~ Experimental study of hydrothermal aging effects on intumescent coating
Transcript of ~ Experimental study of hydrothermal aging effects on intumescent coating
8/18/2019 ~ Experimental study of hydrothermal aging effects on intumescent coating
http://slidepdf.com/reader/full/-experimental-study-of-hydrothermal-aging-effects-on-intumescent-coating 1/14
Author's personal copy
Experimental study of hydrothermal aging effects on insulative properties
of intumescent coating for steel elements
L.L. Wang a, Y.C. Wang a,b,n, G.Q. Li a
a College of Civil Engineering, Tongji University, Chinab School of Mechanical, Aerospace and Civil Engineering, University of Manchester, UK
a r t i c l e i n f o
Article history:
Received 29 September 2010
Received in revised form
9 January 2012
Accepted 17 October 2012
Keywords:
Intumescent coating
Hydrothermal aging test
Fire test
Effective thermal conductivity
Fire resistance
a b s t r a c t
This paper reports the results of an experimental study of degradation in fire protection performance of
two types of intumescent coating after different cycles of accelerated hydrothermal aging tests.
Intumescent coating (without top coating) was applied to steel plate to make a test specimen. After
subjecting the specimen to the aging test, fire test was carried out to obtain the steel plate temperature.
In order to help understand the aging mechanism of intumescent coating, TGA tests, XPS tests and FTIR
tests were also conducted on the intumescent coating after the accelerated aging test. In total, tests
were performed on 56 intumescent coating protected steel specimens, of which 16 specimens were
applied with type-U intumescent coating and the other 40 with type-A intumescent coating. Results of
the degradation mechanism study reveal that the hydrophilic components of intumescent coating
move to the surface of the coating and can be dissolved by moisture in the air, which can destroy the
intended chemical reactions of these components with others and deter formation of the desired
effective intumescent char. The consequence of this is reduced expansion of the intumescent coating
and increased effective thermal conductivity. Compared to specimens without hydrothermal aging,
after 42 cycles of hydrothermal aging (to simulate 20 years of exposure to an assumed exposure
environment), the effective thermal conductivity of type-U intumescent coating was 50% higher andthat of type-A intumescent coating 100% higher than that of the fresh coating. These increases in
effective thermal conductivities resulted in increases in steel temperatures of up to 150 1C and 220 1C
higher than the steel temperatures of the specimens without hydrothermal aging for the type-U
intumescent coating and type-A intumescent coating specimens, respectively.
& 2012 Elsevier Ltd. All rights reserved.
1. Introduction
Among different forms of fire protection to steel structures,
intumescent coating is particularly favored by architects because it
allows the attractive steel structural form to be exposed. Intumescent
coatings are now widely used as passive fire protection to steel
structures and in countries such as the UK, the use of intumescent
coating dominates the passive fire protection market [3]. The coat-ings, which usually are composed of organic components contained
in a polymer matrix, are designed to decompose and expand when
subjected to high temperatures so as to provide an insulating, foamed
char to protect the underlying substrate.
When specifying intumescent coating fire protection for steel
structures, the following assumptions are made:
(1) the type and thickness of the intumescent are correctly
specified;
(2) the intumescent coating is correctly applied;
(3) the fire protection performance of intumescent coating does
not degrade in time.
Assumptions (1) and (2) may not be fulfilled in practice, but
the problem is not a technical one. Assumption (3) deals with
durability of intumescent coating. Since most of the chemical
components in intumescent coating are organic, it would not beunreasonable to expect that they react with the exposed environ-
ment and that the fire protection function of intumescent coatings
deteriorates over time.
There are very few reported research studies in open literature
on durability of intumescent coatings. Sakumoto et al.[8] carried
out some accelerated aging tests according to the standards of
[7,2] to investigate the principal environmental factors that affect
the durability of intumescent coatings; [9,10] carried out accel-
erated aging tests in a SH60CA weatherometer according to [1]
standard. However, despite progresses made in these studies,
there was no quantification of how the fire resistance perfor-
mance of intumescent coatings reduces over time.
This paper reports the results of a comprehensive experimental
study to provide some quantitative information on reduced fire
Contents lists available at SciVerse ScienceDirect
journal homepage: www.elsevier.com/locate/firesaf
Fire Safety Journal
0379-7112/$ - see front matter & 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.firesaf.2012.10.004
n Corresponding author at: University of Manchester, School of Mechanical,
Aerospace and Civil Engineering, PO Box 88, Manchester, M60 1QD, UK
E-mail addresses: [email protected],
[email protected] (Y.C. Wang).
Fire Safety Journal 55 (2013) 168–181
8/18/2019 ~ Experimental study of hydrothermal aging effects on intumescent coating
http://slidepdf.com/reader/full/-experimental-study-of-hydrothermal-aging-effects-on-intumescent-coating 2/14
Author's personal copy
protection performance of degraded intumescent coatings to
steel structures. Two series of tests have been conducted. In series
one (to be referred to as fire test), intumescent coating specimens
were subjected to different cycles of accelerated aging and then
tested in fire. The measured data include final expanded thick-
nesses and the substrate steel temperatures. From these tests, the
effects of hydrothermal aging on effective thermal conductivities(related to the original intumescent coating thickness) of intumes-
cent coatings were obtained. In the accompanying series of tests
(to be referred to as chemical analysis tests), the aged intumescent
coatings were subjected to TGA, XPS and FTIR tests to measure
their mass loss, change of element contents and migration of
components in the intumescent coating system. The tests help to
explain the degradation processes.
2. Fire tests
2.1. Specimen preparations
A total of 56 specimens were tested. Each specimen was madeof 16 mm thick steel plate coated with 1 mm or 2 mm Dry Film
Thickness (DFT) intumescent coatings on all sides. Of the test
specimens, 16 were protected by type-U intumescent coating (to
be referred to as type-U specimens) and 40 were applied with
type-A intumescent coating (to be referred to type-A specimens).
The two different types of intumescent coating were supplied by
two different manufacturers operating in the Chinese market. The
principal components of type-U and type-A intumescent coating
are APP-MEL-DPER (aided with zinc borate) and APP-MEL-PER,
respectively; the acid resins of type-U and type-A intumescent
coatings are ethylene benzene–acrylic and single component
acrylic, respectively.
For the 16 type-U specimens, four replicate tests were per-
formed for each of the following 4 cycles of accelerated aging: 0
(no aging), 11 cycles (simulating 5 years in service), 21 cycles
(simulating 10 years in service) and 42 cycles (simulating 20
years in service). All specimens were coated with 1 mm DFT. For
the 40 type-A specimens, 20 were coated with DFT 1 mm coating
and the other 20 with DFT 2 mm coating. For each coating
thickness, four replicate tests were performed for each of the
following 5 cycles of accelerated aging: 0 (no aging), 4 cycles
(simulating 2 years in service), 11 cycles (simulating 5 years in
service), 21 cycles (simulating 10 years in service) and 42 cycles
(simulating 20 years in service). In all cases, the substrate steel
plate measured 200 mm by 270 mm by 16 mm thick. A primer
was applied to the steel surface first to act as an aid to adhesion of
the intumescent coating; this was then followed by different
layers of intumescent coating to achieve the desired DFT.
However, no top coating was applied. For each specimen, DFTwas measured and recorded before the accelerated aging test.
Three thermocouples (2.0 mm diameter, type K) were embedded
in each steel plate. Table 1 lists the main specimen parameters
and Fig. 1 shows the specimen dimensions, where d is the
intumescent DFT.
2.2. Hydrothermal aging test
Intumescent coating aging is an extremely complicated pro-
cess of physical and chemical interactions between the chemical
components of intumescent coatings and the external environ-
ment. Whilst it would be ideal to carry out real time aging test,
this process would be extremely long, running into many tens of
years. An alternative is to conduct accelerated aging test, in whicha real environmental condition over a long period of time is
represented by a short cyclic exposure of the intumescent coating
to a concentrated dosage of the environment. During any accel-
erated aging test, it is necessary to determine the environmental
conditions that the product (intumescent coating) will be exposed
to, the length of time of the exposure and the performance
criterion based on which the effect of aging is assessed.
The accelerated aging test was performed according to the
European guideline ([5]). In this guidance, four types of environ-
mental exposure are simulated: (a) type X for all conditions;
(b) type Y for internal and semi-exposed conditions; (c) type Z1
for internal conditions which have above zero temperatures and
high humidity; and (d) type Z2 for internal conditions that have
above zero temperatures but humidity conditions that are not in
class Z1. The accelerated aging test reported in this paper adoptedexposure condition Z1, simulating the more severe exposure
condition of application around the coastal provinces in China.
Table 1
Main test parameters.
Coating
type
Coating
DFT (mm)
No. of cycles of
accelerated aging
Simulating time
in service (years)
Specimen
ID
U 1 0 0 UI-1–00-i
(i ¼1–4)11 5 UI-1–11-i
(i ¼1–4)
21 10 UI-1–21-i
(i ¼1–4)
42 20 UI-1–42-i
(i ¼1–4)
A 1 0 0 AZ-1–00-i
(i ¼1–4)
4 2 AZ-1-04-i
(i ¼1–4)
11 5 AZ-1-11-i
(i ¼1–4)
21 10 AZ-1-21-i
(i ¼1–4)
42 20 AZ-1-42-i
(i ¼1–4)
2 0 0 AZ-2–00-i
(i ¼1–4)
4 2 AZ-2-04-i
(i ¼1–4)
11 5 AZ-2-11-i
(i ¼1–4)
21 10 AZ-2-21-i
(i ¼1–4)
42 20 AZ-2-42-i
(i ¼1–4)
Fig. 1. Specimen dimensions.
L.L. Wang et al. / Fire Safety Journal 55 (2013) 168–181 169
8/18/2019 ~ Experimental study of hydrothermal aging effects on intumescent coating
http://slidepdf.com/reader/full/-experimental-study-of-hydrothermal-aging-effects-on-intumescent-coating 3/14
Author's personal copy
For exposure condition type Z1, each cycle of exposure is as
follows:
8 h at (4073) 1C and (9872)%RH;
16 h at (2373) 1C and (7572)%RH.
According to ETAG018, 21 cycles of accelerated aging is
equivalent to 10 years in service. Based on this correlation, 0 cycle,
4 cycles, 11 cycles and 42 cycles correspond to fresh coating,
2 years, 5 years and 20 years in service.
2.3. Surface appearance
After the accelerated aging test but before the fire test, the
specimens were checked for their coating surface appearance.
Figs. 2 and 3 show typical appearance of type-U and type-A
specimens after having gone different cycles of accelerated aging
test. Type-U specimens did not appear to suffer any change in
appearance after 11 and 21 cycles of hydrothermal aging tests
(Fig. 2(b) and (c)). After 42 cycles, wrinkles can be clearly seen
(Fig. 2(c)). In contrast, type-A specimens experienced noticeable
changes in appearance after every accelerated aging test. After
Fig. 2. Type-U coating appearance after different cycles of hydrothermal aging test. (a) UI-1-00, (b) UI-1-11, (c) UI-1-21 and (d) UI-1-42.
Fig. 3. Type-A coating appearance after different cycles of hydrothermal aging test. (a) AZ-1-00, (b) AZ-1-04, (c) AZ-1-11, (d) AZ-1-21 and (e) AZ-1-42.
Fig. 4. Furnace door with observation holes.
L.L. Wang et al. / Fire Safety Journal 55 (2013) 168–181170
8/18/2019 ~ Experimental study of hydrothermal aging effects on intumescent coating
http://slidepdf.com/reader/full/-experimental-study-of-hydrothermal-aging-effects-on-intumescent-coating 4/14
Author's personal copy
only 4 cycles, the surface of type-A specimens appeared uneven
(Fig. 3(b)). After 11 cycles of accelerated aging test, bumps
appeared on the surface (Fig. 3(c)). After 21 and 42 cycles, the
specimen surface was uneven, with very large bumps
(Fig. 3(d) and (e)). As will be shown later in this paper, there is
strong link between the surface appearance and fire protection
performance of an intumescent coating. Coating surface appear-
ance may be explored when determining a replacement strategy
in real applications.
2.4. Fire test
After the specimens were subjected to hydrothermal aging as
described in the previous section, they were placed in a furnace
(Fig. 4) and exposed to fire. The furnace temperature was
measured by four thermocouples and the average furnace tem-
perature was regulated according to the ISO 834 ([6]) standard
temperature–time relationship. The ISO 834 standard
temperature–time curve, or a very similar one, is followed world-
wide in assessment of fire resistance of construction elements,
including intumescent coating protected steel structures.
Fig. 5 shows four specimens tested together in the furnace. The
steel temperature was measured by three thermocouples
embedded in the steel plate and recording was made every
minute continuously. Four observation holes were placed on the
furnace door to enable the fire tests to be observed and pictures of the surface of the specimens to be taken. Each test was continued
until the steel temperature reached 700 1C.
3. Test results
3.1. Experimental phenomena
When exposed to flame, intumescent coatings for all speci-
mens underwent the following main steps of chemical reaction:
(1) melting of the acid base;
(2) expansion due to release of gas by the blowing agent;
(3) char formation;
(4) char degradation due to oxidation.
Depending on the composition of the chemical components
and the fire exposure condition, these reactions may happen in
sequence or together. Type-A intumescent coatings began to
expand earlier than type-U intumescent coatings. In the intumes-
cence (expansion) stage, bubbles that appeared on the surface of
type-A intumescent coatings were much larger than that those on
the surface of type-U intumescent coatings.
Intumescent coatings for both types are highly ‘‘engineered’’ to
pass the standard fire resistance test when freshly applied.
Hydrothermal aging causes some chemical components in the
intumescent coatings to migrate to the surface, altering the
chemical reactions. In the intumescence stage, the blowing agent
in the intumescent coating decomposes to produce gas, a fraction
of which is trapped within the molten matrix to cause the coatingto expand. From the pictures taken of the specimens through the
observation holes on the furnace door, many bubbles appeared
Fig. 5. Specimens in furnace. (a) Specimens hung on steel beams and (b) Specimens laid flat on steel beams.
Fig. 6. Bubble appearance on the surface of type-A specimens. (a) AZ-1-00, (b) AZ-1-21 and (c) AZ-1-42.
L.L. Wang et al. / Fire Safety Journal 55 (2013) 168–181 171
8/18/2019 ~ Experimental study of hydrothermal aging effects on intumescent coating
http://slidepdf.com/reader/full/-experimental-study-of-hydrothermal-aging-effects-on-intumescent-coating 5/14
Author's personal copy
during the intumescence stage on both types of intumescent
coating after 0 or 4 cycles of hydrothermal aging tests, as shown
in Figs. 6 and 7(a). This means a large amount of gas was
produced due to decomposition of the blowing agent. After 11
and 21 cycles of hydrothermal aging tests, the number of bubbles
decreased drastically and the bubble distribution was much less
uniform, shown as Figs. 6 and 7(b). After 42 cycles, bubbles werealmost non-existent, see Figs. 6 and 7(c).
The observed phenomena for type-A specimens with both
1 mm and 2 mm DFTs were generally similar.
The most important parameters that directly reflect the fire
protection performance of intumescent coatings are the final
expanded thickness and internal structure of the char [11].
Fig. 8 shows the expanded heights of both types of coatings after
different cycles of aging test. These figures also give someindication of the consistence of the char.
Fig. 7. Bubble appearance on the surface of type-U specimens. (a) UI-1-00, (b) UI-1-21 and (c) UI-1-42.
Fig. 8. Cross sectional view of expanded intumescent char after different cycles of accelerated aging test. (a) AZ-2-00, (b) AZ-2-42, (c) AZ-1-04, (d) AZ-1-42, (e) UI-1-11
and (f) UI-1-42.
L.L. Wang et al. / Fire Safety Journal 55 (2013) 168–181172
8/18/2019 ~ Experimental study of hydrothermal aging effects on intumescent coating
http://slidepdf.com/reader/full/-experimental-study-of-hydrothermal-aging-effects-on-intumescent-coating 6/14
Author's personal copy
It can be seen from Fig. 9 that the expanded thickness
decreased greatly for both types of intumescent coating after 42
cycles of aging test. In addition, the integrity and consistency of
the char for the specimens after 42 cycles of aging test are poor
Fig. 8.
Table 2 lists the measured DFTs, the measured final thick-
nesses and the expansion ratios of the different specimens. Itmust be pointed out that due to unevenness of the final char, the
measured final thickness in some cases is around the average
value. Fig. 10 plots the expansion ratio as a function of the
number of cycles of accelerated testing. It can be seen that the
expansion ratio decreases considerably after only a few cycles of
aging test. After 21 cycles (simulating 10 years in service), the
expansion ratios of all three groups of intumescent coatings were
about 60% of those without aging. The expansion ratio may be
used to give a measure of the effective thermal conductivity of
intumescent coating. This means that after 21 cycles, the effective
thermal conductivity of the intumescent coating is about 1.7 times
(1/0.66) the effective thermal conductivity without aging. After 42
cycles, the expansion ratio was about 1/3rd of that without aging.
3.2. Temperature results
The average furnace temperature followed the ISO 834 stan-
dard temperature–time relationship.
Fig. 10 presents the measured steel substrate temperature–
time curves for the replicate tests of each specimen.
It can be seen from Fig. 10 that the replicate tests give
generally consistent results even though some discrepancies
exist. The average values of temperatures of the steel substrates
will be used. Figs. 11 and 12 compare the average steel
temperature–time relationships to show the effect of aging on
steel substrate temperature.
It can be seen from Figs. 11 and 12 that compared to speci-mens without aging, there is sharp increase in steel substrate
temperature after a certain number of cycles of aging test. For
type-A coating (Fig. 11), 11 cycles (representing 5 years in service)
appear to mark the beginning of sharp increase in the steel
substrate temperature. For type-U coating (Fig. 12), the increase
in steel temperature appears to be more even over the entire
range of aging cycles.
Figs. 13 and 14 present the steel substrate temperatures after
different cycles of aging at the same time when the specimens
without aging reached 400 1C, 500 1C and 600 1C. Furthermore,
Tables 3 and 4 present fire resistance times that may be achieved
by the different specimens if the steel limiting temperature is
400 1C, 500 1C, 600 1C and 700 1C.
The increase in steel temperature or the reduction in fireresistance time, due to degradation in intumescent coating
performance is very high. A question may arise on when intu-
mescent coating should be replaced. On the assumption of
replacing degraded intumescent coatings after suffering a loss of
20% in its fire resistance rating, then type-A coating (Table 3)
would need replacing after 11 cycles (representing 5 years in
service) and type-U coating (Table 4) would need replacing after
21 cycles (representing 10 yes in service). If no loss of fire
resistance rating is allowed, the only alternative solution would
be to specify intumescent coating DFT based on the degraded
intumescent coating performance. For example, look at the results
in Fig. 14 for type-U intumescent coating. Suppose the steel
limiting temperature is 604 1C. If it is desired to use the intumes-
cent coating after 42 cycles (corresponding to 20 years in service),then when specifying the fresh intumescent coating DFT, a steel
limiting temperature of 500 1C should be used.
3.3. Thermal conductivity
The simple quantity of effective thermal conductivity (refer-
ring to the initial, not expanded, intumescent coating thickness)
may be used to indicate the overall effects of hydrothermal
aging on fire performance of intumescent coatings. The effec-
tive thermal conductivity may be obtained from the following
0
12
24
36
48
0
number/cycles of hydrothermal aging
e x p a n s i o n r a t i o
U group
AZ-1 group
AZ-2 group
11 22 33 44
Fig. 9. Reduction of expansion ratio with number of cycles of hydrothermal aging.
Table 2
Expansion ratios for specimens.
Specimen Initial thickness(mm) Final thickness(mm) Expansion ratio
Type-U specimens (U-group)
UI-1–00 0.95 28.00 29.47
UI-1–11 1.01 22.00 21.78
UI-1–21 1.04 19.00 18.27
UI-1–42 0.94 10.00 10.64
Type-A specimens with 1 mm coating (AZ-1 group)
AZ-1–00 1.02 47.00 46.08
AZ-1–04 1.05 42.00 40.00
AZ-1–11 1.10 35.00 31.82
AZ-1–21 1.08 28.00 25.93
AZ-1–42 1.06 10.00 9.43
Type-A specimens with 2 mm coating (AZ-2 group)
AZ-2–00 2.20 85.00 38.64
AZ-2–04 2.16 76.00 35.19
AZ-2–11 2.09 69.00 33.01AZ-2–21 2.18 52.00 23.85
AZ-2–42 2.22 35.00 15.76
L.L. Wang et al. / Fire Safety Journal 55 (2013) 168–181 173
8/18/2019 ~ Experimental study of hydrothermal aging effects on intumescent coating
http://slidepdf.com/reader/full/-experimental-study-of-hydrothermal-aging-effects-on-intumescent-coating 7/14
Author's personal copy
equation ([4]):
l p,t t ð Þ ¼ d p V
A p c ara 1 þ f=3
1
yt ya,t
Dt
" #
½Dya,t þ e
f=10
1
Dyt ð1Þ
where
Dya,t is the increase in steel temperature during the time
interval Dt
l p,t is the effective thermal conductivity of intumescent
coating during the time interval Dt
d p is the initial DFT of intumescent coating
c a is the specific heat of steel
ra is the density of steel A p/V is the section factor of the protected steel section
yt is the furnace temperature at time t
0
200
400
600
800
0 20 40 60 80
time (minute)
AZ-1-00-1
AZ-1-00-2
AZ-1-00-3
0
200
400
600
800
0 12 24 36 48 60
time (minute)
AZ-1-42-1
AZ-1-42-2
AZ-1-42-3
0
200
400
600
800
0 14 28 42 56 70
time (minute)
AZ-2-00-1
AZ-2-00-2
AZ-2-00-30
200
400
600
800
0 12 24 36 48 60
time (minute)
AZ-2-42-1
AZ-2-42-2
AZ-2-42-3
0
200
400
600
800
0 20 40 60 80
time (minute)
t e m p e r a t u r e o f s t e e l
s u b s t r a t e ( ° C )
t e m p e r a t u r e o f s t e e l
s u b s t r a t e ( ° C )
t e m p e r a t u r e o f s t e e l
s u b s t r a t e ( ° C )
t e m p e r a t u r e o f s t e e l
s u b s t r a t e ( ° C )
t e m p e r a t u r e o f s t e e l
s u b s t r a t e ( ° C )
t e m p e r a t u r e o f s t e e l
s u b s t r a t e ( ° C )
UI-1-00-1
UI-1-00-2
UI-1-00-30
200
400
600
800
0 12 24 36 48 60
time (minute)
UI-1-42-1
UI-1-42-2
UI-1-42-3
Fig. 10. Replicate temperature–time relationships of steel substrate.
0
200
400
600
800
0
time (minute)
AZ-1-00
AZ-1-04
AZ-1-11
AZ-1-21
AZ-1-42 0
150
300
450
600
750
0
time (minute)
t e m p e r a t u r e o f s t e e l
s u b s t r a t e s ( ° C )
t e m p e r a t u r e o f s t e e l
s u b s t r a t e s ( ° C )
AZ-2-00
AZ-2-04
AZ-2-11
AZ-2-21
AZ-2-42
14 28 42 56 7020 40 60 80
Fig. 11. Effect of aging on steel substrate temperature–time relationship for type-A specimens.
L.L. Wang et al. / Fire Safety Journal 55 (2013) 168–181174
8/18/2019 ~ Experimental study of hydrothermal aging effects on intumescent coating
http://slidepdf.com/reader/full/-experimental-study-of-hydrothermal-aging-effects-on-intumescent-coating 8/14
Author's personal copy
ya,t is the steel temperature at time t
Dya,t is the increase of furnace temperature during the time
interval Dt
f ¼ c pr pd p A p
c araV
Dt r30 s
Since the specimen was exposed to fire on all sides, the section
factor was calculated as A p/V ¼142 m1.
In theory, the temperature dependent specific heat and density
of intumescent coating should be used when calculating the
effective thermal conductivity of intumescent coating using
Eq. (1). However, since the amount of heat stored inside the
intumescent coating is very small and may be considered to be
negligible compared to that in the steel substrate, FE0 so Eq. (1)
may be simplified to Eq. (2) below:
l p,t t ð Þ ¼ d p V A p c ara 1yt ya,t
Dt
" # Dya,t ð2Þ
For each specimen and for each time interval, the intumescent
coating temperature y p may be taken as the mean of the steel and
fire temperature so that
y p ¼ yt þya,t
2 ð3Þ
Figs. 15 and 16 present some of the results of coating thermal
conductivity–temperature curve.
It can be seen from Figs. 15 and 16 that the effective thermal
conductivity of both types intumescent coating starts to fall
sharply after the temperature of intumescent coating reached
about 100 1C, indicating chemical reactions starting at about
100 1C. The effective thermal conductivity of the coating becamestable after reaching temperatures over 400 1C, a clear indication
that the coating had reached full expansion. Afterwards, the
effective thermal conductivities increase with temperature, which
may be explained by increased radiation inside the bubbles at
increasing temperatures [11]. It is the stable, fully expanded stage
of intumescent coating that is providing the fire protection
function, and the discussions below will focus on this stage.
It can be seen from Figs. 15 and 16 that there are some small
discrepancies between the results for the same three nominally
identical specimens. Nevertheless, the three replicate tests gave
generally consistent results. In the discussions to follow, the
0
200
400
600
800
0
time (minute)
t e m p e r a t u r e o f s t e e l
s u b s t r a t e s ( ° C )
UI-1-00
UI-1-11
UI-1-21
UI-1-42
14 28 42 56 70
Fig. 12. Effect of aging on steel substrate temperature–time relationship for type-
U specimens.
350
400
450
500
550
600
650
700
750
10 20 30 40 50 60 70 80
time (minute)
t e m p e r a t u r e o f s t e e l
s u b s t r a t e s ( ° C )
AZ-1-00
AZ-1-04
AZ-1-11
AZ-1-21
AZ-1-42
350
400
450
500
550
600
650
700
750
10 20 30 40 50 60 70 80
time (minute)
t e m p e r a t u r e o f s t e e l
s u b s t r a t e s ( ° C )
AZ-2-00
AZ-2-04
AZ-2-11
AZ-2-21
AZ-2-42
530°C
592°C
506°C
400°C
567°C
643°C
500°C
711°C
600°C
675°C
443°C
621°C
530°C
400°C
556°C
641°C
500°C
600°C
658°C
450°C
420°C
630°C
424°C
532°C
630°C
Fig. 13. Temperatures reached when the specimens without aging reached 400 1C/500 1C/600 1C. (a) Type AZ-1 specimens and (b) Type AZ-2 specimens.
L.L. Wang et al. / Fire Safety Journal 55 (2013) 168–181 175
8/18/2019 ~ Experimental study of hydrothermal aging effects on intumescent coating
http://slidepdf.com/reader/full/-experimental-study-of-hydrothermal-aging-effects-on-intumescent-coating 9/14
Author's personal copy
average effective thermal conductivities for intumescent coating
temperatures between 250 1C and 750 1C at 50 1C interval will be
used. In order to help clarify discussions, the raw results in
Figs. 15 and 16 are smoothed and the average results are shown
in Fig. 17.
The changes in effective thermal conductivity of all coatings
follow a consistent pattern as a function of the number of cycles
of hydrothermal aging when the temperature increases. There-
fore, the change in average thermal conductivity at the intumes-
cent coating temperature range of 650–750 1C, which correspond
to the practically interesting steel temperature range of 500–
600 1C, may be used. The results are presented in Fig. 18.
The effect of hydrothermal aging is to cause the hydrophilic
components of intumescent coating to move to the surface of the
coating and then be dissolved by moisture. Hence, although type-
A coating of the same DFT (1 mm) performed better than type-U
coating if neither suffers any aging effect (as can be seen in Fig. 18
by the lower thermal conductivity of type-A coating at 0 cycle),
type-A coating suffers from hydrothermal aging much more
quickly than type-U coating due to it having a lower water
resistance as explained in Section 2.3. Whilst the thermal con-
ductivity of type-U coating increases very gradually as the
number of cycles increases, type-A coating suffers significantly
more loss in performance after only 4 cycles of hydrothermal
aging.
Comparing the performance of 1 mm DFT type-A coating with
that of 2 mm DFT type-A coating, the same degradation processoccurs, but the thicker DFT 2 mm coating delays the process
slightly so that the large change occurs between 11 and 21 cycles
of hydrothermal aging for the thicker DFT instead of after 4–11
cycles of hydrothermal aging for the thinner DFT.
4. Chemical analysis tests
A number of chemical analysis tests were carried out to further
examine the degradation process in more detail. The chemical
analysis tests included TGA test, XPS test, FTIR test and SEM test.
4.1. TGA test results
TGA test gives mass loss as a function of temperature. The TGA
tests were conducted using a Pyris diamond TG/DTA instrument
under nitrogen at a heating rate of 20 1C mini within a tempera-
ture range of 25–800 1C. Fig. 19 presents the measured mass loss
results.
It can be seen from Fig. 19 that there is little difference in the
TGA test results after different cycles of hydrothermal aging. This
indicates that the hydrothermal aging tests did not cause the
chemical components to be any different. However, the optimum
matching of chemical components in the intumescent coating
changed due to migration of the hydrophilic components to the
surface. Hence, the expansion ratios of the fire test specimens
were different. This also suggests that when detecting changes in
intumescent coating performance over time, the TGA test would
not be suitable.
4.2. FTIR test results
FTIR (Fourier transform infrared spectroscopy) test was used
to investigate the migration of chemical components for both
types of specimens with different cycles of hydrothermal aging.
The FTIR test was conducted on samples extracted from the
surface layer of intumescent coating, using the EQUINOXSS/HYPERION2000 device. The experimental results are presented
in Fig. 20.
350
400
450
500
550
600
650
700
750
10 20 30 40 50 60
time (minute)
t e m p e r a t u
r e o f s t e e l
s u b s t r a t e s ( ° C )
UI-1-00
UI-1-11
UI-1-21
UI-1-42
483°C438°C
400°C
525°C
547°C
500°C
600°C
703°C
423°C
604°C 624°C
649°C
Fig. 14. Temperatures reached when the specimens without aging reached 400 1C/500 1C/600 1C.
Table 3
Specimen Limiting steel temperature(1C)
400 500 600 700
(a) Fire resistance times for type AZ-1 specimens (in minutes)
AZ-1–00 30 41 51 63
AZ-1–04 27 38 48 59
AZ-1–11 25 35 44 53
AZ-1–21 22 30 37 47
AZ-1–42 17 23 30 40
(b) Fire resistance times for type AZ-2 specimens (in minutes)
AZ-2–00 33 43 54 66
AZ-2–04 30 40 50 62
AZ-2–11 28 37 48 59
AZ-2–21 23 31 39 50
AZ-2–42 19 25 31 40
Table 4
Fire resistance times for type U specimens (in minutes).
Specimen Limiting steel temperature(1C)
400 500 600 700
UI-1–00 23 32 41 52
UI-1–11 21 30 39 49
UI-1–21 20 28 37 47
UI-1–42 17 24 32 41
L.L. Wang et al. / Fire Safety Journal 55 (2013) 168–181176
8/18/2019 ~ Experimental study of hydrothermal aging effects on intumescent coating
http://slidepdf.com/reader/full/-experimental-study-of-hydrothermal-aging-effects-on-intumescent-coating 10/14
Author's personal copy
Although the FTIR test results can only be used to gain a
qualitative understanding of the effects of aging, they can still
give some indication of the extent of aging in service.
For type-A coatings (Fig. 20a), the troughs at 3321 cm1,
3176 cm1, 2957 cm1 and 1668 cm1 indicate N–H bonds
contained in MEL and APP, O–H bonds in PER, C–H bond in acrylic
acid resin and PER, and C¼O bonds in acrylic acid resin,
respectively. The trough at 1428 cm1 is the overlapping peak
of the absorption peak of CH2 group contained in acrylic acid resin
and PER and the absorption peak of triazing rings which are the
main structure of MEL. The troughs at 1253 cm1, 1078 cm1,
1013 cm1 and 889 cm1 indicate P¼O bonds in APP, C–O–H
bonds in PER, C–O–C bonds in acrylic acid resin and triazing rings
in MEL, respectively. Similar wave numbers of these bonds can be
observed in Fig. 20(b) for type-U coating.It is clear from Fig. 20 that the absorption peak of the above
mentioned different chemical bonds contained in PER and APP are
enhanced with increasing number of cycles of aging. This indi-
cates that PER and APP migrated from within the coating to the
surface of the coating after different cycles of aging.
Compared to AZ-1–00, the absorption peak at 1428 cm1 of
AZ-1–11 was weakened whereas the width of the peak increased.
This indicates that the polymer binder degraded under the effect
of water and oxygen and some of the CH2 groups were oxidated
into C ¼O groups. This enhanced the absorption peak of C¼O
bonds at 1668 cm1 with increasing number of cycles of aging.
Degradation of the polymer binder (acrylic acid resin) was
present during the whole process of aging. The absorption peaks
at 1428 cm1and 889 cm1 were also enhanced, indicating
migration of MEL from within the coating to the surface of the
coating after different cycles of aging.
It is observed from the above analysis that the degradation of polymer binder (acrylic acid resin) and migration of flame
retardant system (APP-MEL-DPER) happened at the same time
0
0.05
0.1
0.15
0.2
0 200 400 600 800
temperatrure of coating (°C)
e f f e c t i v e t h e
r m a l c o n d u c t i v i t y
W / ( m
° C )
UI-1-00-1
UI-1-00-2
UI-1-00-3
0
0.05
0.1
0.15
0.2
0.25
0 200 400 600 800
temperatrure of coating (°C)
e f f e c t i v e t h e
r m a l c o n d u c t i v i t y
W / ( m •
° C )
UI-1-42-1
UI-1-42-2
UI-1-42-3
Fig. 15. Effective thermal conductivity (l p)–coating temperature (y p) relationships for type-U specimens. (a) Type U specimens with 0 cycles of aging and (b) Type U
specimens with 42 cycles of aging.
0
0.05
0.1
0.15
0.2
0 200 400 600 800
temperatrure of coating (°C)
e f f e c t i v e t h e r m a l c o n d u c t i v
i t y
W / ( m • ° C )
AZ-1-00-1AZ-1-00-2
AZ-1-00-3
0
0.03
0.06
0.09
0.12
0.15
0 200 400 600 800
temperatrure of coating (°C)
e f f e c t i v e t h e r m a l c o n d u c t i v
i t y
W / ( m • ° C )
AZ-1-42-1
AZ-1-42-2
AZ-1-42-3
0
0.06
0.12
0.18
0.24
0.3
0 200 400 600 800
temperatrure of coating (°C)
e f f e c t i v e t h e r m a l c o n d u c t i v i t y
W / ( m • ° C )
AZ-2-00-1
AZ-2-00-2
AZ-2-00-3
0
0.05
0.1
0.15
0.2
0.25
0 200 400 600 800
temperatrure of coating (°C)
e f f e c t i v e t h e r m a l c o n d u c t i v i t y
W / ( m • ° C )
AZ-2-42-1
AZ-2-42-2
AZ-2-42-3
Fig. 16. Effective thermal conductivity (l p)–coating temperature (y p) relationships for type-A specimens. (a) Type AZ-1 specimens with 0 cycles of aging, (b) Type AZ-1
specimens with 42 cycles of aging, (c) Type AZ-2 specimens with 0 cycles of aging and (d) Type AZ-2 specimens with 42 cycles of aging.
L.L. Wang et al. / Fire Safety Journal 55 (2013) 168–181 177
8/18/2019 ~ Experimental study of hydrothermal aging effects on intumescent coating
http://slidepdf.com/reader/full/-experimental-study-of-hydrothermal-aging-effects-on-intumescent-coating 11/14
Author's personal copy
0
0.03
0.06
0.09
0.12
0.15
200
temperature of coating θp(°C) e f f e c t i v e t h e r m
a l c o n d u c t i v i t y λ
p W
/ ( m • ° C ) AZ-1-00 AZ-1-04
AZ-1-11 AZ-1-21
AZ-1-42
0
0.03
0.06
0.09
0.12
0.15
e f f e c
t i v e t h e r m a l c o n d u c t i v i t y
λ p W / ( m • ° C )
AZ-2-00 AZ-2-04
AZ-2-11 AZ-2-21
AZ-2-42
0
0.02
0.04
0.06
0.08
0.1
e f f e c t i v e t h e r m a l c o n d u c t i v i t y
λ p W / ( m • ° C )
UI-1-00
UI-1-11
UI-1-21
UI-1-42
300 400 500 600 700 800 900
200
temperature of coating θp(°C)
300 400 500 600 700 800 900
200
temperature of coating θp(°C)
300 400 500 600 700 800 900
Fig. 17. Effects of aging on effective thermal conductivity of intumescent coatings. (a) Effect of aging on effective thermal conductivity of Type AZ-1 coating, (b) Effect
of aging on effective thermal conductivity of Type AZ-2 coating and (c) Effect of aging on effective thermal conductivity of Type U coating.
0
0.017
0.034
0.051
0.068
443322110
temperature of coating θp(°C)
e f f e c t i v e t h e r m a l c o n d u c t i v i t y
λ p W / ( m • ° C )
6 5 0 ° C ~ 7 5 0 ° C
Type-U Type-A (1mm)
Type-A (2mm)
Fig. 18. Effect of aging on effective thermal conductivity of intumescent coating.
0 100 200 300 400 500 600 700 800
40
50
60
70
80
90
100
w e i g h t l o s s ( % )
temperature of coating (°C) temperature of coating (°C)
AZ-1-00
AZ-1-11
AZ-1-21
AZ-1-42
0 100 200 300 400 500 600 700 800
40
50
60
70
80
90
100
w e i g h t l o s s ( % )
UI-1-00
UI-1-11
UI-1-21
UI-1-42
Fig. 19. TGA curves of intumescent coating after different cycles of aging.
L.L. Wang et al. / Fire Safety Journal 55 (2013) 168–181178
8/18/2019 ~ Experimental study of hydrothermal aging effects on intumescent coating
http://slidepdf.com/reader/full/-experimental-study-of-hydrothermal-aging-effects-on-intumescent-coating 12/14
Author's personal copy
during the process of aging, which resulted in the reduced fire
protective properties of intumescent coatings.
In practical application, when examining the effects of aging
on intumescent coatings, if on-site FTIR test shows little change in
the absorption peaks, then there is high confidence that the
effects of aging are minimal.
4.3. XPS test results
XPS (X-ray photoelectron spectroscopy) test gives information
on the amount of chemical elements being examined. For example,
Fig. 21 presents the amounts of Carbon and Nitrogen existent on
the surface layer of both types of intumescent coatings after
different cycles of aging. The XPS test was conducted using an
elemental analyzer VARIOEL 3.
C element is contained in MEL(C3H6N6) which acts as the
blowing agent and in DPER/PER and (C(CH2OH)4) acting as the
charring agent; N element is contained in MEL and
APP(NH4)nþ2PnO3nþ1) which act as the catalytic agent. Table 5
lists the percentage of Carbon and Nitrogen elements obtainedfrom the XPS tests for representative samples of both types of
intumescent coatings.
It can be seen from Table 5 that compared to specimens
without aging, the contents of C and N elements on the surface
layer of type-A and type-U specimens increased with increasing
number of cycles of aging. The change in N element is much more
3500 3000 2500 2000 1500 1000 500
1428
88910781253
166829573321 3176
AZ-1-42
AZ-1-21
AZ-1-11
AZ-1-00
Wavenumber (cm-1)
3500 3000 2500 2000 1500 1000 500
Wavenumber (cm-1)
1435
1078 8901251
1661
294031153325
UI-1-42
UI-1-21
UI-1-11
UI-1-00
Fig. 20. FTIR test results. (a) FTIR test results for Type-A coating and (b) FTIR test results for Type-U coating.
10000
1000
2000
3000
4000
5000
6000
7000
8000
O1s
N1s
C1s
N ( E )
Binding Energy (eV)
AZ-1-00
0
1000
2000
3000
4000
5000
6000
7000
8000
O1s
N1s
C1s
N ( E )
AZ-1-21 AZ-1-42
0
1000
2000
3000
4000
5000
6000
7000
8000
O1s
N1s
C1s
N ( E )
0
1000
2000
3000
4000
5000
6000
7000
8000
O1s
N1s
C1s
N ( E )
UI-1-00
0
1000
2000
3000
4000
5000
6000
7000
O1s
N1s
C1s
N ( E )
UI-1-21
0
1000
2000
3000
4000
5000
6000
7000
8000
O1s
N1s
C1s
N ( E )
UI-1-42
800 600 400 200 01000
Binding Energy (eV)
800 600 400 200 0 1000
Binding Energy (eV)
800 600 400 200 0
1000
Binding Energy (eV)
800 600 400 200 0 1000
Binding Energy (eV)
800 600 400 200 0 1000
Binding Energy (eV)
800 600 400 200 0
Fig. 21. XPS test results.
L.L. Wang et al. / Fire Safety Journal 55 (2013) 168–181 179
8/18/2019 ~ Experimental study of hydrothermal aging effects on intumescent coating
http://slidepdf.com/reader/full/-experimental-study-of-hydrothermal-aging-effects-on-intumescent-coating 13/14
Author's personal copy
sensitive to the change in C element. The current research is not
sufficiently comprehensive, but further extensive testing should
be done to ascertain whether it would be possible to link the
changes in C and N elements to the changes in fire protection
performance of different intumescent coatings.
4.4. SEM test results
SEM test gives some information on the change in internal
structure of chars after different cycles of aging. The SEM micro-
graphs of chars obtained from type-A coating after 11, 21 and 42
cycles of aging are presented in Fig. 22.
Both Fig. 22(a) and (b) show the expected honeycomb struc-
ture, but the pore size in Fig. 22(b) is much larger than that in
Fig. 22(a) and the number of pores decreases. For Fig. 22(c),
although the coating can still expand to form a char structure
after 42 cycles of aging, the aging process has damaged its
expanding effect and the ‘‘honeycomb’’ structure of intumescent
char does not exist.
5. Conclusions
This paper has presented the results of a series of fire tests on
intumescent coating protected steel plates after the intumescent
coatings have been exposed to different cycles of hydrothermal
aging according to exposure condition Z1 in European guide ETAG
018 Part 2. The numbers of cycles were 0, 4, 11, 21 and 42,
corresponding to 0, 2, 5, 10 and 20 years of nominal service. The
results have been presented in terms of the expansion ratio, the
steel temperature and effective thermal conductivity. Surface
observations were made and additional chemical analysis (TGA,
FTIR, XPS and SEM) tests were also carried out. The followingconclusions may be drawn:
1. Bumps with different degrees of unevenness appeared on
the surfaces of specimens applied with type-A intumescent
coatings after different cycles of hydrothermal aging tests.
But no obvious change was observed for specimens applied
with type-U intumescent coatings after 11 and 21 cycles of
aging tests. Slight wrinkles appeared on the surfaces of
specimens applied with type-U intumescent coatings after
42 cycles of aging. The surface appearance can be used to
give a visual guide to the effectiveness of intumescent
coating performance in service.
2. Both types of intumescent coating suffered considerable
reduction in performance after 42 cycles of accelerated agingtest (corresponding to 20 years in service under the assumed
exposure condition). For example, the expansion ratio
reduced by over 70% and the steel plate temperature was
increased by about 200 1C compared to the steel tempera-
ture of 500 1C with fresh intumescent coating.
3. The results from TGA test, FTIR test and XPS test show that
the aging process did not cause the chemical components to
be any different, but the optimum matching of these
components in the examined intumescent coatings changed
due to migration of the hydrophilic components to the
surface of the coating when exposed to the hydrothermal
aging environment. This damaged the expanding ability of
the intumescent coatings. From the chemical analysis test
results, the TGA test is not suitable for detecting changes foraging effect. The FTIR test can detect the qualitative changes
of aging. The XPS test may be used to quantify the aging
Table 5
Contents of C/N.
Element
contents(%)
Specimen
Element AZ-1–00 AZ-1–21 AZ-1–42 UI-1-00 UI-1–21 UI-1–42
C 61.6 64.1 65.5 63.7 63.8 64.9N 7.9 8.7 10.5 9.9 10.1 13.4
Fig. 22. SEM micrographs of type-A intumescent chars after different cycles of aging (a) 11 cycles; (b) 21 cycles; (c) 42 cycles.
L.L. Wang et al. / Fire Safety Journal 55 (2013) 168–181180
8/18/2019 ~ Experimental study of hydrothermal aging effects on intumescent coating
http://slidepdf.com/reader/full/-experimental-study-of-hydrothermal-aging-effects-on-intumescent-coating 14/14
Author's personal copy
effects, but much more extensive testing is required before a
quantitative relationship between XPS test results and fire
protection performance results (e.g., changes in expansion
ratio/effective thermal conductivity) can be established.
4. The SEM test is destructive but the results can be used to
indicate that the effects of aging.
It should be pointed out that intumescent coatings aretop-coated in practice to protect them from environmen-
tal damage. Their durability will be much better.
References
[1] American Society for Testing and Materials (ASTM) (2005) ASTM D 4587-91,Practice For Conducting Tests on Paint and Related Coatings and MaterialsUsing a Fluorescent UV-Condensation Light and Water-Exposure Apparatus,ASTM, Philadelphia.
[2] British Standards Institution (BSI) (1992) British Standard BS8202, Coatingfor Fire Protection of Building Elements, Part 2: Code of Practice For theAssessment and Use of Intumescent Coating System for Providing FireResistance, British Standards Institution, London.
[3] J. Dowling, Fire protection costs for structural steelwork, New steel Constr.
(UK). Sep/Oct (2003) 8–9.[4] European Committee for Standardization (CEN) (2002) ENV13381, Test
Methods for Determining the Contribution to the Fire Resistance of StructuralMembers, Part 4: Applied Protection to Steel Members, European Committeefor Standardization, Brussels.
[5] European Organization for Technical Approvals (EOTA) (2006) ETAG018,Guideline for European Technical Approval of Fire Protective Products, Part2: Reactive Coatings for Fire Protection of Steel Elements, European Organi-zation for Technical Approvals, Brussels.
[6] International Standards Organization (ISO) (1975) ISO 834: Fire ResistanceTests-Elements of Building Construction, Geneva.
[7] Institute Fur Bautechnik(1977). Procedures for Testing and Approval of Intumescent Coating on Steel Members to Comply with F30 Fire ResistanceCategory of DIN 4102, Berlin.
[8] Y. Sakumoto, J. Nagata, A. Kodaira, Y. Saito, Durability evaluation of intumes-cent coating for steel frames, J. Mater. Civ. Eng. 13 (1) (2001) 274–281.
[9] Z.Y. Wang, E.H. Han, W. Ke, Effect of nanoparticles on the improvement infire-resistant and anti-ageing properties of flame-retardant coating, Surf.Coat. Technol. 200 (2006) 5706–5716.
[10] Z.Y. Wang, E.H. Han, W. Ke, Fire-resistant effect of nanoclay on intumescentnanocomposite coatings, J. Appl. Polym. Sci. 103 (2006) 1681–1689.
[11] J.F.,Yuan (2009). Intumescent Coating Performance on Steel Structures underRealistic Fire Conditions, Ph.D. Thesis, University of Manchester.
L.L. Wang et al. / Fire Safety Journal 55 (2013) 168–181 181