Strange results of devices

[mst-parkhachev]

Dear colleagues,

Let's give an example of a simplified project. Here we have a two-story building, and a small fire outside. The wall material is inert.
test.txt

The volume inside the building is not connected to the outside. However, DEVCs inside the building register strange values. This is especially noticeable on the top floor.

1. The temperature becomes lower than ambient.
2. The pressure is constantly increasing. It grows quite noticeably.
3. The density of oxygen and carbon dioxide increases.
4. All "RADIATIVE HEAT FLUX GAS" devices show a NEGATIVE value.

Why does this happen? In particular, where does the extra gas in the building come from? Why does temperature drop when pressure increases? Why is heat flux negative?

[drjfloyd]

See 21.1 in the User's Guide

This is due to velocity penetration error with the default FFT pressure solver on the zero-thickness OBST. You could:

Make the walls one cell thick. This greatly reduces this error for the FFT solver.
Decrease the VELOCITY_TOLERANCE and increase MAX_PRESSURE_ITERATIONS
Use another pressure solver like ULMAT

[mst-parkhachev]

The manual says about VELOCITY_TOLERANCE: "The default value is δx/2, where δx is the characteristic grid cell size."
How to convert this value from length to speed?

[drjfloyd]

The speed is dx/2. Consider it as a characterisitic length of dx/2 and time of 1 s.

[mst-parkhachev]

Thank you!