# y+ calculation

 Register Blogs Members List Search Today's Posts Mark Forums Read

 September 5, 2019, 05:54 y+ calculation #1 New Member   Sai Krishna Join Date: May 2018 Posts: 13 Rep Power: 3 IAM doing a conjugate heat transfer analysis of a channel and I need to know my mesh refinement is sufficient enough to capture the heat transfer happening near the wall. For that I read about y+ and how to calculate it using the formula y+ = y * u-star/(kinematic viscosity) Where u-star is sqrt of wall shear stress and density. My question is how to get this wall shear stress, as it needs velocity gradient along normal direction of the flow as Tau = mu * (du/dr) @ r=R and I don't know the velocity profile. If at all I have to take local values which one I should consider, values near the wall or mass weighted average value of that boundary. I am doing Thanks in advance.

 September 5, 2019, 11:33 #2 Senior Member   Lucky Tran Join Date: Apr 2011 Location: Orlando, FL USA Posts: 3,735 Rep Power: 45 You don't know the wall shear stress beforehand, you can only take your best guess. At the end of the day, your grid resolution is up to you. If there are local areas where the y+ exceeds your criteria then, it is what it is.

 September 6, 2019, 09:47 #3 New Member   Sai Krishna Join Date: May 2018 Posts: 13 Rep Power: 3 I really appreciate ur help Lucky Tran. I made mesh with 50 microns near wall and when I ran the simulation, I got y+ around 1.5 near wall and 0.03 in the core. So I refined it to 10 microns and simulating again. I had one.more doubt. For my entry free stream conditions, flow is laminar. But during heating flow reynolds number will iase and becomes transition and turbulent inevitably. So now which model I have to use now, laminar it turbulent. I read for heat transfer analysis, as they take place near wall it's always better to choose k-w SST model. But considering turbulence from the entry won't be over predicting the solution. Thank you

 September 8, 2019, 13:41 #4 Super Moderator   Sijal Join Date: Mar 2009 Location: Islamabad Posts: 4,539 Blog Entries: 6 Rep Power: 49 This way of comparing meshes is very much inaccurate. Choose some integrated quantities such as heat transfer coefficient etc. Or local Nuselt number. Secondly are you sure that you are properly resolving thermal boundary layer? Moreover having very Y+ values will also induce errors in simulation in addition to round off errors. Try to keep maximum Y+ under 10 which is required by latest hybrid wall functions in these softwares praphul likes this.

 September 11, 2019, 15:30 #5 New Member   Sai Krishna Join Date: May 2018 Posts: 13 Rep Power: 3 Thank you Far for ur reply. I checked mass averaged values from reports> surface integrals , the above mentioned parameters heat transfer coefficient ( both wall and surface ) and nusselt. Number. But to my surprise they were all zeroes for all intermediate planes in the flow direction. When I plot it's giving some values for those parameters. Ur second suggestion, resolving thermal boundary layer. How to know it's thickness. I read it's thickness will vary from 0 to Radius in internal flows. So upto whatsl height I need to resolve the mesh. IAM beginner in thermal analysis, so it will be helpful if anyone suggest what to follow for this kind of problems. Thank you.