CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   wall roughness in enhanced wall treatment (https://www.cfd-online.com/Forums/fluent/101445-wall-roughness-enhanced-wall-treatment.html)

lobstar May 4, 2012 05:10

wall roughness in enhanced wall treatment
 
Hi,

Does anyone know the value of wall roughness used by fluent when the enhanced wall treatment option is selected? I'm aware that you can't set a value when you select this option. what I'd like to know is what value Fluent uses when calculating the flow or how it calculates that value. I ran a series of flows through a thin pipe using an axis-symmetric k-e model with enhanced wall treatment. I also ran a simulation with standard wall functions and set the value to zero. The results were different which suggests that fluent gives the walls a roughness value. I also carried out some theoretical calculations for the flow rate with different wall roughnesses for comparison to my enhanced results. It appears that the value of wall roughness used by fluent changes with the size of the geometry. If anyone could offer any guidance I would be very appreciative.

LuckyTran May 4, 2012 12:41

Quote:

Originally Posted by lobstar (Post 359252)
Hi,

Does anyone know the value of wall roughness used by fluent when the enhanced wall treatment option is selected? I'm aware that you can't set a value when you select this option. what I'd like to know is what value Fluent uses when calculating the flow or how it calculates that value. I ran a series of flows through a thin pipe using an axis-symmetric k-e model with enhanced wall treatment. I also ran a simulation with standard wall functions and set the value to zero.

Rough walls cannot be used at all for enhanced wall treatment for any of the epsilon based models. So how did you run that simulation?

lobstar May 8, 2012 09:13

I am unsure as to what you mean. I ran a k-e with enhanced wall treatment of the flow through a pipe and also ran a simulation of the same geometry also with the k-e model but this tim eusing the standard wall function. I then set both the wall roughness and the roughness factor to zero to simulate a smooth wall. As the results were significantly different I concluded that the enhanced wall treatment doesn't treat the walls of the geometry as smooth. I'd like to know how it does treat them as from my undertsanding the roughness of a wall will affect the flow over it. so my question is how does fluent determine the roughness of walls when enhanced wall treatment is enabled.

LuckyTran May 8, 2012 12:06

Quote:

Originally Posted by lobstar (Post 359932)
I am unsure as to what you mean. I ran a k-e with enhanced wall treatment of the flow through a pipe and also ran a simulation of the same geometry also with the k-e model but this tim eusing the standard wall function. I then set both the wall roughness and the roughness factor to zero to simulate a smooth wall. As the results were significantly different I concluded that the enhanced wall treatment doesn't treat the walls of the geometry as smooth. I'd like to know how it does treat them as from my undertsanding the roughness of a wall will affect the flow over it. so my question is how does fluent determine the roughness of walls when enhanced wall treatment is enabled.

with any of the epsilon based models, if enhanced wall treatment is used, you cannot apply the wall roughness (not allowed by Fluent). so how did you do something that was impossible? perhaps it was allowed in older versions?

Anyway, the enhanced wall function approach is not compatible with wall roughness. If you want to use wall roughness, you should be using the standard or scalable wall function approach. Using the enhanced wall treatment, if you even somehow magically were able to do it, will lead to errors.

the enhanced wall treatment does treat the geometry as hydraulically smooth, same as the standard and scalable wall functions.

lobstar May 9, 2012 08:31

I'm not sure you understand my question. I am aware that you can't apply wall roughness when using enhanced wall treatment, only when using standard or scalable. However I am under the impression that for a turbulent flow through a pipe, the wall roughness has an effect on the flow. I'd like to know how fluent estimates the wall roughness to determine this effect. Can you please define hydraulically smooth? I ask because like I said I ran two separate simulations. One with enhanced wall treatment, one with standard wall treatment but with a wall roughness height of zero. I got very different results between the two. My understanding that a pipe is hydraulically smooth if the roughness height is roughly 0.5% of the diameter. For my pipe of 2mm this should be around 10 microns, or 0.1e-5m as a maximum. What I'd like to know is how exactly does fluent determine the roughness to use in the flow.

LuckyTran May 9, 2012 11:29

Quote:

Originally Posted by lobstar (Post 360097)
I'm not sure you understand my question. I am aware that you can't apply wall roughness when using enhanced wall treatment, only when using standard or scalable. However I am under the impression that for a turbulent flow through a pipe, the wall roughness has an effect on the flow. I'd like to know how fluent estimates the wall roughness to determine this effect. Can you please define hydraulically smooth? I ask because like I said I ran two separate simulations. One with enhanced wall treatment, one with standard wall treatment but with a wall roughness height of zero. I got very different results between the two. My understanding that a pipe is hydraulically smooth if the roughness height is roughly 0.5% of the diameter. For my pipe of 2mm this should be around 10 microns, or 0.1e-5m as a maximum. What I'd like to know is how exactly does fluent determine the roughness to use in the flow.

If you use a roughness height of 0, then the rough wall is a smooth one and there is no change to your wall function. Fluent does not estimate the wall roughness, that is a user input. The determination of the type of roughness is done based on a parameter very similar to y+ of the roughness height, k+ or k*. It just makes the appropriate adjustments to the wall modelling based on the roughness height by computing the wall offset and then applying either the linear law or the rough law of the wall.

lobstar May 10, 2012 08:42

Ok I think I understand, but do you have any idea why there is such a significant change in the mass flow rate? With EWT I got a flow rate of 5e-3 kg/s but only changing the wall treatment to standard and the roughness height to zero resulted in a flow rate of of 3e-2 kg/s. Is k+ or k* anything to do with the Von Karman constant?

lobstar June 29, 2012 04:43

I have been testing out different settings on fluent to try and find the difference between the treatment of wall roughness using the standard wall settings and enhanced wall treatment. However I've found that no matter what I set the wall roughness to on the wall boundary condition under standard wall, there is no change in flowrate. I've gone from zero to ridiculous heights such as 1 metre. Does anyone know of a reason why the simulation is ignoring the wall roughness?


All times are GMT -4. The time now is 17:43.