[Mr.Hz]

Hello,

I am running a dyson bladeless fan simulation using the simpleFoam solver with k-omega turbulence modelling. The computational domain is a (2m x 1.64m x 0.5m) box with top, bot, left, right, and back using zeroGradient boundary condition, while the front uses symmetry boundary condition. The inlet is located at 1/3 of the geometry length. You can see the sketch of this settings in the attached compDomainHalf1.pdf. After running the simulation for t = 1000s, the result shows flow coming out from the bottom region of the computational domain, disrupting the flow generated from the geometry, as seen in figure dyson60.pdf. You can also find the 0 file attached as zip. I expect the flow to have vorticity in the middle of the geometry.I am thankful for the help.


Mr.Hz

[shock77]

I dont fully understand the problem. The flow will always follow the pressure gradient. Depending on where you set which boundary conditions you impact your solution. But for me its not clear which BC are applied at which patch.


Maybe you can add a sketch to clarify things.






Kind regards,
shock77

[Mr.Hz]

Originally Posted by shock77

I dont fully understand the problem. The flow will always follow the pressure gradient. Depending on where you set which boundary conditions you impact your solution. But for me its not clear which BC are applied at which patch.


Maybe you can add a sketch to clarify things.






Kind regards,
shock77

Thank you for the reply. You can see the boundary conditions in the compDomainHalf1.pdf. Please take a look.


Kind regards,


Mr.Hz

[shock77]

I have seen your attachments, but I cant see all the BCs. Or is everything zeroGradient?


Your 0 folder says, that you have definied a outlet with a fixed pressure and a inlet with a fixed velocity. This is typical for subsonic cases. But in your image it looks like everything is zeroGradient.


Also the image with the vectors makes it hard to understand what you mean and what the real issue is.


An image with the patch names and the BCs for atleast U and p would be helpful for me to understand the problem correctly.









Kind regards,
shock77

[Mr.Hz]

Update on the simulation, I have manage to "fix" the unwanted flow direction by changing the internal field in the 0/U into (-1 0 0). What it does is, it forces the flow direction into what I perceive to be the correct flow direction. You can see the simulation result in the attached figure.