I am trying to create a cone top optimize the flow to a filter. it is essentially a cone with a screen at the end just before the filter. When I run the analysis, it shows the flow going off to one side. Is this even possible? I'm not really convinced that this would happen.
Any opinions?
Could you provide a simple sketch including your Boundary Conditions too please?
Chances are, if this is set up correctly, you do not have sufficient mesh - so an image showing the mesh on a cut-plane would also be very useful if possible.
Kind regards,
Jon
I had added an image but it appears to have not uploaded. I'll try again.
As far as boundary conditions... I have 10LPM inlet and 0 PSI outlet. The thin part at the large end of the cone is a 10% open area screen.
T
I suggest refining your mesh - also could you share the resistance settings? Do you just have FAR or 0.1 in each of the 3 directions? If so, that is OK.
The ideal here is to run this as a 2D axisymmetric analysis, lots of mesh and far faster!
I have a FAR of 0.1 allowing flow only in the K direction.
I am unfamiliar how to run a 2D simulation. Can you direct me to some help on that?
Sure, take a look in the help and Wiki but the basics are:
Sketch a 2D slice of your model on the x-y plane (this is important) - use either the x or y axes for the actual part axis as you are going rotate around as you will use this in CFD.
Then when in CFD you apply everything the same but volumes are now surfaces and surfaces are now edges.
Also be sure to select 2D axisymmetric in the top left (Units) part of the screen.
Then run it. You have the option here for a far better mesh and hugely reduced runtimes.
The downside is that you cannot use traces as you have here - the results are strictly 2 dimensional.
Does that help?
Thanks,
Jon
Do I need to sketch each part in the cross-section? Also If you'll notice how I have the 'filter' on the left side (close-up below). Because our filter is essentially a bundle of straws, we section it this way to prohibit flow in the non-K direction. Would I need to create the individuatl sections of the filter here too?
You can sketch it is a new part within the assembly but then just launch it on its own.
You should be able to change the resistance settings to prevent flow in the directions you do not wish it to travel.
I tried loading a sketch and got an error that the Granite kernel could not launch 2D geometry and to try the Mechanica launch method. When I do this, it actually doesn't launch at all. Any ideas?
Tony,
I agree here with Jon. Knowing some of the geometry you've dealt with in the past, you might be able to look at portions from a 2D axisymmetric aspect. However if you were going to run in 3D (quarter/half symmetry, or full model) I would also suggest refining the mesh in the cone as you can see that teh flow field is following the coarser elements and as you refine you'll start to see tht the results become more symmetric.
I would also encourage you to run ADV5 (in 2014/2013) or ADV2 (otherwise) as it will help with the results in applications like this.