Learning Forum

[jasonsum25]

Hi all, 

I am setting up a transient simulation modelling the air flowing through a single stage tesla valve for learning its flow physics.

Setup: k-omega SST model, PISO algorithm with 2nd order upwind

B.C.: pressure inlet of 1000 Pa total pressure and pressure outlet of 0 Pa gauge pressure - the delta p driving the air flow across the valve

Apart from viewing velocity and pressure contour plots, I want to straightly identify air entering without including the air originally staying in the cell zone, so that I can track how the incoming air develops along the winding paths (as if tracing flow with ink in real life experiments)

I tried to set up a User Defined Scalar and set such scalar in the inlet as 1, yet the result did not look like a tranisent air scalar behavior. Therefore I am not sure is it the right way to do so. (animation screen cap attached below)

The visualised result I want is something like this below (reference from Holzman CFD Youtube video):

https://www.youtube.com/watch?v=6fgfr2BZUF8

It is sincerely hoped for help! Thanks so much!

Jason

 

[Rob]

That looks like the flow didn't develop very well. Can you post the velocity contour so we can see what's going on? As an aside the angle and shape of these sorts of valves is very important, small deviations can have significant effects on the results: power fluidic devices use this to control flow. 

[jasonsum25]

My trial on this T45 single-stage tesla valve is just for myself to understand the flow physics behind when there is pressure difference driving the inflow. 

Below are result of transient velocity and pressure contour plots under convergence:

My concern is, the passive scalar of incoming air shown in previous comment gradually filling up the valve, but not similar to the result from the Holzman CFD Youtube video (only "concentration/portion" of incoming air tracing). I would like to know is sth wrong about my setup of UDS. Thank you!

[Rob]

Is this for water, and what is the scalar diffusivity? 

[jasonsum25]

It is air flowing from inlet of the valve (single phase). I set scalar diffusivity as 1, following some online videos on setting, not sure if it is not making sense. I would like to know the way to obtain the visualised result like the one from Holzman CFD Youtube video as shown above. Thanks for the guidance!!!

[Rob]

Ah, that'll explain it. The scalar is diffusing far too much. Treat the scalar like a species, and divide the typical value for that species diffusivity by the fluid density (scalars don't really have units). Liquids will be diffusivity of 1e-9 to 1e-10 or so - look it up as I usually have to.