This paper was published in Studies in Conservation, 52, 199-210 (2007)
Original publication is avaible at: http://www.iiconservation.org
Impact of indoor heating on painted wood: monitoring the
altarpiece in the church of Santa Maria Maddalena in
Rocca Pietore, Italy
Ł. Bratasz, R. Kozłowski, D. Camuffo, E. Pagan
1
Impact of indoor heating on painted wood:
monitoring the altarpiece in the church
of Santa Maria Maddalena in Rocca Pietore, Italy
Łukasz Bratasz a, Roman Kozłowski a, Dario Camuffo b, Emanuela Pagan b
a Institute of Catalysis and Surface Chemistry, Polish Academy of Sciences,
ul. Niezapominajek 8, 30-239 Kraków, Poland, e-mail: [email protected]
National Research Council of Italy, Institute of Atmospheric Sciences and Climate (CNR-
ISAC), Corso Stati Uniti 4, I-35127 Padova, Italy. e-mail: [email protected]
Abstract
In the church of Santa Maria Maddalena in Rocca Pietore, Italy, the dimensional response of
the wooden altarpiece to variations in indoor temperature and relative humidity (RH) was
monitored between December 2002 and March 2005. Measurements demonstrated that only
the external layer of wood, several millimetres thick, continually absorbs and releases water
vapour following external variations in RH. For massive elements this leads to strong
gradients in the moisture content through wood, a restraint of the dimensional change and a
development of stress, which is the main threat to the integrity of wood and the decorative
layer. Particularly strong RH variations and related high stress levels were produced by the
intermittent heating system based on the inflow of warm air. To incorporate requirements for
preservation, heating systems must provide a localised comfortable temperature in the area
where people are without changing the natural climate of the church as a whole.
Introduction
Originally historic churches had no heating. The demand for heating in churches increased
with the improvement of heating at home. Most churches are heated now and a number of
heating systems have been developed, the choice usually being made between warm air
emitted from floor or wall, radiant heaters - electrical or heated by gas combustion, water-
filled radiators heated from a boiler, underfloor and pew heating [1,2]. Two fundamental
strategies for heating churches can be distinguished: stationary heating, when a church is kept
at a constant temperature, and intermittent heating when a church is heated occasionally over
a short time. A combination of the two strategies is sometimes adopted when a church is
continually kept at a primary low temperature of 8-12 oC and is heated rapidly shortly before
and during services to a comfortable temperature of 15-20 oC [3, 4].
It has been recognised that unheated churches generally preserved their artworks in good
condition over centuries, while rapid signs of degradation were found after heating had been
introduced. Therefore, many conservation authorities conclude that no heating is best.
However, a constant low-level heating might be preferable in countries with mild, rainy
winters as a simple method to avoid exceedingly high humidity and related mould growth.
Another aim of low-level background heating is the reduction of condensation risk on cold
indoor surfaces specially during spring.
2
The indoor climates of unheated buildings are essentially governed by the outside climate
modified by the building envelope which impedes heat exchange between outdoors and
indoors, and buffers the external climatic variations by taking up and giving off heat and
humidity. Generally, churches possess large buffering capacities and, if properly maintained
and used, low rates of air exchange. Therefore, their indoor conditions slowly follow the
average outdoor climate. The seasonal cycle is attenuated and the short-term variations are
significantly attenuated. Furthermore, works of art made of moisture-containing materials
sensitive to variations of thermal and humidity conditions, like wood, have adapted over the
centuries to the local climate pattern. This adaptation might have involved a certain degree of
permanent change, as deformation or fracturing, releasing internal tensions in the materials
generated by the variations of relative humidity (RH) and temperature (T). Should a work of
art be brought into a different microclimate – even a well-controlled museum environment –
or should the variations of RH and T increase, it will suffer damage.
The concept of a local natural climate, to which the objects have adapted over their long-
term exposure, was first derived from scientific observations [5, 6] and then explicitly
expressed in the Italian Standard [7] on choice and control of the indoor environmental
conditions favouring conservation of sensitive historic materials. The standard stresses that
for the best preservation of materials sensitive to moisture, the recommended RH ranges
should replicate the long-term local climate and that the RH fluctuations centred on the local
RH level must be kept to a minimum.
Heating can introduce serious destabilisation to the natural indoor climate in a church. A
stationary heating regime may bring the indoor temperature to a high ‘comfortable’ level
causing a low and variable RH indoors as the cold air outside is drawn in and heated up. The
intermittent heating for liturgical services or cultural events may in turn cause periodic
fluctuations of RH corresponding to the heating episodes when RH drops first from high to
low levels and then returns to high RH after the heating system is switched off. Two general
principles in church heating, therefore, could be formulated to reduce the adverse effects
described:
- heat a church as little as possible during the cold season, or adjust carefully the heat
input just to reduce the excessive dampness and the drop of temperature in winter
- provide localised heating to the areas where people are and maintain a natural or
approximately natural climate in the rest of a church
A broader European research programme [8] was undertaken to develop a novel heating
system consisting of low temperature radiant sources located in pews which would provide
direct confined heat just to people sitting and leave undisturbed the church as a whole. The
work focused on the church of Santa Maria Maddalena in Rocca Pietore, situated at 1143 m
above sea level in the Italian Dolomites. Previously, the church had had an intermittent
heating system based on a forced inflow of warm air. The system operated usually a number
of times a week, for a short period during services. A detailed study of the internal climate of
the church recognised a particularly negative impact of this heating system on the church
fabric and its contents as it generated short-term temperature peaks accompanied by drops in
RH [9].
An important part of the present investigations was a continuous in situ monitoring of the
dimensional response of the main altar, a polychromed and gilded wooden triptych, executed
in 1516-17 by Ruprecht Potsch from Bressanone (Figure 1). The altar was carved in lime
wood (Tilia sp.). Due to its size and location in the church, the altarpiece was particularly
endangered by the microclimatic variations caused by the heating episodes. A few years after
restoration, new deep cracks appeared in the triptych, explicable by the repeated desiccation
of the wood subjected to blown warm air (Figure 2). The inventory of damage made during
the current project revealed an ongoing deterioration of the altarpiece, especially on its upper
3
part. The monitoring of the altarpiece recorded not only dimensional changes of carved
elements of different thicknesses, but also the closing and opening of a large crack in the
wood (Figure 3). The results of the monitoring, which ran for more than two years, between
December 2002 and March 2005, are presented here and discussed to provide direct insights
into the response of complex historic wooden objects to short-, medium- and long-term
variations of T and RH which make up the environment of a heated church. A more general
assessment of the idea of the intermittent heating of historic churches will be attempted.
Monitoring the dimensional response of wood
The dimensional changes occurring in two carved elements of differing thickness were of
particular interest. One element was a head – a massive element of cylindrical shape 15.5 cm
in diameter, and the other a finger – a fine element 4 cm long and 0.5 cm thick. The locations
of these elements on the altarpiece are marked in Figure 1. Triangulation laser displacement
sensors were used as they allowed fast, precise, non-contact field measurements. The
principle and features of the triangulation technique, details of the measuring system and
accuracy of the obtained data were described in detail in [10].
The width of a crack in the saint’s head was monitored parallel to the dimensional change
of the head as a whole. The crack was about 2 mm wide and ran 2-3 cm deep into the wooden
structure. A small inductive EX-110V Keyence sensor was attached to one internal surface of
the crack and a metallic reference plate to the other (Figure 3). The sensor recorded changes
of inductivity resulting from the displacement of the reference and thus allowed precise
monitoring of the crack movement within a 2 mm range with uncertainty of about 6 µm. The
measurements were possible for RH values below around 60 % because higher RH caused
significant narrowing of the crack, which disturbed the measurements.
The selection of the carved elements for monitoring was guided by an attempt to follow
the mechanism of the deterioration process for the Rocca Pietore altarpiece driven by the RH
variations. Massive wooden elements - like the head of the sculpture - are most endangered on
the fall in RH. As the moisture diffusion out of wood is not instantaneous, the external layer
of wood dries more quickly than the interior. The gradient of moisture develops and leads to a
differential shrinkage through the wood. The dry external layer is restrained from shrinkage
by the substrate of still wet and swollen wood beneath. This creates mechanical stress and
eventually wood can crack if the tension goes beyond the strength of the material [11, 12, 13].
The described mechanism is not pertinent to finer, freely moving wooden elements which
experience a uniform moisture distribution on drying or wetting due to their small size, and
hence little internal restraint and stress. It was important to precisely know the response of the
altarpiece to the natural climatic conditions prevailing in the church to which the object had
adapted throughout its history. The wooden sculpture was in perfect condition some ten years
ago, when it was restored. The subsequent recent damage can be attributed to the use of
intermittent warm air heating.
Results and discussion
The seasonal cycle of the local climate
The climate in the church in Rocca Pietore is essentially governed by the outside climate,
the protective and buffering properties of the building envelope and the use of the church. All
indoor air parameters were measured at many sampling points within the church, and the
nearest point to the altar was a probe placed in its central part at 3.5 m above the floor. The
4
natural ventilation rate in the church, when the heating was off, was 1.3-1.4 air change per
hour (ACH) which can be regarded as comparatively high. For example, ACH of around 0.1
is typical for churches with airtight plastered stone vaults and 0.5 – 0.75 for churches with a
wooden vault [1]. A high air exchange between the interior and the outside is illustrated in
Figure 4 by plots of indoor and outdoor moisture content in 2003 expressed as humidity
mixing ratio (MR), i.e. grams of water vapour per kilogramme of dry air. The MR inside
varies between about 2 and 13 g/kg which is in accordance with the external seasonal
variations. Figure 5 shows plots of T and RH inside and outside during 2003. The variations
both in T and RH indoors are evened out when compared with outdoors due to impeded air
exchange and the buffering effect of the building.
The seasonal cycles of both T and RH indoors result from the external seasonal variation
and heating the church in the cold period. The interior temperature varied seasonally from
about 1 oC in January to 24
oC in August with a marked increase in the cold period caused by
heating. The heating episodes are visible as upward blips in the indoor temperature record.
RH shows the same yearly cycle as temperature which ranges from about 35 % in March to
72 % in June. Average RH is about 55 %. It should be stressed that the low-high RH cycle
indoors is caused by heating in winter as the RH outside does not show any distinct seasonal
variation and fluctuates irregularly around a calendar year average of 70 %. The winter
heating brings down the indoor RH by approximately 15-20 % into the range between 30-50
% instead of 50-70 % during the unheated period. However, as shown below, short episodes
of quick warming-up, giving rise to drastic drops in RH, are the main cause of damage to the
wood rather than the average drop in RH during winter.
The seasonal cycles in the indoor RH are reflected in dimensional changes of wooden
elements as illustrated in Figure 6. The figure compares plots of the dimensional response of
wooden elements of the altarpiece with the T and RH variations. The diameter of the thick
wooden head varied within a range of about 0.4 % - a contraction of 0.2 % on the maximum
decrease in RH to 34 % in March was followed by an expansion of 0.2 % on the rise of RH up
to 73 % in summer, if the mean head diameter was assumed as a zero level.
Wood is anisotropic and its moisture-related dimensional changes vary in its three
principal structural axes – longitudinal, or parallel to grain, radial and tangential. The
monitored sculpture can be viewed as carved from a cylindrical tree stem; therefore the
diameter of the head most likely coincides with the radial direction in the wood. The
laboratory measurements of the radial moisture-related expansion was for lime wood
1,38±0,05*10-3
per percent by weight uptake of water. If an equilibrium moisture content in
wood had been attained, the RH variation of 39 % during the seasonal cycle would have
caused water content variation of 7.5 % by weight. This should have led to a wood movement
within a range of about 1 % i.e. 2.5 times larger than actually observed. The smoothing of the
response of the element with respect to the RH change was due to the slow moisture diffusion
into and out of the wood and the resulting dimensional restraint which did not allow for the
equilibration of a 15 cm thick element even to the very slow, year-long variation of RH.
The movement of the fine wooden finger was confined to a much broader range of about
2.5 % - the contraction of 1.75 in winter was followed by the expansion of 0.75 in summer.
The observed movement even exceeded the range of 1.4 % calculated from the radial swelling
coefficient measured in the laboratory - apparently an extra component due to a movement of
the entire hand disturbed the measurement of the long-term dimensional change of the finger,
in spite of several precautions in the configuration of the experiment to record just the isolated
relative dimensional change of each particular element [10]. However, the dimensional
changes of the finger recorded over shorter periods during which the disturbance was less
relevant were in close accord with the calculated changes. Figure 7 shows an example of
5
measured and calculated movements of the finger in August 2003 in which longer and shorter
RH variations occurred in the church.
At this point it should be added that the thermal contraction or expansion affects the
overall dimensional change of wood as the pattern of the RH and T variability is the same
during the year. As shown below, the variability in temperature rapidly affects the structure of
the wood when compared to the rate of heating, so there is no smoothing effect on the
response to any temperature variation. The thermal expansion coefficient of the lime wood
measured in the laboratory was about 7±1*10-5
per oC for radial as well as for tangential
directions. Therefore, the temperature variation from the minimum of –1 oC in winter to the
maximum of 25 oC in summer caused the dimensional change of 0.2 % as compared to 1-1.4
% caused by the RH variation as calculated above. Therefore thermal expansion had a
comparatively minor effect on the overall dimensional change of the wood.
Short-term variations
The monitoring allowed a comparison of the response of wooden elements to irregular
variations in RH for periods shorter than a seasonal cycle. The plots of the dimensional
response of two wooden elements - the head and the finger – to three such microclimate
variations, recorded typically in the church in Rocca-Pietore, are shown in Figure 8.
The first was a natural weekly variation due to a spell of dry weather lasting several days.
An episode in July 2004 with a relatively strong fall in RH of 10 % followed by the change in
the opposite direction was selected. During the period between December 2003 and March
2005 only about 15 such strong cycles were recorded. The wooden head responded weakly to
this variation and the contraction recorded was 0.1 % as compared to the calculated full
response of 0.6 %.
The second short-term variation selected was an indoor diurnal cycle caused by a strong
external cycle in the summer, when solar radiation was intense, with a drop in RH of 12 %
between night and day. Almost no dimensional response of the head was recorded for this
natural diurnal cycle.
Finally, a short, man-made cycle due to a single heating episode for a service in January
2003 was selected. The heating was operated for a total duration of some 90 minutes and it
generated a quick increase of temperature from 4 to 21oC accompanied by a 27 % fall in RH
from 54 to 27 %, followed by a slow return of both parameters to their initial values when the
heating was switched off.
During this heating episode the head expanded due to the temperature increase but no
shrinkage due to the fall in RH was detected. In contrast, the finger exhibited a rapid,
unrestricted movement which agreed well with the calculated dimensional change. The finger
first expanded, due to the increase in temperature, and then shrank due to the decrease in RH.
Monitoring the crack
A large crack in the saint’s head was continuously monitored for changes in width, i.e.
narrowing and widening. Results for the winter 2003 are plotted in Figure 9 as an example.
The crack width increased on a decrease in RH due to a shrinkage of the external layer of the
sculpture and a resulting tensile stress acting tangentially. Conversely, the crack size
narrowed on an increase in RH as the external layer swelled and a tangential compressive
stress appeared. The expansion and contraction of the crack was rapid and followed the short
microclimate fluctuations as shown in Figure 10. This observation is direct evidence of stress
continuously engendered by climatic variations in the outer zone of the wooden statue. The
movement of the crack was measured as -0,047 mm or 1*10-4
of the circumference of the
sculpture per 1 % of RH change. The entirely unrestricted swelling/shrinkage at the
6
circumference of the sculpture was on average 4*10-4
per 1 % of RH change. Therefore, the
crack is able to reduce by only 25 % the tangential stress appearing in the external wooden
layer during the RH variations. In consequence, the threshold of allowable microclimate
fluctuations which wood can ultimately endure without damage, increases in the same
proportion.
Conclusions
The elements of the altar responded to the variability of indoor T and RH. The moisture
content in the wood varied according to whether RH increased or decreased in the proximity
of the altar. The variation in the moisture content caused dimensional changes in the wood;
the response, however, was characterised by ranges and rates which varied considerably with
the thickness of the wooden elements. Fine wooden element 0.5 cm thick expanded and
contracted quickly and completely even during short-term RH changes. Therefore, the
external layer of all wooden elements of the altar, at least several millimetres thick, will be
strongly affected by any, even rapid, change in RH.
On the other hand, the overall dimensional reaction of the massive wooden element 15 cm
thick was much slower as the uptake or release of the water vapour was too slow to produce a
uniform moisture content through wood. For short-term cycles, natural or due to heating
episodes, practically no change in the overall dimension of the element was observed in
association with changes in RH. The slow overall moisture movement out of or into the wood,
combined with an immediate response of the external layer, led to strong radial gradients in
moisture content. This in turn resulted in a restraint of the dimensional change and stress
development at the external layer of the wood. At a sufficient level of stress the external layer
might suffer from mechanical damage, such as cracking, a principal deterioration feature of
the wooden sculptures.
The level of the stress depends on the amplitude and rate with which the RH varies. A
systematic numerical modelling of the phenomenon for a cylindrical object, imitating a
wooden sculpture, was reported elsewhere [14]. The obtained stress levels for the natural
fluctuations were found to be much smaller than the critical magnitude corresponding to the
elastic limit of the wood, i.e. the yield point. In contrast, the fluctuations produced by the
intermittent heating with the inflow of warm air gave rise to stress levels beyond the
allowable limit. The modelling further predicted that the most endangered massive wooden
elements of the altarpiece could endure significant RH fluctuations of the amplitude of up to
25 % when the initial RH level was 70 %.
The cracking of wood, which has occurred in the past, can increase the allowable
threshold in the microclimate variations [6] as the crack movement will release some of the
stress at the external layer of the wood and facilitate diffusion of the water vapour into and out
of the wood. However, only substantial cracking seems to bring about a significant
‘adaptation’ of a sculpture to the environmental variations.
The results of the present study add new arguments to the discussion of advantages and
disadvantages of intermittent heating in historic churches. Rapid heating of a church for
services is sometimes claimed to be a good strategy because ‘it reduces the relative humidity,
but for such a short time that the painted woodwork does not have time to respond’ [4]. The
lack of an overall dimensional response of massive wooden elements to fast RH changes
gives a false impression of ‘stability’ of the objects. In reality, their external zone, at least to
the depth of several millimetres, continually absorbs and releases water vapour, changing
their size. The resulting gradient of moisture content and stress is the main threat to the
integrity of wood and to the decorative layer.
7
There is a similarity and not a difference between the RH buffering exerted by wood and
by a painted porous wall, described in [4]. The danger in both cases arises from the moisture
movement in the surface layer: in the case of the wall, it will cause salt crystallisation
threatening the adhesion of a wall painting; in the case of wood, it will cause stress leading to
mechanical damage.
Intermittent heating can be economical. It can be also the least damaging strategy for
heating historic churches under the condition that it provides a localised comfortable
temperatures in the area where people are without changing the natural climate of the church
as a whole. The novel heating system developed within the European Project Friendly
Heating (2002-2005) provides direct confined heat to people seating in the pew area. It brings
an enormous improvement in the condition of the altarpiece in Rocca Pietore [15].
Other heat sources like properly positioned electric overhead radiant heaters, warming up
the floor and seats, can be advocated as providing localised heat without adversely affecting
painted walls and the contents of churches. However, care should be taken that sensitive
works of art, responding rapidly to changes in T and RH, like paintings on canvas or wood
panel, are not exposed to intense infrared radiation as they can be damaged by the direct
increase of the temperature and by the associated loss of moisture. Finally, the best form of
localised intermittent heating should be developed individually for each church building as
each constitutes a special case.
Acknowledgements
This research was carried out within the FRIENDLY HEATING project (contract EVK4-CT-
2001-00067), supported financially by the European Commission 5th Framework Programme,
Thematic Priority: Environment and Sustainable Development, Key Action 4: City of
Tomorrow and Cultural Heritage. Thanks are due to Laszlo Bencs (University of Antwerp)
and Henk Schellen (Technical University, Eindhoven) for their information and advice
concerning ventilation. We thank also Mons. Giancarlo Santi and Don Stefano Russo from the
Italian Episcopal Conference in Rome, Don Giacomo Mezzorana of the Diocese in Belluno
and Don Attilio De Zaiacomo, the parish priest in Rocca Pietore, for their assistance in this
study.
Suppliers
Laser displacement sensors: Micro-Epsilon Messtechnik GmbH & Co., Koenigbacher Strasse
15, D-94496 Ortenburg, Germany.
Inductive sensor for monitoring the crack: Keyence Corporation Japan, website:
http://world.keyence.com
References
1 Schellen, H., Heating Monumental Churches, Indoor Climate and Preservation of
Cultural Heritage, Eindhoven Technical University, Eindhoven (2002).
2 Bordass, W., and Bemrose, C., Heating your church, Church House Publishing,
London (1996).
3 Pfeil, A., Kirchenheizung und Denkmalschutz, Bauverlag GMBG, Wiesbaden and
Berlin (1975).
8
4 Padfield, T., Bøllingtoft, P., Eshøj, B., and Christensen, M., Chr., ‘The wall paintings
of Gundsømagle church’ in Preventive Conservation: Practice, Theory and Research,
IIC, London (1994) 94-98.
5 Camuffo, D., Microclimate for Cultural Heritage, Developments in Atmospheric
Sciences 23, Elsevier, Amsterdam (1998).
6 Camuffo, D., Pagan, E., Bernardi, A., and Becherini, F., ‘The impact of heating,
lighting and people in re-using historical buildings: a case study’, Journal of Cultural
Heritage, 5 (2004) 409-416.
7 Italian Standard UNI 10969, Cultural Heritage - Environmental conditions for
conservation. General principles for the choice and the control of the indoor
environmental parameters. Part 1 Microclimate (2002).
8 CNR - Istituto di Scienze dell'Atmosfera e del Clima, Padua, Italy Friendly heating:
both comfortable for people and compatible with conservation of art works preserved
in churches. www.isac.cnr.it/friendly-heating/.
9 Camuffo, D., Sturaro, G., Valentino, A., and Camuffo, M., ‘The conservation of
artworks and hot air heating systems in churches: are they compatible?’, Studies in
Conservation, 44 (1999) 209-216.
10 Bratasz, L., and Kozlowski, R., ‘Laser sensors for continuous in-situ monitoring of the
dimensional response of wooden objects’, Studies in Conservation, accepted for
publication, (2005).
11 Siau, J.F., Wood: Influence of moisture on physical properties, Department of Wood
Science and Forest Products, Virginia Polytechnic Institute and State University,
Blacksburg (1995).
12 International Conference of COST Action E8 on Wood-Water Relations, Copenhagen,
16-17 June 1996, ed. P. Hoffmeyer, Technical University of Denmark, Copenhagen
(1997).
13 Mecklenburg, M.F., Tumosa, C.S., and Erhardt, D., ‘Structural response of painted
wood surfaces to changes in ambient relative humidity’ in Painted Wood: History and
Conservation, The Getty Conservation Institute, Los Angeles (1998) 464-483.
14 Bratasz, L., Jakiela, S., Kozlowski, R., ‘Allowable thresholds in dynamic changes of
microclimate for wooden cultural objects: monitoring in situ and modelling’, in ICOM
Committee for Conservation, 14th Triennial Meeting, The Hague, 12-16 September
2005: Preprints, James & James, London (2005) Vol II 582-589.
15 Bratasz, L., Kozlowski, R., Camuffo, D., Pagan, E., ‘Assessing the impact of heating
systems in churches by monitoring the dimensional response of wooden sculptures’,
Building and Environment, submitted for publication (2005).
9
Fig.1. The main altar of the church of Santa Maria Maddalena in Rocca Pietore, Italy. The
numbers indicate the position of wooden elements monitored for dimensional change:
(1) the saint’s head, (2) the Child’s finger.
1
2
10
Fig.2. A crack in the saint’s head, result of the microclimatic stress.
11
Fig.3. An inductive sensor and a metallic reference plate glued to internal surfaces of a crack
to monitor changes of its width.
12
Fig.4. Indoor and outdoor moisture content in 2003 expressed as humidity mixing ratio
(MR). The plots were smoothed by calculating every five minutes an average of the
data points in the two adjacent 24 hour periods.
13
Fig.5. Indoor and outdoor climate in 2003 as plots of T and RH smoothed as in Figure 4.
14
Fig.6. Relative dimensional change (∆l/lo) of the saint’s head and the Child’s finger during
2003. Records of indoor T and RH are shown for comparison. Averaging as Figure 4.
15
Fig.7. Relative dimensional changes of the Child’s finger recorded every five minutes in
August 2003 are compared with the values calculated on the assumption of full
unrestricted response.
16
Fig.8. Relative dimensional changes (∆l/lo) of the saint’s head (A) and the Child’s finger (B)
in response to three short variations of climate: weekly variation due to a change of
weather outside, indoor diurnal cycle in summer and fluctuation caused by a heating
episode in winter involving an inflow of warm air. Records of indoor T and RH are
shown for comparison.
17
Fig.9. Change in the width of the crack from January to March 2003. Records of indoor T
and RH are shown for comparison.
18
Fig.10. Change in the width of the crack in response to three short variations of climate shown
in Figure 8.
Top Related