesuch - the cart takes 20 steps to cross the tile just outside the trench, before hitting the impulse ramp that gets it up to speed.
3. Since the empty cart is still standing on a ramp, it accelerates _slightly_, rolls into the magma-filled moat and reliably stops there. Since the trench tile is a pseudo-flat tile behind a ramp, it's subject to checkpoint effect, i.e.
a) the cart stops at the very border of the tile and needs only the tiniest push to move on to the next tile (outside of the trench)
b) the cart is charged the checkpoint speed "toll", one step's worth of ramp acceleration opposed to the last bit of ramp acceleration gained. I.e. even on a completely empty tile, the cart would stop, because it only got one step's worth of push on the last turn, and the checkpoint effect neutralises exactly that much.
4. Since the cart is "slow enough" while sitting in a sufficiently magma-filled tile, it is filled with magma.
5. the next empty cart arrives and pushes the cart in the trench.
Carts rolling freely through a water-filled tile like that with low enough input speed ("high" speed roller or less) will fill by themselves. I tried it with magma, too, but it seems that magma's higher friction prevents this from working: in my tests, the cart either passed without picking up magma or stopped in the liquid. The push setup appears to work flawlessly.
I first stumbled upon this way of getting liquid into minecarts in
this thread. The link's mainly for historical interest, i don't think the explanation works that well.
My explorations suggest that
- carts pick up liquid when moving slowly enough _or_ standing under sufficiently deep liquid. _Maximum_ speed appears to be 10 000.
- it seems that they need to be in contact with the floor - carts _falling_ into or through liquid don't fill before they reached the floor
- one-wide trenches are an effective way of filling carts with liquid because they're "incomplete" ramps, providing no acceleration; all you need is the proper speed to deal with the resistance put up by the liquid itself
- dwarven pushes give a speed that's compatible with free-passage filling in 7/7 water, but usually not in 6/7 - 6/7 water has a friction of ~600, 7/7 of ~10 000; only the full tile can slow down the cart enough to get under the speed limit for picking up fluid
To verify that carts really need to be in contact with a surface to fill up, i hurled another wooden cart into the sea. Interestingly enough, a cart stacked atop another is apparently "on floor" enough to pick up water. Currently, there are five minecarts on the ocean floor, 80-130 tiles from land. Oh, and a stepladder; unless i'm misremembering this, flying stepladders also bounce off the water when coming in low enough.
By the by - 6/7 liquid _is_ enough for pickup; just had a cart roll into 6/7 magma, on the next turn the tile changed to 4/7 with a full cart. Problem is presumably that 6/7 provides too little friction and doesn't slow down free-passing carts enough.
Author
Topic: What's going on in your fort? (Read 6223487 times)