Discussion Groups

Simulation CFD

Reply
*Expert Elite*
OmkarJ
Posts: 434
Registered: ‎10-02-2012
Accepted Solution

Growth rate for inflation layer

239 Views, 8 Replies
07-24-2013 07:16 AM

Hi

 

I was wondering if it was possible to change the growth rate of the inflation layers for boundary layer? At the moment, the growth rate seems to be between 1.07 to 1.1 for successive layers ( I deduced it by plot digitiser manually since there is no setting for this).

 

Having a larger growth rate will help in obtaining a smooth transition from prisms to tets with limited number of layers, especially when wall functions are used and boundary layer resolution is not needed since the first point falls on log layer.

 

Thanks

OJ

Please use plain text.
Product Support
Royce_adsk
Posts: 616
Registered: ‎08-24-2011

Re: Growth rate for inflation layer

07-24-2013 07:28 AM in reply to: OmkarJ

Hi OJ,

 

I believe the setting you are looking for is within the 'Advanced Meshing Controls' window.  Look at the line 'Enhancement Growth rate' which as you deducted is approximatly 1.1 by default.

 

On occasion I have seen some need to adjust this value.  Just remember this value is only valid when you have surface refinement turned on.

 

Best regards,

Royce



Royce Abel
Please use plain text.
*Expert Elite*
OmkarJ
Posts: 434
Registered: ‎10-02-2012

Re: Growth rate for inflation layer

07-24-2013 07:42 AM in reply to: Royce_adsk

Oh yes, it was an oversight!

 

But even if you adjust it to say 1.9 after turning the surface refinement on, there is no change in growth? I have attached the cfz.

 

Thanks

OJ

Please use plain text.
Product Support
Royce_adsk
Posts: 616
Registered: ‎08-24-2011

Re: Growth rate for inflation layer

07-24-2013 08:50 AM in reply to: OmkarJ

Did you click on the refine button again after you made the adjustment to the setting?

 

-Royce



Royce Abel
Please use plain text.
*Expert Elite*
OmkarJ
Posts: 434
Registered: ‎10-02-2012

Re: Growth rate for inflation layer

07-24-2013 08:55 AM in reply to: Royce_adsk

Yes, once I change the enhancement growth rate, i autosize and refine (making sure there is no exclamation triangle on either of them) and then mesh it, but it seems to have no effect. Am I missing something?

 

regards

OJ

Please use plain text.
Product Support
Royce_adsk
Posts: 616
Registered: ‎08-24-2011

Re: Growth rate for inflation layer

07-24-2013 10:57 AM in reply to: OmkarJ

Hi OJ,

 

Image on the left is from the large growth rate and the image on the right is from a low growth rate (default advanced settings throughout).

 

7-24-2013 12-50-58 PM.png

 

-Royce



Royce Abel
Please use plain text.
*Expert Elite*
OmkarJ
Posts: 434
Registered: ‎10-02-2012

Re: Growth rate for inflation layer

07-25-2013 02:02 AM in reply to: Royce_adsk

Thanks. I think we two are talking different interpretations of the term growth rate. For the current  cylindrical geometry, SimCFD's definition of growth rate is along the axis. What I wanted is a different growth rate in the direction perpendicular to the wall (ie radial direction).

 

Please refer to the snap below:

 

enhacnment.png

 

In image 1, we see a small growth rate of layers perpendicular to the wall, and hence it necessicisates larger number of prism layers to have smooth transition from prisms to tets. In Image 2, with larger growth rate, we can use smaller number of prism layers to achieve a smooth transition to tets. Image 3 demonstrates the situation that is to be avoided as it uses smaller number of layers and smaller growth rate (1.1) and thus we have sudden expansion of cell size from prisms to tets. Image 2 achives a smooth transition with economical mesh size. 

 

Thus my interpretation is "growth rate of individual layers, Δy2/Δy1". SimCFD's interpretation is "growth in total thickness of layers along the axial direction".
 
Is it possible to change Δy2/Δy1?
 
Thanks
Omkar.
 
PS: I didnt see the change as you showed in your pics because I didnt have different surface mesh sizes on left and right faces of the cylinder. So There can't be any growth along aaxial direction. 
Please use plain text.
Product Support
Royce_adsk
Posts: 616
Registered: ‎08-24-2011

Re: Growth rate for inflation layer

07-25-2013 07:03 AM in reply to: OmkarJ

Hi OJ,

 

I understand now.  What you need to use is the new layer gradation functionality added in 2014.

 

http://wikihelp.autodesk.com/Simulation_CFD/enu/2014/Help/0000-What_s_N0/0001-What_s_N1/0004-Meshing...

 

-Royce



Royce Abel
Please use plain text.
*Expert Elite*
OmkarJ
Posts: 434
Registered: ‎10-02-2012

Re: Growth rate for inflation layer

07-25-2013 07:06 AM in reply to: Royce_adsk
We currently haven't migrated to 2014 and hence haven't explored it. Will give it a go once we start using it more.

Cheers
OJ
Please use plain text.