Comparative experimental study of fire-resistance ratings of timber assemblies with different fire protection measures

Abstract

This article presents the results of a comparative experimental study to investigate the effects of using three different fire protection measures to improve the fire endurance of timber assembly. The three fire protection measures were fire-retardant intumescent coating, gypsum plasterboard and filling the timber assembly void with mineral wool. The fire-resistance period obtained from the fire endurance tests was based on the integrity and the insulation criteria. Compared to the reference timber assembly without any fire protection (fire-resistance time = 37 min), the increases in fire resistance using the three different fire protection measures were 6 min (16%), 37 min (100%) and 142 min (384%) for using intumescent coating (specimen F2), 12-mm-thick gypsum plasterboard (specimen F4) and infill mineral wool (F3), respectively. The specific intumescent coating used in the test failed to expand. Therefore, this specimen (F2) behaved very similarly with the control specimen (F1) without any fire protection. Attaching an additional layer of gypsum plasterboard to the timber assembly on the fire-exposed side improved the fire-resistance rating by about 30 min, which is higher than that obtained from using the current design guidance such as Eurocode EN 1995-1-2. Among these three fire protection methods, filling the void between the top and bottom timber boards gave the best result because the mineral wool not only provided insulation but also stopped direct flame attack of the timber board on the unexposed side.

Keywords

cavity filling fire protection fire resistance gypsum plasterboard improvement charring timber assembly

Introduction

A large number of buildings are constructed with timber in China and in many parts of the world. Timber is often considered more sustainable than other construction materials such as steel and concrete, which has made timber construction more popular. However, timber is combustible and if fire occurs in a timber building, the loss can be very high, see Figure 1 (Wang et al., 2011). It is essential that behaviour of timber construction in fire is thoroughly understood so that timber structures can be designed and constructed with sufficient fire resistance.

Figure 1.

A timber structure in fire (Wang et al., 2011).

A lot of research studies have been conducted over the years to investigate the fire performance of timber structures. Richardson and Batista (2001) investigated the fire resistance of timber decking in older heavy timber buildings undergoing renovation. Their results showed that it was not uncommon to observe 2- to 4-mm spaces between the timber planks, and they suggested attaching either plywood or oriented strand board (OSB) panels to the timber decking in order to prevent fire penetrating through these spaces. Attaching gypsum plasterboard to the underside of the timber decking was observed to improve the fire-resistance rating by 18–20 min. Sultan and Benichou (Benichou, 2006; Sultan, 2008, 2012; Sultan and Benichou, 2003) conducted a series of full-scale fire-resistance tests to investigate the effects of various design parameters on the fire resistance of lightweight frame timber floor board assemblies, including gypsum board screw spacing from the board edge, method of interior insulation installation, interior insulation type, joist spacing and joist depth. It was observed that interior insulation had the most significant effect. Full-scale fire tests were carried out by Bullock et al. (2000) on a six-storey timber frame building. It was observed that gypsum plasterboard played an important role in protecting the timber frame in fire.

Full-scale natural fire tests on gypsum-lined structural insulated panel (SIP) and engineered floor joist assemblies were conducted by Hopkin et al. (2011). Mechanisms for fire spread were identified where fitting details were not appropriately sealed. Shields et al. (1999) concluded from an experimental study that ignition of the wall lining was delayed for fire-retardant walls. Tsantaridis et al. (1999) performed cone calorimeter tests on woods with gypsum plasterboards protection. The test results showed that the time to the onset of charring was more dependent on the plasterboard thickness than the density of the boards. The use of gypsum plasterboard increased the time to reach the charring temperature of timber and also reduced the charring rate. Saxena and Gupta (1990) developed a few fire-retardant intumescent coating for wood and wood-based products. The coating was found to be quite effective in stopping the spread of flame and afterglow combustion. Frangi and Fontana (2005) carried out fire tests on box timber members with mineral fibres and glass fibres filling in the cavity. The specimens with mineral fibres were observed to have better insulation performance than those with glass fibres.

Fire tests were conducted by Ni et al. (2009) to obtain the fire-resistance data of timber members including beams, columns, walls and floors. The results showed that timber members with appropriate fire-protective measures exhibited good fire performance and reached the required fire-resistance rating of the relevant design codes. Xu et al. (2011, 2012) carried out experiments to investigate the fire performance of timber beams with fire-protective coatings both under and after fire. The beams with intumescent coatings were observed to have lower rate of temperature rise and delay the onset of char formation, leading to much higher fire-resistance rating. The same beneficial effect of using lime putty finishing was also observed for timber columns exposed to four-side fire by Li et al. (2010).

Timber behaviour under heating in fire is complex; therefore, fire-resistance design of timber structures is still largely based on fire-resistance testing. Nevertheless, attempts are being made to calculate the fire resistance of timber structures. For this purpose, determining the charring behaviour is the most important. Frangi et al. (2008) proposed a charring model for timber floor assemblies with void cavities, taking into account the influence of high temperatures developed in timber after fire protection has fallen off. Using this charring model, a simplified model for calculating the fire resistance of timber slabs made of hollow core elements was developed based on the reduced cross-section method (Frangi et al., 2009). The model was then used to further develop (Frangi et al., 2010) a design model for verification of the separating function of light timber frame wall and floor assemblies.

The above literature review identified the benefits in improving fire resistance of timber construction by providing interior insulation in the void, using gypsum plasterboard and by incorporating fire-retardant materials. However, direct experimental comparison of the effectiveness of using different fire protection measures, based on the same timber construction, is not available. This is the objective of this article. Furthermore, the fire test results of this article will provide valuable experimental data to help develop calculation methods.

Experimental programme

Test specimens

Four timber floor assemblies were prepared and tested by being exposed to the ISO standard fire. These specimens, referred to throughout this article as F1, F2, F3 and F4, respectively, consisted of the control specimen (F1) without any fire protection, with intumescent coatings (F2), with the cavity filled with mineral wool (F3) and with attached gypsum plasterboard (F4). Figure 2 shows the construction of these four specimens. They were designed following the statistical geometry and design details from a large number of existing floor assemblies in China. The control timber floor assembly consisted of two timber plates of 1650-mm long, by 1500-mm wide and by 15-mm thick, and wood joists of 200-mm deep by 50-mm wide spanning across the shorter dimension of floor at 400-mm spacing. Specimen F2 had fire retardant–treated wood joists, and both the top and bottom surfaces of the wood floors were protected with fire-retardant intumescent coatings. In specimen F3, both the wood joists and floors were protected with fire-retardant intumescent coatings, and the cavity between the joists was filled with mineral wool; in timber assembly F4, one layer of gypsum plasterboard was attached to the bottom side of the floor assembly.

Figure 2.

Geometry and detailing of the specimens (all dimensions in millimetres).

Material properties

Red pine was selected for the wood joists. Its measured density was 450 kg/m³, and the measured moisture content was 16.8%. The wood floor was made of Merbau with no paint, with a measured density of 450 kg/m³ and a measured moisture content of 11.4%. The fire-retardant intumescent coating used in specimens F2 and F3 consisted of inorganic flame retardant, blowing agents, carbonisation, emulsion binder and titanium dioxide. Three layers of the coating were applied, resulting in usages between 0.5 and 1.0 kg/m². The density of the mineral wool in specimen F3 was 110 kg/m³. The thickness of the gypsum plasterboard in specimen F4 was 12 mm, and its ambient temperature thermal conductivity was 0.18 W/(m K).

Fire testing procedure

The fire tests were conducted on the floor furnace in the fire testing laboratory of Shanghai Research Institute of Building Sciences. The dimension of the furnace was 4.2 m × 3.0 m × 1.5 m. The furnace temperature–time curve followed the ISO 834 standard fire curve, controlled by eight K-type thermocouples placed in the furnace. Figures 3 and 4 show the layout of the test specimen in the furnace. Only the middle 500-mm width of the bottom floor was directly exposed to fire due to limited width of the furnace lid. The longer span (with 1650-mm length) was simply supported on the furnace lids, while in the shorter span (with 1500-mm width) direction, the specimen was free from any support. To ensure impermeability of the furnace, the opening part of the furnace was sealed with gypsum plasterboards and ceramic fibres after installation of the timber floor assembly. To prevent passage of the smoke and flame into the cavities of the specimen, the perimeter of the specimen was wrapped with ceramic fibres. A distributed load of 1.2 kN/m² was applied using blocks of 25-kg mass during the heating process, representing a practical working load.

Figure 3.

Test specimen arrangement on top of furnace.

Figure 4.

Test setup.

Instrumentation

Figure 5 shows locations of the thermocouples used to monitor temperatures in the wood joists and floors. They were placed on the wood joists at a distance of 50, 100 and 150 mm from the bottom surface of the top layer of wood floor, about 20 mm away from the central plane of the assembly. Thermocouples were also placed at various locations on the surfaces of the wood floors in the central plane of the assembly.

Figure 5.

Layout of thermocouples.

Test observations

Assembly F1

Three minutes after fire exposure, a small amount of smoke was observed to penetrate from the west side of the timber assembly F1 (Figure 6(a)). The smoke gradually increased with the fire exposure time. Twenty minutes into the test (Figure 6(b)), the ceramic fibre that was used to wrap the specimen became light yellow, and a large amount of smoke rose from the specimen perimeter and from the space between the planks of the top layer of wood. After 25 min, large areas of the ceramic fibre wrapping became yellow. At 35 min, the sides of the ceramic fibre wrapping turned to black, and the smoke escaping from the space between the planks increased (Figure 6(c)). At 37 min, spalling sound was heard from the specimen. After 43 min, the top layer of the wood floor was getting dark (Figure 6(d)), and the spalling sound became more frequent. Forty-eight minutes after the test, flame was observed to pass through the top layer of the test floor for 10 s (Figure 6(e)), indicating integrity failure of the test specimen according to the standard fire testing procedure (EN 1995-1-2, 2004). At this stage, the gas was then cut off immediately, and the fire was extinguished with water. When the furnace lid was opened to examine the test specimen (Figure 6(f)), it was observed that the bottom surface of the top layer and the top surface of the bottom layer of the specimen, in the region without fire exposure charred, and the middle three wood joists were heavily burnt, while the other two joists were slightly burnt.

Figure 6.

Key events of specimen F1: (a) smoke from side of specimen (3 min); (b) smoke from the specimen perimeter and from the space between the planks of the top layer of wood (20 min); (c) the sides of the ceramic fibre wrapping turned to black, and the smoke escaping from the space between the planks increased (35 min); (d) the top layer of the wood floor getting dark (43 min); (e) flame pass through the top layer of the test floor for 10 s (48 min) and (f) surface charring of wood joists and floors.

Table 1 lists the fire-resistance ratings of this specimen according to the integrity condition, the unexposed surface average temperature condition and the unexposed surface maximum temperature condition. The minimum fire-resistance rating was according to the average temperature rise. Nevertheless, the differences among the three values are relatively small.

Table 1.

Summary of the fire-resistance ratings of the test specimens.

Specimen ID	Fire-resistance rating (min)
Specimen ID	Integrity condition: cotton pad test and flame passage	Insulation condition: average temperature increase on the unexposed surface	Insulation condition: maximum temperature increase on the unexposed surface
F1	48	37	43
F2	46	44	43
F3	179	>179	>179
F4	82	74	76

Assembly F2

Specimen F2 behaved in a very similar way to specimen F1. Table 1 lists the fire-resistance ratings according to different conditions, and Figure 7 shows appearances of the specimen at different key stages of the test – 3 min (Figure 7(a)): observation of a small amount of smoke from the bottom side, 25 min (Figure 7(b)): the ceramic fibre wrapping beginning to turn light yellow, 35 min (Figure 7(c)): the surface of the top layer of the floor became light yellow and 46 min (Figure 7(d)): integrity failure due to continuous flaming for 10 s. The flame had to be extinguished with water (Figure 7(e)). After the test, although the specimen remained structurally sound, the specimen was extensively burnt (Figure 7(f)). Examination of the specimen after the fire test suggested that the intumescent coating failed to expand to provide the expected fire protection.

Figure 7.

Key events of specimen F2: (a) small amount of smoke from the bottom side (3 min), (b) ceramic fibre wrapping beginning to turn light yellow (25 min), (c) surface of the top layer of the floor became light yellow (35 min), (d) integrity failure due to continuous flaming for 10 s (46 min), (e) extinguishing fire with water and (f) serious damage of the treated wood joists.

Table 1 lists the fire-resistance ratings according to different criteria. The values are very similar to those of test F1.

Assembly F3

The sequence of events for this specimen was similar to the other tests. However, the timing was considerably delayed compared to the other two tests. The key events and their timings are summarised in the following:

Observation of smoke from all four corners of the specimen: 10 min (Figure 8(a));

Ceramic wrapping turning light yellow: 30 min;

Smoke escaping from the planks of the top layer: 40 min;

Spalling sound heard: 47 min;

Large areas of ceramic wrapping turning black and large amount of smoke: 90 min (Figure 8(b));

Top surface colour change: 120 min (Figure 8(c));

Ceramic fibre wrapping turning red and wood joists starting to burn: 140 min (Figure 8(d));

Integrity failure – observation of flame passing through the top layer of floor for 10 s: 179 min (Figure 8(e)).

Figure 8.

Key events of specimen F3: (a) smoke from all four corners of the specimen (10 min), (b) large areas of ceramic wrapping turning black and large amount of smoke (90 min), (c) top surface colour change (120 min), (d) one ceramic fibre wrapping turning red and wood joists starting to burn (140 min), (e) flame passing through the top layer of floor for 10 s (179 min) and (f) the bottom surface of top floor has not been charred and fire unexposed wood joists have not been charred.

Figure 8(f) shows that finally, the two edge joists were totally burnt and the three middle joists were damaged seriously on the fire-exposed side. However, these three joists kept the original colour and did not char on the top unexposed surface. The infill mineral wool remained intact.

Assembly F4

The key events of this test are again similar to the other tests. Figure 9 shows these events and their timings are as follows:

Observation of smoke from all four corners of the specimen: 7 min (Figure 9(a));

Smoke escaping from planks of the top layer: 55 min (Figure 9(b));

Ceramic wrapping turning light yellow: 60 min (Figure 9(c));

Spalling sound heard: 75 min (Figure 9(d));

Integrity failure – observation of flame passing through the top layer of floor for 10 s: 82 min (Figure 9(e)).

Figure 9.

Key events of specimen F4: (a) smoke from all four corners of the specimen (7 min), (b) smoke escaping from planks of the top layer (55 min), (c) ceramic wrapping turning light yellow (60 min), (d) ceramic wrapping turning black (75 min), (e) flame passing through the top layer of floor for 10 s (82 min) and (f) after extinguish of fire.

Figure 9(f) shows that the bottom surface of the top layer and the top surface of the bottom layer of the specimen, in the region without fire exposure charred, and all the five wood joists were slightly burnt. Table 2 compares the timings at the different key events between the four different specimens.

Table 2.

Comparison of timings at different key events between the four different specimens.

Key event	Timing (min) for specimen
Key event	F1	F2	F3	F4
Smoke from side of specimen	3	3	6	4
Smoke escaping from planks of the top layer	20	25	40	55
Spalling sound heard	37	37	47	75
Flaming on the unexposed side	48	46	179	82

Test results and discussion

Comparison of fire-resistance ratings

Table 1 lists the fire-resistance ratings of the four specimens according to different criteria. Only integrity and insulation conditions are included because there was no failure of the load-bearing condition.

It appears that the specific type of intumescent coating applied to this specimen was not effective. Therefore, when comparing the fire-resistance ratings in Table 1, the only increase using the intumescent coatings was according to the average temperature condition and the improvement was quite small (43 min compared to 37 min). While this may be related to the specific intumescent coating used in the test, the fact that test timber had high moisture content may have contributed to the ineffectiveness of the intumescent coating.

The fire-resistance ratings of F4 (with one additional layer of gypsum plasterboard) are higher than those of F1 (control specimen) by 34, 30 and 33 min. According to Eurocode EN 1995-1-2 (2004), the additional fire-resistance rating (min) for one layer of gypsum plasterboard on the fire-exposed side is 1.4 × h_p × 0.8 × 0.8, where h_p is the gypsum plasterboard thickness in millimetres. The gypsum plasterboard thickness used in the test was 12 mm, giving an increased fire-resistance rating of 10.8 min. Although this is a conservative value, it grossly underestimates the improved fire-resistance rating achieved in the test of 30 min.

Specimen F3 (with infilled mineral wool within the cavity) performed best, achieving a fire-resistance rating of 179 min. Only this specimen had cavity infill, demonstrating (as shown in Figure 12) the effectiveness of blocking direct radiation within the cavity space in reducing temperatures on the unexposed side.

Temperature developments

Figure 10(a) to (d) shows recorded temperatures at the top of the exposed timber panel and at the bottom of the unexposed timber panel, each specimen at two locations, for the four test specimens. Timber charring started at about 300°C, so available temperature data above this temperature were curtailed. Although thermocouples 4 and 5 were placed at the same relative positions to the fire exposure surface, they had different measurements due to material variability. So did thermocouples 1 and 2. This difference may also be attributed to the non-uniform temperature heat fluxes inside the furnace. Despite some differences, the measured temperatures were similar. So results from one set of thermocouples (4 and 5) will be used in further discussions.

Figure 10.

Temperature developments on surface: (a) timber assembly F1, (b) timber assembly F2, (c) timber assembly F3, (d) timber assembly F4 and (e) comparison of temperature developments in floor surface between specimens F1–F4.

Figure 10(e) compares temperatures measured by thermocouples 4 and 5 between the four different tests. It can be seen that at similar thermocouple 4 temperatures, the thermocouple 5 temperatures for tests F1, F2 and F4 were similar and much higher than those for test F3. In particular, once the bottom wood panel has burnt through, as indicated by thermocouple 4 recording temperatures of over 300°C, test specimens F1, F2 and F4 experienced accelerated temperature rise on the underside of the unexposed (top) timber panel (measured by thermocouple 5). In contrast, the temperature on the unexposed layer of wood panel in specimen F3 (which had cavity infill) increased slowly even after the exposed side had burnt through. This clearly demonstrates the effectiveness of blocking direct radiation in the cavity space by filling the cavity space with insulation material (F3).

Comparison between results of tests F1, F2 and F3 indicates that temperatures on the exposed side (thermocouple 4) in these tests were very similar. The slower temperature rise in test F4 (with one layer of gypsum plasterboard) clearly shows the benefit of reducing the exposed side temperature using gypsum plasterboard on the exposed side. Nevertheless, once the timber panel protected by the gypsum plasterboard burnt through, the temperature rise on the underside of the unexposed panel accelerated, as happened in tests F1 and F2.

These four tests reached their fire-resistance ratings due to integrity or insulation failure. In some cases, load-bearing capacity may govern the fire resistance of the timber floor assemblies. For determining load-bearing capacity, temperatures in the wood joists should be accurately determined. Figure 11(a) to (d) provides the recorded temperature developments in the joists in the four tests. Figure 11(e) compares thermocouple 9 readings for the four tests. Again, the intumescent coatings provided in test F2 had no effect. Figure 12(a) to (d) shows the recorded temperatures on the unexposed surfaces of the four tests, at the timber joist position and away from the joist. Plotted on this figure are also the temperatures for determining fire-resistance rating based on the maximum temperature rise of 180°C and the average temperature rise of 140°C. It can be seen that the temperatures on the unexposed surface measured at the joists or away from the joints were similar. Therefore, the fire-resistance ratings determined with or without the timber joist were very similar. This implies that when evaluating fire-resistance rating of this type of construction based on insulation performance, the complexity of including the timber joists in the calculations may be dispensed with. This would simplify the calculation procedure considerably because heat transfer, without including the internal joists, would be one dimensional in the thickness direction of the wood floor assembly.

Figure 11.

Temperature developments in wood joists: (a) timber assembly F1, (b) timber assembly F2, (c) timber assembly F3, (d) timber assembly F4 and (e) comparison of temperature developments in internal joists between specimens F1–F4.

Figure 12.

Temperature developments on the unexposed surface: (a) timber assembly F1, (b) timber assembly F2, (c) timber assembly F3 and (d) timber assembly F4.

Summary and conclusion

This article has presented experimental results to compare fire performance of four timber floor assemblies, one without any protection and three with different fire protection measures. This article has provided detailed description of the fire testing procedure and detailed temperature and fire-resistance rating results. Based on the analysis of these results, the following conclusions may be drawn:

The most effective way of improving the fire-resistance rating of a timber floor assembly is to fill the cavities with good insulation materials. Based on this research, filling the cavity with mineral wool increased the fire-resistance rating from 37 to 179 min.

Although not providing as significant improvement on fire-resistance rating of timber assembly as filling the cavity, attaching gypsum plasterboards on the exposed side was also effective, which increased the fire-resistance rating by 30 min.

Failure of the specific intumescent coating to expand during fire exposure used in this study meant that this type of intumescent coating was not effective in improving fire-resistance rating of the timber floor assembly. Further research is clearly required to investigate the effectiveness of intumescent coatings for protection of timber construction.

The method in Eurocode EN 1995-1-2 to calculate improved fire resistance of using gypsum plasterboard (F4) was found to be safe but underestimates the test result by a very large margin. The Eurocode calculation result gave 10.8 min and the test result gave 30 min.

For timber floor assemblies without cavity insulation (F1, F2, F4), the temperature on the unexposed side experienced accelerated increase once the bottom timber panel was burnt. This points out the importance of accurately determining the burning through time. In contrast, when there was cavity insulation, the unexposed side temperature was not sensitive to burning through of the bottom timber panel.

When determining fire-resistance rating of timber floor assemblies according to insulation using calculation, the effect of the internal timber joists may be ignored. This would simplify the calculation considerably, by making the heat transfer process one dimensional.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This work is sponsored by the National Nature Science Foundation of China (No. 51178115), Program of Shanghai Subject Chief Scientist (B type) (No. 15XD1522600) and Shanghai Rising-Star Program (Nos 07QB14031 and 11QH1402100).

References

Benichou

(2006) Predicting the structural fire performance of solid wood-framed floor assemblies. In: Proceedings of the 4th international workshop structures in fire (SIF ’06), Aveiro, 10–12 May, pp. 909–920. Aveiro, Portugal: University of Aveiro.

Bullock

Lennon

Enjily

(2000) The Fire Resistance of Medium-Rise Timber Frame Buildings Summary Report. Buckinghamshire: Chiltern International Fire.

EN 1995-1-2 (2004) Eurocode 5: Design of Timber Structures – Part 1-2: General – Structural Fire Design (British Standards). Brussels: European Committee for Standardization.

Frangi

Fontana

(2005) Fire performance based design of multi-storey timber buildings. In: Proceedings of the IABSE symposium on structures and extreme events, Lisbon, 14–17 September, pp. 99–106.

Frangi

Erchinger

Fontana

(2008) Charring model for timber frame floor assemblies with void cavities. Fire Safety Journal 43(8): 551–564.

Frangi

Knobloch

Fontana

(2009) Fire design of timber slabs made of hollow core elements. Engineering Structures 31(1): 150–157.

Frangi

Schleifer

Fontana

(2010) Design model for the verification of the separating function of light timber frame assemblies. Engineering Structures 32(4): 1184–1195.

Hopkin

Lennon

El-Rimawi

. (2011) Full-scale natural fire tests on gypsum lined structural insulated panel (SIP) and engineered floor joist assemblies. Fire Safety Journal 46(8): 528–542.

. (2010) Experimental research on fire endurance of timber columns exposed to four-side fire. Building Structure 40(3): 115–117.

10.

Qiu

Mehaffey

. (2009) Experimental study on fire protection of timber assembles. In: Proceedings of the application of structural fire engineering, Prague, 19–20 February.

11.

Richardson

Batista

(2001) Fire resistance of timber decking for heavy timber construction. Fire and Materials 25: 21–29.

12.

Saxena

Gupta

(1990) Development and evaluation of fire retardant coatings. Fire Technology 26(4): 329–341.

13.

Shields

Silcock

GWH

Moghaddam

. (1999) A comparison of fire retarded and non-fire retarded wood-based wall linings exposed to fire in an enclosure. Fire and Materials 23: 17–25.

14.

Sultan

(2008) Fire resistance of wood joists floor assemblies. Fire Technology 44: 383–417.

15.

Sultan

(2012) Fire resistance of wood truss floor assemblies. Fire Technology. Epub ahead of print 15 December. DOI: 10.1007/s10694-012-0311-8.

16.

Sultan

Benichou

(2003) Fire resistance performance of lightweight floor assemblies. In: Proceedings of the designing structures for fire conference, Baltimore, MD, 30 September–1 October 2003, pp. 203–214.

17.

Tsantaridis

Ostman

BA-L

Konig

(1999) Fire protection of wood by different gypsum plasterboards. Fire and Materials 23: 45–48.

18.

Wang

Marsden

Kelly

(2011) Challenges of fire fighting in fire engineered built environment (Proceedings of the 5th conference on performance-based fire and fire protection engineering). Procedia Engineering 11: 583–592.

19.

. (2011) Static experimental research on mechanical behavior of timber beams with lime putty finishing after fire. Journal of Building Structures 32(7): 73–79.

20.

Zhang

Shang

. (2012) Experimental research on fire endurance of timber beams exposed to three-side fire. Building Structure 42(12): 127–130.