[Bynn]

Hey all!

I'm writing my project at my uni at the moment and I'm trying to simulate the flow in a pelton distributor with CFX. I've done experiments and obtained the static pressure on 6 different places and I'm trying to find the head-loss through the distributor.

In my model I have simple outlets without the nozzles as these are not what I'm interested in. I've specified inlet pressure and outlet velocity through the "nozzles". My problem is the choice of turbulence model, I've tried with a bunch but I haven't found any clue to what kind of modle is the best for me.

Another question I have is about the post-processing of my simulation. I've created planes where I measure the static pressure and compare the head through the turbine with the results from the measurements. I've been using the "ave" function for pressure and velocity on the planes, but someone said that I should use the "areaAve" function instead since this takes into account the size of the nodes. Problem is that the shape of the head-loss matches my experiments when I use the "ave" function, but not when I use the "areaAve" function. I might be going at it wrong, but I wanna know what you guys think.

Thx!

[ghorrocks]

My problem is the choice of turbulence model

If I was you I would just use SST. If you feel like creating work for yourself I would run k-e as well and compare the difference - in this case probably very little. I suspect the turbulence model is not affecting your results much so your choice of model is not too important.

To get average pressures you should use the areaAve function for exactly the reason you describe. To get average velocities you should use the massflowAve function.