[IsentropicFlow]

Hey all, I am familiarizing myself with Star-CCM+ and using it for a CFD project I am working on. I have a flow geometry modeled with one inlet and two outlets and am noticing a weird discrepancy between the inlet area calculated by Star using a sum report and the measure tool in the CAD modeler window in the geometry step.

The inlet diameter is 0.609" and the area should be 0.291289 in^2. The CAD modeler window correctly calculates the radius as 0.3045" and the area as 0.291289 in^2. However, when I create a report to calculate the area, Star gives the inlet area as 0.2903843 in^2, which is not equal to the area measured in the CAD modeler.

I am creating a User > Sum report and setting the field function to Area: Magnitude and the Parts to the inlet BC under regions. I have not generated a mesh or anything at this point, it is just the raw geometry file and report.

If I proceed with the simulation and set the inlet velocity to the magnitude corresponding to my target mass flow rate (1.782 kg/s), I instead get 1.759499 kg/s, which matches a flow area of 0.287611198 in^2 (still not sure why this is even smaller than the 0.2903843 in^2 value Star gives from the report).

I am a bit perplexed by this and am wondering if anyone has any insight to give. I have no clue how to go about resolving this.

[mCiFlDk]

Hi Frank,

If I understood well, you created (or imported) a geometry in the CAD modeler, and when measured the area of one of the faces you get a certain value. However, when you re-measure it with a report, you get a slight difference.

In order to use a report, the CAD model (igs, step, etc.) has to be converted into a STL file. StarCCM is only able to work with triangles, and my bet is that when you converted the surface into a triangle-based file, you lost some precision.

Could you send a picture or describe what are your tesellation parameters?

You get them via: Geometry > 3D-CAD Models > Your-model-name > New Geometry Part

[IsentropicFlow]

Originally Posted by mCiFlDk

Hi Frank,

If I understood well, you created (or imported) a geometry in the CAD modeler, and when measured the area of one of the faces you get a certain value. However, when you re-measure it with a report, you get a slight difference.

In order to use a report, the CAD model (igs, step, etc.) has to be converted into a STL file. StarCCM is only able to work with triangles, and my bet is that when you converted the surface into a triangle-based file, you lost some precision.

Could you send a picture or describe what are your tesellation parameters?

You get them via: Geometry > 3D-CAD Models > Your-model-name > New Geometry Part

Yes, that understanding is correct. The model I imported was a Parasolid binary file, if that somehow affects how Star handles geometry compared to something like an STL file.

The tessellation parameters I've been using are just the default values Star uses when you generate a new geometry part. I've attached a picture of the part create options I used when creating the part.

• Coincidence Tolerance: 1.0E-5
• Mark Feature Edges: Sharp CAD Edges
• Sharp Edge Angle: 30.0
• Tessellation Density: Medium

[mCiFlDk]

Yes, definitely that might be the issue.

As it seems you need significant precision, I'd firstly reduce the coincidence tolerance some orders of magnitude.

Now take a look to the other tolerances I highlighted. These are the ones for "medium" tesellation density. If you select "Tesellation Density" as "Very fine", you'll see how much they change. And if you still need more precision, you can play with the tolerances a bit more until you feel comfortable with the results you're getting.

Cheers!

[IsentropicFlow]

Do you have any recommendations for the coincidence tolerance and tessellation parameters? I've never actually messed around with this setting and just kind of assumed the default values Star applies works (it worked for the last project I worked on).

EDIT: I looked at tessellation parameters questions from past users and according to this post, the tessellation parameters don't mean anything unless you have meshing with CAD projection disabled. When I go to mesh, I'm using a surface remesher and have the Project to CAD option enabled, so shouldn't tessellation parameters not matter according to this forum post?

[mCiFlDk]

Hi Frank,

Sorry for the late reply. Answering in order:

1) I'd use a coincidence tolerance between 1e-7 and 1e-9 for your specific case.

2) Ok, in that case, have a look to this documentation fragment (pic 1 attached) where it talks about how CAD projection is managed internally. Have you marked all the key/main edges to that the projection is done correctly?

In any case, just have a try deselecting the CAD projection option and tesellate the CAD with the following features (pic 2 attached), or even changing the values "0.3" by "0.1" if it's not fine enough.

Hope it helps.

[IsentropicFlow]

I set Mark Feature Edges to All CAD edges to no avail. It seems like the default option of sharp edges also defines sharp edges as any edge formed between two surfaces with an angle >30 degrees, which is the case for my geometry.

I tried tessellating the CAD with the options specified, and when looking at the report immediately after part generation, the area it spits out is 0.2912879 in^2. This is pretty much the area given when I measure it in CAD, which is great.

But whenever I try to mesh the part and re-measure the area with the report, the area changes down to 0.2876110 in^2. I have mesh with 3D CAD association enabled too (exact same settings as the ones you posted). Is there something else too under mesh generation that can affect the tessellation?

I truly appreciate your help with this though, thanks for all the work you've done trying to help me figure this out

EDIT: The area changes after I mesh the surface. Only executing the surface mesher, the area changes to 0.2898125 in^2, and presumably, lowers itself even more when the volume mesh starts. I've attached the Parasolid geometry file and Star-CCM+ file I've been working with, if you want to take a look some time on your own and mess around with a few settings: MediaFire Link. (I couldn't upload the .zip file directly)

[mCiFlDk]

Hey Frank,

I can't open the sim file because I don't have access to the newest star version. Would you be able to save it as star19 or lower?

[IsentropicFlow]

I don't see an option to save as with an older version of the file. The only file extension available is a .sim file. I also tried exporting and have a few options to select, but nothing that specifically states a legacy version of Star.

[Aeronautics El. K.]

Can you show the definition of the report? Which representation of the part do you use in the report? Do you use the geometry representation or the surface/volume representation?

If you use the latter, refine your mesh.