CHAPTER 5 FLASHOVER OCCURRING IN LARGE-SCALE ENCLOSURES
5.3 R ESULTS AND D ISCUSSIONS
5.3.4 Temperature contours on the surface of enclosures
146
Temperature distributions at the center of the door at the onset of flashover or at 1200s are shown in Fig. 5-9. At the end of 1200s, temperature at the center of the door is extremely low for test F1 and F4. At the onset of flashover, the maximum temperature at the door is about 600 . Thus, it is believed that flame front is at the place where temperature is about 600 . There is little temperature difference between the left door and the right door for test F2 at the onset of flashover. However, for the test with beams constructed, temperature at the left door is distinctly higher than temperature at the right door, especially for test F5 and F6. It is shown that constructing beams under the ceiling changes the direction of flame spread, resulting in the time difference of the flame running out of enclosures through doors. The location of the extrusive flame can be determined by temperature distributions at the center of doors. The flame running out of the door is along the top of the door for test I, F3 and F5. For test F2 and F6, the flame running out the door at the 1.6m above the floor.
(b) F2
(c) F3
(d) F4
389 511372
529 667
593
Distance from left wall (m) S1=120s
805 830764
751 530 426 362 970
0 2 4 6
0 2 4 6
389 372511
529 593
667
Distance from left wall (m) S2=274s
764 805
362 830 751
426 530 970
0 2 4 6
0 2 4 6
537 519
681 835 719
789
Distance from left wall (m) Flashover at 316s
651 970 818 982
795 705 885 887
0 2 4 6
0 2 4 6
181 554479
231 404
Distance from left wall (m) S1=328s
591
466 506 779
569
292
175
0 2 4 6
0 2 4 6
175 421376
208 287
Distance from left wall (m) S2=952s
485
349 501
200
166 370 678
0 2 4 6
0 1 2 3 4 5 6
7 589
324 454 637
604
Distance from left wall (m) Flashover at 1110s
706
561 645
338 625
438 910
0 2 4 6
0 2 4 6
62 176174
84 83
Distance from left wall (m) S1=266s
239
91 179
29 283
72 115
0 2 4 6
0 2 4
6 125
76 130
74 70
Distance from left wall (m) S2=606s
147
73 132
37 161
92
74
0 2 4 6
0 2 4 6
95 486569
449 287
Distance from left wall (m) S3=738s
436
230 542
94 544
280
45
0 2 4 6
0 2 4 6
148 (e) F5
(f) F6
Fig. 5-10 Temperature contours on the ceiling at different times (S1, S2 and S3) ( ) (Flame front is at the place where temperature is 600 for the enclosure with the flat
ceiling and is at the place where temperature is 400 for the enclosure with beams constructed)
The temperature field on the interior surface of enclosures is illustrated by means of the contour plot. Temperature contours on the ceiling, on the left wall, on the back wall and on the right wall were separately drawn. In order to investigate the progress of flashover occurring in the enclosure, three stage plots are produced, time to rapid
407
218 249 435
280
Distance from left wall (m) S1=234s
544
327 514 604
93 556
225
0 2 4 6
0 2 4
6 244
151 211
133 132
Distance from left wall (m) S2=616s
293
162 230
69 198
109 510
0 2 4 6
0 2 4 6
505 633
483 621
508
Distance from left wall (m) Flashover at 710s
823
574 815
331 977
784
503
0 2 4 6
0 2 4 6
72 242247
103 75
Distance from left wall (m) S1=212s
184 295 185 211
67 53 386
0 2 4 6
0 2 4 6
67 150
79 140
70
Distance from left wall (m) S2=606s
157
112 130
55 54 125 303
0 2 4 6
0 2 4
6 584
298 428 610
450
Distance from left wall (m) Flashover at 778s
809
562 777
507 681
507 965
0 2 4 6
0 2 4 6
transition and the time at the peak of the heat release rate were plotted by Matlab. The measured temperature is shown in the figure as well.
As temperature contours are capable of representing flame spread on the surface of enclosures, corresponding to down arrows drawn in Fig. 5-7, the measured temperature and temperature contours on the ceiling are plotted in Fig. 5-10. For the test with the flat ceiling, flame front locates at the place where temperature is 600 . While beams constructed under the ceiling, flame spread is toward to the front wall. It is defined that flame front locates at the place where temperature is 400 for the test with beams constructed. Since interior finish materials are not lined on the ceiling, flame front is confined to the initial fire influenced region at the first 600s for test F1. With the burning of interior finish materials at the initial fire influenced region, flame front moves in a short distance toward to the front wall at 922s. While lining interior finish materials on the ceiling of test F2, flame spread is along the diagonal direction of the ceiling and passes through the center of the ceiling at 274s. At the onset of flashover, flames run out of enclosures through doors. Unlike flame spread in the enclosure with the flat ceiling, flame spread is changed by constructing beams under the ceiling. Flame spread moves along the length of beams to the front wall. Flame spread along the length of beams is faster than along the direction perpendicular to the length of beams for the test with beams constructed. Affected by the heat release rate of the initial fire and the area of interior finish materials lined on the "T" region, the flame covered area is different at the time of S1 for the test with beams constructed. At the time of S2, the flame covered area is significantly decreased. At the onset of flashover, flame front closes to the front wall. However, flame front arrives at the middle of the ceiling at 738s and flashover does not occur for F4.
150
In a parallel manner, the measured temperature and temperature contours are displayed in Fig. 5-11. Flame front is defined at the place where temperature is 600 . As few thermocouples are mounted on the left and on the right wall for test F3, F4, F5 and F6, only the measured temperature at different time stages is shown in the figure.
The high temperature region is concentrated at the "T" region after the initial fire ignited.
As shown in Fig. 5-11(b) and Fig. 5-11(c), rapid transition to flashover, at the time of 274s and 952s, commences when flame front hits the right wall and turns to toward to the front wall. As shown in Fig. 5-11(c), flame front hits on the side wall at 328s (time to the first peak), but rapid transition does not occur as flame spread is reduced while burnout appears. We notice that the flame covered area exceeds 0.2 times of the back wall area at the onset of flashover. This result will be used to define the flashover occurrence in enclosures.
He ig ht (m )
100
He ig ht (m )
He igh t(
m)
100
He ig ht (m )
He ig ht (m )
He ig ht (m )
600
He igh t(
m) He igh t(
m) He igh t(
m)
He igh t(
m) He ig ht (m )
He igh t(
m)
600
He igh t(
m) He ig ht (m )
600
600 He ig ht (m )
He ig ht (m )
He ig ht (m )
He igh t(
m) He ig ht (m )
He ig ht (m )
He igh t(
m)
He ig ht (m )
He ig ht (m )
He igh t(
m)
He ig ht (m )
He ig ht (m )
He ig ht (m )
He ig ht (m )
He ig ht (m )
He ig ht (m )
He ig ht (m )
He ig ht (m )
He igh t(
m) He igh t(
m)
He ig ht (m )
He igh t(
m)
He igh t(
m) He igh t(
m) He igh t(
m)
He ig ht (m )
He ig ht (m )
He ig ht (m )
600
He igh t(
m) He ig ht (m )
He ig ht (m )