EXPLORATORY INVESTIGATION OF FIRE-RETARDANT TREATMENTS FOR
PARTICLEBOARDU.S. DEPARTMENT OF AGRICULTURE • FOREST SERVICE FOREST
PRODUCTS LABORATORY • MADISON, WIS. In Cooperation with the
University of Wisconsin
U.S.D.A. FOREST SERVICE RESEARCH N O T E FPL-0201 A U G U S T
1969
EXPLORATORY INVESTIGATION OF FIRE-RETARDANT TREATMENTS FOR
PARTICLEBOARD
Abstract
More than 80 exploratory experiments were made on the development
of fire-retardant-treated particle- board using 15 fire retardant
chemicals, 3 resin binders, 2 species of wood, and various types of
application. Among the types of application were those employing
solutions of fire retardant salts sprayed on green flakes; and
others employing dry, finely divided fire retardants added after
mixing binder resin and flakes. Of the boards produced and tested,
several using certain borates, AWPA Type C and Type D fire
retardants, and monoammonium phosphate showed promise. Properties
considered were flamespread, smoke production, strength, and
stability. Salts applied in solution to green wood, followed by
drying, gave results superior to those obtained from the use of dry
fire-retardant salts. None of the boards were tested for fire
retardance and strength after cycling through changing humidities.
The results of this series of experiments are released in the form
of a progress report for use by others engaged in similar
studies.
FPL-0201
U.S. Department of Agriculture
Introduction
As a result of increasing concern with fire safety, modern
buildifig codes have developed specific requirements for fire
performance of construction materials. With greater use of
particleboard in building construction there is greater need for
the development of an effective fire-retardant treatment for this
product.
The fire-test results reported here were obtained from a single
test for each set of boards. The data, therefore, should be
regarded only as indicators of treatment effectiveness as a prelude
to a more comprehensive study of promising combinations. The
results of this series of experiments are released in the form of a
progress report for use by others engaged in similar studies.
The value for each relative humidity or soak condition for each set
of boards in the evaluation of dimensional stability is based upon
the test of a single speci- men. Also, none of the boards were
tested for fire retardance and strength after cycling through
changing humidities. This must be considered in inter- preting
these test results.
1Maintained at Madison, Wis., in cooperation with the University of
Wisconsin.
FPL-0201
Use of Fire Retardants in Particleboard
Although there is appreciable knowledge on the use of fire
retardants with solid wood, plywood, and textiles, there is
comparatively little information on their application to
particleboard. The use of fire retardants in this product involves
some specific problems.
The particles composing the board are held together by the resin
binder, not in the form of a somewhat continuous film as in
plywood, but as a multiplicity of fine spots of adhesive which tack
together the contacting particles. There is obviously a much
smaller amount of adhesive per bond area in particleboard than, for
example, in plywood.
The nature of the pressing operation in the manufacture of
particleboard is such that precise control of the curing of the
resin is essential. During the time required for the particleboard
press to close to final raw board thickness, heat is being applied
to the binder resin. The resin must not cure before the press has
closed to board thickness. From a production standpoint, it is
undesirable to slow the curing of the resin sufficiently to require
a long pressing time.
Unfortunately, the chemical nature of the fire retardant and the
quantity required to be effective are such that frequently the
curing properties of the binder resins are affected.
Fire-retardant-treated particleboard often has lower strength than
untreated board.
Study Materials
The particleboards upon which this report is based were made and
evaluated in two separate though similar investigations. To
distinguish between the two groups of boards, since their
fabricating conditions were different, the investiga- tion
concerning the group made first has been designated as Study I and
that concerning the second group of boards has been designated as
Study II. Differ- ences in fabrication variables of the two groups
are indicated in the list of fabrication specifications.
FPL-0201 -2-
Fabrication Specifications
The boards for this investigation were made to the following
specifications:
1. Species: Douglas-fir heartwood in Study I; Douglas-fir heartwood
or aspen in Study II.
2. Density: 40 pounds per cubic foot.
3. Board size: 1/2 inch by 24 inches by 28 inches, rough
dimensions.
4. Particle size: 1 inch by 0.015 inch by random-width
flakes.
5. Resin type: Urea-formaldehyde, 65 percent solids;
phenol-formaldehyde, 43.5 percent solids; and
melamine-formaldehyde, 45 percent solids. The urea and phenol
resins were used in the liquid form as they were supplied by the
manufacturer. The melamine-formaldehyde resin, in dry form, was
dissolved in 1:15 isopropanol-water solution before using.
6. Wax sizing (where used): 1 percent wax solids, based on ovendry
wood weight; applied as supplied, a 46 percent solids
emulsion.
7. Mat moisture content: Adjusted to 12 percent in Study I and to
10 percent in Study II.
8. Press temperature: In Study I, 325° F. for the urea-formaldehyde
and melamine-formaldehyde resins and 350° F. for the
phenol-formaldehyde resin; in Study II, 300° F. for the urea- and
melamine-formaldehyde resins and 325° F. for the
phenol-formaldehyde resin.
9. Press time: In Study I, 15 minutes for all binders: in Study II,
8 minutes for the urea- and melamine-formaldehyde binders and 10
minutes for the phenol-formaldehyde binder.
10. Time to stops: 2 minutes in Study I; 1 minute in Study
II.
FPL-0201 -3-
Selection of Fire Retardants
Numerous fire-retardants have been suggested for wood (14,15 ,16).2
Most of these have been water-soluble inorganic salts, a number of
which have been in use for many years. Among the more recently
proposed fire retardants are organic compounds which seem to lack
some of the disadvantages of the inorganic water-soluble salts.
Some of these newer fire retardants, which have proven quite
effective for textiles and for paper, are too expensive to consider
for particleboard treatment.
The fire-retardant treatments selected for these experiments
include a number of the more common inorganic treatments which have
proven effective for wood, and were considered economically
practical for this application.
The following fire retardants were included in this
investigation:
Monoammonium phosphate Diammonium phosphate Borax Boric acid Borax
- boric acid, 1:1 Monoammonium phosphate - ammonium sulfate, 1:1
Zinc sulfate Aluminum sulfate AWPA type C (4) AWPA type D (4)
11-37-0 ammonium polyphosphate (12) Dicyandiamide - phosphoric acid
- formaldehyde, 1:1:0.15 mole ratio Zinc sulfate - zinc
silicofluoride - urea Ammonium polyphosphate (32 percent
phosphorous content; proprietary
product) Boric acid - disodium borate tetrahydrate, in weight
ratios of 0:100,
10:90, 20:80, and 25:75 (7)
2 Underlined numbers in parentheses refer to Literature Cited at
the end of this report.
FPL-0201 -4-
There are several possible methods of applying fire retardants to
particle- board. Selection of a particular method depends upon such
considerations as the practical requirements of the manufacturing
process and the characteristics of the fire-retardant-treatment
chemicals.
Application of water solutions of fire retardants to the green or
wet wood particles soon after the particles are made is expedient
because this permits penetration of the treatment into the wood,
and the water of the treating solution may be evaporated in the
subsequent drying operation. The solution could be applied either
by spraying or by soaking the particles.
The disadvantage of any application in solution, particularly for
chemicals with rather low solubility, is the need to remove the
added water in the dryer.
The fire-retardant solution could also be added at the blender
where the resin and wax sizing are added to the dry particles. The
disadvantage of this, however, is that the water of the treatment
solution would have to be removed in the pressing operation. In
addition, the added water might dilute the resin and cause
excessive strike-in. It is probably desirable for the fire
retardant to penetrate into the wood and for the binder resin to
stay at the surface. This would be difficult to accomplish by
spraying both materials on the dry particles at the blender.
It may appear convenient to add the fire retardant in solution to
the finished board by soaking, applying to the board surfaces, or
pressure impregnation. However, treatment of the finished board
would require an additional drying operation to remove the
relatively large amount of water added. Thus, it is more practical
to add the fire-retardant solution to the particles rather than to
the finished boards.
The fire-retardant-treatment chemical may be added as a dry powder,
preferably during or immediately after the resin application, so
that it will be retained on the wood particles by the tackiness of
the binder. Dry application obviously eliminates the necessity of
removing the water added by solution application of the fire
retardant. However, many binders lack the requisite tack to retain
the powder. Problems associated with dry application include
lowering of the effectiveness of the binder, sifting of the
chemical through the mat, and difficulty, compared with solution
application, of obtaining even distribution of the treatment.
FPL-0201 -5-
Three methods of adding the fire retardant were investigated in
these studies: (1) application of dry chemicals at the resin
blender, (2) spraying the chem- icals on the wet particles, and (3)
roller application of the solution on the surface of the hot board
just removed from the press.
Application of dry chemicals to the wood particles in the blender
immediately after spraying with resin presented problems of
obtaining even distribution and of adhering all the required fire
retardant to the particles. After fire test results indicated that
application dry was less efficient than in solution, the dry
application was used only for those chemicals having low
solubility. Applica- tion of approximately 30 to 35 percent water
solutions to the wet particles immediately before drying proved to
be simpler, more efficient, and more practical for production, so
this method was used for many of the combinations
investigated.
Preliminary investigations had indicated that some of the chemicals
reacted with the wood when the treated particles were oven dried:
therefore, in Study I all particles treated with fire retardants in
solution were air dried at room temperature. In Study II, a number
of similar boards were made from particles which had been air dried
at room temperature, and some others of particles which were oven
dried to about 5 percent moisture content at 225° F. In produc-
tion, the particles would be exposed to a higher temperature for a
shorter time in a continuous dryer. In order to control the rate
and uniformity of drying in a laboratory oven, however, it was
necessary to use a lower drying temperature.
An attempt was made (board 18) to add the fire retardant,
diammonium phos- phate in this instance, in solution to the surface
of the hot hoard immediately after its removal from the
press.
Fire-retardant-treatment chemical solutions for Study I, and for
Study II except where noted otherwise, were made up at 30 to 35
percent concentration in water heated to 120° F. The solutions were
sprayed on wood particles which were at about 30 percent moisture
content. The amount of fire-retardant mate- rial added was based on
dry weight of the chemical as a percentage of the weight of the
ovendry wood.
All liquid materials--fire-retardant solution, binder resins, and
wax emulsion-- were applied by spraying them into the mass of
particles tumbling in a rotating drum. Dry fire-retardant chemicals
were applied slowly to the resin-sprayed particles in the rotating
drum. All materials applied dry were reduced to pass through a
100-mesh screen.
FPL-0201 -6-
The American Wood-Preservers’ used in this investigation
were:
Association fire-retardant formulations (4)
Diammonium phosphate Ammonium sulfate Sodium tetraborate, anhydrous
Boric acid
10 60 10 20
Zinc chloride Amnnonium sulfate Boric acid Sodium bichromate
35 35 25 5
Monoammonium Phosphate- Ammonium Sulfate treatment
High smoke density is a problem with particleboards treated with
monoammo- nium phosphate. This may be reduced through the addition
of ammonium sulfate. In Study I equal amounts of these chemicals
(1:1 weight ratio) were used in three series of boards (boards 37,
38, and 39). In Study II, to attempt to reduce the effect of the
sulfate on the resins, a 3:1 weight ratio of monoammonium
phosphate-ammonium sulfate was used (boards 64 and 65). This
treatment was added at a 13-percent level so that the level of the
monoammonium phosphaete would be about 9 percent.
Boric Acid-Disodium Octaborate Treatment (7)
The boric acid-disodium octaborate solutions were made up in
proportions of 0:100, 10:90, 20:80, and 25:75 at a solution
concentration of 38 percent in water heated to 145° to 165° F. The
disodium octaborate tetrahydrate (Na B O
2 8 13. 4H.O) was dissolved first, followed by the boric acid. The
solution was main-
tained and sprayed at 145° to 165° F. Application was 18 percent,
based on oven- dry wood weight.
FPL-0201 -7-
Zinc Sulfate- Zinc Silicofluoride- Urea Treatment
This treatment, which has been recommended as a means of
insolubilizing a zinc compound within the wood, was applied at a
rate of 2 pounds of zinc per cubic foot of board. The following
formulation was used:
Component Weight Percent
Water 55.0 Zinc sulfate 16.5 Zinc silicofluoride 18.1 Urea
10.4
100.0
The zinc content of this formulation is 7.5 percent.
After the green (wet) wood particles were sprayed with this
solution they were placed in tightly closed polyethylene bags and
heated at 175° to 180° F. for 15 hours. This treatment is intended
to react the zinc compounds with ammonia derived from decomposition
of the urea. The particles were removed from the polyethylene bags
and dried at 225° F. to about 5 percent moisture content.
Treatment with Ammonium Polyphosphate (32 Percent P)
This material is a low-solubility ammonium polyphosphate
proprietary product with a phosphorous content of 32 percent.
Because of the low solubility of this chemical, it was necessary to
apply it as a dry powder following resin application.
Dicyandiamide-Phosphoric (9,10)
Acid Treatment
A 20-percent solution phosphoric acid, and application was used
in
was formaldehyde treating
the wood particles.
amide mole
(1-cya ratio.
no- A
guanidine) 10-percent
The dicyandiamide was dissolved in water which had been heated to
120° F. With constant stirring, the phosphoric acid and
formaldehyde were added in turn and the temperature was increased
to 180° F. The sprayed particles were dried in a 180° to 190° F.
oven to approximately 5 percent moisture content. Care was taken to
insure that the oven temperature did not go above 190° F. and that
the particles were not overdried.
FPL-0201 -8-
Evaluation of Board Properties
Three boards were made for each treatment. Four half-boards, each
13-3/4 inches by 23 inches, and one piece 13-3/4 inches by 4 inches
were required for each fire test. Two half-boards were cut into
specimens for mechanical and dimensional stability tests. All test
specimens were unsanded.
At the beginning of Study I, control boards without fire retardants
were made for reference. Table 1 gives test values ofproperties of
the control boards made with urea-, phenol-, and
melamine-formaldehyde resins. Because specifications are not
available for fire-retardant-treated particleboard, arbitrary
values were established for treated boards fabricated to the
control specifications to assist in identifying the most promising
treatments.
Fire Performance
The fire performance of the boards was evaluated by the
8-foot-tunnel furnace test for surface flammability (1,8) Figure 1
shows the specimen side of the test furnace. All board specimens
were conditioned to equilibrium at 30 percent relative humidity
before fire testing.
The 8-foot-tunnel method (1) has been developed by the Forest
Products Laboratory as a research technique for measuring surface
flammability. It generally ranks the flammability of materials in
the same order as the 25-foot- tunnel furnace test (3), but the
actual index values may differ. Both methods use index values
relative to flame-spread distances traveled with respect to red oak
at 100. However, because of longer exposure times in the 8-foot
tunnel, fire-retardant-treated wood at the lower range of the scale
frequently has a flame-spread index of about 40 as compared to 25
on the same material in the 25-foot-tunnel furnace. Therefore, one
of the criteria selected for an adequate treated particleboard was
a flame-spread index in the 8-foot-tunnel furnace of 40 or less.
The heat contribution of similar materials generally correlates
with the flame-spread distances involved in the tests, and
therefore a heat contribu- tion index of 40 or less was also
selected as an acceptance criterion.
Smoke density index values are established for both methods
relative to the smoke density produced by untreated red oak lumber
in the respective tests. In the 25-foot tunnel with a large
ignition burner, both untreated red oak and fire-retardant-treated
materials burn under a flaming exposure condition. However, in the
8-foot tunnel, where most of the heat is supplied by a radiant
source, exposure results in flaming combustion for the untreated
red oak, but nonflaming exposure for fire-retardant-treated
products. Nonflaming exposure of cellulosic products results in
much greater smoke production than does
FPL-0201 -9-
flaming combustion. In previous experience with the more effective
fire- retardant chemicals, smoke index values as high as 400 were
obtained in the 8-foot tunnel for products rated as 50 or less for
flaming combustion in the 25-foot tunnel. Therefore, an acceptance
criterion for boards produced in this study was established as a
smoke index value of 400 or less in the 8-foot-tunnel test.
Strength Properties
Following methods specified in ASTM D 1037-64 (2), strength
properties of the boards were evaluated by static bending and
tension perpendicular to the surface. An initial strength loss of
approximately 40 percent as a result of treatment was arbitrarily
applied in choosing the minimum strength values in table 1 for
treated boards. The 40 percent reduction was chosen as a reasonable
indicator of boards showing promising strength properties after
treatment. Other values could be used. For example, selectionof an
approximate 20 percent initial strength reduction would reduce the
number of treated boards showing promise to three (boards number
66, 67, and 68).
In Study I, nine static bending specimens and 24 tension
perpendicular to the surface (internal bond) specimens were tested
for each treatment. In Study II, six static bending specimens and
eight tension perpendicular to the surface specimens were tested
for each treatment. Specimens were cut from each set of three
boards as shown in figures 2 and 3.
Dimensional Properties
Dimensional stability test specimens 3/4 inch by 22 inches were
made for each set of boards. All specimens were first brought to
equilibrium in an atmosphere of 30 percent relative humidity at 80°
F. Then one specimen from each set of boards was exposed to each
moisture treatment as follows: One oven dried to moisture freeness
in a 220° F. mechanical convection oven for 24 hours; one soaked in
water for 38 days; and one each exposed to 30, 60, 80, or 90
percent relative humidity at 80° F. for 30 days. A vacuum-pressure
soak test (11) was performed on a specimen of each board.
Measurements were made on each specimen after the initial
conditioning at 30 percent relative humidity and again after
reaching equilibrium with the find condition. Length measurements
to ±0.001 inch were taken from two small holes 20 inches apart in
the face of the specimen, and thickness measurements were taken at
five marked, equally-spaced points between the holes (11). The
tabulated values are based upon a calculated ovendry
condition.
FPE-0201 -10-
Included for controls in both Studies I and II were boards made
without fire-retardant treatments (boards 1 through 4 and 44
through 49). These included a board made with each of the three
resins, and in Study II control boards were made both with
Douglas-fir and with aspen. One board (board 2) was made with the
addition of 1 percent (solids) wax.
The relationship between properties of untreated and
fire-retardant-treated commercial production boards may differ from
that of the boards made in this study because of certain basic
differences in the laboratory and the plant processes. For example,
the higher drying temperatures and the required storage of the
fire-retardant-treated and resin-sprayed particles in production
will probably increase the effect of the chemicalson the wood and
on the binder resin. The rapid press closure to stops used with the
laboratory press may not be possible with most production presses.
Increases in linear movement and thickness of approximately 50
percent as a result of treatment were arbitrarily allowed in
choosing the values given in table 1.
Supplementary Investigations
Two brief investigations were made to further consider the
relationship of the fire-retardant-treatment chemicals and binder
resins. One involved several cure plate tests of resins alone and
of mixtures of resins and fire-retardant chemicals. The other was
an investigation of the effect of storage of fire-
retardant-treated and resin-sprayed wood particles upon board
properties.
Cure Plate Tests
Cure plate tests were performed to study the effect of several of
the fire- retardant-treatment chemicals on the curing of the binder
resins. Immediately before each test, a 10.0-gram sample of the
liquid resin was placed in a small beaker, a weighed amount of the
fire-retardant-treatment chemical was added, and the mixture was
thoroughly stirred. About 2 to 4 gram of the mixture were spread on
the hot surface of the plate and stroked with a spatula until the
curing point was detected.
The amounts of fire-retardant chemicals in the test samples were
equivalent to 2, 5, and 10 percent in the board, with the 10.0
grams of resin as equivalent to 8 percent resin solids in the
board.
The apparatus used in this test was simply a laboratory hotplate
with accurate thermostat control (±2° F.). A stainless steel sheet
was clamped to the surface of the hotplate. The required
temperatures were set by a thermocouple placed beneath a weight on
the plate surface.
FPL-0201 -11-
A high and a low temperature were used in these tests. A low
temperature of 225° F. for all resins was intended to be the
equivalent of the interior board temperature, and a high
temperature of 325° F. for the urea and melamine resins or 350° F.
for the phenol resin was intended to be the equivalent of the board
surface temperature during the pressing operation.
Effect of Storage
In the laboratory, particles are sprayed with resin and almost
immediately formed into a mat and pressed into a board. In a
particleboard plant under production conditions, it is necessary to
hold resin-sprayed particles in a surge bin before the forming
line. Therefore, it is desirable to determine the effect of storage
of the fire-retardant-treated, resin-sprayed wood particles upon
board properties. For this investigation, monoammonium phosphate
was applied in solution at a 10 percent level for the
fire-retardant treatment, and urea- formaldehyde resin was used as
the binder. The resin-sprayed particles were stored at room
temperature in closed polyethylene bags until they were formed into
mats and pressed. Fabricating conditions follow those given for
Study II of this investigation.
Observations
Results of the evaluation of the boards made in this investigation
are given in tables 2 through 4. Tables 2 and 3 respectively give
results for the boards made in Studies I and II. Table 4 contains
data extracted from table 3 to more clearly indicate the effect of
species as shown by the difference in properties of similar boards
made with Douglas-fir and with aspen. Results of the brief study of
the effect of storage of treated, resin-sprayed particles on board
properties are given in table 5. Cure plate tests of the three
binder resins in mixtures with several representative
fire-retardant treatments are presented in tables 6, 7, and 8.
Table 9 gives 10 combinations (extracted from tables 2 and 3) that
exhibited properties equal or superior to the arbitrary limits
shown in table 1.
Fire Test Properties
As the results given in tables 2 and 3 indicate, several of the
fire-retardant treatments gave adequate fire test properties to the
particleboards.
Monoammonium phosphate at a 10-percent level, when applied in
solution to the green particles gave adequate flame-spread and heat
contribution values,
FPE-0201 -12-
but smoke density tended to be too high. The smoke density
evidently may be reduced with the addition of ammonium sulfate
(boards 37, 38, 39, 64, and 65), however this salt is a strong
catalyst for the urea resin.
Diammonium phosphate, applied in solution to the green wood
particles, at a 5-percent level (boards 16, 62, and 63), gives good
flame-spread and heat con- tribution values, but the smoke density
is too high.
Treatments with boric acid and disodium octaborate in the
proportions used in this investigation (boards 66, 67, 68, and 69),
applied in solution at a 10- percent level, provided adequate fire
test properties. However, the treatments with boric acid and borax
(sodium tetraborate) in a 1:1 weight ratio (boards 19, 20, 21, 22,
and 23) gave inadequate flame-spread properties.
Particleboards had adequate fire test properties when treated with
either AWPA Type C or Type D formulations applied in solution at a
15-percent level (boards 29, 30, 31, 33, 34, and 35). When these
treatments were applied dry at a 15-percent level (boards 32 and
36), the flame-spread was too high.
For the combinations of Fire retardants and binder resins used in
this study, mechanical strength and dimensional stability of the
boards were affected by the addition of the tire-retardant
chemicals. Though additional factors may be involved, it is
reasonable to conclude that the lower quality of the
fire-retardant- treated boards is due mainly to interference of the
treatment chemicals with the resin binder. Boards made with the
urea-formaldehyde binder were particu- larly affected by the
presence of the fire-retardant-treatment chemicals.
In an earlier exploration, the combination of borax or boric acid
with phenol formaldehyde resin resulted in poor quality boards.
Test specimens often split apart as they were being prepared,
indicating that tensile strength perpendicu- lar to the surface was
very low. The board made with the phenol-resin binder and the
borax-boric acid fire-retardant treatment (board 22) had low
strength properties, and it failed in the vacuum-pressure soak
test. Phenol-resin-bound hoards containing the AWPA Type C and Type
D fire-retardant formulations (hoards 30 and 34) were of such poor
quality that test specimens failed in preparation. This may be due
to the boric acid and sodium tetraborate, though some of the other
chemicals in these formulations may be involved.
FPL-0201 -13-
ddobson
Line
ddobson
Line
The problem of using phenol-formaldehyde resins with boron
compounds has been noted before (6) but the exact nature of the
interference is uncertian.
The interference with the urea-formaldehyde resin by the
fire-retardant treatments containing ammonium salts is probably due
to the pH change caused by the decomposition products of the
ammonium compound when exposed to heat of drying or pressing
(5,13). A reaction between the free ammonia and the formaldehyde
may also take place.
Results of the cure plate tests indicated that
fire-retardant-treatment chemicals affected the curing of the
particleboard binder resins. Though curing conditions in the cure
plate test are somewhat different from those in the particleboard
in the press, nevertheless the results of the cure plate tests
indicate that the cure time is shortened by the presence of several
of the fire- retardant-treatment chemicals. Therefore, it appears
probable that the lower strength of boards containing fire
retardants is due to the resin starting to harden before the press
has closed to stops. This would account for the especi- ally low
values for tensile strength perpendicular to the surface shown by
many of the boards.
Cure plate tests of the phenol-formaldehyde resin mixed with
borax-boric acid and with AWPA Types C and D were not possible
because the resin hardened as soon as it was mixed with these
materials.
Method of Application of Fire Retardant
From the limited investigation of methods of adding fire retardants
to the wood particles, it appears that the best procedure is to
apply the chemicals in water solution to particles having a
moisture content at about the fiber satura- tion point. This method
of application necessitates that water added as the solvent be
removed in the subsequent drying operation, and the possible effect
of heat of the dryer upon the treatment chemical should be
considered. It has been suggested (7) that application in solution
to the green particles is more efficient because the chemical
penetrates into the wood rather than being con- centrated at the
surface as it would be in dry powder operation.
The effect of drying temperature was investigated for the ammonium
phos- phates since it was assumed that these would tend to
decompose with the heat of drying or pressing. Results showed that
the boards made from particles which had been dried at room
temperature after application of monoammonium phosphate solution
were usually slightly stronger than those boards made from
particles which had been dried at 225° F. after treatment (compare
boards 50 and 52, 51 and 53, 54 and 56).
FPL-0201 -14-
Boards made from particles which had been treated with dry
fire-retardant chemicals after binder resin application had poorer
flame-spread and heat con- tribution values than those made from
particles treated with the same chemicals at the same level but
applied in solution (for example, compare boards 5 and 10, 6 and
11, 21 and 24, 29 and 32, and 33 and 36). Some chemicals, however,
have such low solubility that it is not possible to apply them in
solution. The ammo- nium polyphosphate product (boards 74, 75, 76,
and 77) is such a material.
En the application of dry fire-retardant chemicals in this
laboratory investiga- tion it was difficult to retain all of the
required amount of the chemical on the particles, Under plant
manufacturing conditions, in which the treated particles would
receive much rougher handling than they do in laboratory board
fabrica- tion, retaining the dry chemicals on the wood particles
would be an even more difficult problem.
Four boards made with application of the fire retardants as dry
powders had acceptable flame-spread values, though smoke density
was too high for all four of them. No other boards made in this
investigation with dry application of fire- retardant chemicals had
acceptable flame-spread properties. Fire-retardant treatments for
the four boards were monoammonium phosphate at levels of 10 and 15
percent (boards 11 and 12), and ammonium polyphosphate at a 10-
percent level (boards 74 and 77).
An attempt to add the fire retardant in solution to the surface of
a hot board (board 18) was not successful. Penetration of the
chemical into the board during cooling was inadequate, and, due
either to evaporation or to absorption of water into the board,
much of the solid fire-retardant chemical was deposited on the
board surface. In plant production it may be possible to apply the
fire-retardant treatment at this point by soaking the boards in the
treatment solution as they are removed from the press. However, the
necessity of removing excess water from the board as well as the
concentration of the treatment in the board sur- face layer where
it will be removed in sanding makes this approach rather
impractical.
The strength properties of the boards containing the boric
acid-disodium octaborate formulations (boards 66, 67, 68, and 69)
were much better than those of other fire-retardant-treated boards
made in this investigation. Additional investigation is necessary
to be certain of the reason for this, but it should be noted that
the pH of this strongly buffered formulation is in the range of 6.0
to 6.8 and apparently more favorable to the proper curing of the
urea-formalde- hyde resin.
FPL-0201 -15-
Two boards were made with the addition of paraffin wax sizing to
determine its effect upon fire test results. The 1 percent wax
addition increased smoke density slightly in the untreated control
board (board 2; compare with board 1) and considerably more in the
board containing the monoammonium phosphate fire retardant (board 7
compare with board 6).
The dimensional stability test results for all boards containing
fire-retardant- treatment chemicals were poorer than those for the
untreated control boards. Though these test results probably
indicate combinations which would perform poorly in use, those that
appear satisfactory in these tests may not perform well in service.
Cycling of the specimens through various humidity conditions, as
would usually occur in application, might causefailure in a board
that showed relatively good dimensional stability properties in
these tests.
Some indication of the effect on board properties of a storage
period for fire-retardant-treated particles following application
of the binder resin was given by the brief supplementary
investigation (table 5). A more complete investigation is required
for definite conclusions, but it appears that, with the use of a
monoammonium phosphate fire-retardant treatment and a urea-resin
binder, the internal bond strength of the boards is decreased when
there is a storage period of 2 hours or more.
To briefly study the possible effects of species on the
fire-retardant-chemical and binder resin relationship, several
boards in Study II of this investigation were made with aspen
flakes. For each of these aspen boards there was a corresponding
Douglas-fir flakeboard fabricated under the same conditions (see
table 4).
Some general observations may be made in comparing the properties
of similar aspen and Douglas-fir boards. Flame-spread was higher
for all aspen boards than for corresponding Douglas-fir boards. The
aspen boards had slightly higher flexural strength. The untreated
aspen boards had somewhat lower tensile strength perpendicular to
the surface than that of the correspond- ing Douglas-fir boards,
but for the monoammonium-phosphate-treated boards this was
variable.
Combinations of fire retardant, concentration, species, and binder
resin yielding laboratory-made particleboards considered to have
adequate properties before exposure to typical use conditions are
shown in table 9. Four additional combinations made with
monoammonium phosphate and diammonium phosphate (boards 50, 51, 54,
and 63) were adequate, except for smoke density being slightly
high.
FPL-0201 -16-
Conclusions
1. It is apparently possible to achieve adequate fire test
properties with several of the treatments investigated. The most
satisfactory fire-retardant treatment found in this investigation
was boric acid-disodium octaborate. Fire- test properties were
adequate. Tensile strength perpendicular to the surface and modulus
of elasticity were approximately the same as those of the untreated
board. Modulus of rupture was about 75 percent that of the
untreated board.
2. Application of the fire retardant in solution to the particles
before they are dried apparently is more efficient than application
as a dry powder to the particles in the blender immediately after
resin spraying.
3. For solution application, 10 to 15 percent of
fire-retardant-treatment chemical (baaed upon the ovendry wood
weight) is required for adequate fire test properties.
4. The critical problem in the development of
fire-retardant-treated particle- board is the effect of the
treatment chemicals upon the resin binder. Though a number of
factors may be involved, the problem appears to be caused mostly by
the lowering of the pH of the system by the fire retardant or its
thermal decomposition products.
Since in many instances tensile strength perpendicular to the
surface was greatly reduced, it seems a reasonable conclusion that
the resins were being catalyzed by the lire-retardant chemicals to
harden before the press was closed to board thickness.
5. It is not possible with the sort of information obtained from an
exploratory investigation of this nature to predict the strength
properties of a specific fire- retardant-treated board in relation
to those of a similar but untreated board. Generally, the results
of this investigation indicate that the addition of most of the
fire retardants used in this study causes an appreciable decrease
in bond strength, apparently by interfering with the proper curing
of the resin. Consider- ation of the possible nature of this
interference leads to the presumption that the strength properties
of such board products could be improved by chemical modification
of the fire-retardant-treatment chemical or, possibly, the binder
resin.
FPL-0201 -17-
Literature Cited
1. American Society for Testing and Materials 1965. Tentative
method of test for surface flammability of building materials
using an 8-foot (2.44 m.) tunnel furnace. ASTM E 268-65T.
2.
1964. Standard methods of evaluating the properties of wood-base
and particle panel materials. ASTM D I037-64.
3. 1967. Standard method of test for surface burning
characteristics of build-
ing materials. ASTM E 84-67.
4. American Wood-Preservers’ Association 1968. AWPA Book of
Standards. Standards for Fire-Retardant Formulations.
P10-68.
5. Bramhall, George 1966. Fire retardant treatment of wood. A
section of Fang, J. B.,
MacKay, G. D. M., and Bramhall, G. Wood fire behavior and fire
retardant treatment; a review of the literature. Canadian Wood
Council, Ottawa, Ontario.
6. Deppen, H. J., and Lux, B. V. 1967. The use of inorganic
compounds in the production of particleboard
materials of low flammability. [German] Holz-Zentralblatt 93(107):
1671.
7. Draganov, S. M. 1967. Borates as fire retardants in particle
board (2nd Ed.). U.S. Borax
and Chemical Corp., Los Angeles, Calif.
8. Forest Products Laboratory, Forest Service, U.S.D.A. 1967. Small
tunnel-furnace test for measuring surface flammability.
Research Note FPL-0167.
9. Goldstein, I. S., and Dreher, W. A. 1961. A non-hygroscopic fire
retardant treatment for wood. Forest Prod.
J. 11(5): 235-237.
FPL-0201 -18-
10. Goldstein, I.S., and Dreher, William 1964. Method of imparting
fire resistance to wood and the resulting
product. U.S. Pat. 3,159,503.
11. Heebink, B. G. 1967. A procedure for quickly evaluating
dimensional stability of particle-
board. Forest Prod. J. 17(9): 77-80.
12. Johnansen, R. W., and Crow, G. L. 1965. Liquid phosphate fire
retardant concentrations, Fire Control Notes
26(2): 13-16.
13. Raknes, E. 1963. Gluing of wood pressure-treated with water
borne preservatives and
flame retardants. J. of the Inst. of Wood Sci. (11): 24-44.
14. Truax, T. R., Harrison, C. A., and Baechler, R H. 1935.
Experiments in fireproofing wood. 5th Prog. Rep., U.S. Forest
Prod.
Lab. Rep. 1118.
15. Weiner, J., and Byrne, J. 1965. Flameproofing. The Institute of
Paper Chemistry Bibliographic
Series, No. 185, Supp. 1. Appleton, Wis.
16. West, C, J., Stringham, E., Roth, L., Weiner, J., and John, B.
1959. Flameproofing. The Institute of Paper Chemistry
Bibliographic
Series, No. 185. Appleton, Wis.
1.0-20FPL-0201 -19-
Ta bl e
2. --R
es ul ts of ev al ua tio ns of fir e- re ta rd an t-t re at ed pa
rti cl eb oa rd s,
St ud y
Ta bl e
2. -- R es ul ts of e va lu at io ns of fir e- re ta rd an t-t re
at ed pa rti cl eb oa rd s,
St ud y
Ta bl e
3. -- R es ul ts o f ev al ua tio ns of fir e- re ta rd an t- tre
at ed p ar tic le bo ar ds , St ud y
II
Ta bl e
3. --R
es ul ts of ev al ua tio ns of fir e- re ta rd an t-t re at ed pa
rti cl eb oa rd s,
St ud y
Ta bl
e 4.
an d
bo ar ds
(I nf or ma ti on fr om ta bl e
2)
5. -- Ef fe ct of st or ag e
of tre at ed pa rti cl es on pa rti cl eb oa rd pr op er tie
s
Ta bl e
re si n1
Ta bl e
re si n1
Ta bl e
9. -- Pa rti cl eb oa rd s ha vi ng in iti al fir e
re ta rd an t
an d
m ee tin g
re qu ire m en ts of ta bl e
1
Figure 1.--Specimen side of FPL 8-foot-tunnel furnace. 1, Gas
supply to main burner; 2, firebox; 3, clamp to hold down cover over
test speci men; 4, gas supply to igniting burner; 5, cover over
test specimen; 6, hood to collect combustion gases for temperature
and smoke measure ment; and 7, photoelectric cell for
smoke-density measurement.
ZM 110 169
Figure 2.--Cutting diagrams for specimens used in Study I for fire
performance and strength property tests. Three boards were made
with each fire-relardant treatment. Tension perpendicular to the
surface (internal bond) specimens are indicated by the abbreviation
I.B. M 136 175
Figure 3.--Cutting diagrams for specimens used in Study II for fire
performance and strength property tests. Three boards were made
with each fire-retardant treatmen-t. Tension perpendicuIar to the
surface (internal bond) specimens are indicated by the abbreviation
I.B. M 136 176
daJl
aJ
UU