Confused about Surface Nusselt number evaluation in Fluent

- January 12, 2020 at 7:57 pm
  
  khalidb77
  Subscriber
  
  Hello
  
  I am really confused about Nusselt prediction in Fluent. I aware that there is a method to predict it from the surface integral>wall fluxes>surface Nusselt number. I tried to evaluate Nusselt on another method based on the temperature gradient in which :
  
  q'' = -k (dT/dx)
  
  then h= q''/(T-Tref) = -k(dT/dx)/(T-Tref)
  
  finally Nu=hL/k = -(dT/dx) * L/(T-Tref)
  
  I found the Nusselt predicted from the two methods is not the same why? knowing that I set Tref at the reference values
  
  Thanks
  
  Khalid
- January 13, 2020 at 5:58 am
  
  DrAmine
  Ansys Employee
  
  It depends in the reference values in for Tref and Lref.
- January 13, 2020 at 6:46 am
  
  khalidb77
  Subscriber
  
  I was set Tref & Lref correctly into reference values. My question if I set the local Nusselt number on a custom field function as below :
  
  Nu = -dT/dx*(Lref/(T-Tref))
  
  why it gives different Nusselt from that obtained by wall fluxes > surface Nusselt number? knowing that I was estimate both numbers on the same line?
  
  Khalid
- January 13, 2020 at 9:08 am
  
  DrAmine
  Ansys Employee
  
  Fluent uses the wall heat transfer coefficient in the definition of Nusselt number. And the latter is build up with the conservative wall heat flux.
- January 13, 2020 at 12:35 pm
  
  khalidb77
  Subscriber
  
  So do you mean that they are not the same?
  
  Khalid
- January 13, 2020 at 12:41 pm
  
  DrAmine
  Ansys Employee
  
  What does alpha(HTC)*Lref/heat conductivity provide if put in CFF?
- January 13, 2020 at 12:46 pm
  
  khalidb77
  Subscriber
  
  Excuse me what do you mean with the term alpha(HTC)?
- January 13, 2020 at 12:55 pm
  
  DrAmine
  Ansys Employee
  
  Surface Heat Transfer Coefficient
- January 13, 2020 at 3:45 pm
  
  khalidb77
  Subscriber
  
  What does alpha(HTC)*Lref/heat conductivity provide if put in CFF?
  
  With this relation it gives the same value as the method of wall fluxes but not the same as the method whuch define Nu in terms of derivative dT/dx?
- January 13, 2020 at 4:27 pm
  
  DrAmine
  Ansys Employee
  
  Yeah as I expected. That is the method which I know moreover heat transfer is not only due to conduction but to convection to.
- January 14, 2020 at 11:32 am
  
  khalidb77
  Subscriber
  
  which method do you mean that heat transfer due to convection and conduction?
  
  Khalid
- January 14, 2020 at 11:42 am
  
  DrAmine
  Ansys Employee
  
  Fluent calculates his Surface Nusselt Number as already mentioned and the way it is documented in the User's Guide.
  
  If you want to have your own definition of Nusselt Number do not compare with the one of Fluent. Your definition does only take into account for thermal conduction and not for convection.
- January 14, 2020 at 12:01 pm
  
  khalidb77
  Subscriber
  
  You mean Surface Nusselt Number in Fluent based on conduction and convection right?
- January 14, 2020 at 1:15 pm
  
  DrAmine
  Ansys Employee
  
  right
- January 14, 2020 at 6:17 pm
  
  khalidb77
  Subscriber
  
  Thanks Mr. Abenhadj for the response. May you tell me please the relation that Fluent uses in evaluating the Total Surface Heat Flux because I failed to find it on the User's Guide?
  
  Khalid
- January 15, 2020 at 5:07 pm
  
  Rob
  Forum Moderator
  
  Have a look in the field variables part of Field Function Definitions.
- January 17, 2020 at 5:44 pm
  
  khalidb77
  Subscriber
  
  Have a look in the field variables part of Field Function Definitions.
  
  I found the relation at last that Fluent uses to predict the surface heat flux
  
  for horizantal wall q=-k dT/dy
  
  for vertical wall q=-k dT/dx
  
  I remarked this post as solved now