4 ms·

So I have one of these in my office. I 3d printed it and tested it. It doesn't really work, but it does give you an interesting flow curve.

brk 7y ago
IANAFDE (I am not a fluid dynamics engineer). My gut tells me that for this to work as designed, you would need to ensure the fluid moves exactly as intended, primarily smoothly and evenly. I know nothing about your 3D printer, but if the output is like most I have seen (even high-end ones), I wonder if the surface aberrations in most 3D-printed objects would be enough to impact the fluid flow such that it would be turbulent enough to overcome the intended friction of the channels? If I had the time, I'd make up a 3D model of this that I could cut on my CNC machine out of a block of aluminum or hard plastic to test.
- serf 7y ago
  from playing with 3d printed tesla valves , I think that you're right that surface distortions hinder correct flow. The way that I created a 3d printed tesla valve that worked was by messing with overall scale. Bigger works better, until suddenly it doesn't work at all. The idea I had previously about why scale may help was the idea that the scale is gradually turning down the amount that the flow error affects the entire system until eventually the mass of the fluid induces more 'error' into the system until the whole thing destabilizes. In other words : with 3d printed tesla valves it's vital to find the sweet scale spot between surface-level distortion causing chaos when the valve itself is too small and liquid mass causing chaos when the valve is too big. That margin between chaos and stability may be made wider by using a finer manufacturing technique, like milling or even SLA/SLS versus FDM. I bet that'd give a wider range of usable sizes. all that said : even professionally made tesla valves must be made with certain scale constraints in mind. I believe those constraints are first-and-foremost a property of the liquid that's going to be transported.
  - jakobegger 7y ago
    I'm pretty impressed what you figured out by just experimenting. What you are describing is the transition between laminar flow and turbulent flow. The point where this transition occurs is governed by the "Reynolds Number". As you correctly discovered, the Reynolds number is influenced by scale and fluid properties (viscosity and density). The final missing piece is the speed of the fluid, which explains your observation: "Bigger works better, until suddenly it doesn't work at all."

brk 7y ago

IANAFDE (I am not a fluid dynamics engineer). My gut tells me that for this to work as designed, you would need to ensure the fluid moves exactly as intended, primarily smoothly and evenly. I know nothing about your 3D printer, but if the output is like most I have seen (even high-end ones), I wonder if the surface aberrations in most 3D-printed objects would be enough to impact the fluid flow such that it would be turbulent enough to overcome the intended friction of the channels? If I had the time, I'd make up a 3D model of this that I could cut on my CNC machine out of a block of aluminum or hard plastic to test.

serf 7y ago
from playing with 3d printed tesla valves , I think that you're right that surface distortions hinder correct flow. The way that I created a 3d printed tesla valve that worked was by messing with overall scale. Bigger works better, until suddenly it doesn't work at all. The idea I had previously about why scale may help was the idea that the scale is gradually turning down the amount that the flow error affects the entire system until eventually the mass of the fluid induces more 'error' into the system until the whole thing destabilizes. In other words : with 3d printed tesla valves it's vital to find the sweet scale spot between surface-level distortion causing chaos when the valve itself is too small and liquid mass causing chaos when the valve is too big. That margin between chaos and stability may be made wider by using a finer manufacturing technique, like milling or even SLA/SLS versus FDM. I bet that'd give a wider range of usable sizes. all that said : even professionally made tesla valves must be made with certain scale constraints in mind. I believe those constraints are first-and-foremost a property of the liquid that's going to be transported.
- jakobegger 7y ago
  I'm pretty impressed what you figured out by just experimenting. What you are describing is the transition between laminar flow and turbulent flow. The point where this transition occurs is governed by the "Reynolds Number". As you correctly discovered, the Reynolds number is influenced by scale and fluid properties (viscosity and density). The final missing piece is the speed of the fluid, which explains your observation: "Bigger works better, until suddenly it doesn't work at all."