@PhilippoChannel I don't know. I set the viscosity to a realistic value for air at room temperature if I remember correctly. Radiation is not taken into account and there is no conduction through the walls. I dont think the neglection of radiation is a big problem.
@davidosterberg hey.. how about this.. the CFD is it 2d or 3d? Because the window and heater won't likely go the length of the room (in the dimension not shown).
That would probably promote more turbulence, which would appear more natural.
@PhilippoChannel The simulation is 2D, for sake of simplicity. In a sense it is accurate because the external wall in my old apartment (of witch this is a model) consisted mostly of windows and heaters. The world is 3D and the eddies will be 3D. In this case I am sure the eddies will have 3D characteristics. As I said above: this was mainly a demonstration of the flexibility of the Elmer solver. And also, I am by no means an expert on CFD.
@edinbox there is a tutorial, in Swedish, at durotec.se/2011/01/naturlig-konvektion-elmer-och-open-source-cfd/. The Elmer case file is available at the bottom and Google translate should work for the blog post.
what it likely doesnt take into account is radiation..
but..
perhaps interms of accuracy... do you think the air would be less viscous than it appears...?
this sort of seems too "fluid"
PhilippoChannel 3 days ago
@PhilippoChannel I don't know. I set the viscosity to a realistic value for air at room temperature if I remember correctly. Radiation is not taken into account and there is no conduction through the walls. I dont think the neglection of radiation is a big problem.
davidosterberg 3 days ago
@davidosterberg hey.. how about this.. the CFD is it 2d or 3d? Because the window and heater won't likely go the length of the room (in the dimension not shown).
That would probably promote more turbulence, which would appear more natural.
PhilippoChannel 3 days ago
@PhilippoChannel The simulation is 2D, for sake of simplicity. In a sense it is accurate because the external wall in my old apartment (of witch this is a model) consisted mostly of windows and heaters. The world is 3D and the eddies will be 3D. In this case I am sure the eddies will have 3D characteristics. As I said above: this was mainly a demonstration of the flexibility of the Elmer solver. And also, I am by no means an expert on CFD.
davidosterberg 2 days ago
Can you make tutorial about this simulation?
edinbox 5 months ago
@edinbox there is a tutorial, in Swedish, at durotec.se/2011/01/naturlig-konvektion-elmer-och-open-source-cfd/. The Elmer case file is available at the bottom and Google translate should work for the blog post.
davidosterberg 2 days ago
It took about 20 hours on a intel i7 processor. But i dont think the simulation is very accurate. It was more of a test of the possibilities
davidosterberg 5 months ago
hello, how long did your simulation take ?
kuryba00 5 months ago