AND

FIRE OFFICES' COMM ITTEE

JOINT FIRE RESEARCH ORGANIZATION

•
FI RE RESEARCH NOTE

NO. 586

THE IGNITION OF WET AND DRY WOOD BY RADIATION

by.

D. L. SlMMSand MARGARET LAW

This report has not been published and

should be considered as confidential
advance information. No raferance shoulcI
be made to it in any publication
Without the written consent of the
Diractor of Fire Research.

April. 1965.
Fira Research Station.
. Boreham Wood.
Herts.
('phone ELStrlZe 1341)
© BRE Trust (UK) Permission is granted for personal noncommercial research use. Citation of the work is allowed and encouraged.
 o

...

.,
 F. R. Note No. 586

MINISTRY OF TECHNOLOGY
AND
FIRE OFFICES' COMMITTEE
JOINT FIRE RESEARCH ORGANIZATION

THE IGNITION OF WET AND DRY WOOD llY RADIATION

by

.. D. L. Simms and Margaret Law

SUMIlARY

The effect of varying the moisture content an bath the pilot and
spontaneous ignition times of different woods of areas 7.6 cm square and 15 cm
square has been measured over a wide range of intensities of radiation. Moisture
increases the energy required for ignition; it also increases the minimum
intensity for ignition though with pilot ignition its effect is only marked for
moisture contents above. 40 per cent. .

Results have been correlated on the assumption that the material is inert
and ignites at a fixed temperature. Simple heat transfer theory has been used
to calculate this temperature with values for the thermal properties appropriate
to the given moisture content and with allowance made within the term for thermal
capaci ty for removing the water; the effect'} '(f moisture migration have been
neglected, fbllowing the results of Williams'7).
oC,
:Fer pilot ignition the correlating temperature is found to be 380
corresponding to a critical intensity of 0.31 cal cm-2s-1, except for fibre
insulating board which appears to ignite at a somewhat lower temperature of
330°C. Earlier experiments, with smaller specimens, gave a similar result, of
360 oC, with the results for fibre insulating board included. The present
correlation extends to much longer times (up to 59 min). The results show that
the chOice of 0.3 cal cm-2s-1 as the maximum accreptable level of radiation for
building regulation purposes, gives a larger margin of safety than was originally
~.. thought.

For spontaneous ignition the correlating temperature is found to bo 545°C,

f- the same as found previously for smaller areas, corresponding to a critical
intensity of 0.74 cal cm-2s~1. The present correlation extends to much longer
times (up to 16 min) and the results suggest that the empirical correction
necessary for the area effects is linked with dtms:l.ty •

•
 TIlE IGNITION OF WET A1ID DRY WOOD BY RADIATION

by

Do L, Simms and Margaret Law

1• Introduction
•. It is important to lmow how moisture content can affect the behaviour
of wood in a fire and experiments have been carried out to investigate the
ignition of woods with varying moisture contents exposed to a wide range of
intensities of radiation o For oven dry wood both pilot and spont~eo~s
igilition occur when the heated surface reaches a fixed temperature\.1 ;2) and
-,
the ignition time can be calculated using the values of the thermal properties
of the dry wood, which is assumed to be inert. This note discusses the effect
of moisture on the thermal properties and shows how an allowance for this effect
enables the model assumed for dry wood to be applied to wet wood without further
modification o It then estimates how moisture can be expected to affect the
igilition of wood in practical conditions.
2. Experimental method and results

Since it is now known th~t the spontaneous igilition time decreases as

the area irradiated increases\.2) it is desirable to experiment with as large a
specimen as possible but clearly the size chosen is limited by the size of the
radiaiii \In source
0 For the se experim.ents the source was a 30 em s quare radiant
panel D) , and the spe cimens of wood could be no larger than 7.6 em square i f they
were to be uniformly irradiated for intensities. as high as t.e
cal cm-2s-1.
This size gives results close to those for an infiilite area\.2). A few results
from ano.ther series of experiments were already available for specimens 15 cm
square but these had been exposed. to lower intensities of radiation.

The specimens used in the pilot igilition experiments were also 7.6. cm square,
and these, as for spontaneous ignition, should give results close to those for an
infinite area.
The specimens of wood were oven dried and then conditioned to moisture
contents of approximately 20, 40 and 60 per cent of oven-dry weight by
repeatedly dipping t)J.em in water and allowing the water to sink: in. This
undoubtedly leaches out some vol~t:l-le material but the effect on the igilition
time is unlikely to be importantl4). The densities of the woods were obtained
from their volumes and weights when oven dried at 95°C Variations in density
0

between individual specimens of the same type of wood were found to be small.

-- The details of the specimens 'are given in Table 1 for the spontaneous
igilition and Table 2 for the pilot ignition experiments.
The experimental procedure was similar to that described earlier(5) except
that a modified water cooled Moll thermopile was used to measure radiation
instead of the Thwing type radiation pyrometer. The pilot flame was put.as
near as possible to the top front edge of the irradi~t~d surface, (Fig. 1.),
since this position gives the short.est igilition time \ 2) • Specimens were
exposed to radiation from the panel until they Lgrri.ted , If ignition occurr-ed,
the next specimen was exposed to a lower intensity of radiation until eventually
it was reasonable to assume that igilition would not occur, i.e •.until the rate
of production of volatiles became very low. With both pilot and spontaneous
igilition, flame could often be seen to appear in the volatile stream and then
flash down to the wood where it persisted on the surface With a few specimens
0

during pilot igilition, there was intermittent flashing between the pilot flame
and the surface for several seconds before flaming finally persisted, and with
a few others, the flame reached the wood but was intermittent before becoming
established o For these the ignition time was recorded as the first appearance
of flame. Specimens of European whitewood, other than those that were oven-
dry, 'behaved differently from the other woods; the moisture emerged from
the surface in small jets of steam. Sealing the edges of the larger specimens
appeared to have little effect on the ignition time. The results for pilot
ignition are shown in Fig. 2. The results for spontaneous ignition are given
in Table 1. The minimum intensities at w~ch ignition took place are shown in
Figs (3) and (4).

Earlier experiments on the effect of mo~sture content on spontaneous ignition

were carried out with smaller specimens 5 cm x 5 cm in area. The procedure was the
same except that the moisture content was controlled by storing the specimens over (6)
saturated solutions of various salis. The results are reported in detail elsewhere
and are compared with the pres~nt results in a later section of this paper.

Experiments had also been carried out on the pilot ignition of dry specimens,
5 cm x 5 cm in area and these results(1) are also discussed later.

/
!

2 •.
. i
 Table 1
Details of woods used' and experimental
results for spontaneous ignition

-.
Area Density Moisture Intensity of Ignition
Thickness
Wood oven dry content radiation time
cm 2 cm -dcm3 -% -cal cm-2 s -1 - s
"
-. Fibre insulating
board 7.6 x 7.6 1.3 0.~5 . oven-dry 1.00. 25
..
0.97 32
0.95 -
,
20 1.50 15
1.30 21
1 .25 33
1022 28
1021 33
1.20 ~

40 1.50 27
1 .40 30
1.30 37
1 .27 -
60 1.45 45
, 1.45 -
Columbian Pine T.6 x 7.6 1 .9 0.4 oven-dry 1020 53
(Pseudosuga 1.12 70
Taxifolia) 1 .10 70
1 .10 -
20 1.50 37
1.20 88
1.20 64
, 1.17 -.
40 1.50 47
1.45 68
1.43 55
1 .41 45
~-.
1.40 -
60 1.60 45
.... 1.58 50
1.55 -
Oak 7.6 x 7.6 1.9 0.7 oven-dry 1 .80 75
(Quercus sp. ) , 1 _20 93
1.16 -
20 1.50 62
1.40 72
/ - 1 .30 105
/
1 .28 -
, -
40 1.60 74
, 84
- 1.55
,, 1.52 80
1.50 -
60 1.60 96
1.58 94
1.57 119
1.55 •.
- 3 -
 Table 1 continued

I Area Thickness Density Moisture Intensi ty of' Ignition

Wood oyen dry content radiation time
cm 2 cm -gjcm3 -% -cal cm-2 s -1 - s

European 7.6 x 7.6 1.9 0.45 oven-dry 1035 45

Whitewood 1.3.0 48
(Picea Abies) 1.22 -
20 1.50 52 .-
1 045 56 -«, -~
1.4.2 55
1.40 -
40 1.60 50
1.58 -
60, 1.70 45
1.68 51
1.65 61
1 .62 -
..
1.60 -
15 x 15 2.5 0.5 oven-dry 1.1 79
- 10 1 .1 650
1.cO 1 .1 100,
0.6 oven-dry 1 .1 521"
Larch 15 x 15 2.5 0.5 oven-dry 1.0 130
(Larix Decidua)

Abura 15 x 15 2.5 0.5 oven-dry 1 .0 680

(Mitragyna spv.) " 360
" 8401

Makore' 15 x 15 2.5 0.6 oven-dry 1 .0 890

(Mimusops II
10~
Heckelii) " 12
1~~
II.

"
,

~
Baltic redwood 15 x 15 2.5 0.7 oven..,dry 0.7
(Pinus sylvestriS:) "
0.9
1 .1 r¢
. ".
I Edges and back surface' painted with flame retardant paint.

- 4 ..,
 Table 2
Details of woods used in pilot ignition experiments
and range of experimental results
Area = 7.6 x 7.6 em 2

Density Moisture Range ofintensities Range of ignition

Wood Thickness
gjcm3 content cal cm-2s-1 times
em % - s

Fibre 0.25 1.3 dry 0.25 - 0.40 610 - 84

insulating 20 0.32 - 0.50 555 - 80
board 40 0.34 - 0.35 535
60 0.37 - 0.45 465 - 365
;. '''t ,~

l' .9 dry 0.20 - 0.30 1440 - 215

20 0.25 1010
40 0.24 - 0.30 2540 - 760
Oak 0.66 1.3 dry 0.38 - 0.50 415 - 140
(Quercus sp. 20 0.55 605
40 0.59 - 0.65 635 - 530
" 60 0.69 - 0.75 510 - 435
0.80 1'.9 ' dry 0.38 - 0.40 1260 - 1115
, 20 0.50 - 0.55 1020 - 630
40 0.41 - 0.45 2580 - 2020
,
Columbian 0.46 1.3 dry 0.46 - 0.50 430 - 160
Pine 20 0.54 - 0.55 460 - 500
( Pseudo auaa 40 0.63 - 0.70 380 - 180
Taxifcili9,) 60 0.74 - 0.76 310 - 140
0.71 1.9 dry 0.39 - 0.40 2130 - 1440
20 0.43 - 0.45 1940 -1770
40 0.39 - 0.41 3540 - 2230
European 0.46 1.3 dry 0.45 - 0.50 240 - 180
Whitewood 20 0.52 - 0.60 610 - 370
(Picea abies) 40 0.57 - 0.60 550 - 260
60 0.67 - 0.70 350 - 270
,
1.9 dry 0.37 - 0.39 2380 - 1520
20 0.40 - 0.52 1800 .., 300
40 0.42 - 0.50 1520 - 530

. ~.

- 5 -
3. Discussion of results

The effect of moisture is to increase the ignition time and energy and the
minimum intensity for both spontaneous and pilot ignition. For example, the
intensity of radiation required for the pilot ignition of dry wood is about half
that for wood with a moisture content of 60 fer cent. Similarly, at the same
intensity of radiation, (say 0.5 cal cm-2s-1) the pilot ignition time may be
increased from 2~ min for oven~dry oak, to 24 min for oak at a moisture content of
40 per cent. The minimum intensity required for pilot ignition in these experi-
ments with 7.6 em square specimens appears to be independent of the density of the
wood except for fibre insulating board which as before(1) is much easier to ignite
(Fig. 3). There appears to be a sharp increase in the minimum intensity between .'
40 and 60 per cent moisture content.

The increased size of the specimens used for the experiments on spontaneous
ignition has resulted in the minimum intensity for ignition being lowered with a
consequent increase in the times taken to ignite. For example dry wood 2.5 em
square ignites in about 30 s at a minimum intensity of about 1.2 cal cm- 2s-1,
7.6 em square in: about 70 s at 1.1 cal cm-2s-1 and 15 em square in about 800 s
at 1.0 cal cm-2s-1•

4. Discussion of the effects of moisture

Heat transfer within the wood affects its temperatureorise and hence its
ignition time and the presence of moisture increa.ses the ignition time by
changing the heat transfer in at least three ways:

(1) moisture increases the values of the thermal properties Le. the therma.l
conductivity and the VOlumetric specific heat;

(2) heat is transferred directly by molecular diffusion of the water;

(3) evaporation cools the hotter regions and condensation heats the
cooler regions.

Water vapour in the atmosphere is an inerting gas, but the effect of this is
negligibly small compared with its effect on the moisture content of materials.

Williams(7) measured the temperature-time profiles in oven-dried woods, and

in woods of different moisture contents of up to 30 per cent at intensities of
irradiation between 2.0 and 3.1 cal cm-2s-1, using a graphite panel furnace.
He found, as expected, a plateau near 1000C, the duration of which increased with
increasing depth; this he suggested was due to the lower rates of heat conduction
and consequently smaller temperature gra.dients. On the assumption that the water
converted the wood into an opaque solid, the dry wood being diathermanous, Williams
found. it possible within a wide scatter to correlate his r esu.Lt s , for the temperature
rise, prior to local exothermic heating, using the thermal properties corresponding
to the appropriate moisture contents and making an allowance for the desorption of
water. This suggested that the temperature histories, both near the surface and ~
within the body of the wood, were unaffected by moisture migration. Williams
thought, however, that the heat released when migrating steam from the surface
condensed in the interior was nearly enough. to vaporise the liquid present there, '.
and that it was for this reason he was able to neglect its effect.

Gardon(8) also found little effect due to moisture migration and correlated
his results using a similar technique to Williams. Gardon pointed out that this
kind of correlation is based on calculating the extra enthalpy required to raise
the temperature of the wet wood by the same amount as the dry wood, Le. on
differences, and so that the method is not a sensitive one.

Williams(7) also calculated the rate at which the zone of vaporization moved
into the solid, by's method based on the work of Von Neumann•.; this divided the
solid into
. two regiorts, separated by an isothermal plane, the 100 0C plateau, the
plane of vaporisation. The rate at which this plane moves at any depthJC was
assumed to depenn only upon the net rate of heat transfer by conduction to that
depth. This led to the following trial equation for the depth X. 'of the plane
at time t
.See for example, Crank J., The ma.thematd.cs of diffusion, Oxford, 1954.
- 6 -
 i
(£1)2 ierfc x = constant (1)
K 2 (kt)i

where is the thfjerma~ffusivity

k
2~1
, ~d ierf~e = --t e-
z2
dz J.
, ,
(-n) /3 ~
0C
When Williams analysed his data on the times at which the 100 plateau
reached a given depth, he found that equation (1) f.r~ted trnnreasonably well;
this was additional evidence that the effects of moisture migration might be
neglected. A further set of experiments carried out using woods containing
220 per cent moisture gave similar results except that no desorption effect
had to be assuaed; p6ssibly because the wood was saturated.

Fons (9) has suggested a simple means of allowing for the effects of
moisture on ignition on this basis. This enables three effects, the change
in the value of the thermal constants, the heat of wetting and the latent
heat to be, considered within the term for specific heat, viz.

c = Co +.AW +. 0,01(L +. go) K'

m
gIl
(10) -1-1
Where c is th~ specific heat of the dry wood = 0.34 cal g degC
o
and c is the specific heat of the wet wood
m
K is' the moisture content expressed as, a percentage of the dry weight.

Aw is the heating ofwetting(11) = 16cal/g,

L is, the latent heat of steam = 540 ca~g

0C,
go is the temperature'rise from ambient to 100 at which temperature
all the water is assumed to evaporate

IIp is the surface temperature at ignition

The effect of moisture oontent on the thermal conductivity K, of moist

wood has been measured and is given bye 12)

Km = 10-'4 jpo(4.78 +. 10.? m) +. 0.5J7/~ ,., •..•. (3)

K
where m = 100
f,o is density of dry wood
and suffix m denotes value when the moisture content is K.
Its effect on densi ty,(O, can be estimated from the method of mixtures
neglecting any chang~' in volume.

i.e. . • • ... .• (4)

The variations in: bo'th K and rdue to moisture are smaller than the variation
in 0, so that for simplicity the values for the wet wood can be assumed to
apply throughout the heating period up to' the time of ignition, although
striotly, once the water has been driven off the values for dry wood apply
and a we:i:g~ted average for the heating period should be taken.

- 7-

. \'
 The ignition time, t, of oven-dry Viood in the form of a semi-infinite
solid, aasumed to be inert., opaque and totally abaor-b.Lng. irradiated on one face
by intenSity,(I~ and losing heat from that face by Newtonian cooling m~ ~e
obtained from. 2)

It
1
= (3/ (1 - erpr 2 erfcf) • •• • • • . . .. (5)
fO (kt)"2 Q.p

H (kt)!
where (3:= K

Ht
= 1
the cooling modulus
fC (kt)2,.
,

and H is the Newtonian cooling coefficient corresponding to QF'

By substituting values of Km,f'm and cm given by equations (2), (3) and (4)
and inserting the experimental values of I and t in equation (5) we shall
attempt to find a value of OF. to give a correlation.

5. Correlation of results

5.1. Pilot ignition

Previous results for du wood, with the pilot flame in a nearly identical
position, were oorrelated(1) using the following values, 9F - 340 degC, H = 8 x 10-4
cal cm-2a-1 deg c-1 , Co := 0.34 cal g-1 deg C-1 o The correlation was satisfaotory
except for 1fibre insulating board at low values of I, i.e. at high values of
Ht~c (kt)2 where the points l~ below the line. The results for the new series
of~oven-dried materials are shown in Fig. 5; the~e.is a reasonably close fit
between the experimental points and aqua tion (5) using the same constants but
a better correlatiqn is· obtained by ignoring the fibre insulating board results
for high t/pc (kt)'2 and using GF = 360 deg C, H = 8.6 x 10-4 cal cm-2s-1 deg C-1•
The experimental results for the specimens of different moisture contents and
densities are plotte~in Fig. 6 using equation (5) and ~ =360 deg Co There is
reasonable agreement between the experimental points and equatio"n (5) except for
fibre insulating board, for which a better correlation is obtained by using
~ := 310 deg C, H = 7.4 x 10-4 cal cm-2s-1 deg C-1 • This leads to a value for
the oritical intensity of wood (1 0 = H~) of 0 ..31 cal cm-2s-1 and for fibre
insul8.ting board of 0.23 cal cm-2s-1•
The correlation shown in Fig. 6 appears to account for both the effects of
different moisture contents and of different densities on the pilot ignition
time of the woods. The results for fibre insulating board are a little below
those for wood, as was found earlier(1), but th~ residual effect associated with
density for the remaining woods found earlier( 1 ) with smaller specimens is absent
here. Addi tionally, although the ignition times are much longer tha.n thOse
studied previously ~ nearly an hour in one experiment and several specimens took
about 40 minutes to ignite - there is ncr sign of a trend away from the correlation
even at the highest value~ of Ht/pC(~t}2. This suggests that any effect due to:
self-heating (an earlier e'ip*anationl 1)) is small even for fibre insulating board.

8 -
 The temperature found to correlate this and the earlier results is similar
and this confirms that any effect of area on pilot ignition time is small,
but since the residual effect associated with density has been removed, it
also suggests that the effect cf area on pilot ignition time as well as on
the minimum intensity for pilot ignition may be associated with the density
of the wood. The rise in the value of the correlating temperature from
340 deg C to 360 deg C is due to the exclusion of the results for fibre
insulating board.

Assuming that the ambient temperature is 200C, the surface temperature

at ignition is 380 0C for the species of wood tested and 330 0C for the fibre
-. insulating board.

5.2. Spontaneous ignition

Previous results(2) for dry wood were correlated using the following
values, ~ = 525 deg C, H = 1.4 x 10-3 cal cm-2s-1 deg c-1 , Co = 0.34
cal g-1 deg C-1. The results for the new series of oven dried woods
7.6,cm square and 15 cm square are shown in Fig. 7. There is a reasonably
close fit between the experimental points and e quation (5) using the same
constants, except for the results for Baltic Redwood which ignited for' very
low intensities. Since there appeared to be no difference between the
results for the other species, these have not been plotted separately.
The results for the 7.6 cm S~lare specimens of different moisture contents
and densities are also plotted in Fig. 7 and there is again a reasonably
close fit between the experimental results and equation (5), although the
results tend to lie above the curve at long times and below it at short
times. '

Previous'results(6) for woo~ cut 5 cm square with densities ranging

from 0.24 to 0.66 gfcm3 and moistu.re contents ranging f'tom a to nearly 30
per oent and covering a range' of inteIlilities of radiation from 1.5 to 2.4
cal cm-2s-1 have been analysed in a similar way, The best fit :!-s obtained_
wi th a value of 9F of 525 deg C and for H of 1.4 x 10-3 cal cm-2 deg C-1s 1
and the results follow the trend of the curve satisfactorily (Fig 8).
However, although the 'effect of moisture content has been absorbed by the
correlation, the effect of density has not ; the denser wood,s lie below
the lighter ones, i.e. they appear to ~g~te at a lower temperature.
This is a similar effect to that found~1) for pilot ignition of specimens
of the same size, viz, 5 cm square.
No correction to the igmtion times for the effect of area(2) has been
applied to these results. Nonetheless, the fact that the residual effect
due to density is similar to though somewhat larger than that found for
pilot ignition with specimens 5 cm squ.are and that it is not found with
either form of ignition with those 7.6 cm square suggests that the area
effect, does depend on density. There is some evidence of a systematic
effect of density in the correlations for spontaneous ignition of thick
solids in the data given in reference (2);! there 'the majority of the
data had been. corrected for the small area' of irradiation using an
average correction factor irrespective of density,' Assuming that the
ambient temperature is 200 e the surface temperature at ignition is 545°C
for the speoies of wood tested. -.

6. APPLICATION 01 RESULTS

6.1. Incidence of fires

The rate of heat transfer from burning fabrics(13) and small\petrol

flames(14) and from flames of similar size, such as matches, is about
0.5 cal cm-2s-1" and the time for which this heat output is maintained is
usually short. For a short ignition time the intensity mu.st be well
above the minimum and thus the present experiments show that for a large
proportion of the sources of ignition responsible for starting fires both
indoors and outdoors, the rate of heating is only just sufficient to ignite
thick cellulosic materials in the presence of a small flame. Hence, a

- 9 -

\.
decrease in moisture content of about 5 or 10 per cent in a wood, resulting in a
similar decrease in the total amount of heat required for pilot ignition; could
increase the probability of fire occurring. Further, the moisture content of
fabrics exposed to the direct radiatipn from the sun .may be lower( 16) than that
expected from the temperature and humidity of the surrounding air because they
are likely to be hotter. The effect would probably be less marked with thicker
materials because the moisture diffusion time 'constant is longer, but this is a
possible physical explanation for the variation of the number of fires with the
hour-s of sunshine (15}.

6.2; Safe separation of bUildings

I~ d~vising safe separation distances from burning bUildingS(17) and timber

stacks~18) in order to reduce the risk of fire spreading to neighbouring property
the critical intensity for pilot ignition of wood was taken as 0.3 cal cm- 2s-1;,
(I = H9F = 0.31 cal cm- 2s-1). In none of the present results did pilot ignition
occur in less than 20 minutes at an intensity of radiation less than 0.37 cal
cm- 2s-1 • Thus, the lowest value at which t he pilot ignition is likely to occur
is above 0.35 cal cm-2s-1. Even at an intensity of radiation of 0.4 cal cm- 2s-1,
the ignition time of dry w;ood will be at least 5 minutes and at the lowest likely
moisture content of 10 per cent, the ignition time will be 10 minutes. Such an
intensity of radiation near the extreme distance at which ignition may occur will
only be reached some time after the outbreak of the fire: the fire brigade should
therefore have ample time to arrive and protect the exposed property~ Notwithstand-
ing this, the dry J!!.8terial presents the greatest danger and regulations have been
based on the hazard of the dry rather- than the wet material.

7. Conclusions ,/
Increasing the moisture content increases the time for both pilot ignition
and spontaneous ignition for any given intensity of radiation. The increase has
been accounted for satisfactorily by using the values of the thermal properties
appropriate to the different moisture contents, and allowing for the effect of
heat of wetting and latent heat of evaporation within the term for specific heat,
as in equation (2). This confirms Williams' calculations which showed that it
was possible to neglect the effeot of moisture migration on the temperature rise.

The pilot ignition time for the positi,on of the pilot flame shown in fig 1
may be oalculated using equation (56 assuming a fixed surface temperature
criterion of 380 0C for wood and 330 C for fibre insulating board, This
oorrelating temperature for pilot ignition is the same for speoimens 5 om and
7.6 om square and 1his suggests that the effect of area on pilot ignition time
is small. The new correlation also appears to be more satisfactory than that
found earlier(1) since there is no residual effect linked with density, and
although fibre insulating board, the lightest material, again appears to ignite
at a slightly l,ower temperature, there is little sign of any systematic departure
from the curve at high values of Ht which wo.uld denote self-heating,
',' i
pd(kt)2
Moisture contents of up to about 40 per cent by weight appear to ha~e little
effect on the minimum intensity at Which ignition occurs, although there mSii' be
an increase above this level, but the energy required for ignition inoreases
markedly for all moisture oontents, The rate of heating from, and the energy
contained in most small souroes, is about the minimum required for ignition so
that even small fluctuations in moisture content may affect the number of fires
started. The experimental results also i suggest that there is an amply safe
margin in the"choioe of 0.3 cal om-2s-:1 as the maximum acceptable level of
radiation used in the Building Regulations to determine' separation of buildings.

The spontaneous ignition time can a Lso be calculated from equation (5)
assuming a fixed surfaoe temperature critenon of 5450C. This correlating
.temperature for the larger specimens (7 cm and 15 om square) is about the same
as that given ,earlier for smaller specimens and the range of ignition times for
whioh it is applicable has been extendec. to at least 130s. However, for the
results of the smaller specimens, '5 om square, analysed :In this paper, there is
a residual effeot due to density similar to that found for pilot ignition for
the same size of speoimen which suggests that the area correction factor is
assooiated with the density both for ignition time and for minimum intensity
of ignition.
- 10 -
 References

SIMMS, D. L. On the pilot ignition of wood by radiation, Combust. and

Flame, 1963, 7, (3) 253-261.

SIMMS, D. L. Ignition of cellulosic materials by radiation, 293-300;

Combust 0 and Flame 1960, ~ (4)0

SIMMS, Do L. and COlLEY, J. Eo Radiation characteristics of a gas-

fired panel. Brito J. Appl. Phys' 14, 292-294, 19630 .
-.
SIMMS, D. L. and ROBERTS, Valerie Eo Effect of prolonged heating on
the subsequent spontaneous ignition of oako J. Wood Sci. 1960 (5)
29-370

LAWSON, Do 1. and SIMMS, Do L. The ignition of wood by radiation.

Brit. J; Applo Physo 19523, (9) (288-292).

(6) THOJ4AS, Po R., SIMMS, Do L o and LA'jV, Margaret. The effect of moisture
content on the spontaneous igqition of wood by radiation. Joint Fire
Research Organization.F.R. Note No. 280/19550

WILLIAMS, G. Co Damage initiation in organic materials exposed to high

intensity thermal radiation. .Fuels Research Laboratory, Massachusetjs
Institute of Technology Technical Report No., 2, Cambridge, Massachusetts,
1953.

(8) GARDON, R, Temperatures obtained in wood exposed to high intensity

thermal radiation. Fuels Research.LabDratory, Massachusetts Institute
of TechnolollY !echnical Report No o 3. ·Cambridge, Massachusetts,. 1953.

(9) FONS, Wallace L, AnalYl!.is of fire spread in light forest fuels.

J 0 Apjric, Res. 1946 72 (3) 93-121.

(10) DUNLAP, Fo The specific heat of wo.od.U.S.Depto of Agrigulture.

Forest Service Bulletin No. 110 Washington 1912.

KRU'IT, R" and MODDERMAN, J. G. Heats of adsorption and wetting.

International Critical Tables, Vol. 5, p , 143.

(12) MACLEAN, s , D. The thermal conductivity of wood. Trans. Amer. Soc.

Heat Vent. Engrs o 1,941.!:il. 1.1840

WEBSTER, C. T., w;RAIGHT, a, and THOMAS, Po H. Heat trans fer from

burning fabrics. J 0 Text. Inst. 1962, (53) No.1 T29-37.

(14) SIMMS, D. L. and HINKLEY, P. L. Protective clothing against Flames

and Heat, Fire Research Special Report No. 30 London, H;M;S.Oo
1960.

(15) HOGG, Jane M. The relationship between fire incidence and climato-·
logical variations 1951-1961 Joint Fire Research Organization;
F.R. Note No. 522.

(16) HOLMES; F. H. Shirley Research Institute (:E'rivate communication).

(17) LAW, Margaret. Heat radiation from fires and building separation.
Fire Research Technical Paper No.5. London, HoMoS.O. 1963.

(18) LAW, Margaret. Spacing from timbf,lr Jrtacks to reduce fire spread.
IoF.E. Quarterly 196312 (49) 68-71.

- 11 -
 : ..
~'.

\il 1·25cm

1·25cm

Specimen

FIG.1. POSITION
.~
OF PILOT FLAME

o
on
t::
1.:

:>
n
D
~
. ..
- -.
., .'
"

0·8
Densit y - 9 Icm J 0'66 0·79
Thickness - c m 1 '3 1 ·9

0·7 <.
• v ....
Moisturrz content
per- cent
dry 0
•
~ <, 20 /)"
•
,•
III)
0 40 0
';IE 0·6 o ....
u·
<, 60 V
u
o
\...
A
~. A

-,
I 0·5
z
o
~
o 0·4 ... --- - -
~
Q:

u,
~ • "

o
>- 0·3
l-
(/)
Z
W
t-
z 0.2

0·1

o 10 20 30 40
IGNITION TIME (t) -min

FIG.2a. PILOT IGNITION OF OAK

 -...Jc- r;
I , U I I"'.;.JOO
"/ .' ':. .' ', , "

o· n

oensl ty 9 I em 3 0'47 0"47

Thickness - cm 1·3 1·9
Moisture content ,;. ..
"." "

-
"

o·7 Der cent

~ dry
20
0
6.
•
A
o·6 <, A
40 0
•
.~
60 V ~
o
~
u
o·5 .
'"'
~~. ~'~
z
0
I-
<{
Cl
<t:
o· 4
' ~

-- '
t:J:: ::!::
• -;
cr
LL
0 O· 3
>-
l-
V)
Z
W O· 2
I-
Z

O· 1

o 10 20 30
IGNITION TIME (t) - min

FIG.2b. PILOT IGNITION OF EUROPEAN' WHITEWOOD

1/6154 FR.586
'. , 'I.
, , .' . \

0-8
Dlll'lsity g/crrf

.>
0-46 0·71
Thlckness>- em 1·3 1·9
0- 7 Moisture content

~t - -
per- oent
dry 0
•
"iIII O. 6
20 fj,
•
~E
u
~ r-,
40
60
0
'V
••
o
u 0.5 ... ,....
I ~
z =-- A
o
I-
<! 0'4
o
- • -
• ,-
<!
a::
IJ.
o 0'3
>-
l-
V>'
Z 0·2
W
I-
Z

0·1

o
.

1) 21J ;jU 40 ' 5'0 61

IGNITlON TIME (t) - min

f"1 G. 2c. PILOT IGNITION OF COLUM BIA N PINE

1 /6153 F.R .586 .., ' .....
. . s,

I
I . 3
Density - 9 l ctt; 0·25 0·25
Thickness -cm 1· 3 1 ·9
Till
0'5
.... Moisture content

,\ -.
.- per oant
';'E
u
0 .
dry
20
0
t::"
•
A
u 0-4 40 0
•
~~t----
~ I
z , 60 V y
0
~
0

,.
"
~ 0·3
"

\ 0
<i
0:
-
~~-

.' -P-- ~
...
-• -
IJ..
0 0·2
-
>-
I-
-
If)
Z 0·1
W
I-
Z

o 10 20 .30 40
IGNITION TIME ( t ) - min

FIG.2d. PILOT IGNITION OF FIBRE INSULATING BOARD

 "t\ll 0·8
(J
';'IE
u )
'E'
u 0- 6 ./
z
o
1=
I

z 0- A
o
(~
~
- ..>:
(l

cr
(

-, ~~

e o· 2
~
lJ)
z
w
~ 0
20 40 60
~ MOISTU RE CONTENT OF WOOD - per cent
~
z
z t:. Fibre Insulating board
2 v Columbian pine
o EuroplZan whitewood
n Oak

FIG.3. THE MINIMUM INTENSITY FOR PILOT IGNITION

(SPECIMENS 7· 6 em' SQUAR~ )
. TI/}
_'~6 2-0 ,

o ,-.
u ( .-JI ~
I 1.5
z I.-

o
t= ~
Z
(!) 1·0
a:
6:
"..
. -
?=' 0·5
l J)
... z:
-.. ~ -,
z·
o 20 40 60
MOl sru RE CONTENT OF WOOD - per cent
-
A Fibr<Z Insulating board
V Co.lumbian pine
o Europ<zan whitewood
(] Oak

FIG.4. THE MINIMUM fNTENSITY FOR SPONTANEOUS

)

,) IGNfrrON (SPECIMENS 7-6 em SQUARE)

 ·'

3
sxto
(0)

/ (b)
u (e)
U'I
~
'0

.....
-
~

~ -
~

" 1cr

2 -I
em S degC cal-

e·
. f H
-2 -I -I
Curve deg C cal em S dflg C

-4
(a) 360 8· 6 x 10
-4
( b) 340 8 ·0 x 10
-4
(e ) 310 7·4 x 10 .

0 European Whitewood
li Columbian pine
0 Oak
V Fibre lnsutctlnq board

Curves qlvcn by equation (5 ) wi th values assigned to Sf and H

FIG.5. prLOT· IGNITION OF DRY WOOD 7·6 em SQUARE

<D
~

-lO
,'"
 ."

. a. OAK

5·01-------+-----+------_+_---___;

+I " E
joo.oju
~E
ar 2- 0 I------+--Qr~:::...;:;;:;:--_+_-----_+_-__:_-___;

1- 0
0·5 1'0 ' 2-0 5·0 10'0
. Ht
PmCm( I<m t >!

,b. EUROPEAN WHITEWOOD

5·0
-
-IN
.....,
.:i£E
+'
1-1
-
uE
Q..E
(1)"'"

2-0

1 . O'""--...-~...................----~--l----"-....a..........&.. .......................
0-5 1-0 5-0 10·0

Symbol Moistu~ content

oer cent
0 dry
6 20
lD '0 40
~ V 60
cr
IJ.: -4 -2 -I -I
Curve qlven by equctton 5 with Qf= 360 dcq C H =8~,6 xl0 cal ern s dC2g C

r-,
1.0
FIG.6A .. CORRELATION OF PILOT IGNITION RESULTS
w
-
~
FOR SPECIMENS 7-6cm SQUARE
 c. COLUMBIAN PIN

5'0
-IN
-
+'
E
~
...,
..... uE
~E
ar .2·0

1'0 2'0 5'0 10·0

Ht
----I
I'mcm(k m t)2"

d. FIBRE INSULATING BOARD /

/
./ CD
5· O t - - - - - - - + - - - - 1------~_+__7I'_--___I
-IN
......-..
E
-
.::i:.

2· 01------t-~___'lf__"""'_+------+_---__1

,"
2'0 5·0

:!
Ht 1
PmCm( k,ntl2

D
:0 Symbol Molsture content
0- p<lr eent
l::
1..: 0 dry
o /1 20
o
D 0 40
.... V 60
-4 -2 -I -I
Curve ~iwn by equctlon 5 with 8t = 360 deg C H =8"6 x10 eal em s dcq C ~
-4 -2 -I -I
- - - - - - - 9t=310 de9 C H=7'4 x 10 cal crn js dt2g C

FIG.6B. CORRELATION· OF PILOT IGNITION RESULTS

FOR SPECIMENS 7'6cm SQUARE
 1/6159 F.'R.586 . ~ .*-\ :- .:--, .», . • • I' •• I· _."

,.
9·Qr-----------..,.-------,--------,---------r-------,

,
5-0
Moisture Arl20
content-
per cant 7·6x7·6 1S-0x1"5'O

-
-iN
o+J
e' 0
: dry ~ '.
crif
0
cm2

•
-
~
20, II
•
.•
otJ
E
...... u, 40 0
60 'l
.sr:
~ 2·0
0),

/
~Oltic
redwood
1 . Ot--------~--~-----:--=-+-+------+--------+--------1

0·5 1·0 Ht 2·0 5'0 10-0

Pme,rt. k m t )~
-3 -2 ·-1 :-1
The curve Is given by equotlon 5 with 9t=525~ deg C , H ..1·4 x10 col cm 5 d129 C

FIG·? SPONTANEOUS· IGNITION OF EIGHT SPECI.ES OF WOOD 7·6cm SQUARE

AND 15cm SQUARE SPEct.MENS
 3·0 r-----.,...--------..-----.----'""T"""'"""-~

2·0
0

.....
t-1
t!
E
U
1·5

~ 1· 0
e
0.7 _--'_ _--lo_~_...L.._...I-..O~__I,_...J"",..L---~
0·15 0·2 0·4 0·6 1·0 1·5

Ht

( b) MoistUN content ~ 5 per cent

. . -3 -2 -I -I
Curve Is qlven by equation 5 with 9t 525 dcq C, H=1 '4x10 cal em s. dzg C
Density
Wood (oven dry) Symbol
••• _ c •• ,_ .. . .. _ ..... , .• _." .0., ....... . 9rn I ern? _.

Fibre Insulating board 0·24 . •

Crzdar 0·37 0
Mahogany 0·52 0
Columbian pine 0·55
•
Oak 0·66
•
£ FJ-G.8A. SPONTANEOUS IGNITION OF Scm SQUARE

-
(i)
r:
..
......
.~
~

'. .
SPECIMENS
 3-0 ....--------r"----..----........---

2 -0 I--_f---------+----+-------,,---+-_~

1 - 5 t--_I-----"""O"'+------... .~-+--~

•
1.0 I--_f-----~=I__--l----~--....:.....--+----__I

••
0-7 ......._I-.-_----L_----L_~--I.. ........__""_--I......I...-_ ____I

0·15 0'2 0'4 0·6 1·0 1-5

Ht
PmCmj"k;f

( c) M oi st Ur'(2 con ten t =:,8 per cent

2-0

~.
1 ·5
..... E
H u
Q..E
~ 1 -0

0-4 0'{3 1·0 1-5

Ht

(d ) Moisture. content-s-tz per cent

~ . F o r key St2CZ fig.8A.

~ -3 -2 -I -I
,Curve is given by equation 5 with 9 f 525 deg C H"'1-4x10 cal cm s dC29 C

FIG.8B. SPONTANEOUS IGNITION OF Scm SQUARE

SPECIMENS
 2· Ot---I--------t
__---+--~+Ww_-~

1 . 5 t--------!I--------+ ~~""'---_+_-____f

1 . 0 ......----.,1__------+---+----+-----1

•
o ·7 .......
- - . j l - - _ . . . . . . . . l . _ - - - . l . _ - ' - -.........---...L.--&.........r......a-._---...

0·15 0·2 0'6 1·0 1·5

Ht
PmCm\p

(<Z ) Moisture content ~15 per cent

3·0r-~-------r--..,.------,...._-___,

2·0 ~--t-----+---+----=-jt------i

1 . 51-----+------f,o__~l::J---".II::.----1I__-___4
..., E
~ u

-
~E
,0>
1 ·0.---+-...... , . . . - - 4 - - - - + - - - - 1 - - - - - - 4

••
o .-15""'-0......-2_ _...a-..._"O""-----'---..._L......I----.&_
0·7 _- - ~
,·0'1·5
0'4 0'6

.•-'
Hi

o.
. ..' I'm CmJkm t
0,
io ..
.c ( f) Moisture content ~ 18 per cant
Fibre, insulat,lng board moisture content -28 per cent
:\I
o'
(5
.... For key see fig. SA·
-.
,
. -3 -2 -I -I
cur-ve Is given by equntlon 5 with St 525 da9 C H -1 '4x 10 cal em s deg C

FIG.Be. SPONTANEOUS IGNITION OF- Scm SQUARE

SPECIMENS

, "
•

~.