P 3 .3~ i ~U
.P 3 .33110
U. S. Department of Agriculture, Forest Servic e
FOREST PRODUCTS LABORATOR Y
In cooperation with the University of Wisconsi n
MADISON, WISCONSIN
THE "FIREPROOFING" OF WOOD
By T . R. TRUA X
Senior Wood Technologis t
Published i n
PROCEEDINGS OF TH E
NATIONAL FIRE PROTECTION ASSOCIATIO N
May, 1931
'-t
THE "FIREPROOFING" OF WOOD 1
(Published in the 1931 Proceedings of th e
National Fire Protection Association )
By T . R . TRUAX, Senior Wood Technologis t
The recent revival of interest in the "fireproofing" of woo d
has come from many and widely divergent sources . Forest and lumber in dustries are directly concerned because of the general trend in th e
building industry away from wood toward so-called "fireproof" construc tion . Wood-preserving and chemical industries see in fireproofing a
possible field of enlarged activities . Architects, contractors, an d
builders, in the hope of finding cheaper and more adaptable fireproo f
building construction ; the fire underwriters, city building commis sions, and other governmental agencies, because of their concern ove r
fire hazards and the reduction of loss of life and property ; and the
consuming public, with a vision of greater safety and reduced losses ,
are all interested in the possibility of fireproofing wood . The For est Products Laboratory, in common with other governmental agencies ,
is concerned both with reducing the large, yearly losses of life an d
property from fire and with the bearing fireproofing has on the economical utilization of forests and forest products .
"Fireproof" Misleading
The terms "fireproof" and "fire--resistant" have come int o
rather general use to designate materials and constructions that var y
considerably in their susceptibility to damage by fire . No materia l
commonly employed in buildings and other structural fields is strictl y
"fireproof ." Likewise, the term "fireproofed wood" can not be take n
in a literal sense, for no treatment of wood, so far as is known, pro tects wood from destruction under high temperatures . "Fire-resistant "
or "fire-retardant" expresses more accurately the properties of woo d
treated with fireproofing chemicals, and the author has used the term s
"fireproof," "fireproofing," and "fireproofed" in this article in th e
more general sense because of established usage .
-Presented at the 35th Annual Meeting of the National Fire Protectio n
Association, in Toronto, May 11-1 L1, 1931 .
R9 .9
Competition Stimulates Fireproofing Processe s
I.
The use of wood in buildings is without question a factor i n
the yearly loss of property and life from fire, but on the other han d
there is danger of over-emphasis of the hazard of wood in certain de tails of construction . Statistics shoal that something like one-thir d
to one-half of the annual fire loss to property occur in frame buildings and their contents . Whether this loss is out of proportion t o
property values and the extent to which it could be reduced by the us e
of more fire-resistant materials are not known . Undoubtedly, the us e
of more fire-resistant materials would reduce these losses somewhat ,
but fire losses wouluk ill be large even if wood were entirely elimi ;'- ch as bette r
nat . tfiiiat
pt'i
r OONV
.er 1'11n-r ol"
s~aat ,. ]mss
spKimproved deter of construction ,
ktL i
tY7~
In
~os.a,hazardous . Nevertheless, a
&mile
Ik
to
maw
'als o_f construction, suc h
_i
as se~3, conch 3 tagitdt
& the more con.nd t
RR
r "ood has
gested Meas of- 'Camp
i
3
ire
is
a
growing
'uen,ncy
to exalre
i
~
to rural
d h a- ,
naturally
9901 .0r..eral '
0
_
berer~ .i
a
•r_ufaa
a
m
W
It*
tat
&~
fs.r
tap elealafte a sea •h for way s
a- can b e
swop" wean& 10 at*-ipplz
leyox~;d that 2 tbse
*Igo
viable or,
s ad k'r. uh. !1
method of
'
tow, if~1,~~ ~L *_ a
maobil& 'imply
isle
and effect rv
4
.
9&111L
r'
ar ds , bu t
ss inflemgeneral
'We of surThe two
required ,
Ibizod.
of woof are b ,
y~ s originated hundering materials for
fire resistance of
led to impres s
ugh 1 '~
should b e
origit 1 ' _
near] 4
*►aod aka
R_
1_~tf
*
are of more recen.countries go back
ercial treatment o f
_ua .
aken about 35
years ago, when the first real demand for fireproofed wood came fro m
the United States Navy for use in battleship construction . A few
years later, in 1899, the City of New York gave further impetus to th e
fireproofing industry by adopting a revised building code, which required that wood, used in the construction of buildings over 150 fee t
in height, must be treated to make it fireproof . Testing methods wer e
adopted by the Bureau of Buildings of the Borough of Manhattan t o
judge the effectiveness of treatments .
The early experience of the United States Navy with fire proofed wood was unsatisfactory, and its use was discontinued in 1902 ,
after a 7-year trial . The discontinuance is reported to have bee n
based, not on a lack of original fire resistance, but on the corrosiv e
action of the chemicals on metal fastenings, on their hygroscopi c
properties, their effect on paint and varnish coatings, and their lac k
of permanence on decks and other parts exposed to the weather or t o
frequent washing . In contrast to the experience of the United State s
Navy, it is interesting to note that the British and •Japanese governments are at present reported to be using fireproofed wood in the construction of naval ships . The early experience of the British Admiralt y
is reported to have been similar to that of our own Navy, but late r
developments brought about improvements so that fireproofed wood wa s
later adopted and has been used in British ships for some 15 or 20 years .
In New York City, where fireproofed wood is used principall y
for interior trim, flooring, and window sash of buildings, there ha s
apparently been no widespread or serious criticism of the treated woo d
for the reasons that made it unacceptable to the United States Navy .
However, the conditions in the two classes of uses are very differen t
and might account, at least in part, for this difference in results .
Slow Growth of the Industr y
On account of the demand created by the Navy and the Cit y
of New York, several companies were formed to treat wood with fire retardants . Some of these met early failure, while a few continued to operat e
for several years . The entire industry had a severe setback during th e
World War, but at its conclusion and also since then several new plant s
have come into existence . During recent years, chiefly because of extensive building in the New York area, there has been a slow growth i n
the industry, and it appears to be on a sounder basis than before .
There are several apparent causes for the comparatively slo w
growth of the industry and the failure of various companies, chief o f
which are the following :
1 . No general recognition of the value of fireproofe d
wood by engineers, building commissioners an d
R949
-3-
others exists, chiefly because of a lack o f
knowledge of its properties, and the absence o f
adequate standards .
2 . The use of objectionable chemicals, which affec t
adversely the properties of the treated wood .
3.
Inadequate treatments, because of carelessness o r
a lack of fundamental knowledge of treatin g
processes .
it- . The cost of treatment .
There is at the present time a more wholesome desire to lear n
the properties of fireproofed wood and to put out a satisfactory product .
It is becoming more generally recognized that accurate and reliabl e
testing methods must be found and clearly defined, that performanc e
standards must be set up, that the chemicals used must not be objection able, that the treatments must be thorough and properly made, and tha t
the cost of the treated wood must be reduced . In other words, the industry is apparently putting itself in a position to make better progres s
and to meet the demands of the consumer . Much work must be done, how ever, before the properties of fireproofed wood will be fully known an d
generally recognized .
j
Work at the Forest Products Laboratory
Fireproofing studies were undertaken at the Forest Product s
Laboratory about 1913, but the work was discontinued during the Worl d
War . Some of the results of this work were reported in 1915 . E
Th e
work was again taken up in 1927 and is still under way . Recent studie s
at the Laboratory have been confined chiefly to two lines of endeavor ,
a study of testing methods, and the relative effectiveness of variou s
chemicals as fire retardants in wood .
A brief statement of some of the fundamental characteristic s
of wood that are important from a fire-resistance standpoint is desirable here . Wood, in common with other organic materials, decompose s
when subjected to high temperatures and gives off inflammable gases .
Research in the fields of wood distillation and combustion indicate s
that the decomposition of wood subjected to increasing temperatures consists of the following steps : Vaporization of the moisture in the woo d
2 Tests on the inflammability of untreated wood and wood treated wit h
fire-retardant compounds, by R . E . Prince, National Fire Protectio n
Association Proceedings, Vol . 19 : pp .lOS-58 ; illus .
(1915) .
R949
L
(up to 212° F .) ; volatilization of extraneous materials (200° to 300 °
F . or higher) ; scorching and the slow evolution of inflammable gase s
(300° to 400° F .) ; charring with more rapid evolution of inflammabl e
gases, accompanied by glowing and eventual flaming (400° to 700° F .) ;
and quick ignition of inflammable gases and glowing of charcoal (700 °
It is important to note that the decomposition of woo d
to 950° F .) .
substance may occur at temperatures substantially below those prevailing in going fires .
Heat from an external source is necessary to ignite wood ,
but, once ignited, the exothermic heat, if not dissipated, is sufficient to carry on the process of combustion . The ignition temperatur e
and other fundamental decomposition characteristics of wood do no t
vary greatly with different species . Treating wood with fire-retardan t
chemicals increases the ignition temperatures somewhat but the orde r
of magnitude of this increase has been shown by various investigator s
to ba anly something like 200° F . with even the most effective chemicals . On the other hand, moisture or suitable chemicals injecte d
into the wood interfere materially with the process of combustion an d
if present in sufficient quantities may prevent it entirely unles s
suffict heat is applied continuously from an external source .
Present known fireproofing methods, however, do not prevent the de composition or charring of wood at high temperatures and in thi s
s
''ter riut be regarded as making wood fireproof .
Wood has good insulatim4 w
tOMBlumit heat readily and it thv1 th4. _
and Ala
other 4ilipprImeme
,-*en
!I
_ -•
%a.10 Om Omni
p
vlt$h
at it does no t
ical propertie s
IT' due t o
buckle, twist ,
ated steel an d
e also responsiadesirabl e
mperature s
-U eu s4
on p r
thu s
tnder4l.
■ it peon
e. _
ar d
vant
Wos
tendency
;
Mel ta,i
tur e s .
,F Wy al
t
.
160
;
.
Fire-fesisTance Tree;
Tests on the
;rou. ,ed into two cl
end tests on
i•1~
fire -
iw Or small pi
uil
.
of wo_ - '
p esentat i
time-tempe a
sive and ar e
nts nor t o
tests are in oximate fire
conditions, while tests on small samples are useful as measures of on e
or more of the properties of fireproofed wood and as measures of th e
effectiveness of treatments in plant operations .
Fire resistance in the broad sense embraces several propertie s
or characteristics of a material, such as heat insulation, the exten t
to which the material gives off inflammable gases when heated, resistance to the spread of flame, and to the building up of temperatures ,
resistance to flame penetration, and the effect of high temperatur e
on strength . Many attempts have been made to measure the fire resistance of wood through one or more of these properties .3- Some investigators have measured the volume and the character of gases evolve d
up on heating wood or the temperature and the time of ignition of 'h e
gases . So r eral 'tests, such as the shavings, crib, timber, splinter ,
muffle, cinder ; and het-plate, are based on the probability of ig_ :ition from a give :« heat source within a specified period or upon th e
persistence of f__oming and/or glowing of the wood after ignition an d
removal from the source of heat . Measuring the time required for a
flame to penetrate a piece of wood is a common test . The depth o f
flame penetration in a board, the unburned cro s-sect :oval area of a
timber, and the time required for a loaded timber to fail when subjected to a flame are different measures of the same essential property .
Heat insulation, which is closely associated with flame penetration ,
has been included in some flame-penetration tests . In other test s
the height of flame from the burning wood is taken as a measure o f
inflammability .
Most of the 7arious tests show differences in fire resistanc e
between untreated and fireproofed wood, but the results are often o f
limited value, because of many uncontrolled conditions and factors i n
the test and because of the lack of definite and significant units o f
measurement . For example, the density of the wood is perhaps as important a factor in flame penetration tests as is the presence o r
absence of fireproofing chemicals . Ignition temperature, of course, i s
related to fire hazard, but it fails to measure the behavior, of woo d
under fire conditions . Likewise, the time required for ignition an d
the tune of flaming and glowing a :ter ignition are affected by so man y
conditions that they are of doubtful value as accurate measures o f
fire resistance . None of these tests appear to measure in definit e
units the tendency of fireproofed wood to resist the spread of flam e
and the building up of temperature under fire conditions .
The Fire-Tube Apparatu s
In an attempt to develop a more accurate and significant tes t
that would overcome some of the limitations of past test methods a
3
-For a summary of testing methods se e ,Measuring Fire Resistance of Wood, "
by T . R . Truax, American Society of Mechanical Engineers' Transactions, Vol . 52 (No . 17), p) . 33-34 (May-Aug . 1930), illus .
R9 4 9
-6-
rsimLt -.to
m
tube apparatus was designed &Ie Forest Products Laboratory .- The
atus (Fig . 1) consists essentially of an open, galvanized-iro n
ich a specimen of wood is burned while suspended from a
Ulanc€llm . A gas flame is applied to approximately the lower quarte r
spec
,
4 minutes . To the balance arm is attached a
d.icates o• a scale the percentage loss in weight as th e
. The
t4•erature at the top of the tube is measure d
of pyrometer . As the wood is decompose d
caught in the bottom of the tub e
ow is noted .
11
*ontrolled so as to give a temperatur e
lured with a No . ILI- Awg. thermo and 356° ± 9° F. (1800 ± 5 °
the top of the tube without
btained with a Bunsen
ate74' 11 inches in heigh t
ed is about 50
,hes . The
about
sition with its low :r end 1 inc h
lion of the tube indicates tha t
}4ower end of the specimen is subfs■4
Ff'
ii +order o
0° F ., and the uppe r
ction i n
1p t t1out 375 0 tv
F
n;
tsnper
f
r
the to
~
up p er half of the
tpecimm . as # ! M *~
'~'7~
peratures rangin g
fro. timed, OlliSt-Ilad
106 approximatel y
f
~ $J IrtlTruil
the l tii¢'
WORM"' lider thee conditions untreate d
-e
of th e
11!e04 biro! Il.ickl -,
spreads
. • iiii .- • Ssm,
rapid7.,y~~
tureS i. •
350° to
L#4 1os 3.*
up
• e w an 3 o
minutes . O n
1$~
t ial of
ecimen ignites with difficulty ,
p during destruction, it does no t
if ma 414 fi g" 'fix
,
siting flame, and it causes littl e
Bet .ieiim
iammo
%ho
Or W tSMIPWW
f11
of the tube .
4'
n-weight and temper : a7UM
t'ff'► TIMMS At dusts in Oft Ilift.0edft et. wood of different degrees o f
fur.
plotted aglrf~l~ ILtle in test . These curves illustrat e
f7tained o• $l
iiual specimens ranging from untreate d
hut
,
gnated wall► natively large amounts of an effec ti
!1l~eretarda~f.
.
fire-tube apparatus se e
ance of Wood," by T . R .
ty for Testing Materials ,
illus .
8949
4
7:1
'Frith the fi
a. .aratus developed to the point where i t
was considered reason
a e and reliable and a useful measure o f
VIM resistance, test yJ ._ undertaken to determine the effectivenes s
Nit 4ous q icalrar
tin wood . The object of thes e
sa Mad cher
cl;};nct fireproofing
pP0110M
N►ad anel that
'axe 4kalit a'will be effective i n
flows ;
mod. EfteallOg4tfthe
work yi.ded
pa
' s7
*tensg!
-7e
9possibilities as fire
10 !Ma and e.
promising
D 0110sMassilr Oat*
,
7obtain a good degre e
lne this "OosdLO,
ecimens were impregsatid 'try'
yi* MIMI* k as* a
Wain complete penetratts* of t
a
-.
.-to
m
1 t friPe serve test pstitih.,
by 3/4 by 0
two car
raons of each chemical
tiarilm Ala Adze,
so air gtlvv eutaimPpriltins
'
tely pound and 5 pounds of
dry
1 roar gnftkia (few%
weli.,y
ies of effectiv e
fire*
to indicate the
pmu 6.t.11itI - a! the ck.idge4ili s4- *
:agent . The pieces
Team- died
treater
i b
dity and temperature
e
l
percent mois .
ti l
ftelli
hen tested in
Ali
~~ •~~e . 0%0110a sihme
ned with 33 chemi cals . tt 12 "°par
ire io m.ftep
"-arence ik
fir; 7,
sista,ncdr -Of '11■iy. &
PT64p d tilts
chemi
. that a
fez of 401000as stand$ fit above t' e
T
.
oh. ha:war
being ma* to learn
.r
more &Ali t
stop the spread o f
fi■mP-aml the bu
chemicals include d
thowed is c t
extensive
joldleir
s wou
expect to have
the
, to. int er e
is us e in service ,
zip. To 'ha
'boards, appro In tbi,af. gorge s
1aa t el
siftwigV-11
nated under
pressure with ruseiligim .4e
d to give a range of
absorptions from about 1 to 7 or ."'
Of dry chemical per cubi c
foot of wood . The boards wer e
treatment and were then
nd-ard #s t specimens, 3/S b
J inch"
• d tested. Figur e
tity of an effectiv e
oss in weight of th e
1 weight, and als o
40 4,iii. O.&4POW
fire-tub e
during test . The character oJ.
consistent relationship in two different k i
1
'
increase per minute are plotted against absorption of
wood . In both figures a general
although there is obviously con the-arious chemicals on the rat e
0 percent o th e
. 41 re
aide
NIP.
tfi o
- ch
1W
Pq.
r
'gha t
ermo
mpr
fix ,r
sta
081
ac t i c* cl
e ni, .
blaract4tililip o r
fire safety . The con
+1
of flame are other properties o f
ttion tom_ the fire--ttbe tes t tdoe y
f-IfIRIMpait
e
gluing
Iteiiie
*0*10i
there
the ..
$
fire.
.tube apparatus i n
measuring the tendenc y
t .. an be express e
e tendenc y
up of temperatu4*
m the standpoin' df
hrough proper
can be made suff,ilig
fireproo f
fire-tube
es all th e
od from - e standp oint o f
esistance to penetration k
n fireproof construe . Still other proeiesr
ode metals, finishing an g ,
of treatment, and effec t
rent chemicals i n
bility of such _
y:
of
re by
testing .
~lar
ore th e
wn. and appreciated, i t
ilities in reducing th e
is of building construe erous structural
would have a
da fI.
and.,
sustai
t.■
most of the loss of
this direction,
tea
ti
in th e
telly have a mos t
mmercial construction, wher e
fers unusual opportunities i n
.
•p . - ..~.. L
A broad at tac
pS
sary . Requirin g
sive fireproof construction in buildings that can not under any
9
R 4+9
Per
(The published
of the annual meeting of the National Fir e
Protection Association contain Mr . Truax t s paper, together with
the discussion that followed its presentation . Reprints of the
paper and the discussion may be obtained from the Associatio n
Secretary at 60 Batterymarch St ., Boston, Mass ., for 20¢ each .)
uI
T
I1IIR
i!!
-
_
'
~~
1
1
1
'
'
, 1
II I 1
1
1
~I
y
R
5~
R
r
is
0~
w
M
M11,
■r
a
=MIMI
°
1N3a t
s7
ie
3m1 N I01 m
go
a
3da1IY)1k)J nrnM/2VW
UM =I
..
.
._
-.
alKj
■ • ■■■■■■■1
N
■■■■~■■■■■■■f
,
E
■.■■■
-_ ■
ip
nn.■■ ~°
~■
■■
■■■
.■■
■■■■
0
~
WWi
W
opQR
a
•
y
r
4
¢
a~y~ a
wt
a F aW a
y ~aC C
-~
F
EoWO a
E
F ZOZ i
4
O
I
~
0 •O■
o
0
0
0
o
o
n
I' .
yH
QR
d
'
•
n
0
•
•
•
~
E
C31nN/N/
m
'0
f
$
V
a
n
Mad 1N39 M3d3 9N/ldV7d ON/Y/70 1N9/3M N/ S3`07 dO 31VM
li .
0
_
-P
.
- No
-
•
-
i
Am
t
yU ~
4tii '~ a
4e p
;
W
4
Jz anE
qJ xT
'•
lh
e .o :o■
a
_
■
•
(31/11/W b'3d 30F/D'0/1137 13311500 3SV7N9N/ Id01VM3dW31 d0 31V 8
J
-
PI
0
E
i fl