Fire tests of Magnesium Oxide board lined light gauge steel frame wall systems

Abstract

Recently, Magnesium Oxide (MgO) board has been widely used in LSF wall systems because of its improved acoustic properties, impact resistance, structural strength and serviceability. However, their thermal properties and fire performance have not been fully investigated. Therefore, in this research study thermal properties of two different types of MgO boards available in Australia were measured and their fire performance was investigated using three full-scale fire tests of LSF walls lined with two types of MgO board. Although the tests were conducted on two different types of MgO boards with different configurations, the fire test results gave a fire resistance level (FRL) of 30 min, in which the failure was initiated by integrity of the board with either board cracking or board joint opening. This paper presents the details of the thermal property tests and the three full-scale fire tests, and their results. In addition, the effects of different MgO boards, joint configurations and compounds, noggings, screw fastening techniques and cavity insulation on the fire performance of LSF walls are also presented.

Full text

Fire Tests of Magnesium Oxide Board Lined Light Gauge Steel
Frame Wall Systems
Mohamed Rusthi a, Anthony Ariyanayagam a, Mahen Mahendran a,* and Poologanathan
Keerthan a
a Queensland University of Technology (QUT), Brisbane, Australia
* Corresponding author’s email address: m.mahendran@qut.edu.au
Abstract
Recently, Magnesium Oxide (MgO) board has been widely used in LSF wall systems
because of its improved acoustic properties, impact resistance, structural strength and
serviceability. However, their thermal properties and fire performance have not been fully
investigated. Therefore, in this research study thermal properties of two different types of
MgO boards available in Australia were measured and their fire performance was
investigated using three full-scale fire tests of LSF walls lined with two types of MgO board.
Although the tests were conducted on two different types of MgO boards with different
configurations, the fire test results gave a fire resistance level (FRL) of 30 minutes, in which
the failure was initiated by integrity of the board with either board cracking or board joint
opening. This paper presents the details of the thermal property tests and the three full-scale
fire tests, and their results. In addition, the effects of different MgO boards, joint
configurations and compounds, noggings, screw fastening techniques and cavity insulation on
the fire performance of LSF walls are also presented.
Keywords: Magnesium Oxide board, Thermal properties, Fire test, Fire performance, Light-
gauge steel frame wall systems
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

1 Introduction
Conventional light-gauge steel frame (LSF) wall systems are made of cold-formed steel
stud wall frame, gypsum plasterboard and cavity insulation materials as shown in Figure 1.
When LSF walls are exposed to fire, the mechanical properties of cold-formed steel studs
degrade rapidly beyond about 500°C, which may cause premature structural failure of the
wall compared to ambient temperature. Therefore, it is necessary to provide sufficient fire
protection to the steel stud wall frame to avoid direct fire exposure as well as delay the
temperature rise of steel to improve the fire resistance of LSF wall systems.
Figure 1. Conventional LSF wall configuration
When there is fire on one side of the LSF wall, gypsum plasterboard lining on the fire
side keep the stud temperatures below the failure limits. Also, the wall board temperatures on
the unexposed side are kept below their limits. In addition to these, plasterboards provide
lateral restraint to the steel wall frame, all of which leads to increased stud failure times and
hence improves the fire resistance levels (FRLs) of LSF wall systems. The FRLs are defined
2
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38

in AS 1530.4 [1] based on three criteria when LSF wall panels are exposed to the standard
fire time-temperature profile on one side. They are: 1) structural: wall must continue to carry
the design loads, 2) integrity: wall's integrity is not affected to allow the passage of hot gases
or flames through the wall and 3) insulation: wall's insulation to restrict heat passing through
the wall (i.e., the change in ambient or unexposed wall surface temperature should not exceed
140 ºC on average or 180 ºC at any point.
LSF wall fire performance is mainly dependent on the fire resistant wall board lining,
which delays the heat transfer through wall and the temperature rise in steel studs. Recent
research studies performed by Gunalan et al. [2] and Ariyanayagam and Mahendran [3]
focused on full-scale fire tests of load bearing LSF walls made of conventional lipped
channel section (LCS) studs and gypsum plasterboard lining. Test wall panels (2.1 m × 2.4
m) were exposed to standard and realistic design fires in their studies. Kesawan and
Mahendran [4] and Jatheeshan and Mahendran [5] performed full-scale standard fire tests on
LSF walls made of hollow flange channel section studs and gypsum plasterboard lining while
Nassif et al. [6] also conducted a full scale standard fire test of LSF wall made of LCS studs
but used a larger LSF wall panel (3 m × 3 m). These fire tests provided comprehensive time-
temperature profiles for various LSF wall configurations made of different stud sections,
gypsum plasterboard lining and cavity insulation materials that can be used to predict their
fire performance.
Recently, Australian building industries have been using new wall lining materials in LSF
wall system to enhance their acoustic properties, impact resistance, structural strength and
serviceability. However, the fundamental understanding of the new lining materials such as
Magnesium Oxide (MgO) boards is lacking in relation to their performance in fire. In
contrast to the LSF walls lined with gypsum plasterboards, there are only limited research
studies on LSF walls lined with MgO boards. In addition to that, there are no full-scale fire
3
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63

test time-temperature profiles for MgO board lined LSF walls and the thermal property
variation with temperature is also not available for Australian MgO boards. Chen et al. [7-9]
performed full-scale fire tests on load bearing LSF walls made of LCS studs and lined with a
combination of gypsum plasterboard and MgO board. These LSF walls (3.4 m × 3.0 m) were
tested under standard fire time-temperature exposure. Their fire test results showed that the
combination of one layer of MgO board and one layer of gypsum plasterboard gave increased
failure time than two layers of gypsum plasterboard. However, the LSF wall configuration
with only MgO board was not tested in their experimental study.
Hanna et al. [10] carried out full-scale standard fire tests on load bearing LSF walls (3.0
m × 2.7 m) made of LCS studs and MgO boards and the time-temperature profiles of the LSF
wall components were measured at various locations. Although cracks were observed on the
ambient side at about 40 minutes, the test was continued until the cracks were wide open at
about 50 minutes to allow hot gases to pass through the cracks that could ignite a piece of
cotton on the unexposed surface. Post-fire observations indicated that partial wall board fall
off and bowing occurred on the fire exposed surface. As Hanna et al. [10] focused on the
post-fire behaviour of the steel studs, no further details are provided on the fire performance
of MgO board. As this is the only full-scale fire test with MgO board lining reported in the
research literature, further studies are necessary to understand the fire performance of MgO
board lined LSF wall systems.
Full-scale standard fire tests were therefore conducted in this study to gain additional
information on the fire performance of MgO board lined LSF wall systems. In addition,
thermal properties including specific heat, mass loss and thermal conductivity of these boards
at elevated temperatures were also measured to support the fire test findings. This paper
presents the details of the experimental study on the thermal properties of two types of MgO
4
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87

boards (referred to as Types 1 and 2 in this paper) available in the Australian market and the
three full-scale fire tests conducted on MgO board lined LSF wall systems.
2 Properties of Magnesium Oxide (MgO) Boards
2.1 Chemical Composition
The main constituents of the boards are Magnesium oxide (MgO) and Magnesium
chloride (MgCl2) as shown in Table 1 based on their manufacturer data sheets. In contrast to
Type-1 board, Type-2 board has higher MgO and lower MgCl2 contents. Other constituents
such as Perlite, woodchip and fiberglass are added to enhance the properties such as
lightweight, insulation, workability and acoustic properties. Instead of Perlite, additional
fiberglass is used in Type-2 board. The chemical composition of MgO board can be written
as Mg(OH)2.MgCl2.H2O.
Table 1. MgO board chemical composition
Chemical name Weight percentage
Type 1 Type 2
Magnesium oxide (MgO) 40% 50-53%
Magnesium chloride (MgCl2) 27-35% 20-28%
Perlite 5-10% 0%
Woodchip 15% 2-10%
Fiberglass 5-8% 18-19%
2.2 Thermal Property Tests
Thermal properties are required to understand the thermal behaviour of wall lining
materials, especially when they are used in fire rated wall systems. Thermal properties that
include specific heat, mass loss and thermal conductivity were measured using representative
samples collected from the wall board component. Thermal properties of Type-1 and 2 MgO
boards were measured in an experimental study using simultaneous thermal analyser
(NETZSCH STA 449F3), which is capable of measuring the specific heat variation using
differential scanning calorimetry (DSC) technique and the mass loss or relative density using
5
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107

thermo gravimetric analysis (TGA) technique. Thermal property tests were conducted in a
purged Nitrogen environment within a temperature range of 50 ºC – 1200 ºC at a constant
heating rate of 20 ºC/min. Tests were conducted according to the standard procedures
stipulated in ASTM E1269 [11] using platinum (Pt) crucibles lined with Alumina (Al2O3)
liners and pin holed lids.
The measured specific heat variations with temperature are shown in Figure 2. The
specific heat peaks for Type-1 and 2 boards are considerably different from each other. There
are two major peaks observed in both MgO board types with Type-1 board having higher
specific heat values compared to Type-2 board. Although both boards exhibit similar peak
patterns, an increase in specific heat for Type-1 board can be observed at about 200 ºC, which
is about 3,000 J/kg/ºC higher than Type-2 board. There are five specific heat peaks as shown
in Figure 2 and they are attributed to the chemical reactions and dehydration processes given
in Equations 1 – 5 [9].
The specific heat peaks are observed at the same temperature for both types of MgO
boards. For Type-1 and 2 boards, the first two peaks occurred at 180 ºC and 230 ºC, which
were induced by the dehydration of magnesium oxychloride (Equations 1 and 2) where
evaporation of water can be observed during this process. The third and fourth peaks
occurred at 400 ºC and 475 ºC, which indicate the hydrolysis (Equations 3 and 4) and
pyrolysis reactions where HCl is released. The final peak was observed at 520 ºC, which is
the further release of chemically bound water in the MgO board. These observations are
similar to those made by Chen et al. [9] in relation to the MgO board they used.
6
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128

0 200 400 600 800 1000 1200
0
2000
4000
6000
8000
10000
Type-1 Type-2
Temperature (ºC)
Specific Heat (J/kg/ºC)
Figure 2. Specific heat variation with temperature
5Mg(OH)2.MgCl2.8H2O  5Mg(OH)2 . MgCl2 + 8H2O (1)
3Mg(OH)2.MgCl2.8H2O  3Mg(OH)2 . MgCl2 + 8H2O (2)
5Mg(OH)2.MgCl2  4Mg(OH)2 + 2MgO + 2HCl (3)
3Mg(OH)2.MgCl2  2Mg(OH)2 + 2MgO + 2HCl (4)
Mg(OH)2  MgO + H2O (5)
The measured mass loss variations of the tested boards are shown in Figure 3. Both
boards exhibit similar pattern in mass loss with Type-1 board showing higher mass loss than
Type-2 board, which can be related to the higher specific heat values seen in Type-1 board.
Type-1 board shows a mass loss of 20% at about 300 ºC, whereas in Type-2 board similar
mass loss occurs at about 375 ºC. At the end of the test, the final mass loss is about 48% and
44% in Type-1 and 2 boards, respectively. The higher mass loss observed in these tests
(>40%) may cause shrinkage cracks of the board, which will eventually affect their fire
performance as well as that of walls lined with these boards.
7
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143

0 200 400 600 800 1000 1200
0
20
40
60
80
100
Type-1 Type-2
Temperature (ºC)
Mass Loss (%)
Figure 3. Mass loss variation with temperature
Thermal conductivity of Type-1 and 2 MgO boards were measured using laser flash
analysis (LFA) equipment from 50 ºC to 500 ºC. Representative samples of 10 mm × 10 mm
× 2 mm were cut from the boards and were used to measure the thermal conductivity. Figure
4 shows the measured thermal conductivity value, in which Type-2 MgO board exhibits
higher thermal conductivity than Type-1 MgO board. This is mainly due the difference in
composition in these two boards. The initial temperature (50 ºC) thermal conductivity values
are 0.39 and 0.47 (W/mºC), respectively for Type-1 and 2 boards. The thermal conductivity
values continue to reduce at higher temperatures, where they are reduced by almost 50% of
the initial value at 500 ºC.
0 50 100 150 200 250 300 350 400 450 500
0.0
0.1
0.2
0.3
0.4
0.5
Type-1 Type-2
Temperature (ºC)
Thermal Conductivity (W/mºC)
Figure 4. Thermal conductivity variation with temperature
8
144
145
146
147
148
149
150
151
152
153
154
155
156

2.3 Small-scale Furnace Tests
Small-scale tests of Type-1 MgO board were conducted to understand its behaviour at
elevated temperatures before performing full-scale fire tests, so that the performance of the
board itself can be investigated. Tests were conducted in an electric furnace using small-scale
samples of 450 mm × 450 mm steel frame lined on one side with 10 mm thick MgO board as.
The test specimens were placed inside the furnace with the top side exposed to heat from the
furnace coils. The furnace was closed so that there is no heat loss to the environment, and
then the test specimens were heated up to 450 and 750 oC over a period of 30 and 45 minutes,
respectively.
The small-scale test specimens were taken out of the furnace after cooling them to the
ambient temperature. Figure 5 shows the test specimens at different temperatures. The
specimen that was heated to 450 ºC showed a crack at the centre which extended further
towards the screw positions. This crack might have initiated after the completion of
dehydration reactions in the MgO board. At 750 ºC the specimen showed significant cracking
of the board. This is due to the very high mass loss discussed earlier. Since these tests were
conducted inside an electric furnace without direct fire exposure, full-scale fire tests are
needed to investigate this behaviour of MgO board lined LSF wall systems, which will be
discussed next.
(a) Test specimen at 28 oC
9
Screws at 200 mm
spacing
Steel frame
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174

(b) Test specimen at 450 oC
(c) Test specimen at 750 oC
Figure 5. Small-scale test specimens exposed to different elevated temperatures
3 Full-Scale Fire Tests of Non-Load Bearing LSF Walls Lined with MgO Board
3.1 Test Wall Configurations
The full-scale fire tests consisted of three non-load bearing LSF walls of 3.15 m × 3.15 m,
which were tested in accordance with AS 1530.4: 2005 [1] using a propane gas furnace
programmed to follow the standard fire time-temperature curve. The fire time-temperature
curve of the furnace was regulated by eight mineral insulated, metal sheathed (MIMS)
thermocouples with an overall diameter of 3mm inside the furnace. Since these tests were to
be performed as non-load bearing walls, six hydraulic jacks were used directly under the steel
studs to support the self-weight of the test wall panel. Additional load exerted on the studs
10
Crack
175
176
177
178
179
180
181
182
183
184

during the fire test due to thermal expansion was released by the hydraulic jacks to ensure no
load acted on the steel studs.
Table 2 shows the tested wall configurations. The LSF wall in Test 1 was built with 10
mm thick Type-1 MgO board while the other two test walls were lined with 10 mm thick
Type-2 MgO board. Both MgO board types have fibre glass mesh layers on both sides to
protect the board and to provide sufficient strength. Test 3 wall had glass fibre insulation
materials inside the wall cavity.
Table 2. Test wall configurations
Test
No. Configuration MgO
board type
Insulation
material Noggings Board
joints
1 Type-1 None Noggings at
600 mm c/c Horizontal
2 Type-2 None None Vertical
3 Type-2 Glass fibre None Vertical
The LSF walls were built with six stiffened channel studs (92 × 35 × 15 × 1.15 mm) with
a yield strength of 300 MPa and the spacing was 600 mm c/c. The studs were fixed between
two tracks at the top and bottom, and then lined with 10 mm thick Type-1 or Type-2 MgO
boards with or without cavity insulation (Table 2). Figure 6 shows the full details of a test
wall configuration (Test 2). Test 3 wall panel configuration is similar to that shown in Figure
6. However, Test 1 wall panel had noggings at 600 mm nominal spacing and thus included
horizontal joints between MgO boards. In all three tests, the board was only fixed to the
studs.
11
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200

Figure 6 LSF wall configuration of Test 2
3.2 Specimen Construction
3.2.1 Test 1
The test wall construction was carried out in two stages. Stage-1 consisted of building the
steel frame made of studs, tracks and noggings and then the first layer of MgO board was
fixed to the steel frame on one side as shown in Figure 7. The studs were aligned vertically at
a spacing of 600 mm between the top and bottom tracks and then fixed with 8-gauge, D-
Type, flat head, self-drilling and non-corrosive screws. The size of the MgO board used in
this test was 1200 mm × 3000 mm. The noggings were fixed at 600 mm spacing between the
studs to support the horizontal board joints.
12
201
202
203
204
205
206
207
208
209
210
211

Figure 7. Test 1 wall configuration with one side board and noggings during Stage-1
Figure 8 shows the details of the horizontal board joints. In order to enhance the bonding
and seal the gap between the board and the steel, a fire rated sealant was applied on the steel
to form a uniform layer as shown in Figure 8a. This fire rated sealant is an acrylic based
sealing compound that can accommodate a maximum joint movement of +12.5%. The boards
were fixed to the steel frame by maintaining a gap of about 5 mm between the boards at the
horizontal joints as shown in Figure 8b. Then the boards were screwed to the studs at 200 mm
spacing with an edge distance of 50 mm from the horizontal joints. Once the board was fixed,
all the joints were sealed with the fire rated sealant to ensure bonding between the joints as
shown in Figure 8c, which was required to avoid joint failure during the fire test by
preventing hot gases passing through the joint.
13
600 mm wide MgO board
Noggings
Track
Studs
212
213
214
215
216
217
218
219
220
221
222
223

(a) Sealant between board
and steel (b) Gap between the boards (c) Joint filled with sealant
Figure 8. Test 1 joint details
Stage-2 consisted of fixing thermocouples to the wall as shown in Figure 7. At first,
thermocouples were fixed to the board and studs as shown in Figure 9 at various locations to
measure transient time-temperature variation during the fire test. All the horizontal joints in
the test specimen were supported by noggings at 600 mm spacing to avoid joint failure during
fire test. The second layer of MgO board was fixed to the steel frame on the other side after
applying a uniform layer of fire rated sealant using the methods used for the first layer of
MgO board.
(a) On the board (b) On the stud
Figure 9. Thermocouples fixed to the board and studs
3.2.2 Test 2
Test 2 LSF wall was made of 10 mm thick Type-2 MgO board without noggings and
cavity insulation material. Figure 10 shows the steel frame lined with MgO board on one
side. Type-2 MgO board had recessed edges that allowed the board joints to be constructed
14
MgO Board
Stud
Sealant
50 mm
5 mm gap Sealant
224
225
226
227
228
229
230
231
232
233
234
235
236

along the studs without noggings, thus no noggings were used in this steel frame. The boards
were fixed with their joints over the studs using non-corrosive screws at 200 mm spacing as
shown in Figure 11a. The joints were filled with a layer of gypsum based joint plaster and a
fibre mesh was used to reinforce the joint as shown in Figure 11b. Finally another layer of
joint plaster was applied on top of the fibre mesh to complete the joint.
Figure 10. Test 2 wall configuration with MgO board lining on one side
(a) Screws at 200 mm c/c (b) Joint plaster and fibre mesh
Figure 11. Test 2 joint details
15
MgO board
Studs
200 mm
Fibre mesh
Joint plaster
237
238
239
240
241
242
243
244

3.2.3 Test 3
Test 3 wall configuration was similar to Test 2 except the inclusion of cavity insulation
and additional joint protection using a high temperature mortar. This high temperature mortar
is a ready to mix, cement based and rapid setting material that can withstand temperatures up
to 1200 ºC. In addition, screw fastening was carried out after predrilling the board at screw
locations to ensure uncracked board edges. As shown in Figure 12 the board joints were filled
with layers of a high temperature cement mortar that can withstand a temperature of 1000 ºC
and fibre mesh laid between the cement mortar layers to reinforce the joints. The cement
mortar was applied to the board joints after wetting the surface with water to ensure good
bonding between the board and the mortar.
(a) First mortar layer application
after wetting the surface
(b) Second mortar layer application
after placing a fibre mesh
Figure 12. Test 3 joint details
The test wall was insulated with 75 mm thick glass fibre cavity insulation with a density
of 11 kg/m3, which is a common component in LSF wall systems to achieve enhanced
thermal comfort (Figure 13). Test 3 will provide experimental results on non-load bearing
cavity insulated LSF walls lined with MgO board in comparison with a non-cavity insulated
16
Joint
First mortar
layer Fibre
mesh
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259

configuration in Test 2 as well as provide some test results on the fire performance of the
high temperature joint mortar with improved joints.
Figure 13. Test 3 wall configuration with glass fibre cavity insulation
3.3 Test Set-Up and Instrumentation
The fire tests were carried out using a 3 m × 3 m gas furnace with a depth of 300 mm. A
hydraulic jack was used under the studs to support the self-weight of the LSF wall specimen.
The LSF wall specimen was placed in front of the furnace within the loading frame as shown
in Figure 14. The temperatures on the studs and the board surfaces were measured using
Type-K cable thermocouples, whose locations are shown in Figures 15 and 16. The
thermocouples attached to the fire side flange of the studs were named as hot flange (HF)
thermocouples and the unexposed side thermocouples attached to the flange of the studs were
named as cold flange (CF) thermocouples. In addition to the temperature measurements,
linear variable differential transformers (LVDTs) were used to monitor the lateral and vertical
displacements of the wall due to thermal bowing and expansion effects. Finally, the gap
between the furnace and the LSF wall specimen was completely sealed with insulation
materials to prevent heat escaping around the edges of the specimen to the environment,
17
Cavity
insulation
Track
Studs
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276

which may affect the furnace performance. The furnace was programmed to follow the
standard fire time-temperature curve. The furnace pressure was not recorded during the fire
tests.
Figure 14. Test wall and instrumentation set-up (Test 2)
Figure 15. Thermocouple locations on the studs
18
Cold flange (CF)
Hot flange (HF)
277
278
279
280
281
282
283

Figure 16. Thermocouple locations on the MgO board surfaces
4 Test Observations and Time-Temperature Profiles
4.1 Test 1
After about 10 minutes of fire exposure, water vapour was observed around the specimen.
This is due to the dehydration processes in the MgO board. At about 25 minutes, small cracks
started to appear on the ambient side of the wall, which opened up rapidly within another 5
minutes. During this period, loud cracking sound was heard from the fire side. With the
ambient side cracks becoming larger in size, the test was stopped by turning off the furnace
and then it was pushed back after 32 minutes of fire exposure. There was no smoke detected
during the test due to non-flammable nature of Type-1 MgO board. The ambient side board
exhibited large cracks as shown in Figure 17 that can be attributed to the higher mass loss and
shrinkage in Type-1 MgO board. However, the horizontal joints were still undamaged during
the fire test because of the flexible fire sealant used between the joints to accommodate any
movements in the joint due to thermal expansion. Post-fire observations on the fire side
19
750 mm
2250 mm
1500 mm
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298

showed delamination, bowing and cracking of the MgO board, which was directly exposed to
fire (Figure 18). The board deformation pattern aligned with the steel frame layout using the
studs and noggings, which is in the form of 600 mm × 600 mm square shapes (see Figure 7).
This suggests that the fixings also contributed to the failure of the board by providing
additional restraints. Partial fall off of the fire exposed MgO board was also observed.
Figure 17. Cracks on the ambient side after the test (Test 1)
Figure 18. Delamination, deformation and cracking on the fire side (Test 1)
20
Horizontal
joints
299
300
301
302
303
304
305
306

The failure of load bearing LSF wall systems depends on the critical HF temperatures as
discussed by Gunalan and Mahendran [12]. These HF temperatures can be obtained from
either load bearing or non-load bearing wall fire tests and then related to load acting on the
studs using load ratio. The load ratio is defined as the ratio between the applied load on the
stud and the ambient temperature ultimate capacity. The proposed load ratios (LR) and the
respective limiting HF temperatures based on [12] are; 1) LR=0.2: 600 ºC, 2) LR=0.4: 500
ºC, 3) LR=0.6: 300 ºC and 4) LR=0.7: 200 ºC. Based on these suggested LR values and
critical HF temperatures, the FRL of a load bearing LSF wall system can be obtained from
the time-temperature profiles obtained from non-load bearing test. Figures 19 and 20 show
the HF and CF temperature profiles measured during the fire tests. At 30 minutes, the stud
temperatures were 400 ºC and 275 ºC on the hot flange (HF) and cold flange (CF),
respectively. Hence, the respective LR for the HF temperature of 400 ºC measured during this
test is 0.5, which is representative of a FRL of 30/–/–.
The maximum deviation between the recorded fire curve vs the standard time-
temperature curve is about 70 ºC during the experiment, which is well within the AS 1530.4
recommended deviation of +100 ºC. The measured temperature on the ambient side was
about 95 ºC, which is well below the standard insulation failure criterion of 160 ºC on the
unexposed surface. However, the failure was initiated under the integrity failure criterion due
to cracking of the board that allowed hot gases to pass through the wall. This ignited a cotton
pad held closer to the crack within about 30 seconds. The FRL of this LSF wall system will
be –/30/– under integrity failure criterion.
21
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327

0 5 10 15 20 25 30 35 40
0
200
400
600
800
1000
Standard Fire Curve Test Fire Curve
HF CF
Time (minutes)
Temperature (oC)
Figure 19. Measured average stud time-temperature profiles in Test 1
0 5 10 15 20 25 30 35 40
0
200
400
600
800
1000
Standard Fire Curve
Test Fire Curve
Fire Side Cavity
Time (minutes)
Temperature (oC)
Figure 20. Measured average MgO board time-temperature profiles in Test 1
4.2 Test 2
Test 2 LSF wall was lined with Type-2 MgO Board with joints being aligned vertically
along the studs without noggings. When the test wall was exposed to standard fire on one
side, water vapour was observed around the test specimen after 3 minutes of fire exposure.
The water evaporation continued for about another 10 minutes, which lead to larger quantity
22
328
329
330
331
332
333
334
335
336

of water pouring from the test wall cavity on to the testing floor (Figure 21). This is due to
the condensation of evaporated water from the test wall, which is from the chemically bound
water in the board.
Figure 21. Water dripping from the test wall (Test 2)
At about 25 minutes of fire exposure, discolouration of one of the vertical joints along the
studs on the ambient surface was observed. This discolouration continued for another 5
minutes and then that joint started to become wider, while the other joints on the ambient
surface also started to discolour. This was followed by further widening of that joint due to
bowing and shrinkage of the ambient side board, which eventually caused cracking of board
at the vertical joint screw locations (Figure 22). When the vertical joint cracks became quite
large, they allowed hot gases to pass through the joint that ignited a cotton pad held closer to
the crack for about 30 seconds. Then the furnace was switched off at about 35 minutes of fire
exposure. Based on the observations, the test wall failed under integrity failure criterion. The
test wall also bent towards the fire side during the fire test. Post-fire observations showed
further cracking along the vertical joints, board bowing and cracking on the fire side as
shown in Figure 23. The board bowing pattern (Figure 23) aligned well with the stud
arrangements and screw fastener locations for Test 2 shown in Figure 10, which is different
to the bowing pattern observed in Test 1 (see Figure 18), where the steel frame had horizontal
23
Water
dripping
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355

noggings (see Figure 7). In comparison to Test 1, there were no board cracks observed in Test
2 during the fire tests. In Test 1 with noggings and more flexible horizontal joints, board
cracks dominated the failure.
(a) Test wall with cracked vertical joint (b) Region-A
Figure 22. Vertical joint cracked on the ambient side (Test 2)
24
A
356
357
358
359
360

Figure 23. Bowing and opening of the fire side board joint after the test (Test 2)
The reason for the Test 2 wall failure under the integrity failure criterion was due to the
opening up of the joint. This was caused by the shrinkage and bowing actions of the board
during the fire test, which eventually caused the board edges to crack at the screw locations.
The boards were screwed to the studs at the joints with an edge distance of 5 – 10 mm due to
the limitations in the stud flange width, which was about 35 mm. In addition, the board does
not have the fibre mesh on one side at the edges that was removed to reinforce the joint
compound as shown in Figure 24. Therefore, limited edge distance for screwing and the
removal of the fibre mesh along the recessed edges might have caused the joints to fail.
Another cause for the joint failure might be attributed to the use of conventional joint plaster
and a fibre mesh that also failed during the fire test. However, the major cause of wall failure
is excessive bowing and shrinkage of the board, which eventually triggered the joint cracks.
25
Bowing of board
towards fire side
Wide-open
board joint
361
362
363
364
365
366
367
368
369
370
371
372
373

Figure 24. Recessed board edge without fibre mesh
The time-temperature profiles measured during Test 2 are shown in Figures 25 and 26.
The average stud temperatures at about 30 minutes were 400 ºC and 240 ºC measured on the
HF and CF, respectively. As discussed for Test 1, when this test is considered as a load
bearing wall, the respective LR for the HF temperature of 400 ºC measured during this Test 2
is 0.5, which is representative of a FRL of 30/–/–. The average temperature of the ambient
side surface was still around 100 ºC, which is well below the insulation failure criteria limit
of 160 ºC, so that the insulation failure criterion was not reached during this test. The test
wall failure was due to the cracking of the joints, which allowed hot gases to pass through the
wall. Therefore, when this test wall considered as a non-load bearing wall, the failure can be
classified under integrity failure criterion with a FRL of –/30/–.
26
374
375
376
377
378
379
380
381
382
383
384
385

0 5 10 15 20 25 30 35 40
0
200
400
600
800
1000
Standard Fire Curve Test Fire Curve
HF CF
Time (minutes)
Temperature (oC)
Figure 25. Measured average stud time-temperature profiles in Test 2
0 5 10 15 20 25 30 35 40
0
200
400
600
800
1000
Standard Fire Curve
Test Fire Curve
Fire Side Cavity
Ambient Side Cavity
Time (minutes)
Temperature (oC)
Figure 26. Measured average MgO board time-temperature profiles in Test 2
4.3 Test 3
Test 3 wall had 10 mm Type-2 MgO board lining and glass fibre cavity insulation. The
joints were reinforced with high temperature mortar to enhance the joint performance. After 3
minutes of fire exposure, water vapour was observed and at about 10 minutes water started to
pour down the wall, which continued for about another 15 minutes. This is due to the
27
386
387
388
389
390
391
392
393
394

dehydration of the chemically bound water in the MgO board. At about 35 minutes one of the
ambient side vertical joints started to crack and then continued to increase in size until about
40 minutes, where the joint opening was significant as shown in Figure 27. This allowed hot
gases to pass through the wall, which ignited a cotton pad held closer to the joint within about
30 seconds. Therefore the wall failed at about 40 minutes under integrity failure criterion.
Although the test wall failed at 40 minutes, the test was continued for another 10 minutes to
monitor the behaviour of the boards. During this period the board started to bow further and
the joint materials fell off due to wide open vertical joints as shown in Figure 27. The post-
fire observation of the fire exposed side showed bowing and partial fall off of MgO board,
whereas the glass fibre insulation has completely melted and lost its integrity due to excessive
deformation as shown in Figure 28. The melting temperature of the glass fibre insulation was
about 700 ºC.
(a) Ambient side of the wall
Figure 27. Ambient side of the test wall after the test (Test 3)
28
A
C
B
395
396
397
398
399
400
401
402
403
404
405
406

(b) Region A (c) Region B (d) Region C
Figure 27. Ambient side of the test wall after the test (Test 3)
Figure 28. Bowing and partial board fall off on the fire side after the test (Test 3)
Figures 29 and 30 show the average stud and board time-temperature profiles,
respectively. The stud temperatures are 475 ºC and 210 ºC on the HF and CF at about 30
29
Melted glass
fibre insulation
407
408
409
410
411
412

minutes as shown in Figure 29. Cavity insulation blocks the heat passage through the cavity,
thus the HF gets heated higher than the CF. This will eventually cause early failure of the
studs at lower load ratios when this wall is considered as a load bearing wall. The HF
temperature of 475 ºC can be related to a load ratio of 0.4 with a FRL of 30/–/– at this load
ratio. When the board temperatures are considered, the ambient surface temperature was
about 90 ºC at the end of 40 minutes of fire exposure, which is well below the insulation
failure criterion temperature of 160 ºC. However, the joints on the ambient surface cracked
and become wide open, allowing the hot gases to pass through the wall. Thus the failure of
the wall can be classified under integrity failure with a FRL of –/30/–.
0 5 10 15 20 25 30 35 40 45 50 55 60
0
200
400
600
800
1000
Standard Fire Curve Test Fire Curve
HF CF
Time (minutes)
Temperature (oC)
Figure 29. Measured average stud time-temperature profiles in Test 3
30
413
414
415
416
417
418
419
420
421
422
423

0 5 10 15 20 25 30 35 40 45 50 55 60
0
200
400
600
800
1000
Standard Fire Curve Test Fire Curve
Fire Side Cavity Ambient Side Cavity
Ambient
Time (minutes)
Temperature (oC)
Figure 30. Measured average board time-temperature profiles in Test 3
4.4 Comparison of Average Board Temperatures
The fire side cavity (FireCavity) and the ambient side (Ambient) board temperature
profiles of all three fire tests are shown in Figure 31. The ambient side temperature values
follow a similar pattern during the fire test, and this temperature at failure of each test was
measured to be about 100 ºC, which is well below the failure temperature under insulation
failure criterion as discussed earlier. The fire side cavity temperature variations for Test 1 and
2 almost match each other. However, the fire side cavity temperature of Test 3 is much higher
than in other two tests, with a temperature difference of about 240 ºC at 30 minutes. This is
due to the presence of cavity insulation in Test 3, which blocks heat transfer through the wall
cavity, thus the temperature on the fire side cavity is rapidly increased. This effect will affect
the fire performance of the LSF wall system by increasing the stud HF temperature to cause
early failure in a load bearing wall configuration. Therefore, cavity insulation will affect the
fire performance of load bearing LSF walls. However, it is included often to achieve greater
thermal comfort during service.
31
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439

0 5 10 15 20 25 30 35 40 45 50 55 60
0
200
400
600
800
1000
Standard Fire Curve Test-1-Fire Cavity
Test-2-Fire Cavity Test-3-Fire Cavity
Test-1-Ambient Test-2-Ambient
Test-3-Ambient Time (minutes)
Temperature (oC)
Figure 31. Average fire side cavity and ambient side board temperature profiles
4.5 Comparison of Average Stud Temperatures
Figure 32 shows the HF and CF time-temperature profiles from the three full-scale fire
tests. Test 3 shows the highest HF and the lowest CF temperatures compared to other two
tests. This is due to the presence of cavity insulation, which blocks the heat through the test
wall as discussed earlier. At about 30 minutes of fire exposure the differences between HF
and CF temperatures are 105 ºC, 153 ºC and 259 ºC in Tests 1 to 3, respectively. When this
difference is higher, the stud will fail earlier due to the reduced mechanical properties of the
HF compared to CF, which will eventually cause additional deflection and moment in the
studs when subjected to an axial compressive load. Therefore, cavity insulation affects the
fire performance of axially loaded fire rated load bearing LSF wall configurations. These
observations were also reported by Gunalan et al. [2] based on their experimental studies of
LSF walls lined with gypsum plasterboard.
32
440
441
442
443
444
445
446
447
448
449
450
451
452
453

0 5 10 15 20 25 30 35 40 45 50 55 60
0
200
400
600
800
1000
Standard Fire Curve Test-1-HF Test-2-HF
Test-3-HF Test-1-CF Test-2-CF
Test-3-CF Time (minutes)
Temperature (oC)
Figure 32. Average stud HF and CF temperature profiles
4.6 Comparison of lateral displacements
Another important observation during the fire tests was the lateral displacement of the
wall towards the furnace. Although the test wall studs were not subjected to any loads, the
test wall deflection was observed due to the thermal expansion of the LSF wall components.
The measured lateral displacements at the mid-height of all three test walls are shown in
Figure 33. The lateral displacement was the highest in Test 3 with cavity insulation compared
to other two tests without cavity insulation. Test 3 shows a lateral deflection of 40 mm
towards the furnace at about 30 minutes of fire exposure compared to 25 mm in other two
tests. This is mainly due to the larger thermal gradient and the reduced steel mechanical
properties caused by the heat being blocked by cavity insulation.
33
454
455
456
457
458
459
460
461
462
463
464
465

0 5 10 15 20 25 30 35 40 45 50 55 60
0
10
20
30
40
50
60
Test-1 Test-2 Test-3
Time (minutes)
Lateral Displacement (mm)
Figure 33. Average lateral displacements measured during fire test
4.7 Test on evaporated water
After the fire test, a pH was conducted on the evaporated water from the MgO board that
was accumulated in the exhaust duct system. The pH indicator revealed that the evaporated
water was acidic with a pH level between 4 and 5. Based on the MgO board constituents the
chemical formulae of the MgO board can be given as Mg(OH)2.MgCl2.H2O. When heated,
the chemicals are released in the form of Mg(OH)2, MgO, HCl and H2O. Therefore, it can be
concluded that the acidity of the evaporated water is due to the presence of HCl released
when the board is exposed to fire.
5 Findings
The main findings of the experimental study reported in this paper are as follows.
34
466
467
468
469
470
471
472
473
474
475
476
477

The MgO board lined LSF wall configurations considered in this study yielded a FRL
of 30/30/30, considering the test observation that insulation failure criteria was not
reached.
All three fire tests failed under integrity failure criterion, where Test 1 with Type-1
MgO board failed due to board cracking and Tests 2 and 3 with Type-2 boards
exhibited board joint cracking that allowed hot gases to pass through the wall.
Although Test 3 joints were reinforced with high temperature mortar, the joint
cracking could not be avoided. However, until the board bent and the cracks opened
up, the joints were able to reduce the heat transfer to the stud considerably in
comparison to the heat transfer through the board. The cracks occurred mainly due to
the excessive shrinkage of the board caused by dehydration reactions (mass loss) and
bowing of the board. Reducing the amount of chemically bound water in the MgO
board will enhance its fire performance. Another way of mitigating the board bending
deformation will be to reduce the stud spacing (currently 600 mm c/c). Closely spaced
studs will enhance the fire performance by reducing the bending action of the board,
thus eliminating joint cracks. However, this will increase the cost of the wall panels.
This observation implies that FRL of MgO board lined walls will depend very much
on stud spacing.
Removal of fibre mesh on the edges to prepare the recessed board edges affects the
strength of the board when exposed to fire. Therefore, alternative manufacturing
processes is needed to retain the fibre mesh on the board edges.
The average edge distances of the screws are about 5 – 10 mm due to the limited
flange width of about 35 mm in Tests 2 and 3. This is not sufficient to withstand
higher movement in the joints due to the bowing of MgO board. Therefore, increased
flange width or back to back stud arrangements are recommended to achieve higher
35
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502

joint strength in fire exposures, which will eventually increase the cost of the wall
system.
Pre-drilling is recommended for screw fastening the board to the stud. This will
prevent cracking of the board when the screws sink to position.
Excessive board cracking was observed in Test 1 with noggings compared to Tests 2
and 3 without noggings, which might have restricted thermal expansion of the board.
Therefore, it is recommended to avoid noggings in LSF wall steel frame, where larger
deformation of the wall board lining is expected when exposed to fire.
Type-2 MgO boards used in Tests 2 and 3 showed no cracks on the board, whereas in
Test 1 board cracking was observed due to the noggings restraining the bowing action
of board. Another reason is higher mass loss in Type-1 MgO board compared to
Type-2 MgO board. Manufacturers recommend the use of fire rated sealant between
the board and studs as well as to fill the joints. However, fire tests showed that such
sealants accommodate movements only to a certain extent. When there is excessive
movement of the joint due to board bowing action and shrinkage, these sealants may
not provide any further protection. Tests demonstrated clearly the importance of joint
detailing on the FRL of LSF wall systems, but most importantly the need to reduce
bending/bowing action and cracking of MgO board caused by significant mass loss
during their dehydration reactions.
Post-fire pH test of the evaporated water from the board during the fire test revealed a
pH level of 4 – 5, i.e. acidic. This is due to the presence of MgCl2 in the MgO board
composition, which releases HCl during fire exposure. Therefore, MgO board lining
is not recommended for structures exposed to severe weather conditions.
36
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526

6 Conclusions
This paper has presented the details of thermal property tests and full-scale fire tests of
MgO boards and MgO board lined LSF walls systems with different configurations. Thermal
property tests conducted on two types of MgO boards have shown that the MgO board had
very high mass loss at elevated temperatures, which is about 50% and 40% loss of initial
mass at elevated temperatures compared to 15% mass loss in a commonly used wall lining
material such as gypsum plasterboard. This higher mass loss will affect the fire performance
of the MgO board lined LSF wall system and will cause cracking of MgO board.
These findings were supported by conducting three full-scale standard fire tests on MgO
board lined LSF wall systems. These fire tests exhibited bending and cracking of MgO board.
All three tests failed in integrity failure criterion, in which Test 1 with Type-1 MgO board
failed due to cracking of board and Tests 2 and 3 with Type-2 MgO board failed due to joint
opening and cracking. The board cracking was due to the higher mass loss as observed during
the thermal property tests while joint failure was caused by bowing and shrinkage of the
board. Therefore, the MgO board lined LSF system will have an FRL of 30/30/30, in which
the wall may fail predominantly under integrity failure criterion.
Fire test observations and results reported in this paper are only applicable to the two
types of MgO boards and the wall configurations used in the tests. This research has
demonstrated that the FRL of MgO board lined LSF wall system depends on the type of MgO
board used in relation to its mass loss characteristics, joint detailing and stud spacing.
Appropriately designed and manufactured MgO board with suitable joint details and stud
spacing have the potential to provide higher FRLs than those reported in this paper.
37
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549

Acknowledgements
The authors would like to thank Queensland University of Technology (QUT) for
providing the necessary research facilities, and Australian Research Council (ARC) and QUT
for providing the financial support to conduct this research project.
References
[1] Standards Australia (SA), AS 1530.4: 2014, Methods for fire tests on building
materials, components and structures, Part 4: Fire-resistance tests of elements of
building construction, 2014, Sydney, Australia.
[2] S. Gunalan, P. Kolarkar, M. Mahendran, Experimental study of load bearing cold-
formed steel wall systems under fire conditions. Thin-Walled Structures, 65 (2013),
pp. 72-92.
[3] A. Ariyanayagam, M. Mahendran, M. Experimental study of load-bearing cold-
formed steel walls exposed to realistic design fires. Journal of Structural Fire
Engineering, 5 (4) (2014), pp. 291-330.
[4] S. Kesawan, M. Mahendran, Fire tests of load-bearing LSF walls made of hollow
flange channel sections. Journal of Constructional Steel Research, 115 (2015),
December 2015, pp. 191-205.
[5] V. Jatheeshan, M. Mahendran, Experimental study of LSF floors made of hollow
flange channel section joists under fire conditions. Journal of Structural Engineering,
ASCE, 142 (2) (2016), 040215134.
[6] A.Y. Nassif, I. Yoshitake, A. Allam, Full-scale fire testing and numerical modelling
of the transient thermo-mechanical behaviour of steel-stud gypsum board partition
walls. Construction and Building Materials, 59 (2014), pp. 51-61.
38
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572

[7] W. Chen, J. Ye, Y. Bai, X.L. Zhao, Full-scale fire experiments on load-bearing cold-
formed steel walls lined with different panels. Journal of Constructional Steel
Research, 79 (2012), pp. 242-254.
[8] W. Chen, J. Ye, Y. Bai, X.L. Zhao, Improved fire resistant performance of load
bearing cold-formed steel interior and exterior wall systems. Thin-Walled Structures,
73 (2013), pp. 145-157.
[9] W. Chen, J. Ye, Y. Bai, X.L. Zhao, Thermal and mechanical modelling of load-
bearing cold-formed steel wall systems in fire. Journal of Structural Engineering,
ASCE, 140 (2013), A4013002.
[10] M. T. Hanna, J. C. B. Abreu, B. W. Schafer, M. Abu-Hamd, Post-fire buckling
strength of CFS walls sheathed with magnesium oxide or ferrocement boards.
Proceedings of the Annual Stability Conference, Structural Stability Research
Council, Nashville, Tennessee, March 24-27, 2015.
[11] ASTM E1269-11, Standard test method for determining specific heat capacity by
differential scanning calorimetry, ASTM International, West Conshohocken, PA,
2011.
[12] S. Gunalan, M. Mahendran, Fire performance of cold-formed steel wall panels and
prediction of their fire resistance rating. Fire Safety Journal, 64 (2014), pp. 61-80.
39
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591