Loopy Pipe Network Converts back-and-forth flow into one-way flow | Hacker News Reader

Loopy Pipe Network Converts back-and-forth flow into one-way flow

physics.aps.org

103 pointssohkamyung22 comments

This seems to be a known technology. Here's a paper: "THREE-DIMENSIONAL SIMULATION OF A VALVELESS PUMP" J. Shin and H.J.Sung", Kaist (2010) [1] It's not quite the same, though. Rather than using turbulence, it uses an impedance mismatch as an obstacle to flow. The idea is to have two impedance mismatches in opposite directions from the pinch point, but at different distances. You get some reflection from each impedance mismatch. So if you choose the right driving frequency, you get the reflections opposing movement in one direction and helping it in the other.

This is a lot like the way some antennas are driven. RF also has impedance mismatches. But not turbulence. (Fortunately, or RF engineering would be really hard.)

[1] flow.kaist.ac.kr/bbs/download2.php?bo_table=pro_domestic&wr_id=93&no=0

There's also the Tesla valve, which of course isn't a pump by itself but uses similar principles

https://en.wikipedia.org/wiki/Tesla_valve

It would be interesting to see if something like this could improve efficiency in valveless pulse jets. Valveless pulse jets rely on different inertia in two air paths to pull fresh air back in one of the exhaust pipes to refill the combustion chamber.

Valveless pulse jets are loud and inefficient, but they're reliable, tolerant of manufacturing defects, and cheap to produce. Of course, acoustic effects providing the only compression above ambient pressure is going to fundamentally limit combustion pressures and temperatures, and thereby thermodynamic efficiency, but I think valveless pulse jets operate significantly below their Carnot / Brayton efficiencies.

Isn't this the same principle behind a Tesla valve?

https://www.youtube.com/watch?v=lFIYKmos3-s

I can see where you're coming from, being on it relying on differential pressures and turbulence to ensure certain flow characteristics... but then again, I wouldn't really say there was any particular unique concept at work here.

Same same but different? Good video of a Tesla valve:

https://www.google.com/search?q=youtube+%22Tesla+Valve+Expla...

There is still so much science to do with simple systems. Even if fundamental physics gets solved, there will be plenty of new discoveries to make.

I wonder if this is interesting to look at topologicaly? Flow dynamics picks out the network connectivity/handedness/etc... Maybe I am over thinking it while simultaneously under thinking it. ;)

Paging V. I. Arnold just in case.

Wonder if it works with liquids

At least some of the tests were with water, so, yes?

does this flow effect work with the water analogy of electricity?

I had to google it. It seems like it's at least plausible that it might: https://pubs.acs.org/doi/pdf/10.1021/nl070935e

This relies on inertia so probably not.

Inductance is similar to inertia in the water model of electricity.

Yes, but it's not directional. You don't get to make a T-joint of three coils and have the current follow the top bar as it's a straight line.

Diodes don't work based on inductance.

Diodes would be something along the lines of a check valve.

The circulator[1] is an analogous component in rf, it send the incident power from any port down to the next port, so you can take one port to be the bidirectional connection, and it splits the signal into transmit and receive connections.

[1] https://en.wikipedia.org/wiki/Circulator

Just realized if it did this would make a very simple rectifier. I guess you could test it by putting ends of a wire with a galvanometer on it to different parts of an antenna and see if you get a current.

Seems related to Tesla valves? Fascinating stuff!

A liquid rectifier?

Reply on news.ycombinator.com