Why areaAve and ave make a difference for 2D simulation

- June 18, 2020 at 3:24 am
  
  AlexC
  Subscriber
  
  I got different mean velocities using ave and areaAve for 2D geometry, which I don't know why.
  
  Please see the figure below. Inlet velocity is 0.1m/s. I want to find out mean velocity (u component) at outlet.
  
  I got different mean values using ave and areaAve:
  
  (1) ave(Velocity u)@outlet=0.0677 m/s
  
  (2) areaAve(Velocity u)@outlet=0.1 m/s
  
  Since inlet velocity is 0.1m/s, so I know the result using areaAve is correct. But what causes the difference?
  
  Based on my understand, ave is simply arithmetic average, and areaAve is area-weighted average. Yet in this 2D simulation, I don’t understand why ave and areaAve make a difference.
  
  Please advise. Thank you very much.
- June 19, 2020 at 3:04 am
  
  AlexC
  Subscriber
  
  I think I know the reason now.
  
  The reason comes from the non-uniformity in the mesh. Because mesh is non-uniform (finer near wall and coarser otherwise), the weight factor (mesh size) of areaAve function makes a difference compared with the ave function.
  
  Still would like to hear some inputs from other people.
  
  Thank you.
  
  Alex
- June 19, 2020 at 11:53 am
  
  Karthik Remella
  Administrator
  
  Yes, your reasoning is correct. Area-Weighted Average takes into account the face area. So, think of it as if the velocity crossing each face were being normalized by the area of the face. So if you have non-uniform distribution, the two values will not match and area-weighted ave would give you a better estimate.
  
  Thanks
  
  Karthik
- June 19, 2020 at 1:44 pm
  
  AlexC
  Subscriber
  
  Thank you Karthik! That helps. Appreciate you taking time to reply my question.
  
  Best,
  
  Alex