Wildfire Ember Transport Modeling

Photo by Max Whittaker on NY Times

Overview

This project is about the numerical modeling of firebrand showers in wildfire simulations.

  • Problem: Firebrands can spread wildfire through the ignition of spot fires, yet there is a gap in knowledge on where they will land due to turbulent wind

  • Solution: Model high-resolution turbulent boundary layers at various turbulence intensities, and release firebrands in those domains to study the effect of small-scale turbulence

  • Results: Novel implementation of firebrand transport model coupled with wildfire simulation WRF-SFIRE for comparison between large-scale and small-scale transport

Problem

  • Firebrand showers are the fastest and most complex form of wildfire spread, by generating spot fires in random locations.

  • There is a gap in knowledge on where firebrands land due to turbulent wind.

  • There is no existing coupled firebrand-wildfire simulation with complex firebrand shapes.

Alt text

Firebrand shapes

Experimental data has shown that firebrands are made of 3 basic shapes: compact, plate, and rod. Studies have shown that particles of different shapes have different flight trajectories. And in the wildfire research community, these difference have not yet been fully explored.

Small & Large-scale Turbulence

Modern wildfire simulations use large grid sizes in their computational meshes (around 250-300 m). This leads to only large-scale turbulence seen in the flow. Understanding the transport of plate and rod shapes in small-scale turbulence is crucial for understanding large-scale transport in wildfire simulations. With this knowledge, researchers will better understand how spot fires are generated.

Turbulent Boundary Layers

Setup

ParametersMesh ResolutionSimulation Cases
U = 2.23 m/s∆x = 0.025 m4% turbulence intensity
Re = 284,000∆y = 0.025 m7% turbulence intensity
L = 2 m∆z = 0.025 m
𝓥 = 1.568E-5

Validation

Velocity profile of turbulent boundary layer simulations was validated with experimental data from Tohidi 2016. Power spectral density of turbulent boundary layer simulations validated with Kolomorgov -5/3 Spectrum to show a fully developed flow.

Results

Q-Criterion Iso-surfaces of the vortices in the 4% turbulence intensity simulation case

Firebrand Transport Simulations

Small-scale Transport

A series of 32 tests were conducted in the high-resolution turbulent boundary layer simulations in uniform and turbulent velocity fields. Plates and rods were released at 4 different heights

Antonio Cervantes
Antonio Cervantes
Ph.D. Student in Civil and Environmental Engineering

My research interests are in wildfires, particle transport, and turbulence