[hamidciv]

hi every one
I found The following sites for calculate first cell size in order to achieve desired yplus in my work.
Is the order of cell size is same element size in cfx mesh?
i have problem in replacement values as i shown follow, please guide me.
i have only one velocity value in inlet of model, while in following image Has been asked velocity in free stream, how do this? also i dont understand, What is the purpose of the reference length?
I also know, desired yplus value in finally determined based on grid sensitivity but calculating cell size could be useeful.
thanks in advance

[Antanas]

Originally Posted by hamidciv

hi every one
i I found The following sites for calculate first cell size in order to achieve desired yplus in my work.

i have problem in replacement values, please guide me.

These are characteristic values and they depend on the problem.

[hamidciv]

dear antanas
if your possible, please explain more.

[Antanas]

Originally Posted by hamidciv

dear antanas
if your possible, please explain more.

For ex. in the case of flow around flate plate characteristic V will be velocity of free (incoming) stream and characteristic dimension will be the lenght of the plate. In the case of flow in a tube they will be velocity on tube axis and tube diameter.

[hamidciv]

dear antanas, thanks for reply.
I almost understand your mean, but beacuse i make sure, in following geometry that shown geometry spillway, What will be the velocity and length?
kind regards
hamid

[Antanas]

Originally Posted by hamidciv

dear antanas, thanks for reply.
I almost understand your mean, but beacuse i make sure, in following geometry that shown geometry spillway, What will be the velocity and length?
kind regards
hamid

I don't know, sorry. But if you already got preliminary results you may plot y+ values to see whether you first near wall node lies within desired y+ value and then modify your mesh accordingly.

[hamidciv]

dear antanas,
Based on the definition of y+ variable, it is distance wall of first node inside boundary layer, thus we can compare it with vertical height in geometry,Does not it?
when y+ value in my geometry is greather than vertical height in my model, What concept can it be?
whether this comparison within y+ and vertical height not correct?
i used sst model, My goal was to simulate the boundary layer and My results are good, But high value of y+ has confused me.
whether In your opinion with regard Conditions of my model , it could be problem?
thanks in advance

[Antanas]

Originally Posted by hamidciv

dear antanas,
Based on the definition of y+ variable, it is distance wall of first node inside boundary layer, thus we can compare it with vertical height in geometry,Does not it?
when y+ value in my geometry is greather than vertical height in my model, What concept can it be?
whether this comparison within y+ and vertical height not correct?
i used sst model, My goal was to simulate the boundary layer and My results are good, But high value of y+ has confused me.
whether In your opinion with regard Conditions of my model , it could be problem?
thanks in advance

y+ is dimensionless distance, you can't compare it with dimensions of your geometry directly. But it's connected with real (dimensional) distance between wall and first near wall node of mesh. Different turbulence models have recommended ranges of y+. Depending on turbulence model chosen and your problem you choose desired y+ value. Based on this and estimated characteristics of the flow you calculate y - dimensional distance (search CFX help materials) to generate your mesh. After performing calculation of your problem you should plot y+ values on walls to ensure that first near wall nodes lie within desired y+ value and modify your mesh if needed.

[-Maxim-]

Originally Posted by Antanas

y+ is dimensionless distance, you can't compare it with dimensions of your geometry directly. But it's connected with real (dimensional) distance between wall and first near wall node of mesh. Different turbulence models have recommended ranges of y+. Depending on turbulence model chosen and your problem you choose desired y+ value. Based on this and estimated characteristics of the flow you calculate y - dimensional distance (search CFX help materials) to generate your mesh. After performing calculation of your problem you should plot y+ values on walls to ensure that first near wall nodes lie within desired y+ value and modify your mesh if needed.

this is a pretty good summary.
hamidciv, you will probably have to read some more about y+ value and boundary layers to fully understand what's going on. It took me a while too...
You mentioned that you are using the SST model. That model is quite 'robust' regarding your mesh since it automatically switches from wall functions to turbulence model.
Maybe this thread helps you to understand more:
http://www.cfd-online.com/Forums/cfx...value-sst.html

[hamidciv]

dear antanas & maxim
thanks a lot for reply & help.