Models

The sub-models that turn FDS output into agent behaviour, one page each. Each Models page is the specification of its sub-model: the coded form, the parameters and their defaults, the assumptions, the deviations from the literature, and the limitations with their issues. The published laws the models start from are on Fundamentals. Worked runs and the Python API are on the “in practice” pages of the documentation. The evidence that the code solves its equations is on Verification.

What pyFDS-Evac does with each hazard of the Fundamentals figure. A sketch, not to scale: it shows implemented behaviour, not a simulation result. Which models run depends on the configuration and on the slices in the FDS output. Figure inspired by Fig. 1 of the Engineers Australia practice note for tenability criteria (2014).

Walking speed, route smoke, sign legibility and the gas and heat doses read the horizontal FDS slice nearest --smoke-slice-height: 1.6 m by default, the HUMAN_SMOKE_HEIGHT of FDS+Evac. This holds for extinction, the gas species, temperature and, when selected, INTEGRATED INTENSITY. The height is an absolute z in the FDS domain, not a height above each floor. If the nearest slice is more than 0.5 m away, the run logs a warning and carries on at that slice’s height, so check the slice elevations in your case (FDS slice sampling). A deck with slices only at 2.0 m is read at 2.0 m without a warning, and the run manifest does not yet record the height used (#165). --heat-regime layer also reads a TEMPERATURE slice at --heat-layer-height.

Where a value is read depends on the model. Walking speed and the gas and heat doses use the value at the agent’s position. Route smoke is sampled along each candidate route. Sign legibility averages the extinction along the straight line, in plan, from the agent’s cell to the sign, on that one extinction slice (Wayfinding). It is a 2-D line on one slice, not a 3-D ray up to the sign.

• Smoke obscuration (on by default). The extinction coefficient K sets each agent’s walking-speed factor (Smoke-speed model), and the smoke along each route refuses and orders exits (Dynamic route rerouting). For agents not fully familiar with the building, sign legibility through smoke decides which exits they learn (Wayfinding). Smoke is not the only routing input: when the gas FED model is loaded, a route whose projected gas FED exceeds fed_rejection_threshold (1.0) is refused under both cost models for the agent’s current exit, and a route to another exit above 0.9 of it (fed_return_margin).
• Toxic and irritant gases (on by default when the case has CO, CO₂ and O₂ slices). Each agent accumulates a gas FED, irritants included, and by default stops once it reaches 1. The threshold is --fed-threshold (default 1.0); every agent has the same threshold unless --incapacitation-mode probabilistic. With --disable-tenability nobody stops, but the FED is still accumulated. Slowing by irritants (FIC) is opt-in, --enable-fic-speed (Fractional effective dose).
• Convected heat (opt-in, --enable-heat-fed). A heat dose from the gas temperature, kept apart from the gas FED; it can incapacitate an agent by itself. Heat does not slow agents and plays no part in route choice (Heat).
• Radiant heat (opt-in, --heat-fed-method total-flux). A radiant term inside the one heat dose, not a second dose. It comes from the gas at the head by default, from a hot layer with --heat-regime layer, or from the FDS INTEGRATED INTENSITY slice with --heat-radiant-source integrated-intensity (Heat › Total flux).

Sources

• Engineers Australia Society of Fire Safety (2014). Practice note for tenability criteria in building fires, version 2.0, 3 April 2014. Society of Fire Safety, NSW Chapter, Engineers Australia. engineersaustralia.org.au. Fig. 1, p. 7.