Learning Forum

[chandanapdev]

Hi,

I am doing a project on wind pressure evaluation on roof mounted solar panels. The problem is to find the pressure coefficients at different angle of attacks. while calculating , fluent is showing a message that " turbulent viscosity is limited to a range to specif number of cells". at the beginning, few number of cells are affected, later the number of cells increased. it affects my monitoring values at finally divergence is detected. 

After reading certain solutions given here, i tried to fine my mesh. i tried with 7 million cells and even 1 million cells. Skewness is affected since i use hybrid meshing, but it is in range of 0.97 and 0.96. Aspect ratio is very large when i give tetra-meshing in the inner-core. First i tried steady state. Within that, i have tried with reducing the under relaxation factors. it controlled to a limit, but the divergence occurred after 800 iterations. Inlet and outlet boundary conditions were kept default, still the error occurred. 

Then i tried with transient flow. still the same issue occurred. the residual range was fixed as 1e-6. In both cases floating point exception is found. such huge values are seen for lift and drag coefficients. 

When i did the same case in 0 degree AOA, i obtained a result. When i changed the AOA, these errors occurred. since reverse flow was found, i increased the boundary. at 0 degree AOA, I used outlet as 15H, later in 45 degree AOA, i changed it to 25H. side boundaries were also increased by 50%. the model and mesh is shown below. 

can anyone help me please. 

 

 

 

[chandanapdev]

[chandanapdev]

[chandanapdev]

[Rob]

These models can need the viscosity ratio limit increasing a few orders of magnitude (Limits in the Solver Controls) but in your case I suspect the mesh is more the issue.  Fix the high aspect ratio and skew cells if they're anywhere near any flow gradients and see how that improves the solution. 

[chandanapdev]

Hi sir,

I tried to mesh the whole model with hexa mesh, but so much pavings are coming. with similar meshing (Hybrid meshing as shown) i have tried for 0 deg. AOA.  at that time i got the result. 

Is there any suggestion on how can i change this mesh? eg. fully tetra or any other method?

Thank you sir