Well even without the fish moving itself it wouldn't be a perpetual motion machine because the running water would be an input to the system. Agree though that a living, untethered fish would need to expend energy to stay balanced and in the right spot.
Reframe the system to the moving water and put a watermill or a hydroelectric power plant in there in there - perpetual motion and extracting energy. A flowing river is not a closed system.
They're saying that because the fish is dead it has to be tethered, otherwise this wouldn't last for more than a few seconds before the fish becomes unstable and the phenomenon stops. For a living untethered fish, some level of energy would need to be expended to keep the fish stabilized instead of the tether.
Agreed, but the tether is just to keep the fish on track, it's still no "effort" on the part of the dead fish in this experiment (effort == energy spent by the dead fish). The article does mention a living fish will spend energy in order to find the sweet spot of the current, and then the water flow will do the rest.
Also, when the dead fish "swims forward" and hits the obstacle, the tether itself is playing no part. It's 100% the water flow and the shape/flexibility of the corpse.
If I understand the article correctly, the energy expenditure by the fish itself is zero. From TFA:
> "Under just the right conditions, there’s actually a resonance between the vortices and the fish’s body that generates enough thrust to overcome the fish’s drag. This means the fish can actually swim upstream without expending any energy of its own!"
That's correct. I actually recieved a copy of this video about a decade ago from a grad student at the MIT tow tank - they were working on their Robotuna design and we were discussing Thunniform propulsion and this video came up. I've used this video in many talks because it's so cool.
In a nutshell, fish are undulating foils. When an oscillating or undulating foils is submerged in a fluid, a trailing Karmen Vortex Street (1) is generated, which is a set of spatially offset vortices. One of the cool things about that is that as the foil "swishes" from, say left to right, it extracts energy from the vortex - the foil can propel itself forward by essentially "pushing" off of the vortex of spinning fluid. The result is that the vortex rotation slows down (that's where the energy to propell forward primarily comes from).
Side note: This is in contrast to a single rotating propeller that leaves a lot of used energy in the swirling trailing wake. Modern profilers can use things like contrarotating propellers or boss cap fins to recover some of that energy.
In any event, for this "dead fish" experiment, the Karmen Vortex Street (KVS) is being generated by the obstacle in the flow in front of the fish - this is due to the low pressure zone directly behind the obstacle. The flexible foil begins to undulate in concert with these vortices. If you look at the figure of the KVS, the region in the center line of the KVS is actually creating a flow in a direction that is opposite that of the vortices themselves. In other words, there's a flow in the center that's effectively sucking the fish towards the rock.
Nothing magic, no free energy sadly, but definitely some cool science! You can absolutely use this knowledge to design energy harvesters (generators) from flows, like rivers or deep ocean currents.
Yet it makes the fish appear to be making progress in the moving water. Sure it's wriggling side to side, very neat. But keeping up with the current? Sure, if it's held there by a wire.
So maybe it expends very little effort, but not zero.