Thin-Walled Structures 115 (2017) 86–99

Contents lists available at ScienceDirect

Thin-Walled Structures
journal homepage: www.elsevier.com/locate/tws

Fire tests of non-load bearing light gauge steel frame walls lined with MARK
calcium silicate boards and gypsum plasterboards
⁎
Anthony Deloge Ariyanayagam, Mahen Mahendran
Science and Engineering Faculty Queensland University of Technology, Brisbane, Australia

A R T I C L E I N F O A B S T R A C T

Keywords: Light gauge Steel Frame (LSF) wall systems are made of cold-formed steel studs and tracks, and lined with wall
Fire tests lining materials. Conventionally, gypsum plasterboards are used as wall lining material in LSF wall systems. The
Light gauge steel frame walls fire performance of LSF walls is mainly dependent on the type and configuration of wall lining material, which
Calcium silicate board delays the heat transfer through the wall and protects the steel studs from being heated rapidly. Recently,
Gypsum plasterboard
calcium silicate board lining is increasingly used in LSF wall systems because of its improved physical and
Magnesium oxide board
Fire resistance level
thermal properties while being lightweight, cost effective, impact and moisture resistant. However, their fire
performance has not been investigated in detail. Hence two full scale fire tests were conducted on non-load
bearing LSF walls lined with calcium silicate boards. For comparison purposes two fire tests were also conducted
on conventional gypsum plasterboard lined LSF walls. This paper presents the details of this experimental study
on the fire performance of LSF walls and the results including fire resistance levels and time-temperature profiles
across the wall panels. Effects of using calcium silicate board lining are discussed by comparing the fire test
results of LSF wall lined with gypsum plasterboards and previously conducted studies on magnesium oxide board
lined LSF walls. The results showed that the fire performance of calcium silicate board lined walls was similar to
that of conventional gypsum plasterboard lined walls, but was superior to that of magnesium oxide board lined
walls. The failure criterion of these calcium silicate board lined walls was found to be insulation as opposed to
being the integrity failure observed in previous studies.

1. Introduction
Cold-formed Light gauge Steel Frame (LSF) wall systems are
commonly used in buildings as both load bearing and non-load bearing
walls. They are made of cold-formed steel studs and tracks and lined
with boards. During fire events, LSF walls act as separating elements
between rooms/compartments and resist fire spread. In general LSF
walls are exposed to fire from one side. Thus they develop a
temperature gradient across the cross-section, and thermally bow
towards the fire side due to differential thermal expansion. This
temperature gradient will induce non-uniform distribution of strength
and stiffness across the steel stud. Also the thermal-mechanical proper-
ties of cold-formed steel studs deteriorate at elevated temperatures, and
cause structural failures of thin-walled studs in load bearing walls with
continuing fire exposure. Therefore fire protective wall linings are
provided in LSF walls to protect steel studs from direct fire exposure
and to delay their temperature rise. LSF wall fire performance is mainly
dependent on the wall configuration and wall lining material. LSF wall
configuration includes studs lined with single and double boards with
and without insulation. The insulated wall panels can be either cavity
insulated or externally insulated, i.e. insulation layer sandwiched
between boards. The wall lining not only protects the studs from rapid
temperature rise but also provides adequate restraint against buckling
of the studs about their minor-axis and twisting.
The Fire Resistance Level (FRL) of LSF walls when exposed to the
standard fire time-temperature curve on one side is determined based
on three criteria [1]. They are: 1) Structural adequacy: ability to
maintain the stability and carry the design loads, 2) Integrity: ability
to resist the passage of flames and hot gases and 3) Insulation: ability to
keep the temperature on the unexposed surface below the limits (i.e.,
the maximum and average unexposed wall surface temperature should
not exceed the room temperature by 180º C and 140º C, respectively).
Gypsum plasterboards are commonly used as lining material in LSF
walls. Pure Gypsum, known as Calcium sulphate di-hydrate (CaSO4·
2H2O), consists of Calcium sulphate with 4–5% of free water and 15–
18% chemically bound water by weight [2–7]. The free and chemically
bound water in gypsum plasterboard is important because it contributes
to the fire resisting behaviour. When exposed to fire, gypsum plaster-

⁎
Corresponding author.
E-mail address: m.mahendran@qut.edu.au (M. Mahendran).

http://dx.doi.org/10.1016/j.tws.2017.02.005
Received 2 January 2017; Received in revised form 9 February 2017; Accepted 9 February 2017
0263-8231/ © 2017 Elsevier Ltd. All rights reserved.
A.D. Ariyanayagam, M. Mahendran
Fig. 1. LSF Wall configurations used in Fire Tests.
board undergoes reactions in which the water is gradually driven off at
temperatures above 100° C. Significant heat energy is required to
evaporate the free water and for the chemical reaction to release the
water in the crystal structure. Gypsum plasterboard also consists of
glass fibre, vermiculite and perlite additives, which improve the
durability and fire performance. Many experimental and numerical
studies [8–20] were conducted to determine the effect of different wall
configurations on the fire performance of LSF walls, such as thickness
and number of layers of gypsum plasterboards, grade and thickness of
cold-formed steel stud sections, stud shapes and wall configurations,
when exposed to both standard fire and realistic design fire curves.
Recently, calcium silicate boards have been increasingly used in LSF
walls because of their improved physical and thermal properties while
being lightweight, cost effective, impact and moisture resistant. Chen
et al. [21] conducted two full scale fire tests of load bearing LSF walls
(3.38 m×2.98 m) lined with calcium silicate boards. The first fire test
wall panel had two layers of 12 mm thick calcium silicate boards while
the second test panel had a combination of 12 mm thick calcium silicate
87
Thin-Walled Structures 115 (2017) 86–99
board and 12.5 mm thick gypsum plasterboard (face layer). The results
showed that calcium silicate board had explosive spalling at elevated
temperatures, thus resulting in lower FRL than gypsum plasterboard
lined LSF walls. Although the failure in the first fire test was based on
the integrity criteria at 58 min, the second wall panel fire test with the
combination of calcium silicate board and gypsum plasterboard suf-
fered a structural failure at 92 min. The studs exhibited both major axis
flexural buckling and local compression failure of the cold-flange. The
cracking of the ambient board surface i.e. gypsum plasterboard surface
was not observed, thus integrity failure was not the governing failure
criterion in this test. Wang et al. [22] conducted one full scale
(3 m×3 m) and five small scale (1.2 m×1.2 m) tests of LSF walls lined
with 9 mm thick calcium silicate boards focusing on the behaviour of
embedded electrical junction box on the fire performance. The study
was focused on the opening in the wall panel due to the fixing of
junction boxes, thus FRLs of the tested wall panels were not reported.
However, they reported that cracking and spalling of calcium silicate
board and explosive noises were observed during the fire test. The study
A.D. Ariyanayagam, M. Mahendran
Fig. 2. LSF wall frame.
also concluded that the quality of the calcium silicate board is vital in
the fireproof effectiveness of these board systems. Although there has
been significant research in the past on the structural and thermal
behaviour of LSF wall panels lined with gypsum plasterboards [8–20]
very few attempts have been made to study the behaviour of LSF wall
panels lined with calcium silicate boards. Also the fire tests conducted
so far [21,22] were on thinner boards, i.e. 9 and 12 mm calcium silicate
boards.
Thicker calcium silicate boards are likely to perform differently
under fire conditions as opposed to thinner boards. The discrete glass
fibres are evenly spread across the thickness of calcium silicate board.
Thus it is expected that cracking and fall-off of calcium silicate board
particles could be delayed and thicker calcium silicate boards may
perform better than thinner boards. Therefore fire tests of non-load
bearing walls were conducted to investigate the fire performance of LSF
walls lined with 20 mm thick calcium silicate boards. For comparison
purposes two full scale fire tests were also conducted on gypsum
plasterboard lined walls.
This paper presents the details of full scale fire tests conducted on
four non-load bearing LSF wall panels lined with calcium silicate boards
and gypsum plasterboards. The failure times and the measured board
and stud temperatures across the wall panels are presented. Effects of
using calcium silicate boards are discussed by comparing the time-
temperature curves obtained for gypsum plasterboard lined walls from
this study and magnesium oxide board lined walls from previous
studies.
2. Fire test details
2.1. LSF wall components
Four full scale fire tests of LSF wall panels of 3 m width and 3 m
height were conducted for the LSF wall configurations shown in Fig. 1.
LSF wall frames were constructed using six lipped channel section studs
and two unlipped channel section tracks. Both studs and tracks were
made of 1.15 mm thick G500 steel. The 90×40×15×1.15 mm studs of
3 m length and the 92×50×1.15 mm tracks of 3.2 m length were used.
The studs were inserted inside the tracks at 600 mm spacings and screw
fastened to both flanges of the top and bottom tracks using D-type
16 mm long flat head self drilling screws. LSF wall frames were then
lined with 16 mm thick fire rated gypsum plasterboards and 20 mm
thick calcium silicate boards based on the test program shown in Fig. 1.
The gypsum plasterboards (GB) of 1200 mm×3000 mm (density of
88
Thin-Walled Structures 115 (2017) 86–99
13 kg/m2) with recessed edges were used as the lining material for Test
specimens LSF1, LSF3 and LSF4 (only as the face layer). They are
manufactured using calcium sulphate, clay, fly ash, vermiculite, glass
fibres and additives. Test specimens LSF2 and LSF4 (only as the base
layer) were lined with 20 mm thick 1200 mm×2500 mm Calcium
Silicate boards (CS). These calcium silicate boards are manufactured
using Calcium Silicate Hydroxide (Ca6Si6O17(OH)2) also known as
Xonotlite, mineral binder and additives. They had a smooth and dimple
pattern surface finish, and a density of 17 kg/m2. The thermal
conductivity and specific heat of gypsum plasterboard and calcium
silicate board used in the fire tests are 0.2 W/mK and 1.09 J/kg. K, and
0.164 W/mK and 0.84 kJ/kg. K, respectively.
2.2. Construction of test specimens
2.2.1. Test specimens LSF1 and LSF2
Test specimens LSF1 and LSF2 were lined with single layer of wall
boards. Test specimen LSF1 was lined with 16 mm thick gypsum
plasterboard. D-type 25 mm long self drilling screws were used to
fasten the 16 mm gypsum plasterboards at 300 mm spacing along the
studs and, 200 mm staggered spacing over the studs that had gypsum
plasterboard joints along their length. Gypsum plasterboards were only
screw fastened to the studs and not to the top and bottom tracks. The
plasterboard layer consisted of three pieces, 600×3000 mm,
1200×3000 mm and 1200×3000 mm to accommodate two vertical
plasterboard joints on Studs 2 and 4 on the fire side and, on Studs 3 and
5 on the ambient side. The recessed edge plasterboard joints were filled
with two nearly equal thickness gypsum based joint filler compound
and finished to the top level of the plasterboard surface. A paper based
tape layer was sandwiched between the two coats of the filler material.
Test specimen LSF2 was lined with 20 mm thick calcium silicate
boards on both sides of the studs. The same fixing detail as in Test
specimen LSF1 was used in Test specimen LSF2. However, calcium
silicate boards are available in 2500 mm lengths. Thus additional
noggings were placed at 500 mm height to accommodate the board
joints (Fig. 2). However, calcium silicate boards were only fastened to
the studs, and not to the noggings. D-type 35 mm long screws were used
to fasten the calcium silicate boards to the studs. The same gypsum
based joint filler and paper tape were used along the calcium silicate
board joints.
2.2.2. Test specimens LSF3 and LSF4
Test specimens LSF3 and LSF4 were constructed similar to Test
specimens LSF1 and LSF2, but were lined with two layers of wall
boards. Test specimen LSF3 was lined with two layers of 16 mm thick
fire rated gypsum plasterboards on both sides of the studs. The first
(base) layer of plasterboards was installed vertically on both sides of the
LSF wall panel while the second (face) layer was installed horizontally.
The base layer (GB2 and GB3) consisted of three pieces:
600×3000 mm, 1200×3000 mm and 1200×3000 mm to accommo-
date the vertical plasterboard joints on Studs 2 and 4 on the fire side
plasterboard (GB2) and on Studs 3 and 5 on the ambient side plaster-
board (GB3). The face layer (GB1 and GB4) also had three pieces:
3000×1350 mm, 3000×1350 mm and 3000×300 mm and included
two horizontal plasterboard joints. D-type 25 and 45 mm self drilling
flat head screws were used to fasten the first and second layers of
gypsum plasterboards in Test specimen LSF3, respectively. The first
layer of boards was fastened at 200 mm spacing along the studs where
gypsum plasterboard joints exist and 300 mm spacing along the
intermediate studs. The second layer of boards was fastened at
300 mm spacing along the studs.
Test specimen LSF4 was lined with a combination of gypsum
plasterboards and calcium silicate boards. The base layer (CS1 and
CS2) was 20 mm calcium silicate board while the second (face) layer
(GB1 and GB2) was 16 mm fire rated gypsum plasterboard on each side
of the studs. The same fixing detailing as in Test specimen LSF3 was
A.D. Ariyanayagam, M. Mahendran Thin-Walled Structures 115 (2017) 86–99

Fig. 3. Thermocouple locations on LSF wall panels.

used. However, D-type 35 mm long self drilling flat head screws and S-
type 75 mm long bugle head screws were used to fasten the 20 mm
thick calcium silicate boards to the studs. Similar to Test specimen
LSF2, noggings were placed at 500 mm height to accommodate the
2500 mm long calcium silicate boards (Fig. 2).

2.3. Locations of displacement transducers and thermocouples

The lateral (out-of-plane) deflections of two middle studs (Studs 3

and 4) were measured using Displacement Transducers (DT) located at
three levels along the stud heights of 750, 1500 and 2250 mm. K-type
cable thermocouples (accuracy of +/-0.0075T within the range of
−200 to 1100° C, where T is the measured temperature in Celsius) were
used to measure the temperatures in the Test specimens LSF1 to LSF4.
The stud and plasterboard surface temperatures were measured at wall
heights of 750, 1500 and 2250 mm. The thermocouple wires on the
studs were connected to their hot and cold flanges and web. Fig. 3
shows the locations of thermocouples along and across the wall
specimens.

2.4. Test set-up and fire test method

Fig. 4. Fire Test set-up.
The fire test was conducted using the propane gas-fired furnace in
the QUT Fire Testing Laboratory. The furnace (3 m×3 m×0.4 m) has

89
A.D. Ariyanayagam, M. Mahendran
Fig. 5. Fire test results of test specimen LSF1.
six radiant flame burners located at 400 mm away from the face of the
test specimen. Fire test was conducted in a specially designed test rig,
shown in Fig. 4, where a load of 0.5 kN was applied to each of the six
studs from the bottom to support the self weight of the wall panel. The
individual studs were concentrically loaded using six hydraulic rams
connected to a pump. The supporting frame consisted of two universal
columns (UC) bolted to the strong floor, and a universal beam (UB) at
the top and bottom levels of the furnace that were connected to two
universal columns (UC). The wall panel unrestrained on its vertical
sides with the gap between the supporting frame and the panel was
filled with ceramic insulation. The target fire time-temperature curve
was set to the standard fire time-temperature curve as specified in AS
1530.4 [1].
3. Fire test observations and results
The wall panel was pre-loaded in axial compression to 0.5 kN per
stud prior to the start of the test, and this load was maintained during
the entire duration of the fire tests. The wall panel was exposed on one
side to the standard fire time-temperature curve. Figs. 5–8 show the
measured time-temperature curves of plasterboard and stud surfaces
and the lateral deflections of Test specimens LSF1 to LSF4, respectively.
Figs. 9–12 show the wall panels after the fire tests. In all the fire tests
the target standard fire time-temperature curve was achieved within
the limits specified in AS 1530.4 [1].
90
Thin-Walled Structures 115 (2017) 86–99
3.1. Fire test LSF1
Fire Test LSF1 wall panel was lined with single layer of 16 mm thick
gypsum plasterboard on both sides of the studs. After 3 min of starting
the furnace, smoke was visible at the top of the specimen for 7 min.
Water drops were seen along the bottom edges of the loading frame. At
the end of 15 min smoke started to emerge again, and continued for
10 min. Again at 48 min, smoke started to appear from the top left hand
corner of the specimen (near Stud 1) and continued for 5 min. The
appearance of smoke was considered to be due to the burning of
plasterboard paper. Water drops were seen again on the surface at the
end of 53 min. Then at the end of 76 min plasterboard joint filler
material along Stud 3 started to discolour from white to brown colour.
At 78 min ambient side plasterboard (GB2) surface started to discolour
and the plasterboard joint filler material along Stud 3 turned to dark
brown colour.
The average measured plasterboard surface time-temperature
curves across the LSF wall panel are shown in Fig. 5(a). In the early
stages of the fire, the dehydration of plasterboards maintained the
surface temperatures at approximately 100° C, while the furnace
temperature was well above 800° C. As seen in the Fire side GB1-
Cavity plasterboard surface time-temperature curves, after about
22 min the fire side face layer plasterboard (GB1) was completely
dehydrated and the temperature started to increase rapidly. Fig. 5(b)
shows the individual and average thermocouple readings on the
A.D. Ariyanayagam, M. Mahendran
Fig. 6. Fire Test results of test specimen LSF2.
ambient side gypsum plasterboard (GB2) surface. It shows that the
individual surface temperatures measured on the ambient surface did
not go beyond the maximum insulation temperature limit that is 202° C
(ambient temperature at the start of the fire test 22° C +180° C).
However, the average insulation temperature limit of 162° C (ambient
temperature at the start of the fire test 22° C +140° C) was exceeded at
94 min. Hence it is concluded that the insulation failure occurred at
94 min based on AS 1530.4 [1]. However, test was continued until
121 min.
As seen in Fig. 5(c), Stud 4 hot flange temperature showed higher
temperatures than Stud 3. This is due to the presence of fire side
gypsum plasterboard vertical joint over Stud 4. Stud 4 hot flange
temperature was 614° C when the insulation criterion was reached at
94 min, and was 655° C at the end of the test and, Stud 3 hot flange
temperatures were 593 and 625° C, respectively. Fig. 5(d) shows the
lateral deflection curves of middle Studs 3 and 4 measured at 750, 1500
and 2250 mm heights (Top, Mid and Bottom). The studs bowed towards
the hot side due to the non-uniform temperature distribution across the
stud section. But with time they started to reverse or remained constant
with respect to its temperature difference across the section. A
maximum deflection of 35 mm was observed at the mid-height of
Stud 4 at 45 min, and it was 33 mm at mid-height of Stud 3. After
45 min, lateral deflections of Studs 3 and 4 reduced and remained
nearly constant until the end of the test. Fig. 9 shows the test wall panel
LSF1 after the fire test. Visual observations after the fire test showed the
91
Thin-Walled Structures 115 (2017) 86–99
discolouration in ambient side gypsum plasterboard (GB2) surface and
the plasterboard joint along Stud 3. Fire side plasterboard (GB1) joint
along Stud 4 had cracked and opened up and plasterboard pieces had
fallen off inside the furnace.
3.2. Fire test LSF2
Fire Test LSF2 wall panel was lined with 20 mm thick calcium
silicate boards on both sides of the studs. After 34 mins of fire exposure
moisture was visible on the board joints. Also water drops were seen
along the edges of the wall panel at 55 min. At the end of 82 mins board
joint filler cracked and board cracking sound was heard. The integrity
failure test was performed using a cotton pad along the calcium silicate
board joint [1] and, no burning of cotton pad was observed. Water
drops were also seen along the edges of the board, and at the end of
109 mins board cracking sound was again heard. Fig. 6 shows the time-
temperature plots of calcium silicate board and studs and, lateral
deflection versus time curves from Fire Test LSF2. As seen in
Fig. 6(b), the individual surface temperature measured on the ambient
side calcium silicate (CS2) board exceeded 204° C (ambient tempera-
ture at the start of the fire test 24° C +180° C), the maximum insulation
temperature limit after 113 min of fire exposure. However, the average
ambient surface temperature was below the average insulation tem-
perature limit of 164° C (ambient temperature at the start of the fire test
24° C +140° C). Hence it can be concluded that the insulation failure
A.D. Ariyanayagam, M. Mahendran
Fig. 7. Fire Test results of test specimen LSF3.
occurred at 113 min based on the maximum temperature insulation
criterion in AS 1530.4 [1]. However, test was continued until 161 min.
Further cracking or spalling of calcium silicate board was not observed
on the ambient calcium silicate board surface.
As seen in Fig. 6(b) thermocouple placed between Studs 2 and 3 at
2250 mm height on the ambient calcium silicate board surface (see
Fig. 3(a) for thermocouple locations) showed higher temperatures than
other thermocouples on the same surface even within the close
proximity. This suggests that this thermocouple measurement was
influenced by a localized effect. If this critical thermocouple measure-
ment was ignored then the average ambient surface temperature would
govern the insulation failure instead of the maximum ambient surface
temperature giving a higher insulation failure time of 130 min (see
Fig. 6(b)). This is a significant increase in FRL for 20 mm thick calcium
silicate board compared to the 16 mm gypsum plasterboard in Fire Test
LSF1 (94 min).
Fig. 6(c) shows the measured stud temperatures while Fig. 6(d)
shows the lateral deflections of Studs 3 and 4 along the height. Stud 3
hot flange recorded higher temperatures than Stud 4 and it was 519° C
when the insulation criterion was reached at 113 min, and 592° C at the
end of the test. This shows that the stud hot flange temperatures in Test
LSF2 are less than those measured in Fire Test LSF1. Fig. 10 shows the
LSF wall after the fire test. It shows that the fire side calcium silicate
board (CS1) joints have cracked and opened up. This allowed the hot
gases to penetrate the cavity. However, no significant calcium silicate
92
Thin-Walled Structures 115 (2017) 86–99
board fall-off was observed even in the fire side layer (CS1) and the
board remained in-place. Further, integrity failure with bursts, large
cracks and spalling off of calcium silicate board were also not observed
during and after the fire test in both fire side and ambient side calcium
silicate boards.
3.3. Fire test LSF3
Fire Test LSF3 wall panel was lined with two layers of 16 mm thick
gypsum plasterboards on both sides of the studs. After 5 min of starting
the furnace, smoke was visible at the top of the specimen for nearly
6 min. Again at 70 min, smoke started to appear from the right side of
the specimen and the smoke layer was very thick and continued for
15 min. This was due to burning of paper layers on the fire side GB1-
GB2 surface. Smoke appeared again at 120, 141 and 168 min from the
top left hand corner of the wall panel. At 190 min, the joint filler on the
ambient side face layer plasterboard (GB4) started to discolour and
screw heads were visible through the plasterboard joints (Fig. 11). Then
at 200 min the horizontal joints on the ambient side face layer plaster-
board (GB4) started to open up and the joint filler material turned to
dark brown colour.
The average measured plasterboard time-temperature curves across
the test specimen are shown in Fig. 7(a). In the early stages of the fire,
the dehydration of plasterboards maintained the surface temperatures
at approximately 100° C, while the furnace temperature was well above
A.D. Ariyanayagam, M. Mahendran
Fig. 8. Fire Test results of test specimen LSF4.
700° C. After about 25 min the fire side face layer plasterboard (GB1)
was completely dehydrated and the temperature started to increase
rapidly, as seen in the fire side GB1-GB2 plasterboard surface time-
temperature curves (Fig. 7(a)). A closer look at the fire side GB1-GB2
surface plasterboard time-temperature curve revealed a rapid rise in the
temperature at about 190 min. Following this, this surface temperature
increased rapidly and approached the furnace time-temperature curve.
Fig. 7(b) shows the individual and average thermcouple readings
measured on the ambient side plasterboard surface. It shows that the
maximum insulation temperature limit of 210° C (ambient temperature
at the start of the fire test 30° C +180° C) was exceeded at 197 min. The
average insulation temperature limit of 170° C (ambient temperature at
the start of the fire test 30° C +140° C) was reached at about 205 min.
Hence it can be concluded that the insulation failure occurred at
197 min. However, test was continued until 220 min. As seen in
Fig. 7(c) Stud 4 hot flange temperature was 780° C when the insulation
criterion was reached at 197 min, and was approximately 820° C at the
end of the test. Studs 3 and 4 at mid-height showed a maximum
deflection of nearly 21 mm at about 110 min (Fig. 7(d)).
Fig. 11 shows the test panel after the fire test. Visual observations
after the fire test showed that most of the fire side gypsum plasterboard
(GB1) had fallen off and the base layer gypsum plasterboard (GB2) had
cracked with board joints opening up at some locations (see Fig. 11).
3.4. Fire test LSF4
Fire Test panel LSF4 was lined with two layers of boards on both
93
Thin-Walled Structures 115 (2017) 86–99
sides, 20 mm calcium silicate board base layer and 16 mm gypsum
plasterboard face layer. After 30 min of starting the furnace, smoke and
steam were visible from the bottom of the specimen for nearly 2 min.
Also water drops were seen along the edges of the wall specimen. This is
most likely due to the release of water from the gypsum plasterboard
(GB1). At the end of 37 min wet patches were seen on the ambient side
gypsum plasterboard (GB2) surface and smoke started to emerge again
from the top of the specimen and continued for 12 min. At the end of
105 mins board cracking sound was heard. This could have been due to
the cracking of fire side calcium silicate board. At 174 min the joint
filler on the ambient side plasterboard (GB2) started to discolour from
white to brown colour.
Fig. 8 shows the measured board and stud temperatures and lateral
deflections from Fire Test LSF4. It shows that even after 250 min of fire
exposure the ambient surface temperatures (GB2) were well below
100° C (Fig. 8(a) and (b)). The average ambient side plasterboard
surface temperature was only 79° C while the maximum temperature on
this surface was only 88° C after 250 min of fire exposure (Fig. 8(b)).
These are well below the insulation temperature limits specified in
AS1530.4 [1]. Hence the test was terminated after 250 min of fire
exposure. Adding a 16 mm thick gypsum plasterboard layer on both
sides of the wall panel has considerably improved the integrity and
insulation performances. As shown in Fig. 8(c), the stud hot flange
temperatures were nearly 650° C at the end of the test. As observed in
previous fire tests, Studs 3 and 4 were observed to bend towards the
furnace during the fire test (Fig. 8(d)). The maximum lateral deflection
of 43 mm was recorded at the mid-height of Stud 3.
A.D. Ariyanayagam, M. Mahendran
Fig. 9. Test specimen LSF1 after the Fire Test.
Fig. 12 shows the fire exposed side of the LSF wall after the fire test.
Visual inspection of the specimen after the fire test revealed that the fire
side gypsum plasterboard (GB1) face layer had mostly fallen off from
the wall leaving the fire side calcium silicate board (CS1) base layer
exposed to fire. However, the calcium silicate board remained in-place
and resisted the heat transfer to the ambient side boards. This is similar
to what was observed in Fire Test LSF2. Similarly ambient side calcium
silicate board (CS2) and gypsum plasterboard (GB2) remained in place.
The ambient side gypsum plasterboard (GB2) layer was seen in good
condition with the presence of confining paper layers.
4. Discussion
4.1. Comparisons with gypsum plasterboard lined LSF wall fire tests
In gypsum plasterboard lined LSF wall specimens LSF1 and LSF3,
insulation failure limit was exceeded prior to the integrity failure.
Similarly, Test specimen LSF2 with single layer of calcium silicate
board lining also failed based on insulation criterion. However, Test
specimen LSF4 lined with a combination of gypsum plasterboard and
calcium silicate board did not fail even after 250 min of fire exposure.
Integrity failure was not observed in calcium silicate board lined LSF
wall specimens LSF2 and LSF4.
Comparing the results of Tests LSF1 and LSF2, the use of 20 mm
calcium silicate board lining instead of 16 mm gypsum plasterboard
increased the FRL due to insulation failure from 94 to 113 min, i.e.
nearly 20 min increase in FRL. As described earlier, if the failure was
based on the average insulation temperature limit instead of the
94
Thin-Walled Structures 115 (2017) 86–99
maximum temperature limit, the FRL of Test specimen LSF2 would be
130 min. Fig. 13(a) shows the average board time-temperature curves
obtained from Tests LSF1 and LSF2. It shows that calcium silicate board
lined wall showed higher fire resistance to temperature rise across the
wall than gypsum plasterboard lined wall. Average calcium silicate
board temperatures, i.e. fire side cavity, ambient side cavity and
ambient side board temperatures from Fire Test LSF2, are less than
the gypsum plasterboard temperatures in Fire Test LSF1 across the wall
panel. It is to be noted that Fire side board temperatures in these two
tests merged well as seen in Fig. 13(a). However, fire side cavity
temperatures showed a difference from 20 min to reach an average
temperature difference of nearly 110° C at 120 min. In both tests, the
average board time-temperature curves had three distinct phases:
gradual temperature rise, constant temperature and rapid temperature
rise. In the early stages of the fire, dehydration of boards maintained
the surface temperatures at approximately 100° C, while the furnace
temperatures were well above 800° C. The average gypsum plaster-
board temperature in Fire GB1-Cav surface showed a sudden increase in
temperature at about 25 min (Fig. 13(a)). This is after the evaporation
and release of the water bound within the gypsum plasterboard at about
100° C. Similar observations can be seen in Fire Test LSF2 with a rapid
increase in temperature of Fire CS1-Cav surface at about 35 min, most
likely after the dehydration process in calcium silicate board. In both
Fire Tests LSF1 and LSF2 steam was seen to escape from the wall panels
during the early stages of testing. Both fire test results showed similar
behaviour, and the difference in board temperatures could be attributed
to the different board thickness, i.e. 20 mm calcium silicate board
versus 16 mm gypsum plasterboard.
A.D. Ariyanayagam, M. Mahendran
Fig. 10. Test specimen LSF2 after the Fire Test.
Comparison of Fire Tests LSF3 and LSF4 results also shows similar
behaviour, where two layers of gypsum plasterboard lined LSF3 wall
panel showed higher average board temperatures across the wall than
LSF4 wall panel (Fig. 13(b)). However, due to the use of 16 mm thick
gypsum plasterboard as the face layer in both tests fire side board
surface temperatures (GB1-GB2 and GB1-CS1) merged well as seen in
Fig. 13(b). There is a temperature difference of about 150° C across the
cavity in Test LSF3 at about 200 mins of fire exposure. This difference is
only about 40° C in Test LSF4, and Amb Cav-CS2 surface temperature is
higher than the same surface Amb Cav-GB3 in Test LSF3. This could be
due to the smooth finish of calcium silicate board surface (no paper
layer), which has the potential to reflect the heat. Thus the heat
radiated through the cavity is reflected back towards the cavity, which
causes a low temperature difference across the cavity and delays the
temperature rise of the ambient side board surfaces. Except the ambient
cavity face, other board surface temperatures across the LSF wall were
seen to be higher in gypsum plasterboard lined wall than calcium
silicate board lined wall.
Further, it is important to note is that the visual observations of the
fire exposed surface after the Fire Tests LSF2 and LSF4 revealed that
calcium silicate boards were relatively intact with the wall and did not
fall-off (Figs. 10,12 and 14). Only the filler material in the board joints
was seen to have peeled off from the wall panel. No obvious signs were
95
Thin-Walled Structures 115 (2017) 86–99
found of portions of the wall breaking away. This could be due to the
use of 20 mm thick calcium silicate board and the composition of raw
materials used to manufacture the calcium silicate boards. However,
hair-line cracks were visible on the fire exposed side calcium silicate
board surface, as a result of dehydration process (Fig. 14). Gypsum
plasterboard relies on the paper layer on either side of the board to hold
the gypsum board together, and when this is burnt the board loses its
integrity. Calcium silicate boards used in the fire tests do not have paper
layers. They had the discrete glass fibres holding the board together
throughout the board thickness. This seems to be giving higher degree
of integrity for the boards, and protecting the fall-off and collapse of
calcium silicate board pieces than the paper layer in gypsum plaster-
board. Thus in load bearing walls, the calcium silicate boards will
remain in place and act not only as a thermal barrier, but will also
provide lateral and twisting restraints to the studs. However, its
thermo-mechanical properties should be measured at elevated tem-
peratures to quantify the effect of lateral and twisting restraints to the
studs exposed to fire.
In summary both gypsum plasterboard and calcium silicate board
lined LSF walls performed in a similar manner when exposed to fire on
one side. The dehydration process in both boards was also similar. The
variations observed in the time-temperature curves were due to the use
of different board thicknesses, glossy and smooth surface of calcium
A.D. Ariyanayagam, M. Mahendran
Fig. 11. Test specimen LSF3 after the Fire Test.
silicate boards and the differences in chemical composition and thermo-
mechanical properties of the boards.
4.2. Comparisons with other calcium silicate board lined LSF wall fire tests
LSF wall fire tests on calcium silicate boards were previously
conducted by Chen et al. [21], who observed cracking and spalling of
calcium silicate boards at elevated temperatures. This led to the
integrity failure of LSF wall panels lined with calcium silicate boards.
However, the test wall panels in [21] were lined with 12 mm thick
calcium silicate boards. Fire tests in this research were based on 20 mm
thick calcium silicate boards (Fire Tests LSF3 and LSF4) which showed
that calcium silicate boards did not exhibit cracks or spalling. In both
fire tests calcium silicate boards were intact with the wall panel
(Figs. 10,12 and 14).
96
Thin-Walled Structures 115 (2017) 86–99
This could be due to the use of thicker boards, presence of discrete
fibres throughout the thickness of the board. Only the board joints were
observed to open up on the fire side boards. Also, Chen et al. [21] fire
tests were for load bearing wall panels with higher load ratios and studs
failed in flexural buckling. Thus the loss of integrity of ambient side
calcium silicate board in one of the fire tests could have been due to the
flexural buckling of studs. Fire tests presented in this paper were based
on non-load bearing wall panels with thicker boards, and ambient side
boards losing their integrity were not observed. Insulation failure
occurred first in Test LSF2 while in Test LSF4 the integrity or insulation
criterion was not exceeded even after 250 min of fire exposure.
Therefore it can be concluded that in thicker (20 mm) calcium silicate
board lined LSF walls, failure is governed by insulation criterion. Thus
further studies can be now performed on LSF walls lined with calcium
silicate boards using finite element thermal models of different wall
A.D. Ariyanayagam, M. Mahendran
Fig. 12. Test specimen LSF4 after the Fire Test.
configurations, especially on walls with calcium silicate boards as face
layers, since they provide good impact and abrasion resistance and
moisture tolerance characteristics. The measured thermal properties of
the LSF wall lining materials, gypsum plasterboard, calcium silicate
board and magnesium oxide boards used in these fire tests available in
[23] can be used in these studies.
4.3. Comparisons with magnesium oxide board lined LSF wall fire tests
Fig. 15 compares the average board temperatures obtained from
20 mm thick calcium silicate board lined LSF wall Test LSF2 with the
magnesium oxide (MgO) boards lined LSF wall fire tests conducted by
Rusthi et al. [24] for similar LSF wall configuration. Two types of
10 mm thick MgO boards lined LSF wall fire tests (MgO – Tests 1 and 2)
were conducted by Rusthi et al. [24]. The MgO boards used in Test 2
had higher magnesium oxide (MgO) and lower magnesium chloride
(MgCl2) contents and with higher fibreglass constituents. As reported in
[23] both fire tests failed due to integrity failure [1]. In MgO – Test 1,
cracks started to appear on the ambient MgO board surface of the LSF
wall after 25 min, and opened up rapidly along the surface with loud
noise within another 5 min. Similarly, MgO board joints along the studs
started to crack and open up after 30 min of fire exposure. The cracks
fully opened up after about 35 min, with integrity failure at 35 min.
97
Thin-Walled Structures 115 (2017) 86–99
Post-fire observation on the fire side showed delamination and bending
of boards, with large cracks on both fire and ambient sides of MgO
boards. Thus it allowed the hot air to escape through the wall and the
average MgO board surface temperatures were much higher than the
calcium silicate boards lined wall (Test LSF2). After 30 min of fire
exposure cavity facing fire side board temperatures (Fire MgO1-Cav) in
MgO – Tests 1 and 2 are nearly 400° C, and it was only 100° C in
calcium silicate board lined wall Test LSF2. Also the dehydration
process at 100° C, i.e. the delay in the board temperature rise observed
in both gypsum plasterboard and calcium silicate boards lined wall fire
tests were not observed in MgO board tests (see Fig. 15). The average
board temperatures in the MgO board lined tests continuously in-
creased until failure. However, calcium silicate board time-temperature
curves showed a distinct pattern: gradual temperature rise, temperature
plateau and a rapid temperature rise similar to that of gypsum plaster-
board lined LSF walls. Calcium silicate boards in Test LSF2 did not
crack, and allow the hot air through the LSF wall. Therefore the average
calcium silicate board temperatures are less than the MgO board
temperatures. Thus it can be concluded that the calcium silicate boards
exhibited superior fire performance without cracking and spalling in
comparison with MgO boards used in [23].
A.D. Ariyanayagam, M. Mahendran Thin-Walled Structures 115 (2017) 86–99

Fig. 15. Comparison of Average Board Temperatures of LSF Walls Lined with Calcium
Fig. 13. Comparison of Average Board Temperatures - Fire Tests LSF1 to LSF4. Silicate Boards and MgO Boards.

Fig. 14. Cross-section of Calcium Silicate Board and Fire Exposed Surface.

98
A.D. Ariyanayagam, M. Mahendran
5. Conclusions
This paper has presented the details of full-scale fire tests of non-
load bearing LSF wall panels made of thin-walled steel studs and lined
with gypsum plasterboard and calcium silicate boards. The fire test
results were used to determine the differences between the fire
performance of LSF walls lined with conventional gypsum plasterboard,
calcium silicate board and magnesium oxide boards. They showed that
calcium silicate board lined walls performed in a similar manner to that
of gypsum plasterboard lined walls and superior to that of magnesium
oxide board lined walls. No integrity failure was observed in the 20 mm
thick calcium silicate board lining. This is due to the use of thicker
boards and the presence of fibrous material throughout the board
thickness. The failure of the non-load bearing wall panel lined with
calcium silicate board was due to insulation criterion.
Calcium silicate boards stayed intact with the wall panels and
retained their thermal barrier feature for the entire duration of the fire
test. Fall-off and cracking of calcium silicate boards were not observed
during the fire tests as in gypsum plasterboards lined wall fire tests.
Hence using calcium silicate board as the face layer instead of base
layer will further improve the FRL of LSF walls. With the failure
criterion being identified as insulation in non-load bearing walls,
further studies can now be performed to improve the FRL of LSF walls
lined with calcium silicate boards using finite element thermal models
of different wall configurations. The use of calcium silicate boards in
load bearing walls is also expected to be beneficial since it provides
greater resistant’s to minor axis buckling and twisting of the studs than
gypsum plasterboards.
Acknowledgements
The authors would like to thank Australian Research Council for
their financial support, Queensland University of Technology (QUT) for
providing the necessary research facilities and support to conduct this
research project, and Stephen J. Bragg and Sean P. Wrigley, QUT
undergraduate civil engineering students, and QUT Banyo laboratory
staff members for their assistance with fire testing of walls.
References
[1] Standards Australia (SA), AS 1530.4: 2005, Methods for fire tests on building
materials, components and structures, Part 4: Fire Resistance Tests of elements of
building construction, Sydney, Australia, 2005.
[2] G. Thomas, Thermal properties of gypsum plasterboard at high temperatures, Fire
Mater. 26 (2002) 37–45.
[3] L.M. Samuel, G.G. Richard, R.K. Scott, B.L. David, Influence of gypsum board type
99
Thin-Walled Structures 115 (2017) 86–99
(X or C) on real fire performance of partition assemblies, Fire Mater. 31 (2007)
425–442.
[4] L. Wullschleger, K.G. Wakili, Numerical parameter study of the thermal behaviour
of a gypsum plasterboard at fire temperatures, Fire Mater. 32 (2008) 103–119.
[5] K.G. Wakili, E. Hugi, Four types of gypsum plasterboards and their thermophysical
properties under fire condition, J. Fire Sci. 27 (1) (2009) 27–43.
[6] P. Keerthan, M. Mahendran, Numerical studies of gypsum plasterboard panels
under standard fire conditions, Fire Saf. J. 53 (2012) 105–119.
[7] P. Keerthan, M. Mahendran, Thermal performance of composite panels under fire
conditions using numerical studies: plasterboards, rockwool, glass fibre and
cellulose insulations, Fire Technol. 49 (2012) 329–356.
[8] F. Alfawakhiri, Fire resistance of load bearing steel-stud walls protected with
gypsum board: a review, Fire Technol. 35 (4) (1999) 308–335.
[9] M. Feng, Y.C. Wang, J. Davies, Thermal performance of cold-formed thin-walled
steel panel systems in fire, Fire Saf. J. 38 (4) (2003) 365–394.
[10] M. Feng, Y.C. Wang, An analysis of the structural behaviour of axially loaded full-
scale cold-formed thin-walled steel structural panels tested under fire conditions,
Thin-Walled Struct. 43 (2005) 291–332.
[11] V.K.R. Kodur, M.A. Sultan, Factors influencing fire resistance of load-bearing steel
stud walls, Fire Technol. 42 (2006) 5–26.
[12] P.N. Kolarkar, M. Mahendran, Experimental studies of non-load bearing steel wall
systems under fire conditions, Fire Saf. J. 53 (2012) 85–104.
[13] S. Gunalan, P.N. Kolarkar, M. Mahendran, Experimental study of load bearing cold-
formed steel wall systems under fire conditions, Thin Walled Struct. 65 (2013)
72–92.
[14] S. Gunalan, M. Mahendran, Finite element modelling of load bearing cold-formed
steel wall systems under fire conditions, Eng. Struct. 56 (2013) 1007–1027.
[15] A.D. Ariyanayagam, M. Mahendran, Experimental study of load bearing light gauge
steel frame walls exposed to realistic design fires, J. Struct. Fire Eng. 5 (4) (2014)
291–329.
[16] A.D. Ariyanayagam, M. Mahendran, Numerical modelling of load bearing light
gauge steel frame walls systems exposed to realistic design fires, Thin-Walled
Struct. 78 (2014) 148–170.
[17] S. Kesawan, M. Mahendran, Fire tests of load bearing LSF walls made of hollow
flange channel sections, J. Constr. Steel Res. 155 (2015) 191–205.
[18] S. Kesawan, M. Mahendran, Thermal performance of load bearing walls made of
cold-formed hollow flange channel sections in fire, Fire Mater. (2015), http://
dx.doi.org/10.1002/fam.2337.
[19] M.A. Sultan, Effect of gypsum board orientation on board fall-off in fire resistance
test assemblies, in: R.A. Wessel, P.H. Shipp, G.L. Rosenthal (Eds.), West
Conshoshocken, ASTM International, 2015, http://dx.doi.org/10.1520/
STP1588201500132.
[20] A.D. Ariyanayagam, S. Kesawan, M. Mahendran, Detrimental effects of plasterboard
joints on the fire resistance of light gauge steel frame walls, Thin-Walled Struct. 107
(2016) 597–611.
[21] W. Chen, J. Ye, Yu Bai, X. Zhao, Full-scale fire experiments on load-bearing cold-
formed steel walls lined with different panels, J. Constr. Steel Res. 79 (2012)
242–254.
[22] Y. Wang, C. Ying-Ji, C. Lin, The performance of calcium silicate board partition
fireproof drywall assembly with junction box under fire, J. Adv. Mater. Sci. Eng.
125 (2015) 1–12.
[23] M.I. Rusthi, Investigating the Fire Performance of LSF Wall Systems Using
Numerical and Experimental Studies (PhD Thesis), Queensland University of
Technology, Brisbane, Australia, 2017.
[24] M.I. Rusthi, A.D. Ariyanayagam, M. Mahendran, P. Keerthan, Fire tests of
magnesium oxide board lined light gauge steel frame wall systems, J. Struct. Fire
Eng. (2017) (In Press).