Fixing internal long edges

[dinesh27mec]

Hi,

While meshing a part, i am stuck up with internal long edge length ratio of 3.5. As per moldflow guidelines it should be within 2.5,please guide me  how to fix it ? what effect does this has over the final results ?

[kristen.kilroy · Accepted solution]

Hello Dinesh,

The internal edge length is referring to the edge length of the elements going through the thickness of your part. This can have an influence on how the simulation predicts flow of polymer - large edge length results in a longer distance between nodes used to calculate flow, ultimately leading to potential for things like flow pattern to not be captured correctly. 

This edge length can be modified by experimenting with both the mesh's Global Edge Length and the Maximum allowed edgelength through the thickness. If you go into the "Generate Mesh" dialog, when meshing with 3D elements, then select the "Tetra" tab. Here, you can assign the 'Maximum allowed edge length through the thickness". To gain a better internal edge length ratio, I would suggest experimenting with these edge lengths.

I've included screenshots below of the internal elements of a rectangular part below. Both of these have the same global edge length, however I changed the "Maximum allowed edge length through the thickness" so as to better see what the effect is. The max internal edge length ratio for the 1mm mesh is 2.12, while the max internal edge length for the 0.25mm mesh is 1.97. 

[Below: Maximum allowed edge length through the thickness = 1mm]

[Below: Maximum allowed edge length through the thickness = 0.25 mm]

I hope this helps! 

Regards,

[maxmarautodesk]

Dear Kristen,

thank you very much for this insight into the topic. I got curious while reading your post for the following reason:

First, I totally agree that it needs a lot of playing around with the mesh parameters to get an optimal mesh. I am doing a lot of trial and error myself, as I think there is no standard receipe and plastic parts can be very different in shape and thickness. I believe I am speaking on behalf of the community, if I say: if you have some general recommendations, let us know.

Secondly, you don't give a clear statement which of the two shown meshes is to be prefered. One could interpret that because of the numbers you give, the lower one is the better one.

Nonetheless, just by looking at the appearance of the mesh, I am personally quite indecisive and even slightly tend to prefer the upper mesh over the lower one, because it looks much more homogenous. However, the lower one looks of course finer trough the thickness. My main point is though that the max. internal edge length ratio is larger for the upper one (upper: 2.12, lower: 1.97), but that does not say anything about how many tetras are affected by this. I think it is also important to look at the average edge length ratio. I believe a few tetras with high edge length ratio can be acceptable, if the average is significantly lower. This is just a heads up for Dinesh and everybody else and I am looking forward to any further comments - from you or others.

Best regards,
Max

[xusho]

Hi Max,

I would suggest to be careful when fixing 3D meshes.

Moldflow meshes are anisotropic and layered as required by solvers. When 3D Mesh Repair Wizard reports long edges, you need to isolate them into a separate layer and look at them. They may not be too bad.

If you fix them by the wizard, the layered structure may be destroyed.

Sometimes, to re-generate mesh with different parameters may be a better option if you really see very long edges or very large tetras. Otherwise, not necessary to fix.

Shoudong Xu

Moldflow Meshing

[sabarinathan1021]

Hello Dinesh,

Yes..Tetra with internal long edges can cause distorted results.

Temperature at a node, for example, takes into account the temperature of all nodes that it is attached to. Assume flow is from left to right in the Figure 2 above and the aim is to predicted the temperature at node E. Because of the larger distance between A and E, the contribution of the temperature at A to the predicted result for the temperature at E will be overstated in any calculations.

How to fix it....

It is recommended that the number of internal long edges be minimized.

It may not be nessecary to eliminate all internal long edges if they occur in non critical parts of the mold. The Mesh Repair Wizard can help rectify this problem: (Mesh tab> Mesh Repair panel > Mesh Repair Wizard). Alternatively: (Mesh tab > Mesh Repair panel > Remesh Tetras).

This way you can fix the internal long edges.. 

[HNW_CHM_2012]

Xusho,

             I'm also experiencing the same problem but when i separated them in to a layer and had a look they look nothing different from the other elements. So i went on correcting using the tiring option in Mesh wizard which got struck after a while .  

I would like to know what are the possible reasons for this problem . Also please note that the Aspect ratio for the dual domain mesh is 12 and the mesh match percentage is 67% (i know its less because of the gemoetry).

Please advice so that we can avoid such problems .

Thanks and regards ,

 Ram 

[xusho]

Hi Ram,

You are doing the right thing:  look at the diagnostic results before repairing.

Mesh diagnostics is conservative, and it may highlight elements that are not too bad if you look at them individually.

In your case, since they look normal, you can ignore them.

It is NOT compulsory to repair 3D meshes.

[mppkumar]

Hello Kristen

In case 2 (Maximum allowed edge length through the thickness = 0.25 mm), we have finer mesh and number of layers also increased

why so? can you explain (Maximum number of elements through thickness restricted to 6 only)

[mppkumar]

Hello Shoudong Xu

In your posts, you are mentioning "NO mesh correction required for 3D mesh"

If it's not required, why in help files, required / mandatory 3D mesh criteria specified

Is these criteria not required at all or what % of irregularities in 3D mesh is allowed

can you please redefine the guide line for 3D mesh

[xusho]

What I am suggesting is: Look at diagnostic results individually before repairing. Some highlighted elements may not be bad. Not all of them need to be repaired.

As for "compulsory" or not, please use the instructions in help.