Ideas for identifying the ‘inlet’ boundary region

[manash.pb]

Hi all,
I am simulating airflow through a human nose. The geometry is taken from an STL file and I want to identify and set the nostrils as the inlet plane. As the structure of the nostril opening is not a proper geometrical shape (not a circle or ellipse and both the nostrils are very closely placed), I am having difficulty in identifying the regions through an IndicatorFunction. Is there a smarter and better way to define the boundary regions? For ex, maybe a list of coordinates may be read from a file, which corresponds to the face centers of the STL triangles making up the plane? Any suggestions will be helpful!
Thanks and Regards,
Manash

[mathias]

Dear Manash,

you can work with intersections of geometry primitives (indicators) like cylinders, cuboids, etc. with the actualy generated voxel mesh by and stl-file or you may extend the region growing apporach (https://www.openlb.net/DoxyGen/html/d3/d2a/classolb_1_1BlockGeometryStructure3D.html#aa1a30cd27881b163d0486556769fcc70) to the superGeometry implementation, which is only available for blockGeometries at the momenent.

Best
Mathias

[manash.pb]

Hi,
Thanks for your reply. I have been trying with different geometries, but still not successful till now. Regarding the region growing, how can I get the iX, iY, iZ for the seed point (I guess these are the lattice coordinate values). How can I access the lattice coordinate values corresponding to the physical coordinates (which I can get from Paraview for the boundary face)?

Manash

[mathias]

Dear Manash,

You need to write a new method which starts the reagion growing for any point within an idicator, lets say asphere, which intersections with a particular material number, e.g. 2, to start the algirthm. But keep in mind that you need to program the parallel region growing algorithm first. Within a colaboration we could help you further but that needs some efford.

Best
Mathias

[manash.pb]

Hi,
Thanks for the reply. I guess I will try to work with the geometry primitives for now. My knowledge of OpenLB (and LBM methods in general) is still a bit sketchy. I would like to spend some time on improving my understanding (of whatever is available in OpenLB) for now. Maybe, at a later phase, I will think about writing methods of my own. I will also look towards attending a OpenLB summer school training in the future to learn better.

Regards,
Manash

[Author]

Posts