The 10 Commandments of FDS

[Kevin]

1. Thou shalt not use mesh cells that have an aspect ratio larger than 2 to 1.
2. Thou shalt not change grid resolution by more than a factor of 2 at a mesh interface.
3. Thou shalt not create mismatched obstructions or hollow ducts at a mesh interface where the grid resolution changes.
4. Thou shalt not generate high speed flow through a small opening, in particular openings intended to model leakage. There are better ways to model leakage.
5. Thou shalt not create a sudden change in the heat release rate.
6. Thou shalt not use ADIABATIC as a solid phase boundary condition. This parameter should only be used for testing because no solid material has a thermal conductivity of zero. Using this parameter can cause unphysical fluctuations at the solid-gas interface.
7. Thou shalt be careful about the removal or creation of an obstruction, like the opening or closing of a door.
8.  Thou shalt not use a high (>100 g/mol) molecular weight fuel molecule that is not in the FDS database. In such cases, reduce the molecular weight but maintain the atom ratios.
9. Thou shalt be careful making fires in  long, sealed tunnels, in which pressure fluctuations can cause spurious numerical artifacts. 
10. Thou shalt not complain that a previous version of FDS worked when you broke one of the above commandments. It probably worked because two wrongs made a right, but one of those wrongs has since been rectified.

[Kevin McGrattan]

These commandments are like the actual Ten Commandments in that we break them all the time. But I have added this list to the User's Guide as a checklist for those who get a numerical instability in an FDS simulation. These seem to be the top reasons for the instability, but it is difficult for the input logic to detect these issues without becoming overly restrictive. Like I said, we break these rules all the time, but it's good to know when we're skating on thin ice.

[Matt]

Hallowed be thy numerical modelling.