May 24th, 1998
Some years ago I came across an astounding phenomenon on a country road: a mechanical laser. This device consisted of nothing more than a corrugated drainage pipe tilted at a critical angle. Here’s how it works:
Each corrugation inside the pipe creates a tiny puddle. Let’s imagine the laser fully charged: each puddle is full to the brim created by the corrugation. The pipe has several score such puddles arranged like a ladder. Now imagine a drop of water entering the topmost puddle. It will displace one drop of water out of this puddle, which will then move to the next puddle, displacing another drop, and so on, until at last a drop emerges from the bottommost puddle and exits the drainpipe.
But there’s another factor at work: because the drainpipe is tilted fairly steeply, the drop of water brings a bit of kinetic energy with it; it falls into the the topmost puddle with some extra energy, creating a splash. That splash causes a a tiny bit more water to slosh over the top of the corrugation and into the next puddle. Thus, the second puddle receives a bit more than one drop of water. Moreover, the incoming surge of water also has some extra energy -- because the second puddle is lower than the first puddle -- and so once again slightly more water is ejected than entered. This process continues all the way down the pipe; by the time it reaches the bottom of the pipe, it is a mighty surge of water which shoots out the pipe with some energy.
But the laser is now discharged; all the puddles have lost some water and are no longer full. Water entering at the top of the drainpipe will be consumed refilling the puddles; during this recharge time, no water at all will exit the drainpipe.
Thus, this mechanical laser exhibits the following behavior: it sits quietly for a short period, with no water exiting. Then suddenly a burst of water shoots out of it, and the cycle begins anew.
This behavior requires a critical combination of drainpipe tilt and water input rate. If the water input rate increases, the lasing process will be replaced by continuous flow over the corrugations. If the water input rate decreases, then the energy with which a drop enters the drainpipe might be reduced below the critical energy needed to splash water out of the tiny puddle.
Post script: I built a working device based on these principles; here’s the description and video