Experimental study on the early stage of upward flame spread with cross air flow

Deng, Z., et al.: Experimental Study on the Early Stage of Upward Flame Spread ... THERMAL SCIENCE: Year 2018, Vol. 22, No. 6B, pp. 2995-3002 2995 EXPERIMENTAL STUDY ON THE EARLY STAGE OF UPWARD FLAME SPREAD WITH CROSS AIR-FLOW by Zhongkai DENG, Jinfeng MAO *, Zheli XING, and Jin ZHOU College of Defense Engineering, PLA University of Science and Technology, Nanjing, China Original scientific paper https://doi.org/10.2298/TSCI160705200D In this work, experiments were conducted to study the upward flame spread with cross wind at an early stage (prior to the acceleration of pyrolysis spread rate of the wide slab). An exponential model was fitted to the experimental data of the spread rate of pyrolysis front and the dimensionless cross wind speed, which showed satisfactory results. The pyrolysis front tilt angle showed a decreasing trend with a low cross wind speed. However, at high cross wind speeds, the pyrolysis tilt angle exhibited an increasing trend with the maximum value of 45°. The flame lengths increased with the cross wind for narrower slabs, whereas the phenomenon was most pronounced for the narrowest slab. Additionally, the flame length did not exhibit the lengthening phenomenon for the wider slab (0.1 m). Furthermore, the flame tilt angle did not exhibit significant change over time (even along the pyrolysis length). The correlation of flame tilt angle with the cross wind speed and width was also obtained in this study. The flame tilt angle presented a power-law increase with respect to the dimensionless cross wind speed. Key words: upward flame spread, cross wind, pyrolysis front, flame length Introduction Upward flame spread is the most rapid and hazardous mode of flame spread [1], and therefore, is of significant interest in fire sciences [2, 3]. The upward flame is affected by several factors during its spread, including slabs width, sidewalls, altitude, and combustible material. Extensive research efforts have been devoted to investigate the influence of these factors on the upward flame spread. Pizzo et al. [3] performed experiments with slabs of 0.025-0.2 m width to study the effects of width on the spread of upward flame at an early stage. The results revealed that, for widths of greater than or equal to 0.1 m, the flame and the rate of spread were found to be width-independent. However, for the narrower slabs (0.025 m and 0.05 m in width) a transition from laminar to turbulent was observed during the entire observation time. Nevertheless, Tsai [4-6] conducted a series of experiments using polymethyl methacrylate (PMMA) slabs of width 0.1-0.7 m with side walls, and observed width effects for the entire range of widths studied. Compared to flames without sidewalls, the existence of sidewalls produced higher flames and generally less heat feedback. These resulted in higher rates for the spread of flame for narrower flames, and vice versa. Liang et al. [7, 8] studied the effect of altitude on the spread of flame over PMMA slabs by conducting flame spread tests in city of Hefei (at an altitude of 29.8 m) and city of Lhasa (at an altitude of 3658.0 m). Compared with the results from Hefei, the lower ambient pressure in Lhasa transformed the delayed transition to a turbulent flow in a single stage of the * Corresponding author, e-mail: 2996 Deng, Z., et al.: Experimental Study on the Early Stage of Upward Flame Spread ... THERMAL SCIENCE: Year 2018, Vol. 22, No. 6B, pp. 2995-3002 flame spread process, whereas the flame spread rate was found to be about half of that in Hefei. Shih and Wu [9] studied the flame interaction effects on the spread of flame over multiple vertical cellulosic papers for a variety of configurations. The influence of a corner configuration on the upward flame was studied [10, 11], in which the pyrolysis front presented an M-like shape. Most of the previous studies were conducted on the upward flame spread without considering any influence of the external wind. However, the upward flame spread with the external cross wind occurs frequently in fire scenarios. The external wind significantly influences the spread of flame. In this study, the upward flame spread with cross wind was studied experimentally. The experiments were conducted on PMMA slabs with widths of 0.025, 0.05 and 0.1 m, and at cross wind speed lying within the range of 0-1.2 m /s. Various influencing factors, such as the pyrolysis front spread rate, pyrolysis front tilt angle, pyrolysis front temperature gradient, flame length and flame tilt angle were analyzed. Experimental Figure 1 shows the schematic of experiment apparatus, which consists of a combustion platform and a wind tunnel. The wind tunnel had a variable speed fan, which produced stable longitudinal air-flow. A piece of honeycomb cloth was installed about 2 meters from the fan to obtain uniform air-flow within the wind tunnel. The combustion platform was placed at a distance of 0.3 m from the end of the tunnel. FurWind tunnel Fireproof thermore, PMMA was used as the combustion board Video camera slab due to its suitable thermophysical properties. PMMA sample Flame The PMMA slab was 12 mm thick, and was surrounded by a fire-proof board having the same Thermal infrared imager Transition thickness (12 mm). The fire-proof surface was section kept flushed with PMMA slab in order to ensure Front view Side view the surface flame. The fire-proof board was fixed Figure 1. Schematic of the experimental set-up in a steel frame which was placed on a platform. Three anemometers, each having the accuracy of ±0.01 m /s, were placed vertically parallel to the central line of PMMA slab to measure the cross-flow air speed on the PMMA slab. The variable speed fan was adjusted to acquire certain air-flow speeds (0.4, 0.8, and 1.2 m /s in this study) across the PMMA slab. The gears of the fan were recorded, and then, the anemometers were removed. When different experiments were performed, the fan was adjusted to the recorded gears to produce certain air-flow speeds. The surface temperature of the slab was recorded using the infrared camera, whereas the pyrolysis front was tracked as the temperature contour of 350 °C. Even the temperature of the surface suffered from the effect of flame at the front of the slab. However, the determination of the pyrolysis front position was not significantly affected due to the reason that the temperature gradient was very high near the pyrolysis front. The rate of upward flame spread was determined by analyzing the infrared video recordings of each experiment. The accompanying software allows the tracking of pyrolysis front as it moved upwards as the 350 °C contour. The method exhibited satisfactory accuracy, and was used extensively in previous researches [4, 6, 7, 12]. The infrared camera could work in the temperature range of –20 °C to 450 °C with the accuracy of ±2 °C and the thermal sensitivity of less than 0.05 °C. The imaging performance was in the form of 320 × 240 images, wh (...truncated)