Publisher’s version / Version de l'éditeur:
Vous avez des questions? Nous pouvons vous aider. Pour communiquer directement avec un auteur, consultez la première page de la revue dans laquelle son article a été publié afin de trouver ses coordonnées. Si vous n’arrivez pas à les repérer, communiquez avec nous à PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca.
Questions? Contact the NRC Publications Archive team at
PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca. If you wish to email the authors directly, please see the first page of the publication for their contact information.
https://publications-cnrc.canada.ca/fra/droits
L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.
Paper (National Research Council of Canada. Division of Building Research); no.
DBR-P-1284, 1985-03
READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE. https://nrc-publications.canada.ca/eng/copyright
NRC Publications Archive Record / Notice des Archives des publications du CNRC :
https://nrc-publications.canada.ca/eng/view/object/?id=87251c63-cb50-4059-bedf-c92440a59a43 https://publications-cnrc.canada.ca/fra/voir/objet/?id=87251c63-cb50-4059-bedf-c92440a59a43
NRC Publications Archive
Archives des publications du CNRC
This publication could be one of several versions: author’s original, accepted manuscript or the publisher’s version. / La version de cette publication peut être l’une des suivantes : la version prépublication de l’auteur, la version acceptée du manuscrit ou la version de l’éditeur.
For the publisher’s version, please access the DOI link below./ Pour consulter la version de l’éditeur, utilisez le lien DOI ci-dessous.
https://doi.org/10.4224/40001734
Access and use of this website and the material on it are subject to the Terms and Conditions set forth at
Assessment of smoke density with a helium-neon laser
Ser
ITH1
N22
dNational Research
Conseil national
no, 1284
I*
Council Canada
de recherches Canada
c, 2
ASSESSMENT OF SMOKE DENSITY WITH A HELIUM-NEON LASER
by F.R.S. Clark
ANALYZED
Reprinted from Fire and Materials
Vol. 9, No. 1, March 1985 p. 30
-
35I DBR Paper No. 1284
Division of Building Research
L 1 a t t G n u a t i o n p a r l a fumGe d e l a lumiOre monochromatique (0,632 um) Emise p a r d e s l a s e r s 2 llhGlium-nGon e t du s p e c t r e c o n t i n u
a
l a r g e bande & m i s p a r un f i l a m e n t d e t u n g s t s n e chaud e s t examinGe. On Gvalue pour l a s o u r c e monochromatique l ' e f f e t d e l a lumiOre d i f f u s e d i r e c t e e t on d z t e r m i n e dans q u e l l e mesure l a l o i d e Bouguer p e u t S t r e a p p l i q u g e 2 une s o u r c e q u i n ' e s t p a s monochromatique. Des i n c e n d i e s d ' e s s a i ,a
' e c h e l l eAssessment of Smoke Density with a
Helium-Neon Laser
Ferrers R. S. Clark
National Research Council Canada, Division of Building Research, Ottawa, Canada KIA OR6
Smoke-obscuration of monochromatic (0.632 pm) light from helium-neon lPseR and of broad-band b continuum light from a hot tungsten filament is reported. The effect of forward-scattered light is evaluated for the monochromatic s o m e , and the degree to which the Bouguer law is obeyed for a non-monochromatic source is discussed. Practical experience with both small-scale and room-scale fire experiments is desaf1,ed.
INTRODUCTION For the present discussion it is sufficient that b,,,
and b,,,,ti,, are, in general, treated as different func-
Products of pyrolysis and combustion pose two related tion of-Gavelength. scatter of light beams off more
than one particle is neglected. According to Friedlan-
to life during fire: damage the
der; multiple scatter can be safely neglected only
gases and smoke particulates and disorientation by
light-obscuring, scattering effects of the gases and when T,,,~, is <0.1, but this property is diacult to
particles. Only the latter will be discussed. measure.
Clearly, the most direct way to gauge by how much smoke obscures light-transmission is to pass a known light beam through it and measure the attenuation. Attenuation is not. however. a sufficient measure of disorienting effect, since smoke particles absorb and scatter light, and probably the most important effect of scatter is the reduction in apparent contrast between lighted areas (e.g. signs) and dark areas.
The process by which smoke attenuates light is
complex. Vervisch et al.' reported the transmission of
light of wavelengths 0.4-4.5 p m through fire smoke and identified two attenuating mechanisms. The in- ferior envelopes of absorption versus wavelength plots were attributed to attenuation by scattering caused by soot particles; for smoke produced by polyether foam this was the only mechanism operating. For smoke from polypropylene, wood and flax, however, complex band structures were superimposed on the soot scatter curve owing to molecular absorption.
The diameter of soot particles in smoke is generally considered2 to fall in the range 0.3-5.0 pm, much the same as the wavelengths of visible light, although
smaller particles have been ~ b s e r v e d . ~ Mie scattering
theory is applicable for this particle size range. For monochromatic light it may be shown4 that
where light of intensity
Il
is attenuated to intensity I2by smoke of optical thickness 7=1:bdz
REQUIREMENTS
FOR A PHOTOMETRICS M O n METER
Two distinct applications of smoke meters are com- mon in fire research and materials evaluation. First, the smoke-generating propensity of a material can be assessed by measuring the obscuration of a light beam passing through a chamber containing a burning sam- ple. This is the basis of the NJ3S Smoke Chamber Test
(ASTM E662-79).' Such devices are primarily de-
signed to measure substantial attenuation of light by heavy smoke. To encompass the very wide dynamic range of the detector signal and the low levels of light to be measured, strong light sources, sensitive detec- tion and gain-ranging amplification are required.
The second application of smoke meters is the monitoring of relatively low levels of light-attenuation from smoke in full-scale burns in rooms and corridors. Here, interest lies in determining the length-of time an occupancy can remain tenable. The maximum tolera- ble smoke density will depend on the distance to be covered to escape from the fire. Quintiere6 recently reviewed several studies showing that an optical den- sity of about 0.5 per metre (OD/m) reduces visibility to about 2 m; to attain 10-m visibility a density of SO. 1 OD/m is required.
At the Fire Research Field Station recently commis- sioned by the National Research Council of Canada, 60 sensors were needed to quantify the amount of smoke in compartments in a ten-storey experimental
over a path-length z = L. This is the familiar Bouguer tower. To calibrate so many smoke senso; and have
or Beer-Lambert law. The extinction coefficient, b, them function simultaneously .is a formidable task.
has two components Thus, a practical application of the present study was
to develov an inemensive smoke meter that would be
b = bscatter
+
batsorption (2) highly stable and reliable.I
ASSESSMENT OF SMOKE DENSJTY WITH A HELIUM-NEON LASERI
SCA'ITERAND
WAVELENGTHRecently, Mulholland2 warned that errors in the measuring of light-attenuation by smoke might arise if forward-scattered light were to reach the detector. Forward scatter provides light to the detector in excess of that expected on the basis of attenuation of the beam by smoke; its effect may be minimized by ensur- ing that the angle of view of the detector is limited. Such provision is made in the design of the NBS smoke chamber (ASTM E662-79)' but not in the design of the NBS-Bukowski portable smoke meter.'
As has been noted, the Beer-Lambert law is strictly
1.
valid only for monochromatic light. ~ u l h o l l a n d ~ esti-'
mated that errors in smoke mass concentration of up to 25% can also occur if non-monochromatic white
I
light is used.I
EXPERIMENTAL DESIGNBoth laboratory-scale experiments and a full-scale room-bum facility were used. In the laboratory, the NBS chamber5 was modified to accommodate two 0.5mW, random polarization, He-Ne lasers (Aerotech) on the roof of the chamber, as close as practicable to the normal white light beam of the NBS chamber: laser I and laser I1 in the discussion that follows (Fig. 1). The laser beams passed vertically down through protective glass plates on the interior surfaces of the upper and lower walls of the chamber.
The beam from laser 11 passed through two circular
orifices, 0.53 cm i n diameter and 34.3 crn apart, before encountering a H e N e laser line filter (bandwidth 10 nm at half-peak height) and finally a silicon diode
phatdetector of 1 cm2 active area (Silicon Detector
Corporation, Model SD-444-12-12-171). The laser line filter could be removed with negligible effect on the results.
Following Mulholland? forward scatter is negligible if the angle subtended at the aperture closest to the source by the aperture closest to the detector is 0.1 times the first minimum of the Fraunhofer diffraction pattern. For particles 5 p m in diameter, at a wavelength of 0.632 pm, the first minimum is at 0.1545 radian (9 degrees of arc); the subtended angle
r
is 0.1 times this value for the aperture and separations described. The standard NBS smoke chamber beam.1
by contrast, Subtends 0.014 radian (0.8 degree)? The beam from laser I passes through the chamber to anI identical detector, but without scatter restriction. The
angle subtended by the aperture in this case is 1.12 radians (64 degrees); the Fraunhofer criterion is met only for particles of diameter less than 0.07 pm. Each aperture is more than eight times the diameter at which the laser intensity is reduced to l / e 2 times its value at the beam cenwe.
The
detector diodes facing lasers I andII
were both optimized for photovoltaic operation; signals from both were conditioned by di- rect coupling to a conventional operational amplifier at fixed gain.The electronics were bypassed to avoid the manual gainranging of the outdated vacuum-tube amplifier
W H I T E L I G H T
Figure 1. Schematic diagram of the modified NBS smoke chamber: (a) perspective view; (b) plan view, optical system location.
fitted to the available NBS smoke chamber. High tension (-700 V) was supplied from a Fluke Model 415B power supply to the dynode chain of the RCA 931B photomultiplier tube. The anodic output of this tube was filtered with a 0.1-pF tantalum capacitor to ground to reduce high-frequency noise, and a suitable load was provided with a 10-MS1 resistor to ground. The signal then provided the input of an operational amplifier identical to that used for the laser photome- ters.
FIRE AND MATERIALS, VOL. 9, NO. 1, 1985 31
F. R. S. CLARK
Signals from both silicon diodes and the photomul- tiplier tube were directed to a 12-bit AID converter (TransEra Model 752-ADC) controlled by a Tek- tronix 4052 desktop computer. In a typical experiment of 20-min duration the AID converter sampled each photometer channel every 2 s, displaying the result on the computer screen. At the end of the run the data could be stored on a magnetic floppy diskette from which files could be retrieved for manipulation and plotting. In all the cases presented here the data were smoothed with a single three-point moving-average routine, and the clear beam outputs were set to zero absorbance to eliminate the problem of dirty optics (glass exposed to smoke was cleaned with ethyl al- cohol before each run).
Calibration of the photometers was particularly im- portant to the success of the study. Eleven neutral density filters (value 0.07 to 3.080D) were used, checked by spectrophotometry to ensure that their transmission was independent of wavelength and calibrated with a Densichron 10 densitometer. Each working day a calibration was conducted for each photometer, using the signals received by the AID converter. A computer routine directed the operator through the calibration procedure in a precisely timed, controlled manner, leading to improved reproducibil- ity and linearity. Provided the photometers and chamber radiant panel had been allowed to equilibrate over several hours prior to calibration, correlation coefficients of fit to linear regression lines in excess of 0.997 were obtained for each of the photometers. Systems with scatter restriction were slightly less linear in response, probably because of the greater demands of alignment. The equation of the linear regression line was in each case used to estimate the smoke density.
SMOKE DETERMINATIONS
The ASTM E662-79 procedure5 was followed, except that to ensure equal smoke concentrations at the three points of measurement an instrument fan, 110-mrn diameter, was mounted in the right rear comer of the chamber. Smoke was thus drawn from the bottom of the chamber and rapidly and evenly dispersed. The materials were burned in (at least) duplicate under both flaming and non-flaming conditions. Representa- tive smoke-generation curves are displayed in the accompanying figures.
RESULTS
Photometer stability
Figure 2 shows the output of three photometers during an experiment in which no smoke was generated. The white light detector (photomultiplier tube) is more subject to high-frequency noise than are the silicon diodes of the laser-based photometers, but the re- sponse of the latter undergoes a slow, roughly sinusoi-
0
0 6 1 2 18 2 4 30
TIME. m i n
Figure 2. Response of laser I, laser II and white light (W) photometers in the absence of smoke.
dal fluctuation that may be caused by drifting of the laser beam across the detector in response to flexing of the chamber walls. Such an effect would not be ex- pected of the white light system since it was strongly braced. Alternatively, laser power may have varied; the manufacturer suggests that variation of up to 3% may be expected. The observed fluctuations suggest a practical minimum of 0.04 O D to low-density smoke measurements, ample for present purposes. Additional output stability may be provided by feedback circuitry and a second detector directed to the unattenuated output of the laser.8
Ell ect of scatter restriction
I
In general, laser I1 (scatter restricted) indicated higher smoke densities than laser I, provided the smoke density did not exceed 2 O D (Figs 3-6). This result was repeatable for several wood products, polyurethane foam and fire-retardant polystyrene foam. The deviation of the meters (0.080 maximum)
TIME, rnin
Figure 3. Smoke production from spruce, flaming mode; two replications displayed; I is laser I, II is laser II, W is white.
ASSESSh4E%T OF SMOKE DENSlTY WlTH A HELIUM-NEON LASER r 0 0 Z 4 6 8 10 TIME, m i n h
Figure 4. Smoke production from polyurethane foam, non- flaming mode; two replications; I is laser I, II is laser II, W is white.
TIME, m i n
Figure 5. Smoke production from fire-retarded polystyrene foam, flaming mode; two replications displayed; I is laser I, II is laser II, W is white.
0
0 6 12 18 24 30
TIME, m i n
Figure 6. Smoke production from plywood, flaming mode; I is
laser I, II is laser II, W is white.
lies just outside the random error expected, but ran- dom error is unlikely to be the explanation of the phenomenon since the order of the outputs is always the same. The observation is, however, consistent with the effect of scattered light; laser I, having no scatter restriction, receives more light and thus indicates a lower smoke density. The small magnitude of the difference between the two measurements may be caused by either the small amount of smoke or the fact that the smoke particles are small and scatter little.
When smoke density exceeded 2 O D during some runs (in these experiments only for cellulosic materi- als) the reverse of the above trend was observed (Fig. 3). Laser I gave smoke-density values in excess of those given by the scatter-restricted laser 11, but it should be noted that the amplifiers for laser I and the white light were both very close to the limits of their useful range at this smoke density.
Influence of wavelength
As the white light photometer does not use mono- chromatic radiation, one would not expect the Bouguer law (Eqn (1)) to be followed r i g i d l ~ . ~ It is, however, provided that wavelength-independent pro- cesses are responsible for attenuation; calibrations to at least 3.08 OD, using neutral density absorption filters, are as completely linear as those with laser I. Wavelength-dependent absorption may not give this result.
Consider now the difference in extinction of red and white light systems. Differences in smoke densities indicated by laser I1 and the white light photometer (both front-scatter restricted) were small but signifi- cant for all sample types. Differences due solely to the wavelength-dependence of scatter would lead to a roughly inverse relation between wavelength and ex- tinction coefficient, provided that particle diameters are much smaller than the wavelength.'a4 As the peak wavelength of the spectral luminance curve of the white light used is less than the laser wavelength, apparent smoke densities are predicted to be higher for the (shorter wavelength) white light. For synthetic polymers (Figs 4 and 5) and many wood products (Fig. 6) this was indeed the case; extinction was greater for the white light meter than for laser 11, and the differ- ence increased with smoke density. Smoke from plywood (Fig. 6) showed a slight reversal if exposure was prolonged.
A reviewer has suggested that at high smoke con- centrations, coagulation of wood smoke particles may occur; when the particle diameter becomes close to the wavelength of the light, scattering anomalies may give results similar to those r e p ~ r t e d . ~ Alternatively, changes in smoke composition during maturation may cause changes in molecular absorption at wavelengths other than 0.632 Fm. Work is currently under way to test this proposition.
Most smoke photometers use the light from a tung- sten filament, often filtered to mimic the useful re- sponse of the human eye.7 The rationale for this is related to the ability of a person to see through smoke. Such a rationale implies that the light to b e observed is
F. R. S. CLARK
white, but this is commonly not the case. Further, the perceived brightness is not a simple function of the spectral distribution of a light source to be seen through smoke. Visibility of a red or green exit sign, for example, will not, in principle, be any more pre- dictable from measurements of smoke density using a tungsten filament than from those using a He-Ne laser beam.
PRACTICAL DETAILS
The design of a photometer based on white light is necessarily more complex than that of one based on laser illumination. careful alignment of the optical system, which must include at least two convex lenses (or their mirror equivalent), is required. By contrast, a laser system is easily assembled, requiring merely a laser, a detector and perhaps a tube fitted with two pinholes to reduce front-scattered light and other ex- -
traneous light. Many photometers, typified by the Bukowski design: are robustly constructed, with rods to ensure permanent alignment. The necessity for rod assemblies and the generally poor quality of optical design put practical limits on the beam lengths over which such photometers might be operated. Among exceptions is the recently developed PEL-BRANZ smoke meter from New Zealand, which has an achromatic doublet lens to collimate the light and a three-lens collector system.1°
Lasers, on the other hand, lend themselves to oper- ation over long distances, with negligible loss of colli- mation of the beam; even over 10-m path-lengths lasers operate with the beam still completely inter- cepted by a detector with an active surface area of
1 cm2. As each end of the system is independent,
however, alignment can be difficult to preserve, espe- cially where temperature gradients are large. In order to alleviate this problem trials were conducted with a corner cube mirror. In this configuration the detector is attached to the laser housing and the beam is reflected with a constant lateral displacement by the mirror. Contrary to expectation, it was not any easier to align such a system and the effective dynamic range of detection was halved because the beam-length was doubled. It was concluded that the laser and detector must be rigidly secured in environments not subject to heat or vibration.
Extraneous light was revealed as a particularly seri- ous problem with large-scale burns. If the smoke meter axis is intercepted by flames from the lire, which often occurs in rooms in which flashover is imminent, large reductions in the recorded smoke levels are observed. If the detector is designed to have only a narrow acceptance angle, this effect is less likely to occur. Flame light will remain a problem, however, since smoke particles in the beam will scatter light from a flame far from the axis and a portion of the scattered rays will strike the detector. Flame directly in the beam will also influence the results. Photometer light source modulation and phase-lock amplification of the signal will help, but they were not investigated in this study. V -I.>> F I L T E R u" F I L T E R \ I s : " ' " " C I Y l V Y L E T E C T O R i c m
f
-
' ~ e - ~ e L A S E R I N T E R F E R E N C E F I L T E R B U N S E N B U R N E RFigure 7. Apparatus to test the effect of selective filtration of the detector.
As only a minute fraction of the light of a flame has the same wavelength as that of the helium-neon laser, it should also be possible to discriminate between flame light and the laser by the use of an interference filter that allows only light of wavelength 0.632 p m to pass. This hypothesis was tested using a laser (helium- neon, 0.5 mW, random polarization) 0.50 m from a silicon diode detector (Fig. 7). A bunsen burner (30- cm flame) was placed to provide a very bright yellow or a low luminance blue flame in the beam. Smoke was represented in this experiment by a 1.80 OD absorption filter before the flame and a 0.33 OD filter after it. In typical experiments a clear beam was established, the 0.33 O D and then the 1.80 OD ab- sorption filters were added and finally the flame was introduced. The apparent optical density was recorded for two flame conditions, yellow (Fig. 8) and blue (Fig.
9), in both the presence and absence of an interference
filter (10-nrn bandwidth at half peak height).
From Figs 8 and 9 it is clear that introduction of the interference filter substantially reduced the deleterious effect of flame light on the optical density recorded for luminous flames. This method of avoidance of the effects of stray light is only available if monochromatic analysing beams are employed.
It was expected that light passing through heated air would be refracted, reducing the signal from the de- tector if the beam was deflected from the sensitive surface. This proved to be of minor importance in the experiment just described since introduction of a rela- tively non-luminous blue flame did not lead to change
34 FIRE AND MATERIALS, VOL. 9, NO. 1, 1985
> C
-
; 1.5 Y n -1 4 U-
1.0 C a 0 0 . 5 Od
I 0.4 0.8 1 . 2 1.6 2.0 T I M E , m i nFigure 8. Effect of selective detector filtration: high luminance flame; solid line, no filter; dotted line, filtered.
-
-
-
I
I
I
1
- I d N FI
J
i
ASSESSMENT OF SMOKE DENSITY W'IlW A HELIUM-NEON LASER
I
5
(1) the Variation materials in the commonly smoke-generating exceeded any propensity differences of
0 in the measures of smoke-extinction studied.
o 0 . 4 0.8 1.2 1.6 2.0 (2) Molecular absorption or increase in particle size, T I M E . rnin changing with smoke-maturation, may cause dis-
2 . 5 2.0 > C
-
q 1 . 5 - w a 2 4 U 1.0 L oFigure 9. Effect of selective detector filtration: low luminance crepancies between photometers using broad-band flame; solid line, no filter; dotted line, filtered. or monochromatic sources.
(3) Front-scatter restriction is a useful technique to impede the scattered light from either the photo-
in observed optical density. It might be revealed as a meter light source or other light sources such as
problem for large-scale fires even if light-source mod- flames from reaching the detector. The problem
ulation narrow bandpass filtration were implemented, was, however, negligible at low smoke densities
since even a small refractive deflection of the beam for the small-scale experiment described. Particles
could, with a long beam-path, lead to a large transla- of larger diameter, close to the wavelength of the
tion of the beam position on the detector. analysing light, would be expected to cause more
significant scatter.
I I I I the only critical variables. In addition, it is an advan-
F tage to screen the detector from at least some forms of
fl
r--*---
extraneous light, using a laser line filter and/or a front-
I-
scatter-limiting aperture. Complex optical systems are)
CONCLUSIONS
AcknowledgmentsI
-
I
I
-
-
I
Helium-neon laser-based smoke photometers are suit- ~h~ author wishes to thank M~ ~~~~~~d Raviani for invaluable
able for both small- and large-scale smoke measure- assistance in the design of the equipment and for conducting many
men%
for
at least the burningof
the materials used in of the experiments. The assistance of colleagues in the Division ofthis study. The advantage of the technique Physics, National Research Council Canada, particularly Dr Alan
over those employing non-laser sources is that the Robertson, is also acknowledged. This paper is a contribution from the Division of Building Re-
specification of a complete light-measuring System is search. National Research Council Canada. and is ~ublished with
thus avoided.
A second advantage arises from the high level of laser-beam collimation, allowing valuable spatial av- eraging of low smoke densities. The generally low quality of beam collimation in conventional white light smoke meters does not permit such long beam-paths.
The study also revealed qualities and limitations of both white light and laser-based photometry.
very simple; the type of laser and its path-length are the approval of the Director of the ~ i v i s i o n .
1. P. Ve~isch, D. Puechbe* and T. Mohamed, Combust. and Flame 41, 179 (1981).
2. G. W. Mulholland, Fire and Materials 6, 65 (1982).
I 3. C. K. Lee, J. M. Singer and K. L. Cashdollar, Fire and Materials 2, 110 (1978).
4. S. K. Friedlander, Smoke, Dust and Haze, John Wiley ( i g n ) .
,
5. Standard Test Method for Specific Optical Density of Smoke Generated by Solid Materials, ASTM, Test MethodE662-79 (1979).
6. J. G. Quintiere, Fire and Materials 6, 145 (1982).
7. R. W. Bukowski, Fire Technology 15, 173, 271 (1979).
8. J. J. Comeford and M. Birky, Fire Technology 8.85 (1972). 9. W. P. Chien and J. D. Seader, Fire Technology 11, 206
(1975).
10. A. B. Trotter and K. Trotter. An improved smoke meter for the early fire hazard test (Australian Standard 1530, Part 3, 1976). DSIR, New Zealand, Report No. PEL R-793 (March 1982).
Received 13 October 1983; accepted 30 January 1984
T h i s p a p e r , w h i l e b e i n g d i s t r i b u t e d i n r e p r i n t form by t h e D i v i s i o n of B u i l d i n g R e s e a r c h , remains t h e c o p y r i g h t of t h e o r i g i n a l p u b l i s h e r . It s h o u l d n o t be r e p r o d u c e d i n whole o r i n p a r t w i t h o u t t h e p e r m i s s i o n of t h e p u b l i s h e r . A l i s t of a l l p u b l i c a t i o n s a v a i l a b l e from t h e D i v i s i o n may be o b t a i n e d by w r i t i n g t o t h e P u b l i c a t i o n s S e c t i o n , D i v i s i o n of B u i l d i n g R e s e a r c h , N a t i o n a l R e s e a r c h C o u n c i l of C a n a d a , O t t a w a , O n t a r i o ,
KIA
0R6.Ce document est d i s t r i b u 6 s o u s forme de t i r G - 8 - p a r t par l a D i v i s i o n d e s r e c h e r c h e s en b b t i m e n t . Les d r o i t s de r e p r o d u c t i o n s o n t t o u t e f o i s l a p r o p r i 6 t G de l ' g d i t e u r o r i g i n a l . C e d o c u m e n t n e p e u t d t r e r e p r o d u i t en t o t a l i t g ou en p a r t i e s a n s l e consentement de 1 1 6 d i t e u r . Une l i s t e d e s p u b l i c a t i o n s de l a D i v i s i o n p e u t S t r e o b t e n u e en G c r i v a n t 3 l a S e c t i o n d e s p u b l i c a t i o n s , D i v i s i o n d e s r e c h e r c h e s e n b s t i m e n t , C o n s e i l n a t i o n a l de r e c h e r c h e s Canada, Ottawa, O n t a r i o , KIA OR6.
