Learning Forum

[rd_asm]

I am using SST k-omega model for my problem, inflation layers were applied near the solid (bottom) wall.. the wall has a roughness height (k_s) = 0.0009 m and first layer height I used is 2.5e-04.. but for this wall avg Y plus was found to be around 9.0; is it acceptable?

Note- From Literature I was made to understand that for k-omega model avg y plus should be ~1 but due to roughness upto what value of y plus would be acceptable?

Would be grateful if someone provide some insight into it..

[Rob]

This is where a blind adherence to y+ and solver maths falls down. For k-w you really want a y+ of around 1, but that makes the roughness larger than the near wall cell. What are you modelling?

[rd_asm]

Thank you so much for your reply.. my concern is with that value of y plus, will my results be accepted?
I am solving flow over building problem, inflation applied at the ground.

[Rob]

Probably: with building models remember there's a certain amount of uncertainty regarding the ground level, curbs, window ledges, buses and what the airflow does over the neighbouring countryside. I tend to use RNG k-e with enhanced wall treatment (I can give details as I've presented this at an CIBSE external event but can't share the presentation) and focus mesh based on what I need to know. Are you looking at pedestrian comfort, building wind loading, smoke flow etc?

[rd_asm]

@Rob thank you sir for your Reply..
I am solving basically natural ventilation problem..
I have another confusion in my mind now now.. In natural ventilation analysis in buildings, which region is 'critical' regarding application of inflation layers, the ground or the building walls for better solution..

[Rob]

If you're inside you generally want a "good" mesh throughout and 1-2 inflation layers to give some wall refinement. Given the relatively low flow speed and chaotic nature of the flow you'll find y+ to vary over a wall.

[rd_asm]

yes sir y+ is varying and it is very difficult to keep within the 'range' prescribed in standard literature

[Rob]

Welcome to CFD modelling! I tend to use RNG k-e with Enhanced Wall Treatment: it's a bit more relaxed re the y+ which is also only really useful to set the first cell height. Given all the interesting stuff happens in the free stream I'd be less worried about it.