[miquiita]

With the sensitivity study, what would that be required to do?

I was thinking refining the mesh around the blades and inner domain, cylinder. Also not to much around the walls, stationary domain, which I posted earlier. My blades can be changed in geometry, it is only one study I found which has these kind of blades. What do you mean by periodic boundary conditions. Which tutorial would that be? I looked at most of them already.


I uploaded the geometry of the blades here

[miquiita]

I found one tutorial at least, Chapter 17. Multiphase Flow in a Mixing
Vessel. thanks anyway. But I was advised by my supervisor to have the usage of blades...

[ghorrocks]

That example does not show the use of periodic boundaries in rotating machines. Try the axial turbine simulations (14 and 26).

I see your solid model is very simple. There is no blade twist and no camber on the blades. Therefore I would expect this blade to run pretty badly. So I would recommend getting rid of the blade thickness and modelling the blades as thin surfaces. At least that will get rid of the square leading and trailing edges.

[miquiita]

I just get the problem with thin surface modelling that the house of the rotor, round bit in the middle will be hard to do as a thin surface. The blades will not be able to do a bolean on the outer domains.

[miquiita]

I could model the blades as a naca wind foil, would that work better?

[ghorrocks]

Yes, a real foil will work better. Have a look in the modelling tutorials (from the ANSYS community webpage) for thin surfaces if you want to go that way, I do not think you understand what they are based on your previous post.

[miquiita]

what do you mean based on?

Also if I would model it as a thin surface, how would that work with the boolean of the model?

[miquiita]

I am sorry, I was trying to find those tutorials but I could not find any. I think I might be looking at the wrong thing. But I am on the ANSYS community homepage

[ghorrocks]

You don't generate thin surfaces by boolean subtraction. You generate the surface and define it as a thin surface. You will have to look at the documentation about the details of how to do it.

[miquiita]

I meant thin surfaces is not able to do boolean?

Otherwise I had a look today, and talked to some of the ANSYS teachers, they recon my simulation needs a smaller timestep, which should be calculated the smallest mesh on the rotor blades time to travel with the angular rpm by the rotor.

Understand me? Does this seem valid?

[ghorrocks]

All of these suggestions about time step being determined by this and that are just starting points. When the simulation is running you then adjust it to fine tune it. If the simulation is unstable you go smaller and if stable but converging slowly you go faster.

[miquiita]

Hi I understand that. My main issue now was to measure the mesh of the rotor. It does not want me to just show the mesh of the rotor, instead the whole domain.

Could I somehow measure the mesh of the tips of the blade in ANSYS?

[ghorrocks]

What do you want to measure in the mesh?

From ANSYS Mesh the best way to measure mesh quality metrics is to get it into CFD-Post and use the mesh quality tab to generate mesh quality measures. Then you can use the standard post processing tools to view the mesh quality as any other variable.

[miquiita]

Hi,

I wanted to measure the mesh on the tip of the rotor blade.

Would the pressure be very low (2Pa) in the upstream of the duct, before the rotor or should the pressure be around 200 Pa? If the normal velocity is around 1.5 m/s and the rotor rotates with 300 rpm?

I get results of the small pressure now, with a residual that are nearly below e-4 (about 2/3 of the e-3 residual).

[miquiita]

I cant seem to locate the mesh quality tab, just the calculator that does not calculate the distance of a mesh

[miquiita]

I managed to solve the first problem of the duct and the rotor. Thanks for all help.

Now I have got the problem of creating the geometry of the room and fan. I first was thinking to create the room as the duct but rectangular, so this would be frozen extrude and then do a small exhaust (not sure which extrude to use here) to let air out and in, an opening. But my model will not work. The rotor and the cylinder around it is add material and cylinder is frozen.

Any thoughts?

[ghorrocks]

It sounds like you are quite confused about how to set these simulations up. Have you done the CFX tutorials?

No duct is needed, I have no idea what you mean by frozen extrude or a small exhaust. Can you draw a picture of what you mean?

[miquiita]

Hi,

I done some tutorials. But I cant seem to find something that is related to what I am doing. I know I do not need a duct, what I am saying is a create a rectangular domain, around the fan instead of the duct but with a frozen wall.

I already created a room model before but without the duct. Looking like picture I attached and also an other picture of the whole model. The problem is that I have the walls as extrude with add material in the picture of the room with no rotor. If I am trying this approach with a rotor in the smal cylinder domain in the roof, the rotor of the fan disiperse.