4 Replies Latest reply on May 11, 2016 1:31 PM by Siavash Khajehhasani

Size of computational domain for simulation of a mask

Erik Liermann May 10, 2016 10:08 AM

Dear all,

I am trying to run a flow simulation to determine the flow rate of air I need to have to deliver enough filtered air to the user of a mask. Unfortunately I have now idea how to realistically set the computational domain as small changes already result in completely different values.

To explain the system: An airflow (inlet volume flow) streams into the mask and protects the user from inhaling outside air. The outside air has a wind speed of 12 m/sec. I set surface goals to get the volume fraction of outside air at mouth and nose. After determining the flow rate I want to make another time dependent study including the human breath.

Do you have any idea how to size the computational domain for this scenario?

Re: Size of computational domain for simulation of a mask

Amit Katz May 10, 2016 10:16 AM (in response to Erik Liermann)

Changing the simulation domain shouldn't really affect your results that much, unless it substantially changes the cell density in your leakage paths. How are you simulating the leakage, exactly?
Like Show 0 Likes(0)

Actions
- Re: Size of computational domain for simulation of a mask
  
  Erik Liermann May 10, 2016 10:45 AM (in response to Amit Katz)
  
  Hi Amit and thanks for the quick response. The leakage was modeled in the form of holes in the area around the nose and the chin. If I look at the velocity plots I see different turbulences around these parts when I change the computational domain so I guess you can be right. But how do I avoid that?
  
  Like Show 0 Likes(0)
  
  Actions
  - Re: Size of computational domain for simulation of a mask
    
    Amit Katz May 11, 2016 8:38 AM (in response to Erik Liermann)
    
    Are you using any kind of local meshes around those holes? The key is to be sure you have enough cells across the holes and the area immediately downstream of them.
    
    Like Show 0 Likes(0)
    
    Actions
Re: Size of computational domain for simulation of a mask

Siavash Khajehhasani May 11, 2016 1:31 PM (in response to Erik Liermann)

Hi Erik,
The safest and most accurate method to assure how far the computational domain (CD) should be put is to perform a simple check to see if any of the main flow parameter gradients such as velocity and pressure is changing across the CD boundaries.
I hope you find it helpful.
Cheers,
- Siavash
Like Show 0 Likes(0)

Actions

Go to original post