
[Sponsors] 
June 2, 2015, 10:00 
Caracteristic length used for Peclet number

#1 
Senior Member
Laurent DASTUGUE
Join Date: May 2014
Location: Chartres, France
Posts: 111
Rep Power: 4 
Hi foamers,
Does anybody know what is the caracteristic length used when OpenFOAM calculate the Peclet number ? I don't have access these days to OpenFOAM and i have to answer this question. Thanks a lot. Laurent 

June 2, 2015, 10:38 

#2 
Senior Member
Alexey Matveichev
Join Date: Aug 2011
Location: Nancy, France
Posts: 1,437
Rep Power: 25 
Hi,
It seems you have access to the Internet. https://github.com/OpenFOAM/OpenFOAM...d/Pe/Pe.C#L118 https://github.com/OpenFOAM/OpenFOAM...olation.C#L136 So, Pe is calculated in assumption Sct = 1 and length scale is cell size. 

June 2, 2015, 10:51 
Sct ?

#3 
Senior Member
Laurent DASTUGUE
Join Date: May 2014
Location: Chartres, France
Posts: 111
Rep Power: 4 
Hi Alexey,
thank you for the links, and for your answer concerning the length scale. Just one question : What is Sct ? Laurent 

June 2, 2015, 11:00 

#4 
Senior Member
Alexey Matveichev
Join Date: Aug 2011
Location: Nancy, France
Posts: 1,437
Rep Power: 25 
Sct is turbulent Schmidt number (https://en.wikipedia.org/wiki/Schmidt_number). As you can see, in Pe.C everything is divided by viscosity, yet for Peclet one has to divide by diffusivity. If we assume that Sc = 1, we can just divide by viscosity (for turbulent case we have to assume that Sc = 1 and Sct = 1, so nuEff( = nu + nut) == Deff( = nu/Sc + nut/Sct)).


June 3, 2015, 04:56 

#5 
Senior Member
Laurent DASTUGUE
Join Date: May 2014
Location: Chartres, France
Posts: 111
Rep Power: 4 
Ok i understand now.
Is the Peclet number calculated here a massic one ? (Pe=Re*Sc) or a thermic one (Pe=Re*Pr). If this is a massic one, can i say that Pe = Re, since Sc = 1 ? Laurent 

June 3, 2015, 05:08 

#6 
Senior Member
Alexey Matveichev
Join Date: Aug 2011
Location: Nancy, France
Posts: 1,437
Rep Power: 25 
Hi,
If you assume Pr = Prt (turbulent Prandtl number) = 1, then it becomes "termic" And yes, you can call the value calculated by Pe utility Reynolds number. 

June 3, 2015, 05:18 

#7 
Senior Member
Laurent DASTUGUE
Join Date: May 2014
Location: Chartres, France
Posts: 111
Rep Power: 4 
In fact what i want to do is the calculation of the convection coefficient h.
I know that there is a relation between h and the Nusselt number, and i know a relation letting me to have Nusselt, by the use of Re and Pr values. Since i have the value of Prandtl number easily, knowing the fluid caracteristics, i just need the value of Re to have Nu and to have finally the value of h. So reading your posts, i understand the following : using Pe utility, i can do the operation Re = Pe/Pr to have Reynolds number. Am i right ? Or must i apply some operations on Pe utility results before having the Peclet number value which will let me calculate Re ? Laurent 

June 3, 2015, 07:54 

#8 
Senior Member
Alexey Matveichev
Join Date: Aug 2011
Location: Nancy, France
Posts: 1,437
Rep Power: 25 
Hi,
Short answer: it depends. Pe field produced by Pe utility is surfaceScalarField and its value is where L is of order of cell size, u is in fact where S is face normal vector with module equal to the surface of the face. So, in fact Pe is Re. If you are happy with h as a surfaceScalarField, then yes, you just manipulate result of Pe utility; if, for example, you need convection coefficient as volume field, you need additional steps. 

June 3, 2015, 08:07 

#9 
Senior Member
Laurent DASTUGUE
Join Date: May 2014
Location: Chartres, France
Posts: 111
Rep Power: 4 
Hi,
thank you very much for your help. So what we have is : Pe = Pe / Pr, isn't it ? Why is this utility not called Re ??? Laurent 

June 3, 2015, 09:37 

#10 
Senior Member
Alexey Matveichev
Join Date: Aug 2011
Location: Nancy, France
Posts: 1,437
Rep Power: 25 

June 3, 2015, 10:14 

#11 
Senior Member
Laurent DASTUGUE
Join Date: May 2014
Location: Chartres, France
Posts: 111
Rep Power: 4 
Thank you very much Alexey for your time. Have a good day.
Maybe this utility will be called Re in a next future thanks to your explanations ;) Laurent 

Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
simpleFoam parallel  AndrewMortimer  OpenFOAM Running, Solving & CFD  12  August 7, 2015 18:45 
decomposePar no field transfert  Jeanp  OpenFOAM PreProcessing  0  June 20, 2014 05:59 
snappyHexMesh sticking point  natty_king  OpenFOAM Native Meshers: snappyHexMesh and Others  2  April 17, 2014 01:24 
Stable boundaries  marcoymarc  CFX  33  March 13, 2013 07:39 
AMI speed performance  danny123  OpenFOAM  19  October 24, 2012 07:44 