[Jeffzda]

Hi there,

I'm using OpenFOAM with the buoyantPimpleFoam solver to simulate buoyant convective flows that develop by heat transfer to air from the inside walls of an open horizontal cavity. The wall temperature is high, at 723K, and ambient is 300K and the mesh cell size near the walls is about 5mm (see close-up of cavity slice with edges shown). I am currently using a constant transport and constant heat transfer model, though I realise that over these temperature ranges that will have to change eventually. I'm not using a turbulence model at this point.

The 3D geometry of the cavity is shown in the picture called 3d_screenshot.jpg. The geometry of the cavity is shown in silver, and the coloured temperature field is a temperature threshold applied to internal mesh so the extent of the flow can be viewed. As you can hopefully see, the flow has started to develop and part of the flow is leaving the cavity at the top.

This is taken at t = 4; data is recorded every 2 secs. The problem that I am having is that in subsequent timesteps there are no changes in the temperature field. I have verified this by running, for example, "diff 4/T 6/T" and none of the cell data is any different. It's the same for all subsequent timesteps.

The funny thing is that the velocity field continues to change.

The time_4_cavity screenshot shows a slice through the centre of the cavity with temperature contours at t=4. The wallcloseup.jpg file is a close up of the top LH corner seen in time_4_cavity screenshot to give an idea of the mesh size.

I included this because simulations with coarser meshes have presented flows that developed much faster and continued to develop for minutes of simulation time. The only difference has been a more refined mesh near the wall.

I hope someone can help me.

Cheers,

Jeff

[Sherlock_1812]

Dear Jeff,

Where you able to find out why? I'm facing a similar problem.