And, continuing with this latest unexpected discovery (during my 01234 layer investigation) of how yPlus area average significantly affects CD and CL values in wall function meshes…
I am flooding this post with data to see if anyone can see a relationship that I do not… Right now, on what I believe is an extremely good (I think by anyone’s standard, with over 98% of the surface cells in the y+20-300 range) 3 layer wall function meshes, the CD and CL results vary nearly 30% depending in what y+ area average that I chose to mesh it at in the y+30-300 range.
Here is the current results table that shows the issue and here is the extended 01234 layer test project that has been continued into Stage 2 and Stage 3 to include all the meshes and sim runs of these next two tables. In my project, every Sim is reserved for a single mesh that has a fully converged and amazingly results stable Sim Run on them. There is no need to make a copy of the project so you can find the buried meshes that are used in any results table of mine. Please have a look at the project, meshes, sim runs and solution sets to see if you can find a irregularity that I missed:
In that table, I have added a column for a meshes ‘Cell - Max Volume Ratio’ (MVR) (from the meshing log) and I have highlighted meshes with MVR over about 40 as yellow and < about 40 as green… (not sure if that is a good
dividing point but this is a quality metric I have no experience with and I needed to start somewhere)…
Also in that table, I chose to look in-depth at the number of mesh cells <20 and >300. I decided to highlight in RED all meshes that had >1%ofArea <20y+ or >1%ofArea >300y+. If this is a good dividing point for this new proposed metric we can filter out a lot of meshes in the table and the CD and CL results now vary only about 9% over the two green meshes of this new metric.
Then, I also ran a full range of y+ meshes with the 1 layer mesh and here are the results colored with the same metric ranges I discussed above and only 1 out of 8 meshes remain green:
Am I heading in the right direction to make sense of these results
I will edit this post and add all the Residual and results plots and histograms etc. now, which will make this post very long, but sometimes you just need to see these things quickly, all on one place to make sense of things
The impact of the changes between the following images is best viewed by clicking on the first image and then cycling through them with your Left and Right arrow keys.
I will start with the 5 yPlusHistograms of the ‘3 Layers at various yPlus values…’ table:
Now the 8 yPlusHistograms of the ‘1 Layer at various yPlus values…’ table: