Advice for modeling nozzle-like geometry in fluent
 

As part of my graduate research, we designed a nozzle like geometry that vibrates quite strongly when water flows through it. I would like to model this in fluent but I'm not really sure where to begin, I have run some simulations using different techniques such as changing the turbulence model, or modeling it in 3D as well as 2D, but it seems like I get different results with different models and techniques. I attached a picture of the 2D cross section.

Try running the simulation on Sunday. :rolleyes:

Before doing anything investigative, you need to normalize your experience with Fluent first.

I have pretty good experience with the Fluent interface actually. I know how to design in design modeler, creat my mesh and setup the simulation in fluent. I'm just not sure what the best turbulent models, iterations settings, etc. would be.

Hi Nikolai,

When it comes to modeling a nozzle, having a good mesh is critical to obtaining a good solution. The mesh at the nozzle tip should be fine enough to capture the pressure gradient and the shock. The boundary layer needs to be refined with atleast 15-20 cells. Looking at your model, you can obtain a nice Hexahedral mesh aligned in the direction of flow. You just need to decomposed the model.

Try using a density based solver with your time step adjusted based on the Courant number. Review your boundary settings carefully. Set the initial gauge pressure appropriately. I have generally had good experience using the SST K W turbulence model. If you observe divergence behavior initially, start with transient mode and then change over to steady state.

Good Luck