Learning Forum

[pssjn]

My study is to find out the air ventilation rate inside the room. So I need to find out the air pressure and velocity on the windows.

My queries are:

How do I draw an Atmospheric domain with the room inside?   

How do I define the boundary condition of both the domain and room? 

I need to produce the following images.                            

 

[peteroznewman]

ANSYS staff do not open attachments, so please Insert the images from your pdf directly into your post (I am not ANSYS staff).

Here are the instructions.

• In SpaceClaim, put a plane at ground level.
• On the Prepare tab, click Enclosure and specify the distance from the sides and top of the house you want. Ignore the Enclosure going below ground.
• Click the Split By Plane tool, and click the plate and delete the body below ground.
• Pick the Solid body in the outline that is the house and right click to select Suppress for Physics. You only need the air, not the building.
• Click on the name at the top of the Structure outline on the left, and in the Properties window, change Share Topology to Share. This will allow solids and fluids to share nodes at the common interface.
• Close SpaceClaim and open Meshing.
• Pick a face and click N to create a Named Selection for inlet, repeat for outlet etc. Any face not named will automatically become a wall.
• Go to the Connections folder and delete any automatically created Contacts. You don't want them since you are using Shared Topology.

There is a method to have zero thickness walls, then each window only has one face.  See this post for an example. That will get you a model with fewer cells that solves a bit faster. If you do that, you must pick the face that is the open window and name it interior. That will allow air to flow through the window.

[pssjn]

Thank you so much for your reply. It helped me a lot. I am a very new user of CFD Fluent with architecture background. Please pardon me for my silly queries.

I tried according to your suggestion. It looks okay for now. I need to study more for myself. My question is to get the images as follows do I need to follow the zero thickness wall method?

Thanks and regards.

[Rob]

Those images look to have used a thick wall (ie it has a dimension in CAD). In Fluent we don't need that, so we can use a surface only which reduces the cell count in the models.  I nearly always use a thin wall approach, and then use extra models in Fluent if I need to model heat transfer. 

[pssjn]

Thanks for your reply. 

I drew one solid box and drew 2 windows with surfaces.

Then put the enclosure.

Set the shared topology and suppress for physics.

 

In the meshing, I named the faces containing windows “interior” and inlet and outlet face on the atmospheric domain.

I didn't succeed. Could you please explain step by step.

Thanks and regards.

[pssjn]

Could you please explain the zero thickness wall method step by step for my case.

Thanks and regards

[peteroznewman]

The following video results in a mesh with 1 M cells, but the Student license has a 512 K cell limit, so this model could be cut in half and use symmetry down the center.

https://youtu.be/6SfpZRPAwWY

[pssjn]

Thanks, peteroznewman. It helped me a lot. Yet, I couldn't able to pass the air through the window. I followed your video step by step. where am I gone wrong in my steps?? Please help me.

Thanks and regards

[Rob]

Check you made a multibody part in DesignModeler (or share topology in SpaceClaim) otherwise the flow can't pass between the volumes. 

[peteroznewman]

Share Topology is done at 3:20 in the video.

Here is a completed simulation at 1 m/s inlet air velocity on a 1/2 symmetry model.

[pssjn]

Thanks peteroznewman. Could you please let me know the other settings.

Cell zone condition (I have 2 zones)

1. Enclosure (type Fluid-air)

2. Solid (type Fluid- air) is it correct???

Boundary condition

Inlet- pressure inlet

Outlet- pressure outlet

Do I need to set the other boundary conditions??

Please forgive me for my silly questions, as this is my very first project on CFD.

Thanks and regards

[peteroznewman]

The simulation will be easier to converge if you have an Inlet - velocity inlet, which is what I did above.

You have to change the Model from Laminar to k-epsilon turbulent model with Realizable wall function.

I used Standard Initialization from the Inlet, and used Patch to set the velocity to 0 in the Solid (house).

[pssjn]

Thanks peteroznewman. I can’t figure out the windward window sharply (like yours). The Leeward window turned out okay. where am I gone wrong in my data input??

[peteroznewman]

Maybe you haven't used small enough elements? In Meshing, take a Section view through one of the windows. Reply with an image inserted to show the size of the elements. There is a button on the section view to show whole elements, turn that on.

[pssjn]

[Rob]

On the Section plane there's a triangle/tetrahedron button, that'll show full cells which is easier to see.  However, in this case one of the problems is mesh resolution. Put a volume sizing on the room so you've got about 6 cells up the height of the window. That may still not be fine enough but will do for now. 

[Skynet]

.

Hello.. can you please repost the videos in this thread?

.

[peteroznewman]

.

Skynet

.