Taking wall roughness into consideration for simulation (Very urgent)

[Eric.Chua_2015]

Hi,

I am currently working on a project that requires me to apply wall roughness to show the different in heat transfer rate and pressure drop in a heat exchanger. However i am unable to get the different after i apply the wall roughness. I was told that the wall roughness is accounted for in the calculation of the effective wall conductivity only for the k-e turbulence model. However, for both k-e and sst model, i could not get the different in pressure drop and heat transfer rate. I had increase the number of wall layer to 15 as well already.

Thanks!

[Jon.Wilde · Accepted solution]

Only use the ke model, as I think we said already in another thread 🙂

Did you already run a mesh sensitivity study?

Just switching to 15 layers isn't the right thing to do - you need to think about the Y+ value (35-300 for ke) and not just increasing layers.

Hope that helps,

Jon

[Eric.Chua_2015]

Hi @Jon.Wilde,

Thanks for your reply! (:
I will include the step for Y+ value adaptation into the simulation. However how do i ensure the Y+ value to be >35 as the only option is to limit the maximum Y+ value, or is it a automated process that it will be brought up to above 35?
Another question, do i need to turn off the IWF for ke model? or is there any no different for ke model if I turn the IWF off or not?
Thanks and sorry for the trouble (:

[Jon.Wilde]

You likely don't need to adapt, just use the default wall settings, I bet that gets you close 🙂

I think just leave IWF off, but I am not 100% certain on that...

[Eric.Chua_2015]

Hi @Jon.Wilde,

Thanks! I will try it out. 
Just one last question, for the mesh sensitivity study, do i just do it once and the mesh can be reuse for other roughness value or do i have to perform the mesh sensitivity study for every wall roughness value that i input?

[Jon.Wilde]

I would think only once 🙂

[Eric.Chua_2015]

Hi @Jon.Wilde,

Thanks for your help. I have solved the issue. 
Just to close the loop of information, it was as you guys have said, the turbulence model should be k-e model. However, IWF have to be turn on in order to ensure the the distance y+ to be 35< Y+ < 300.