How to analyze the vortex tube flow analysis? It is different from the reality.

[JW_13226379]

I learned a little bit, but I haven't mastered it at all.
It should be cold at one end and hot at the other, but I didn't expect that one end is hot and the other end is even hotter.
Is there expert tell me where I am wrong?

[tmoser]

You'll probably need to provide a little more information on your model.

Is the inlet condition applied to all of the small tubes, or only one of them?

What is the fluid in the model?

Is the model fluids only, or are there solids with conjugate heat transfer?

[JW_13226379]

yes, the inlet conditions apply to all small tubes,  medium is air. the physical model includes turbulence and heat

[KenFarley]

Just looking at your image, I see you have exit boundary conditions of 0.1 Pa and 0.05 Pa. That's pretty much calling for a vacuum (10E-07 ATM) at both ends in comparison with the 8 ATM inlet. Why are they different? Why is the number so low? I would expect something like 1 atm (ambient)?

And again, what does the flow stream field look like? Are the flow lines in the tube as one would expect?

[JW_13226379]

The outlet pressure seem not affect the simulation results. its more focuses on the pressure differential between the outlet and inlet, and the flow stream is fellow reality, except the cold end outlet, he does not output cold air, but suck inward, which is so weird

[cfdgeek]

The pressure difference between inlet/outlet for sure affect your simulation, i believe you need to check your boundary conditions, also you need to use ideal gas law or real gas law to make sure you model the compressible flow,  which is your case, also need to check your chocked condition if there are some throat area, again check your BC, your pressure outlet seems wrong

[cfdgeek]

and your assumption of only p difference matters are wrong too, that only true for incompressible flow, for compressible flow, you for sure value of pressure itself matters, use real working condition or make sure your Mach numbers and Re numbers are correctly match real problem numbers. Anyway if you didn't notice the temperature changes, you may not use the compressible flow at all and if you use ic flow, for sure no temperature changes from your simulation

[KenFarley]

Hard to tell just from what you've got.

What do you see for the flow field? Do the flow streams look like what is seen in renderings by people who have researched these? Is your swirl chamber defined properly to get the flow you need? As I understand it, that's where the "magic" is in designing a vortex tube. As well as the configuration of the exit cone on the hot side.

Had a similar dilemma trying to model the flow of a cyclone in the past. Took a good bit of adjustment to get those luscious bits of solid debris to spin around and drop out of the device the way they were supposed to. Best part is when we built them they worked.