I have been doing some simulation on a very easy part geometry (plate) using Dual Domain and 3D and keeping all other parameter the same.
If I compare both pressure and temperature calculation the difference is more than 100%.
During this study I have tried also very fine and coarse meshes, but the differences stay almost the same.
Can anyone confirm my experience? Are there any certain points I have to take into account that I might have missed?
Solved! Go to Solution.
from the very short info you are sharing, i would focus on:
1. going through tutorials to make sure you are setting up the model correctly on both softwares.
i suppose the meshes are acceptable for both approaches (quality criteria speaking)
thank you for the answer!
To explain in more detail what I mean, I added two pictures. As you can see, for a very simple geometry (using all same process parameter, using same material data, just changing the mesh) the calculated pressure is (in special near the gate) very different for Dualdomain and 3D.
Even changing the mesh density (what I additionally tested) no real changes for 3D or Dualdomain occurred. I have assumed that the pressure calculation for such a simple part should be similar, even when using different meshes.
I hope someone can explain this huge difference to me.
I tried to make a quick model like yours here, comparing dual domain vs 3d and changing basic stuff like: solver, laminae number, inertia, etc. and I keep seeing the same difference in terms of predicted pressure. (I used a regular PC material)
Although it isn't a 100% difference, there is a small variation observed (~15% difference).
A couple of pointers from my experience:
1. Since I can tell that it is a chunky flat surface (8mm thick?), I would recommend to go with the 3D model checking the inertia and gravity effects on the solver parameters; this in order to really capture the flow inside the 'chunky' thickness of the part, and evaluate potential jetting, etc...a dual domain mesh would not tell you exactly what is going on in this area, instead it will potentially give you a different end of fill due to the nature of the mesh.
2. the gate area on these type of meshes is modeled with several cones covering the entire gate area, instead of only 1 cone, but I guess for evaluation purposes it can be neglected, but just FYI. (see tutorials for reference).
Hope I was helpful, and good luck with your analysis...
There is a a key mesh assumption that seems to be getting overlooked in this comparison.
When using only an injection cone on the part, DD and midplane models treat the injection node as a mathematical starting point for the flow calculations.
The 3D Flow solver, on the other hand, requires an implied gate size to provide accurate flow front prediction, more accurate pressure results and better jetting representation around the gate when no actual gate or runner system is modeled. In much older versions this was based solely on the mesh density in the area of the injection node, but in newer versions we have a setting in the advanced options (see image) and the calculation is a little more intricate. More information on this can be found in the Help.
My first recommendation would be to model a simple feed system, or at least a small gate segment using beams, for the comparison. This will ensure a more "apples-to-apples" comparison between the mesh types.
Thank you both for the answers!
I will follow the tips and see, if the results are more compareable then.