This paper investigates the cause of ignition of wooden shields during surface fire fronts. For this purpose an experiment was conducted in which a zone 50 m in length and 10 m in width was chosen as the experimental site. Wooden shields (made of spruce boards) and samples of high-moor and valley peat were placed in this zone. In the experiment it was ascertained that to ensure safety during the surface fire in absence of firebrands and embers it was necessary to clear a perimeter around the zone not less than 5 m in width, and remove cleared fuels. To reduce the risk of fence ignition, it was necessary to make the fences permeable. It was found that grass fires turn into peat fires if there is a conductor of combustion above the layer of peat.

A 3-D mathematical model of fuel bed (FB) ignition initiated by glowing firebrands originating during wildland fires is proposed. In order to test and verify the model, a series of experiments was conducted to determine the FB ignition time by a single pine bark and twig firebrand (Pinus sylvestris). Irrespective of the pine bark sample sizes and experimental conditions, the ignition of the FB was not observed. Conversely, pine twigs, under certain parameters, ignited the FB in the range of densities (60–105 kg m−3) and with the airflow velocity of ≥2 m s−1. The results of the mathematical modelling have shown that a single pine bark firebrand ≤5 cm long with a temperature of T ≤ 1073 K does not ignite in the flaming mode the FB, and only the thermal energy of larger particles is sufficient for flaming ignition of the adjacent layers of the FB. The analysis of the results has shown that the firebrand length is a major factor in the initiation of ignition. Comparison of the calculated and observed FB ignition times by a single firebrand have shown that our modelling accords well with the experimental results.

Probability and conditions for ignition of surface litter (pine needles) caused by firebrands is studied in the laboratory conditions. For modeling of firebrands, pine bark of various sizes 10×10, 15×15, 20×20, 25×25, 30×30 mm2 and 5 mm in thickness is used. The experiment was conducted in the absence of wind and at different wind velocities: 1, 1.5, 2 and 3 m/s. To conduct investigations, an experimental setup was constructed for generation of firebrands and their impact on surface litter. The results of experiments have shown that the increase in air velocity leads to the increase in probability of surface litter ignition. Thus, wind plays a role of catalyst in the ignition process, bringing an oxidizing agent to firebrands and supporting the process of smoldering. However, if the wind velocity is insufficient for ignition, then there is only the process of smoldering. The area of “uncertainty”, where there is smoldering of surface litter without transition to ignition, is found to decrease with increasing the wind velocity. Based on the received results, it can be concluded that the ignition curve of surface liter by firebrands is nonlinear and depends on the wind velocity. At the same time, there is no smoldering and ignition of surface litter for all the wind velocities and the particles with a size of 10 × 10 mm2, regardless of their number.

This paper presents the experimental study results on the effect of heat flux emitted by a standard source on the charring and ignition characteristics of wood construction materials (plywood, chipboard, and oriented strand board) using infrared thermography (IRT) in the narrow spectral ranges of infrared wavelength. The time to ignition (TTI), charring rate and depth were obtained for the samples. In addition, the effect of several fire retardants on the charring rate and depth of the samples and TTI was analyzed. All fire retardants contribute to an increase in TTI, which confirms their main function—fire protection. However, the effect of fire retardants differs noticeably depending on the material. A new experimental technique is suggested, with the infrared imaging of the temperature distribution along the end of a sample under the heat flux effect on its frontal surface. The uniqueness of this approach consists in the registration of the entire process of ignition and combustion of the presented materials, which occurs in real time without contact with high spatial and temporal resolution. Using the infrared camera of the research class, it becomes possible to record the entire process from the occurrence of the temperature exposure region to the deep carbonized crater in the body of the material. The results can serve as additional recommendations in the development of fire hazard testing methods for construction materials and fire retardants.

The probability of structural ignition is dependent both on physical properties of materials and the fire exposure conditions. In this study, the effect of firebrand characteristics (i.e., firebrand size, number of firebrands) on wood ignition behavior was considered. Mathematical modeling and laboratory experiment were conducted to better understand the conditions of wood ignition by a single or group of firebrands with different geometry. This model considers the heat exchange between the firebrands, wood layer and the gas phase, moisture evaporation in the firebrands and the diffusion gases of water vapor in the pyrolysis zone. In order to test and verify the model, a series of experiments to determine probability and conditions for ignition of wood-based materials (plywood, oriented strand board, chipboard) caused by wildland firebrands (pine twigs with a diameter of 6–8 mm and a length of 40 ± 2 mm) were conducted. The experiments investigated the firebrand impact on the wood layer under different parameters, such as firebrand size and quantity, wind speed, and type of wood. The results of experiments showed that the increase in wind speed leads to the increase in probability of wood ignition. Based on the received results, it can be concluded that the ignition curve of wood samples by firebrands is nonlinear and depends on the wind speed and firebrand size as well as their quantity. At the same time, there is no ignition of wood samples in the range of wind speed of 0–1 m/s. The ignition of wood is possible with a decrease in the distance between the firebrands with a decrease in the firebrand length. This result agrees more closely with the model.