Learning Forum

[shal3391]

Hi,

I am trying to model a liquid injection into a air flow.

My domain is a 3D cylinder with a nozzle at the top, a cross section is attached.

I have a fluid velocity inlet at the nozzle of 4.3m/s and an air velocity inlet at the top of the cylinder of 36m/s.

I have currently selected an 'outflow' at the bottom, however the air velocity is increasing exponentially at this outflow. Shown in the figures attached.

What outlet boundary conditions should I use?

And do you have any recommendations regarding turbulence models, relaxation factors or solving schemes. Because I am also struggling choosing these parameters.

Many Thanks,

[Rob]

Use a pressure outlet. Outflow is an old boundary dating back to before compressible flow and GUIs. nRe models and settings have a read through the documentation and come back with some suggestions and reasons for using a model or setting. n

[shal3391]

Thanks Rob,nnAt the moment I am using the standard k-epsilon turbulence model because the majority of my fluid flow is far from any walls. However I am unsure if I need the turbulence dampening provided by the k-omega model. Would a k-epsilon realisable model or a k-omega SST model be an adequate compromise ?nMy goal is to induce fluid breakup in a jet stream, so I want there to be adequate turbulence fro this to happen. At the moment I am using the default values for all relaxation factors, would increasing the turbulent kinetic energy factor (default is 0. provide more accurate turbulence representation ? My understanding is to reduce the relaxation factors if the residuals are increasing, my residuals (attached) aren't necessarily increasing but don't look to be particularly stable.nI am currently using a SIMPLEC solving scheme (based on trial and error) but have read that a coupled scheme is generally better for VOF models. I am using a 'Compressive' volume fraction solver. I have also read that I should tick Pseudo Transient, but am unsure as to what it does.nI am using a variable time step to cap the Global Courant Number at 0.5.nThanks Again

[Keyur Kanade]

Please check nPlease check nAnd nArrayRegards,nKeyurnGuidelines for Posting on Ansys Learning ForumnHow to access ANSYS help linksnn