CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Pointwise & Gridgen (http://www.cfd-online.com/Forums/pointwise/)
-   -   Meshing a Wheel in Pointwise/Fluent problem (http://www.cfd-online.com/Forums/pointwise/82502-meshing-wheel-pointwise-fluent-problem.html)

Marli November 28, 2010 12:38

Meshing a Wheel in Pointwise/Fluent problem
 
3 Attachment(s)
Hi All,

I am attempting to mesh an external flow around an aircraft wheel (closed rim) in Pointwise to run using (at first) the k-e RKE model with Enhanced Wall Function.

My mesh is built up in sections, first is the boundary layer. Using Y+ of 1 (0.0057mm) and expansion ratio of 1.2, for a wheel D of 1.4m, and reynolds number of 6.71x10^6, 70m/s. The quarter-wheel profile is meshed in 2d then rotated/mirrored around to cover the whole wheel. This is then built up into the far field cuboid mesh, 5.6m x 5.6m x 4.2m (infront) x 10m (behind), 3.6 million cells in total. (see attachments). Pointwise Wall Spacing examination corresponds to equivalent ~ 0.6>y+>1.6 around the whole wheel.

Now, when I run it in Fluent, I get Continuity divergence after only about 50 iterations, and when I do a Turblence>YPlus contour in fluent I often see huge Y+ values (10^3), although this seems to vary depending on how long i run it, so I don't know whether due to the divergence this examination is irrelevant.

So in summary, according to Pointwise, my y+ is pretty consistent, but I get continuity convergence quickly in fluent. I have re-meshed loads of times with same divergence. So not sure whether it's a Pointwise or Fluent problem.

Any help would be much appreciated.

cnsidero November 28, 2010 17:43

Luke,

What are the dimensions of the wheel? I ask because you have to make sure it's scaled properly in Fluent, as it works in meters. For example, if you created a wheel with a diameter of 1000 in Pointwise, representing the units mm, you will have to scale it in Fluent by a factor of 0.001.

Since you reported y+ values 3 orders of magnitude higher than you were expecting, my guess is that your Reynolds number is 1000 too large, hence the incorrect scaling.

Let me know if that works.

-Chris

Marli November 28, 2010 19:27

Genius! :) It worked!

Thanks Chris, this had been bugging me for weeks. It was indeed the scaling so Fluent thought my mesh was 10km long! Now I can move on to the proper stuff :)

Many thanks again, I should have posted this ages ago. Yet another reason why CFD online/wiki is an indispensable resource!

Luke

arapha January 22, 2011 22:49

More scaling/units problem
 
Hi,

I am using a k-omega SST model on a flow going through a duct. I should get my y+ between 1 and 5 I believe from what I've seen. I found out about a scaling issues in Fluent thanks to your previous posts. However I am still having difficulties obtaining the right y+ values from my mesh. When I try to adapt in Fluent it halves the y+ value but horribly increases the number of cells. My mesh y+ is about 60 so that I can't obtain the correct y+ value with a decent sized mesh in Fluent. I'm wondering if this is another scaling issue in Pointwise (importing from CATIA to Pointwise, or in Pointwise, do I have to set the units ? If I go to Properties the ratio of Grid/Database is about 600 ??). Or am I using the adapt function in Fluent wrong ?
Any thoughts or suggestions greatly appreciated.

Thank you !!

tobino January 26, 2011 22:15

Dear all,

I am meshing a model of ship. Have anybody known How to create structure mesh? please advise me !


All times are GMT -4. The time now is 01:13.