The Effects of Moisture on the Flammability Characteristics

of Textile Materials

BERNARD MILLER, J. RONALD MARTIN, BHUVENESH C. GOSWAMI,

AND CHARLES H. MEISER, JR.

Textile Research Institute, P. O. Box 625, Princeton, New Jersey 08540, U. S. A.

ABSTRACT
The TRI Flammability Analyzer has been used to determine the effects of moisture on the flammability characteristics
of textile materials. Both moisture in the fabric and in the test environment have been considered for a number of un-
treated and flame-retardant-treated fabrics. The results are reported and interpreted in terms of several measurable
quantities: mass transformation rate (MTR), oxygen sensitivity (ΔMTR/Δ% O
), and extrapolated oxygen index
2
(O. I.)
. Variations in fabric moisture content appear to have a greater effect on the behavior of relatively hydrophobic
0
fibers (such as polyester), as compared to cotton or nylon.

,
Introduction ,
[1, 16].’ It is obvious that there is a great need for
It is generally expected that the presence of moisture quantitative data on the role of moisture content if
in a textile fabric will have some in8uence on its flam- progress is to be made in standardizing the techniques
of measurement and in understanding the factors con-
mability behavior. If nothing else, one would antici-
trolling the relative hazard potentials of various textile
pate a loss of thermal energy to the water as it undergoes materials.
heating and vaporization at temperatures well below Workers at the National Bureau of Standards have
the decomposition threshold of fibrous polymers. This
lost energy would otherwise go toward enhancing igni- studied the effect of conditioning humidity on char
tion or 8ame propagation processes. There has also length for polyester (PET), nylon, nylon/PET blends,
been conjecture that the relative humidity of the and polyvinyl chloride, as well as for a number of
immediate efivironment of a burning fabric should have flame-retardant cottons and rayons [14]. For most of
some modifying effect.
these materials, decreased conditioning humidity (i.e.,
Consistent with the often ex- ’

decreased fabric moisture content at the start of the

pressed viewpoint that flammability must be considered increased
as a systems property, it seems imperative that test) the char length only slightly; however,
specific several of the 9ame-retardant
information be obtained as to the consequences of cellulosics displayed
moisture being an inherent part of a material. and its abrupt and drastic increases in char length when the
samples were conditioned at relative humidities less
’

surroundings. ,

than 60%.
Flammability evaluation techniques that do not in- The Southern Regional Research Laboratories of the
clude the monitoring and controlling of moisture con-
USDA have produced several publications [4, 7, 8]
tent, or that impose unrealistic environmental condi--
tions during testing, may produce- results which can dealing with the effects of fabric moisture content on
lead to inaccurate conclusions regarding the relative flame retardancy.’ Char length (from the verticals
hazard potential of certain materials. Only limited flame test) and oxygen index were determined as a
attention seems to have been paid to this problem when function of initial fabric moisture content for a number
most of the general test methods for fabric flammability
of untreated and flame-retardant fabrics. For all the
were being developed.
materials investigated, oxygen index increased with
Consequently, some studies
have relied on evaluation techniques that prescribe increasing fabric moisture content in a complex manner.
preconditioning at some general median environmental The oxygen index of flame-retardant cottons displayed
a greater dependence on moisture than untreated cot-
condition [2, 15] (e.g., 70°F and 65% RH), while other
investigations. have been based on testing done under 1 It is at this point to comment on the unrealistic
whatever conditions happen to exist at the time (which nature ofappropriatethe "bone dry" sampling requirement. Experience
will vary widely with location and time of year) [5, 6]. in these laboratories has shown that cotton fabric, vacuum dried
at 105°C for 2 h and cooled over P , picks up close to 1%
5
O
2
More recently, methods have been introduced that moisture within few seconds upon exposure to typical
a laboratory
specify &dquo;bone dry&dquo; preconditioning before testing ambient conditions.
328

Downloaded from trj.sagepub.com at The University of Iowa Libraries on June 1, 2015

 329

tons. For the flame-retardant cellulosics, char length directly or through an air-tight humidification chamber
was found to increase with decreasing moisture. For where they can come into contact with water atomized
inherently flame-retardant fibers, moisture had little by jet sprays. The 8ow of gases and water is controlled
effect on char length but did raise the oxygen index by means of two solenoid valves which are in turn
value. regulated by the setting of a feedback humidistat con-
These studies were all related to the moisture initially trol in the main chamber of the Analyzer. For testing
&dquo;&dquo;

present in the fabric before actual testing was begun. ~ in a dry gas environment, the gases are passed directly
However, the test procedures used made it highly likely into the chamber through a bypass line. The arrange-
that fabric moisture contents would be continually ment, shown schematically in Figure 1, provides hu-
changing during the pre-ignition phase of the test, a midity control within f 2% RH over the entire range
phenomenon which has not been accounted for in the from 0 to 100% RH.
studies reported to date. The rate of change of fabric
moisture content could be strongly affected by the
humidity of the gases in the test environment (i.e., the
sample would be gaining or losing moisture depending
on its equilibrium moisture content at the test environ-
ment humidity); no systematic study of this effect has
*

been reported.’ In addition, the studies reported in the

literature deal only with the extinguishment facet of
the flammability problem, which is not necessarily
related to the other aspects of flammability (i.e., igni-
tion, heat emission, flame propagation, etc.).
The investigation reported here deals with the effects
of fabric moisture and test environment humidity on
the burning rates of both untreated and flame-retardant
textile materials. In addition, extinguishment prop-
erties of these materials have been ascertained by H ic. 1. Apparatus for humidity control during flamma-
determining burning rate as a function of oxygen con- bility testing: (A) gas humidification chambei, (B) Flamma-
centration ; the slope and/or intercept of the resulting bility Analyzer, (C) satellite sample for moisture-content
linear relationship can then be used as a measure of the
monitonng.
ease of extinguishment EIOI- Materials studied were For control of. fabric moisture content, samples
cotton, nylon 66, polyester, acrylic, wool, and rayon. mounted on test-wheel frames were first dried in a
With the exception of wool and acrylic, flame-retardant vacuum oven for approx 2 h and subsequently cooled
analogs of all the above have also been investigated. over PtÛ6 for i h in a desiccator. Each sample was
The effect of equilibration to a given fabric moisture then placed in an air-tight conditioning chamber con-
level from both wet and dry sides is described for
’ taining an appropriate salt solution to produce the
several representative fabrics. desired humidity level. Samples were allowed to equili-
brate in the conditioning chambers for 24 h. Each
Experimental Procedure sample was then removed and quickly positioned in the
burning chamber, where the environment had been
Burning rates were determined using the TRI Flam- brought to the desired oxygen concentration and hu-
mability Analyzer [12], a device which measures flame midity level (either nominally dry or at the same hu-
propagation rates by feeding fabric into its own flame midity level at which the sample had been conditioned).
at a rate which keeps the flame at a fixed position in After a short conditioning period (1-2 min), steady-state
space. This is accomplished by mounting the fabric on burning was achieved by igniting one end of the fabric
a circular frame and adjusting the speed of rotation of and adjusting the speed of the wheel so that the flame
this wheel to match the rate of flame propagation. In remained in the desired stationary position. The flame
this manner propagation rates can be determined for
propagation rate (in./min) can be determined from the
any desired burning direction (i.e., horizontal, 45° rotational velocity of the wheel, and the mass trans-
upward, vertically upward, vertically downward, etc.). formation rate’ (oz/min-in.) can be calculated by
The wheel of the Analyzer is mounted in a glass-front
multiplying the flame propagation rate by the areal
cabinet, and gas of any desired composition is passed density of the sample. Previous studies have shown
.

up through this cabinet at a rate of approx 25 I/min. that flame propagation rates are inversely proportional
Facilities are included for controlling the composition
as well as the humidity of the gaseous mixture (usually 2 The term "mass transformation rate"
(MTR) is now used for
the derived value of burning rate (actually the rate at which
nitrogen-oxygen) flowing through the chamber. The material is "transformed" by the advancing flame) to distinguish
nominally dry gases are mixed to the desired proportion this quantity from a directly measured mass burning rate [11]
in a manifold and passed into the burning chamber either which may or may not be the same.

Downloaded from trj.sagepub.com at The University of Iowa Libraries on June 1, 2015

 330

to fabric weights (i.e. areal densities) but are indepen-

’

dent of fabric thickness [12]. Therefore, fabrics of

.
different weights can be compared by normalizing to
masts transformation rates.
’

In addition to the sample which was to be burned,

two small satellite samples were placed in the condi-
tioning chamber. One of these was used to determine
the moisture level of the fabric at equilibrium with the
humidity of the conditioning chamber; at the instant
the main sample was removed from the chamber, this
sample was placed in a sealed bottle for subsequent
determination of moisture content. The second small
sample was used to detect any change in moisture
content occurring in the sample during testing. This
second sample was hung in the chamber of the Flam-
mability Analyzer at the same time that the wheel was
placed into position, at a location where it would not
come in contact with the flame or combustion products

(see Fig. 1). After a steady-state burning condition

was established, this sample was removed and placed
in a sealed bottle for subsequent determination of
moisture content. The moisture contents of the fabrics
are reported either as a range or as the average of the
values. obtained f rom the two small samples.
Moisture content was determined as follows: The FIG. 2. Effect of fabric moisture content the 45°
on upward
sample was quickly hung on a hang-down wire of an burning behavior of cotton sheeting in a dry atmosphere of 18.1,
electronic continuous-readout balance, and its wet 19.8, and 21.8% O2 (arrows indicate the direction of change in
moisture content). ,

weight was determined. A tubular furnace at 125°C

was then raised around the sample, and weight loss
followed as the sample dried. A constant dry weight
was usually obtained after only a few minutes of
,’
It can be seen that there is a significant decrease in
heating the transformation rate of this fabric with increasing
’
moisture content at all three oxygen levels. The rela-
Results and Discussion tionships between mass transformation rate and fabric
moisture content appear to be curvilinear, with the
-

EFFECT OF FABRIC MOISTURE CONTENT ON THE MASS rate changing more drastically at lower moisture con-
TRANSFORMATION RATES OF UNTREATED FABRICS.
tents (2 to 3.5%) than at higher values (3.5 to 5.0%).
The effect of fabric moisture content on mass transfor-
mation rate will be discussed for two representative
’

natural and synthetic textile materials. Similar studies Acrylic. Study of the influence of moisture on the
have been carried out for several o,ther fabric types; burning behavior of an acrylic fabric was carried out
these are described in the subsequent sections dealing with a 6.4 oz/yd’ staple fiber, 2 X 1 twill fabric. The
with the effects of flame retardants and relative hu- results for 45° upward burning in a nominally dry
midity as well as fabric moisture content. In all cases atmosphere at two oxygen concentrations are shown in
the data points plotted represent the average result of Figure 3. Over the moisture range investigated, the
at least two and usually three replicate experiments. mass transformation rate of the acrylic fabric decreased
linearly with increasing moisture content at both the
oxygen levels studied. However, the rate of decrease
Cotton. The mass transformation rates for 3.2 oz/yd2 is faster at 19.8% as compared to that at 21.8% Ot.
cotton sheeting in a dry environment were obtained for
45° upward burning at three oxygen levels; these are EFFECT OF FABRIC MOISTURE ON THE BURNING OF /
plotted as a function of fabric moisture content in FLAME-RETARDANT FABRICS. Considering the poss
Figure 2. The curves are drawn through the mean bility of interactions between fabric moisture and
,
values of the moisture contents obtained before and flame-retardant additives, the effect of moisture on
,
during the burning test as explained above. The the mass transformation rate of flame-retardant fabrics
arrows indicate the direction of moisture content change might be expected to differ from that for untreated
when there was any appreciable difference between the fabrics. This possibility has been explored for flame-
two values. .. retardant cellulosics, polyester, and nylon 66 fabrics.

Downloaded from trj.sagepub.com at The University of Iowa Libraries on June 1, 2015

 331

function of DAP add-on first increases (at low add-on

levels) and then decreases with addition of more DAP
[3, 13]. The six percent add-on has been just about
enough to bring the rate back to the value for untreated
cotton. The flame retardant does make the cotton
more extinguishable, and, in consequence, these rates
had to be obtained in 27.1% Ot.
Results obtained for burning the same pair of fabrics
in the 45° upward and vertically upward modes at
21°~o 02 are shown in Figures 5 and 6. Here again,
for both treated and untreated materials, mass trans-
formation rate decreases with increasing moisture con-
tent. However, in these burning modes the flame-
retardant material actually bums faster (by as much as
a factor of 2) than the untreated material. This un-
expected result has also been observed with a number
of other flame-retardant cotton systems including
THPC and Pyrovatex®-treated cotton.

FIG. 3. Effect of fabric moisture content on the 45° upward

burning behavior of polyacrylonitrile (Orient) staple fiber fabric
in a dry atmosphere of 19.8 and 21.8% O2 (arrows indicate the
direction of change in moisture content).

Cellulosics. The effects of fabric moisture content

on the horizontal mass transformation rates of untreated
cotton sheeting and of cotton + 6% (OW F) diam-
monium phosphate (DAP) are compared in Figure 4.
The points for both the untreated and flame-retardant
material fall along the same curve of slightly decreasing
transformation rate with increasing moisture content.
This similarity in transformation rates for treated and
untreated materials is not altogether surprising, since
other studies have shown that transformation rate as a

FIG. 5. Effect of fabric moisture content on the 45° upward’

burning behavior of untreated and flame-retardant treated (6%
DAP) cotton sheeting in a dry atmosphere of 21% Ot.

This is not the first time that transformation rate

enhancement for flame-retardant fabrics has been noted
C3, 13], and in retrospect it is not completely surprising.
Simplified models have suggested fundamental differ-
ences in burning mechanisms for burning in different
modes [9]. During upward burning char is produced
by a rapidly moving flame, which leaves the char behind
to cool quickly in an environment of fresh air brought
in by convection currents. Unburned material, just
ahead of the flame front, is shielded from air by the
FiG. 4. Effect of fabric moisture content on the horizontal
burning behavior of untreated and flame-retardant (6% DAP) ascending combustion by-products and is preheated by
cotton sheeting in a dry atmosphere of 27.1% 02. convective heat transfer from the flame front. When

Downloaded from trj.sagepub.com at The University of Iowa Libraries on June 1, 2015

 332

Ftc. 6. Effect of fabric moisture content on the vertical up-

ward burning behavior of untreated and 8ame-retardant treated FIG. 7. Effect of fabric moisture content and oxygen con-
(6°Jo DAP) cotton sheeting in a dry atmosphere of 21% Os. centration on the 45° upward burning behavior of untreated and
flame-retardant treated (6% DAP) cotton sheeting in a dry
atmosphere.
a flame downward, it is in contact with char for
moves
a longer period of time, but the char has little chance
to be exposed to any fresh supply of oxygen while it is mode. As before, for a given moisture content the
hot. The material to be burned ,would be heated up transformation rate is higher for the treated material
more slowly (since most of the flame heat is directed over the entire range of oxygen concentration investi-

upward) but will have a better chance of undergoing gated. The oxygen sensitivity is slightly less for the
air oxidation. Horizontal burning represents a condi- untreated material and appears to be independent of
-
tion somewhere between the above two cases. Chemi- moisture for both materials. The dependence of trans-
cal differences have been confirmed by infrared analysis formation rate on moisture content (seen visually by
.
of chars produced under various modes of burning [9]. the spacing between the lines of constant moisture
Since the increase in burning rate with addition of content) is similar for both the treated and untreated
flame retardant is most noticeable for upward burning materials.
modes, it is conjectured that the effect results from the Figure 8 compares the burning behavior of a flame-
preheating of unburned material in an oxygen-deficient retardant rayon (13% PFR, 3.5 oz/yd2) with untreated
atmosphere. rayon challis (3.9 oz/yd2) as a function of both oxygen
Additional information can be obtained from plots concentration and fabric moisture content. In con-
of mass transformation rate versus oxygen concentration trast to the two cotton systems discussed above, the
for different levels of fabric moisture content. This oxygen sensitivity for the untreated material
is signifi-
makes it possible to observe the effect of moisture cantly greater and appears to increase as fabric mois-
content on the extinguishability parameters: extrapo- ture content decreases. The flame-retardant rayon also
lated oxygen index, (O.I.)o, and oxygen sensitivity, shows a significantly larger dependence of transforma-
åMTR/ å% 02 [10]. The former represents the rela- tion rate on moisture content.
tive ease of extinguishability of a material, while the The results presented here all reveal a consistently
latter describes the response of the transformation rate faster upward transformation rate for cellulosic flame-
to oxygen depletion. retardant materials as compared to their untreated
Figure 7 shows the relationship between mass trans- controls. The trend has been observed over a wide
formation rate and oxygen concentration with fabric range of oxygen concentrations, and no crossover was
moisture as a parameter for untreated cotton and found or indicated where the order of the rates would.
cotton + 6% (OWF) DAP burning in the 45° upward be reversed.

Downloaded from trj.sagepub.com at The University of Iowa Libraries on June 1, 2015

 333

. -.

FIG. 9. Effect of fabric moisture content and oxygen con-

centration on the 45° upward burning behavior of polyethylene
terephthalate (PET) in a dry atmosphere.

rate to some extent. At these particular add-on levels,

TBPA < DAP < T23P, the extrapolated oxygen in-
FIG. 8. Effect of fabric moisture content and oxygen con- dices are increased in the same order. Because of the
centration on the 45° upward burning behavior of untreated and differences in add-on levels and amounts of phosphorus
flame-retaidant (13% PFR) treated rayon in a dry atmosphere. or bromine present, it is not possible
to~compare di-
rectly the relative effectiveness of the carious flame
Polyester. The mass transformation rate of a con- retardants. However, it is significant to note that
tinuous-filament, 3 X 1 twill polyester fabric is shown TBPA (containing no phosphorus) has a relatively
in Figure 9 as a function of oxygen concentration with small effect on the flammability behavior as compared
fabric moisture as a parameter. These curves were to the other compounds.
generated from plots of mass transformation rate versus The results in Figure 10 also show oxygen sensitivities
moisture content obtained at several oxygen concen- to be independent of moisture content for the flame-
trations. Figure 10 shows the effect of certain flame retardant polyesters. The response of transformation
retardants on the burning rate of this polyester: rate to changes in moisture content is about equal for
5% (OWF) TBPA (tetra-bromophthalic anhydride), all three treated materials. ’

4.9% (OW F) DAP, and 8.4% (OW F) T23P (tris-2,3-

dibromopropyl phosphate).’ This covers three of the
basic possibilities for flame-retardant additives: TBPA
contains bromine (3.4 wt % add-on) but no phosphorus;
DAP contains phosphorus (1.2 wt % add-on) but no
bromine; and T23P contains both bromine (5.8 wt %
add-on) and phosphorus (0.4 wt % add-on).
.

The untreated polyester (Fig. 9) shows the expected

decrease in transformation rate with increasing moisture
content, the parallel lines of constant moisture indi-
cating an oxygen sensitivity independent of fabric
moisture.
Comparison of the flame-retardant polyesters with
the untreated sample reveals that addition of each
flame retardant lowers the 45° upward transformation

3
DAP, T23P, and TBPA were padded onto the polyester from FIG. 10. Effect of fabric moisture content and oxygen con-
solutions of the flame retardant in water, perchloroethylene, and centration on the 45° upward burning behavior of flame-retardant
benzyl alcohol (plus 1% Triton 100&reg; as a surfactant), respec- polyethylene terephtalates (PET) (5.25% TBPA, 4.9% DAP,
tively. and 8.4% T23P)
10 a dry atmosphere.

Downloaded from trj.sagepub.com at The University of Iowa Libraries on June 1, 2015

 334

~ Nylon. Figure 11 compares the burning behavior of

an untreated nylon 2 X 1 twill fabric (3.2 oz/yd2) with
nylon + 2.7°~0 - (OWF) thiourea prepared by padding
from an aqueous solution. As with the polyester
systems, the addition of 8ame retardant lowers the
mass transformation rate and raises the extrapolated
oxygen index. The response of this rate to changes in
fabric moisture is about the same for the treated,and
untreated materials, but the oxygen sensitivity is sig-
nificantly greater for the Rame-retardant material.

FiG. 11. Effect of fabric moisture content and oxygen con-

centration the 45° upward burning behavior of untreated and
on
flame-retardant treated (2.7% thiourea) nylon 66 in a dry at-
mosphere. FIG. 12. Effect of fabric moisture and equilibration from
:he wet and dry sides on the 45° upward burning behavior of
untreated and flame-retardant treated (6% DAP) cotton sheeting
WETTING VtISUS DRYING. In the above examples, n a dry atmosphere.
fabric moisture levels were approached from the dry
side (i.e., fabrics were dried and then brought to the
desired moisture content). Studies have also been
carried out by equilibrating from the wet side. For
,
these cases’ cotton samples (untreated anp flame-
retardant) were sprinkled and stored overnight to allow
the moisture to distribute itself evenly through the
fabric. The samples were then placed in the condi-
tioning chambers and brought to equilibrium. Mass
transformation rates and moisture content were deter-
mined as described above.
The results are presented in Figures 12 and 13, which
FiG. 13. Effect of fabric moisture and equilibration from
show mass transformation rates (45° upward and hori- the wet and dry sides on the horizontal burning behavior of
zontal) as a function of fabric moisture, reached from untreated and name-retardant treated (6% DAP) cotton sheeting
both wet and dry sides, for untreated cotton and cotton in a dry atmosphere.. ,

-~ 6% DAP. The general character of the mass

,transformation rate-fabric moisture relationship is will not be evaporated as readily. These results show
similar for equilibration from the wet and dry direc- the expected faster 45° upward transformation rate for
tions. However, in both burning modes the curve for treated cotton; in the horizontal mode the flame-
the material equilibrated from the wet side is shifted to retardant, material bums at the same rate as the
,slightly lower rates. This effect is observed to equal untreated.
extents in both flame-retardant and untreated mate-
rials. This is very likely the consequence of a physical EFFECT OF TEST ENVIRONMENT HUMIDITY ON MASS
distinction, since the materials equilibrated from the TRANSFORMATION RATE. In addition to the effects of
dry side have more water at the surface of the fabric fabric moisture content, this study has been concerned
where it can be more easily removed by heating, while with the possible influences of test environment hu-
~ the materials equilibrated from the wet side tend to midity on mass transformation rate. Humidity might
have more moisture deep within the structure where it be expected to influence burning characteristics both

Downloaded from trj.sagepub.com at The University of Iowa Libraries on June 1, 2015

 335

through conceivable reactions of water vapor with

gas-phase reactants and products and through its eff ect
on the rate of change of fabric moisture content
during
the burning process. In studying the effect of hu-
midity, two types of experimental conditions can be
envisioned: those in which the humidity of the test
environment is in equilibrium with the initial fabric
moisture content (herewith termed equilibrium bum-
ing), and those in which the humidity of the test
environment is either greater or less than that which is
in equilibrium with the initial fabric moisture content
(termed nonequilibrium burning), such as burning in a
dry atmosphere. The results presented here deal with
nylon fabric in nonequilibrium burning and wool and
rayon in both equilibrium and nonequilibrium burning.

Nylon. Figure 14 shows the effect of test environ-

ment humidity on the mass transformation rate for a
2 X 1 twill, nylon filament fabric (3.2 oz/yd2), con-
taining 2.8% moisture, burning in the 45° upward
mode under 21.8% O2. The data were taken from a
series of nonequilibrium experiments in which a number FIG. 15. Effect of test environment relative humidity and
of samples fortuitously showed the same moisture con- fabric moisture content on the 45° upward burning behavior of
tent. Thus, it was possible to observe the effect of worsted wool fabric in 23.4% ot.
test environment humidity without the complicating
factor of sample moisture variations. The results show but the curve obtained for burning under equilibrium
a gradual decrease in transformation rate with increas- environmental conditions (wet atmospheres) is shifted
ing relative humidity of the test environment. to lower rates.

Rayon. Figure 16 shows results for

a rayon challis
fabric (3.9 oz/yd2) burned in both dry and humid
(equilibrium) atmospheres in the 45° upward mode

FIG. 14. Effect of test environment relative humidity on the

45° upward burning behavior in 21% O2 of nylon filament fabric
containing 2.8% moisture.

Wool. Figure 15 shows results obtained on a worsted

wool, 2 X 1 twill f abric (7.0 oz/yd2) for 45° upward
burning at 23.4% O2 in both humid (equilibrium) and
dry (nonequilibrium) atmospheres. The dry atmo-
sphere was effectively at 0% RH. The results show a
curvilinear decrease in mass transformation rate with
increasing moisture content in both dry and humid test FIG. 16. Effect of test environment relative humidity and
environments. The general trend of the curves is the fabric moisture content on the 45° upward burning behavior
same for both wet and dry environmental conditions, of rayon fabric in 19.8% O=.

Downloaded from trj.sagepub.com at The University of Iowa Libraries on June 1, 2015

 336

under 19.8% 02. The results show the expected de-

in transformation rate with increasing fabric
’

crease
moisture content in both wet and dry atmospheres.
’
However, in sharp contrast to the other materials in-
vestigated, the rates in humid atmospheres are grcoler
than those in a dry atmosphere. Other rayon fabrics
exhibited the same unexpected behavior. However, in
the horizontal burning mode this unexpected behavior
is not observed; as shown in Figure 17, the horizontal
’

rates for rayon are greater in the dry environment.

FIG. 18. Effect of test environment relative humidity and

fabric moisture content on the horizontal burning behavior of
rayon fabric in 54.5% 02.

These results showed that cotton, in contrast to rayon,

burns slower in a wet (equilibrium) atmosphere. Thus,
FiG. 17. Effect of test environment relative humidity and the anomalous burning behavior of rayon appears to be
fabric moisture content on the horizontal burning behavior of peculiar to rayon, and explanations must be sought in
rayon fabric in Z 1 °~o 02. the specific characteristics of rayon itself.

Summary
As apossible explanation for this behavior, it was
conjectured that, when a wet rayon sample is tested in The results presented here point out several effects
a dry environment at the relatively rapid-burning 45°, that moisture, present either in the fabric or in the
upward position, moisture evaporating from the fabric atmosphere of the test environment, may have on the
effectively cools the sample enough to lower the trans- flammability characteristics of textile materials. The
formation rate. However, in the horizontal mode the TRI Flammability Analyzer has proved to be a most
rate may be slow enough so that the release of moisture useful tool for these evaluations, providing indications
is completed, and the fabric comes back to ambient of moisture effects in terms of several measurable
temperature before the rate is established. In order to quantities: mass transformation rate (MTR), oxygen
test this hypothesis, experiments were carried out in sensitivity (4MTR/ å% 02), and extrapolated oxygen
the horizontal burning mode at oxygen concentrations index, (O.I.)o. The results do suggest that it is useful
’

(54.4%) sufficient to effect a transformation rate com- to consider flammability in terms of several properties.
parable to that in the 45° upward mode. The results The data from the Southern Regional Research
off these experiments (Fig. 18) show that, even at faster Laboratories of the USDA [4, 7, 8J indicated that the
’

transformation rates, the order of the results is not sensitivity of oxygen index values to fabric moisture
reversed (i.e., horizontal burning is still faster in the content was greater for flame-retardant cotton than for
.

dry atmosphere)., No other explanation for this be- untreated cotton. However, the flammability charac-
havior can be offered at this time. teristics reported here (mass transformation rate, extrap-
In order to determine whether this behavior is olated oxygen index, and oxygen sensitivity) show
peculiar to rayon or whether it’ is typical for cellulosic similar moisture sensitivity for both untreated and
materials, the effect of humidity on the burning be- flame-retardant cotton. The USDA data also showed
that the oxygen index of polyester was strongly mois-
,

havior of cotton fabric has also been investigated.

Downloaded from trj.sagepub.com at The University of Iowa Libraries on June 1, 2015

 337

ture-sensitive, while, in contrast, O.I. values for nylon ACKNOWLEDGMENT. The results reported in this
and cotton were only slightly affected. The results paper were obtained as part of a project at the Textile
here also show polyester flammability behavior, in Research Institute, &dquo;Flammability and Combustion
terms of transformation rate and extrapolated oxygen Behavior of Textiles,&dquo; sponsored by a group of eorpo-
index, to be much more dependent on fabric moisture rate TRI participants.
content than is the case for either nylon or cotton.
The flame-retardant rayon (PFR) exhibited a flam- Literature Cited
mability behavior that was more sensitive to fabric 1. AATCC 33-1962, Test for the flammability of clothing textiles.
moisture content than cotton treated with a phos- 2. AATCC 34-1966, Test for the fire resistance of textile fabrics.
phorus-type flame retardant. This is consistent with 3. Coppick, S., in "Flameproofing Textile Fabrics", R. W.
information reported by the National Bureau of Stan- Little, ACS Monograph Series, New York, Reinhold,
1947, pp. 50-54.
dards [14]. This same PFR rayon is also the only 4. Drake, G. L., Jr., et al., How Moisture Affects Flame Re-
example found where the additive has altered appre- tardancy, Textile Chemist and Colorist 4, 43-50 (1972).
ciably the oxygen sensitivity of the transformation rate 5. Fire Code (1967), City of Boston, Sec. 1107.
6. Fire Code, State of California, Para. 1273.3, Title 19.
(compare Fig. 8 with Figs. 7, 9, 10, and 11). 7. Hendrix, J. F., Beninate, J. V., Drake, G. L., Jr., and Reeves,
In general, the presence of moisture in a flame- W. A., Effects of Moisture on Oxygen Index (O.I.) Values
retardant or untreated fabric has a similar effect, for Textiles, Textile Res. J. 41, 854-855 (1971).
independent of whether it was equilibrated from the 8. Hendrix, J. F., Drake, G. L., Jr., and Reeves, W. A., Some
wet or dry side. Increasing fabric moisture content Factors Affecting Fabric Flammability as Measured by
Oxygen Index (OI), Textile Chemist and Colorist 5, 13-17
appears to raise the extrapolated oxygen index but does (1973).
not usually affect oxygen sensitivity. 9. Miller, B. and Gorrie, T., Chars Produced from Cellulose
Results obtained in this investigation reiterate the under Various Conditions of Thermal Decomposition,
previously reported phenomenon ~3, 13J wherein, for J. Polymer Sci., Part C, No. 36, Ed. E. C. Jahn, 3-19
cellulosic materials, the presence of a flame retardant (1971).
can increase the mass transformation rate. For non-
10. Miller, B., Goswami, B. C., and Turner, R., The Concept of
Extinguishability as a Flammability Criterion, Textile Res.
cellulosic materials the upward transformation rate is J. 43, 61-67 (1973).
decreased by flame retardants. 11. Miller, B., Martin, J. R., and Meiser, C. H., Jr., Measurement
The significance of test environment humidity has of the Burning Rates of Unrestrained Fabrics, in press.
12. Miller, B. and Meiser, C. H., Jr., A Method for Measuring
also been demonstrated by comparing mass transfor- the Burning Rates of Fabrics, Textile Chemist and Colorist
mation rates (as a function of fabric moisture content) 3, 55-59 (1971).
for materials burned in a dry atmosphere with those 13. Miller, B. and Turner, R., The Transfer of Flame Retardant
obtained for burning in a wet atmosphere (i.e., humidity Effects, Textile Res. J. 42, 629-633 (1972).
in equilibrium with the moisture content of the fabric). 14. NBS Technical Note 749, "The Flammable Fabrics Program
In general, materials burn more slowly in wet atmo-
1971," December 1972, p. 6, pp. 62-64.
15. U. S. Govt. DOC 4-72, Ignition standard for the flamma-
spheres with the singular exception of rayon which, in bility of mattresses.
upward burning modes, burns faster in a wet atmo- 16. U. S. Govt. DOC FF 3-71, Standard for the flammability of
children’s sleepwear.
sphere. This unexpected phenomenon is peculiar to
rayon and, at this time, defies explanation. Afaxuscript received August 15. 1974.

Downloaded from trj.sagepub.com at The University of Iowa Libraries on June 1, 2015