You are on page 1of 150

Consolidation Studies

Consolidation Studies 243

ABSTRACT

The Getty Conservation Institute, in


THE GETTY ADOBE RESEARCH PROJECT AT FORT SELDEN.
I. EXPERIMENTAL DESIGN FOR A TEST WALL PROJECT

association with the Museum of New Neville Agnew


Mexico State Monuments, began a Getty Conservation Institute
field-test program early in 1988 at 4503 Glencoe Ave. Marina del Rey, CA 90292
Fort Selden State Monument in U.s.A.
southern New Mexico. The p urpo se
is to evaluate, using adobe test walls,
various preservation materials and Introduction
echni
t ques. These include chemical
consolidants (mainly diisocyanates, At the end of 1986 a three month pilot research project was undertaken at the Getty
echni
and silanes); t ques of
application of consolidants (spraying,
brushing, bulk infiltration), surface
Conservation Institute (GCI) laboratories into the use of various chemical
consolidants and other protective measures for the conservation of historic and
archaeological adobe. While adobe was the primary focus of this research, it was
and depth treatments; drainage, implicit that the findings could have application to other earthen structures as well,
shelter designs and materials, and to outdoor heritage sites in general. The report on this work [1] and subsequent
reburial techniques for archaeological research over a nine month period showed that there was sufficient promise in
sites, and some structural reinforcing chemical consolidants to warrant their field testing. The consolidants that tested
materials and methods. Accelerated most satisfactorily against deterioration by water and salts were aliphatic
aging through water spray has been diisocyanate prepolymers and commercially available alkoxy and alkylalkoxysilanes.
used on the walls and their condition Preliminary results have been published [2-5]. Physical measures of protection and
reinforcement of adobe structures were also planned in a comprehensive program
photographically. This
has been monitored stereo­
paper
presents the experimental design.
of field testing.

The present paper describes the research design and implementation of the
Some preliminary results are given in
subsequent papers in the present continuing field testing at Fort Selden. Preliminary results of the field testing are
preprints. given elsewhere in the present publication. After the first phase of research a
collaborative program was established at Fort Selden State Monument, where the
Museum of New Mexico State Monuments ( MN MSM) had already established a test
KEYWORDS wall facility. By late 1987 the design of the experimental system had been completed
and construction of the test walls was finished in February 1988. Integral aspects of
Adobe, consolidation, site the design of the field project were an accelerated weathering (water spray) system,
and regular temperature, moisture, and stereophotographic monitoring. Treatment
stabilization, site shelters,
isocyanates, silanes, outdoor
accelerated aging, stereophotographic
of the test walls was completed by the end of May 1988 and testing began August
1988 with the start of spraying.
1,
monitoring.
The Need for Adobe Preservation Research

A significant component of the world's architectural heritage is built of earth,


and the problem of preservation and protection of earthen structures is one of
world-wide concern. In the late 1960's an Iraqi-Italian Center for the Restoration and
Preservation of Monuments, with the collaboration of ICOMOS and ICCROM, was
set up to study the preservation of unbaked brick. Its brief was to:

"systematically undertake collection of data, chemical and physical analysis of the


bricks, research into the causes, concomitant and otherwise,1 of their degradation and
be able to suggest methods most suited to give lasting results" [6].

Four international meetings on adobe were organized under the auspices of


ICOMOS (Yazd, Iran in 1970, 1976, Santa Fe, USA in 1977 and Ankara, Turkey in
1980). The recommendations of the United States ICOMOS - ICCROM Adobe
Preservation Working Session in Santa Fe in 1977 [7] called for (among other
things) research into:

-Surface and subsurface drainage of mud-brick structures.


-Chemical surface treatment and consolidation.
-Injection and grouting techniques for consolidation or weather-proofing.
- Structural stabilization and reinforcement materials and methods.

In addition the ICOM-ICOMOS mudbrick symposium in Ankara in 1980 [8]


recommended:

- Design studies be made of low-cost shelters that afford protection against


direct erosion by rain or melting snow, provide thermal insulation but avoid the
"greenhouse effect" and preferably be permeable to water vapor; and that shelter
design concepts be tested in the field.
-Pilot field projects be undertaken to test conservation systems on entire
structures.

In his introduction to the 1980 Ankara symposium on mudbrick preservation


Torraca [9] noted the following points and problems for which a satisfactory solution
was not yet in sight:
Adobe 90

-The long-term reliability and efficacy of partial protection methods


including chemical treatment, capping and coating were questioned on the basis of
recent experience and the high cost of these processes.
-Shelters open on the sides are effective at intercepting rain (and so stop this
5' most damaging factor), but they allow a slow process of crust formation to occur.
This is probably due to humidity/temperature cycles which are intolerable for
decorated surfaces.
-There is a trend to either complete physical protection (by enclosures or
reburial) or reliable maintenance routines.

Clifton [10] has reviewed the status, to 1976, of adobe preservation technology.
He concluded that the successful preservation of adobe depends on effective
protection of structures from the elements, especially water. He emphasized the
need for analysis of the unique set of deteriorative factors operating at each different
structure on site; the selection of preservation materials and standardized test
Figure 2 . Sche m atic diag r a m with
methods; and pointed out the unlikelihood of the existence of a single universal
depth of test wall foundations and
preservation material process. Finally, he emphasized the importance of
typical test w all dimensions.
documenting and monitoring preservation processes that have been implemented.

-
topjvlew�
P"'fA1!'W¥W - � l
The research design presented here and the materials chosen for field testing
address most of the concerns listed above - though not necessarily
comprehensively in each case.

Deterioration of Adobe and Preservation Appro


Much has been written about the effects of environmental factors, particularly
aches

water, on earthen structures. These include weathering by rain and snow, capillary
rise of moisture into structures, migration of soluble salts, moisture-induced
swelling and shrinking cycles as the material is melted, dries and so on. These and
other deteriorative factors such as temperature, freeze-thaw cycles, wind erosion,

3. slde vlew
and biological and human factors have been discussed [10, 1 1].

Figure Sche m atic diag r a m of Ukewise, the pros and cons of various treatment options involving chemical
diagonal plastering of o p p osite
sides of a d obe test walls. The
wall shown for the side view is
and polymer systems have been debated. In
particular the question of penetration
by surface applications and permeability of the treated surface has been one of
concern [12]. For these reasons a variety of treatments, including spray, brushing
one of the half walls. and bulk-infiltration using a gravity-fed reservoir system were included in the
design.

Test Wall Construction

After completion of an archaeological survey of the area for the test walls at Fort
Selden, and final approval of
the test wall design, construction began in December
1987. Figure 1 shows the plan view of the test wall site along with some individual
wall treatment / experimental conditions. The standard test wall was 1.52 x 1.52 m
by approximately 0.25 m thick (5 x 5 feet x 10 inches) constructed from nominal
sized bricks 35.5 x 25.4 x 10 em
(14 x 10 x 4 inches). Wall foundations were 20.3 (8 em
inches), or two adobe bricks into undisturbed earth (Fig. 2). Diagonally opposite
sides of the walls were mud-plastered (Fig. 3).
r1 -16 - 17- -18 - -19 -20- - -21 - 2 23 24 25 26 27 28 Covered

1 c:::::::J c:::::::J c:::::::J c:::::::J c:::::::J c:::::::J c:::::::J c:::::::J c:::::::J c:::::::J c:::::::J c:::::::J& c:::::::J
D l l socyanate - treated Control S l 1 ane (SSH)- treated w wlo
Geodra l n

I1 2940 3041 4231 l3243 c:::3344 ::::J 3445 46c:::35-1 ::::J 47--c:::3648-::::J 49 � 50� �5,:,1 N 1
c:::::::J c:::::::J c:::::::J c:::::::J
S l l ane (SSOH)
-treated
ReburI a l
c:::::::J
I
� � 6 1
er ex le o vered

t E3 11 & c:::::::J c:::::::J 1 c:::::::J c:::::::J c:::::::J c:::::::J c:::::::J t


c:::::::J §
Geobar-
r e I nforced
c:::::::J c:::::::J
w wlo
Geotext l l e
Salt RIse
- - _ _I
Pre l I m Inary Test Wa l l s
(ha l f h e Ight )
Control

Figure 4. Photograph showing


chemical consolidant being applied to
Figure
treatment
1. I
M ap view of test walls at Ft. Selden. N.M. showing wall nu mber and type of
experiment. Area enclosed by dashed line represents walls subjected to
accelerated (water spray) weathering.
adobe test wall by spraying.

Chemical Treatments

Chemical preservation is the application of a substance, usually a polymer in


solution or a monomer which converts in situ into a polymer, and confers its
properties of durability, strength (consolidation), weather resistance and so on.
Consolidation Studies 245

From the extensive literature on the preservation of adobe by chemical treatment,


from the evidence of the present research, it has been possible to define the
requirements of an hypothetical "ideal" system which will minimally alter the
inherent properties and appearance, yet confer durability. These have been reported
[3].

Consolidants - Appli cation Techniques

Consolidant solutions were applied to the test walls by spray, brush and from
reservoirs with infiltration tubes into the core of the wall (bulk infiltration) as
shown in Figs. 4-6. Five basic patterns of consolidant application .were used. These
are: surface coating (using minimal material); multiple surface coating (wet-on­
wet) to achieve good depth of penetration; bulk infiltration to fully permeate the test
wall; surface coating together with basal infiltration of the wall footings; and partial
Figure 5. Photogl·aph showing
infiltration of the top and bottom of the wall and footings, with only light surface
chemical consolidant being brushed
treatment on the vertical face. These patterns are depicted schematically in Figs. 7-1 1
onto adobe test wall.
which were chosen t o provide as full a range o f treatment responses t o the
perceived patterns of erosion of adobe walls.

Control Walls

A series of control walls is incorporated into the experimental design. Thus, there is
a master control wall (#40, used also for thermal monitoring) that is not treated, nor
sprayed, but exposed only to natural weathering; a control wall not treated but
sprayed (#21) to monitor the effect of spraying on natural adobe. For the geodrain
and simulated reburial test walls there are control walls without drains and without
geotextile (#27 and #44, respectively).

In general, each wall can be referenced to one (or more) other wall(s) in which
only one variable has changed, for example, the isocyanate (DN3390) - treated series
(#16-20) is compared directly with the silane (Conservare SSH) - treated series (#22-
26), and each series can be compared with the sprayed control (#21) and the master
control wall (#40). Likewise, the effect of physical protection of adobe against
spraying (simulating rain) by using knitted synthetic fabric ("aerotextile") as a shelter
Figure 6. Photograph showing adobe material can be judged by comparing the sprayed control wall (#21) with a sprayed,
test wall being treated with chemical sheltered wall (#39).
consolidant using bulk infiltration
technique. Accelerated Weathering

It can take a long time for results from field testing under natural conditions to
become apparent. To avoid such delay an accelerated weathering water-spray
system at Fort Selden was used. The results of an artificial testing regime such as the
present one are comparative only and are not intended to be absolute, nor will they
necessarily relate in a quantifiable way to natural weathering. However, the results
do allow certain valid conclusions to be drawn when judging one process against
another and when suitable controls are included in the design.

Important criteria to be decided when initiating a spray regime are: frequency


of spraying, time of day of spraying, volume of water per spray cycle, force of spray
jet, uniformity and pattern of spray impinging on the wall. Also, local rainfall and
meteorological conditions need to be recorded, as these obviously add to the
weathering rate.

Figure 7. Sche m atic diag r a m of


pattern of consolid ation for walls
The initial spray program at Fort Selden proved to be too aggressive and after
two months was stopped during a particularly wet s umm
program comprised two 15 minute cycles daily (at 0900 and 1600
er. The initial spray
hr). Each nozzle
treated by single surface coating .
sprayed approximately 22 Durin
liters of water per cycle. Thus, the total water load per wall
per two cycles daily was 88 liters, or over the two months of initial spraying
approximately 1400 gallons. - g this summer approximately 64 mm (25 inches)
of rain also fell. Several walls (#30, 32, 46) collapsed during storms - presumably
because of wind load on nearly saturated walls and footings.

Monitorin& -- Photographic. Temperature and Moisture

A protocol for stereophotographic recording of the condition of the test walls has
been developed. This comprises 6 x 6 cm format stereopairs taken at night with
standardized lighting to avoid the variable effects of daylight . Approximately once
monthly one side only (to minimize the cost) of each test wall is photographed. A
single batch of film is used and standard processing of the negative is carried out. It
is intended that quantitative information will be derived from the negatives by
Figure 8. Sche m atic diag r a m of
pattern of consolid ation for walls
u sing m ultiple surface coating s.
digitization and image processing. Fig. 12 shows a single contact print of one of a
stereopair.
Adobe 90

Surface and interior temperatures of the master control wall (#40) are
monitored for 24 hours at a time on a seasonal basis using thermocouples and a 16-
channel data logger system. This system is shown in Fig. 13. Thermocouples are
embedded on the surface and at depths of 2.5 emand 12.5 on the east and west
em
faces of the wall both on the adobe brick and the mud-plaster surfaces.

Moisture monitoring in selected walls (the control wall, geodrain walls, and
others such as the basally infiltrated walls) is done using a calibrated resistivity
meter and preset pins. Results so far are of doubtful value. Occasionally moisture
determinations have been done by coring and use of a calcium carbide meter. Core
holes are sealed afterward With silicone rubber plugs.

Protective Structural Experiments

Figure 9. Sche m atic diagra m of


pattern of consolidation for walls
treated by b ulk infiltration.
As stated previously the resolutions of the 1980 Ankara conference on mudbrick [7]
proposed lines of research on shelters and temporary low-cost protection. However,
the design, construction, aesthetics, cost and long-term effects of shelters have not
received much attention until recently. In the protection of adobe sites from the
weather, shelters have an important role.

There are three important criteria that should be considered in research into
shelter design and materials. These, in order of importance, are effectiveness, cost,
and visual or aesthetic impact on the scene. Effectiveness should cover all the
desirable requirements mentioned above for site protection; cost is obviously
important, especially in remote or poor areas, and the aesthetic appeal is relevant
where a site is on show to visitors. All shelters will intrude on the landscape to
some extent, but a sensitive design will clearly find more acceptance that one at odds
with the structure it is protecting. Adobe is especially demanding because of its
earth color, rounded and weathered contours and the fact that often it can be
overpowered by modem materials such as steel which lends itself to rectilinear
design and construction.

Figure 10. Sche m atic diagra m of

treated by surface coating &


pattern of consolidation for walls
b asal
A synthetic, knitted shadecloth (or "aerotextile"), ori ginal
ly developed for the
horticultural industry, was chosen for evaluation as a shelter material. Fig. 14
shows the structure of the 70% shade density grade. Shelter designs usually try to
infiltration.
totally protect easily weathered material like adobe. The basic premise in proposing
this open-weave textile for testing was that in many cases a completely water-proof
shelter may not be needed, particularly if the adobe site can be treated, as part of the
site protection program, with a chemical preservative.

The advantage in testing this aerotextile for shelter use are:

-It breaks the force of rain and will prevent snow from accumulating on the
structure.
-It has a wind attenuation factor of 60-70 percent (for the 70 percent density
fabric, as shown in Fig. 15). It thus breaks the force of the wind, without any
possibility of creating dangerous turbulence which is always a hazard with shelter
roofs and walls that are solid. It will also cause wind�bome sand to be deposited if
side screens are used.
-The textile is available in a range of "shade densities" of 40, 50, 70, 90 percent
Figure 11. Schem atic diagr a m of
pattern of consolid ation for w alls
light reduction.
-It can be used to ameliorate thermal stress on fragile or heat-susceptible
structures (e.g., east-facing walls) when an appropriate shade density is chosen.
treated by surface coating and
partial infiltration at top and b ase.
-It will inhibit growth of vegetation on sites by cutting down light. In
ruins growth of weeds and robust plants causes great damage through invasive root
systems. Thus it will also cut down on the need for expensive maintenance and
many

herbicide treatment.
-It is inexpensive; has a 25-year guarantee in any outdoor situation (the
polymer is UV-stabilized polyethylene) and is light in weight and easily transported.
It could find application in remote areas and in third-world countries where
relatively unskilled labor could be used in construction.
-Being light in weight, open-weave, and tear-proof it does not demand the
use of massive (and costly) support structures but can be constructed on light tubular
steel and wire framework. Methods of fixing the textile to free-standing frame­
works are well established. In addition, conventional shelters of steel, bricks or
concrete are normally supported by robust pillars, the installation of which pose a
threat to a site both in the excavation of post holes and in the use of heavy
machinery for the handling and erection of the structure.
Figure 12. Contact print of one of a -It allows circulation of air and thus avoids the stagnant microclimates
pair of stereophotographs taken of characteristic of totally enclosed protective shelters.
test wall #20 (DN3390-treated). Note -It is available in a variety of colors, beige, brown, black, green, white --- the
cracking of outer plaster surface from beige probably being most suitable for an adobe site.
accelerated (water spray) weathering. -If handled sensitively in shelter design it need not intrude unduly on the
Consolidation Studies 247

historic landscape of the site it is protecting.

There are apparent disadvantages in the use of this material:

••
Being open-weave it is not water-proof.
It is not fire-proof and is easily vandalized..

It will tend to sag slightly between supports and cause drips at particular
points. These will erode holes in adobe.

The main disadvantage of the textile, that of being non-waterproof, can


probably be overcome by treatment of the material being protected with a chemical
preservative; deep surface impregnation rather than bulk infiltration may be
sufficient.

Technology transfer of existing materials - such as aerotextiles and geotextiles


(as discussed in the next section) - from the horticultural and civil engineering fields
respectively, to applications in adobe site-preservation research, indeed, heritage
Figure 13. Photograph of control wall sites in general, has many clear benefits. The development work has been done and
(#40) with thermocouples hooked up the materials are being marketed internationally and have proven acceptance and
to data logger for exterior surface and durability. Furthermore, they are often available in a range of grades which allows
internal temperature monitoring. matching with site requirements. Thus, by comparison with steel-roofed
constructions, material such as aerotextiles would appear to have great advantages
in cost-effectiveness, ease of erection, and aesthetic acceptance with minimal
intrusion on the historic site landscape. Two aerotextile shelter designs were used at
Fort Selden and are shown in figure 16. These are an A-frame shelter (walls #38
39) and a puckered hexagonal ring structure (wall #37).
&
Ground-Water Control and Water-Proofing Foundations.

The problem of ground-water control has been reviewed thru 1979 by Clifton and
Davis [13]. These authors also discuss the possibility of water-proofing foundations
and underlying soil and conclude that the approach appears worthy of evaluation in
field testing.

In general, ground-water control is best handled by installation of drainage


systems. Trials are being carried out at Fort Selden on drains of advanced design.
These are less intrusive of soil (and archaeological deposits) around the base of
adobe structures than traditional gravel-filled drains (Fig. 17). Stripdrain and
Cordrain are such drains. They consist of a non-woven polyolefin geotextile sheath,
as a filter to prevent clogging, surrounding a dimpled plastic core (40 rom wide in
the case of Stripdrain). This material is superior to conventional slotted agricultural
drain-pipe on the following counts:

Fimatgureeria14.l wiStthructureshofadeaerdensotextitiyle.
70%

It is more efficient as a filter in preventing blocking.

It is able to intercept water flow in several subsoil layers and enables
drainage of high volumes.
(50 Ekmstlimharea)toientdefrxmeotinleatnomafwiatnerdiabslepehinded ,
A




It is tolerant to grade inconSistencies over long lengths.
It can be installed in narrow trenches with minimal excavation of subsoil.
It is stiff vertically and can bridge small variations in the trench base.
It is flexible and can be bent around comers.
�4O
l�-g -g 2030 � ".::;;.0"h::...0 .... ..f:.......$';'..... �
�:..
B

Reburial

�:E1l 100oWind20sP"pe d40in fron60t of ma80terial100'



Many authorities consider reburial of an excavated site - adobe as well as other
categories of archaeological sites - as the only responsible measure when protection
and preservation are not possible [14]. Adobe sites, in particular, deteriorate rapidly
on exposure as previously stated [14]. Yet little work has been done on the
Fig ure 15. Graph showing effect that
different density g r ad e s of aerotextile
if is
effectiveness or otherwise of reburial as a preservation method. Soil disturbance on
excavation increases permeability, and the excavation material is used as backfill
the percolation of water and aq:ess of air increased. Thus, reburial may, in some
B.
80\, 40\, 40\, 50\, 70\,
Curves A ,
den sity grades of
and
C, D, E
material has on mean wind spee d .
correspond to

res pectively. Solid line


instances, cause unwitting accelerated deterioration.

a
A common practice in reburial is to use layer of clean sand as a marker in
represents measured wind spee d ,
dashed line represents predicted wind
speed. ( I nform ation provided by
Technisearch, Melborne, A u stralia).
the event of future re-excavation. In
remote areas sand is not always available and
transport costs may be high. Sand is ineffective in covering and delineating vertical
surfaces that are being reburied. These considerations have prompted consideration
of the need for assessment of geotextiles as a reburial shroud for adobe (and other)
structures. This fabric is available in different weights and is used as the external
sheath in the geodrain systems already discussed. The fabric is permeable and will
not create wet, anaerobic conditions around structures that plastic sheeting - which
has been used on some sites - may cause. It will conform to the profile of a structure
and provide physical protection on both burial and re-excavation.

Two walls (#43, 44) have been used for this experiment, in which half the
Adobe 90

wall was built below grade (see


Fig. 18). The experimental wall was draped with
Mirafi 140 NS geotextile and then buried until only the top of the wall showed. The
control wall was was not draped but was similarly buried. Both walls have been
water sprayed.

Structure Reinforcement

Damaged adobe sites may result from weathering or seismic activity. The
effectiveness of synthetic internal reinforcing bars ('geobars') on test walls at Fort
Selden is being assessed. The bar is a rigid, fiberglass rebar 12.5 mm in diameter,
externally ribbed. It is non-rotting and non-corroding. Test wall configurations
similar to that shown in Fig. 19 are being tested. Two methods of inserting the rods
Figure 16. Photograph showing
adobe test walls (#38, 39) protected
by A-frame, aerotextile-covered
have been used. In both a 13 mm is
pilot hole is first drilled into the adobe wall. This
is then filled with water and the resulting mud extruded using a 50 mm auger bit
(with extensions as necessary). The geobar is inserted and in one case the hole is
shelters, and (#37) by hexagonal filled with mud, in the other polyester resin was poured to fill the hole, thus giving
shaped 'puckered' shelter with a 50 mm diameter rigid rod. Two walls (#41, 42) were reinforced with the geobars;
aerotextile roof. however wall #42 has collapsed leaving one geobar-reinforced wall for evaluation.

Conclusion

The Fort Selden field experiment was designed and set up as a large-scale outdoor
laboratory experiment to assess certain treatments as a preliminary to real-site
implementation.In order to achieve results within a reasonable time frame an
accelerated weathering spray system was used. Results of two and a half years of
is
combined natural and artificial weathering are appearing. It expected that final
comparative evaluation of the various measures implemented on the test walls will
be possible by the end of 1991.

Acknow ledgements
The author acknowledges with gratitude the contributions of his co-workers at the
GCI, Dr. Charles Selwitz, Richard Coffman, & Dr. Frank Preusser. Special
recognition is due James Druzik (GCI) and Thomas Caperton (MNMSM) who
helped initiate the project. Michael Taylor (NPS formerly with MNMSM) who
worked on site in May 1988 with the author and has been an active collaborator
throughout. Frances Gale of Pro So Co also worked on site applying the Conservare
H & OH consolidants generously donated by her company. Mobay Corporation has
unstintingly supported the project through donations of diisocyanate prepolymers,
and Mike Jakstis of Mobay has provided much useful guidance. Paul G. McHenry Jr.
(designer of the test site) and his sons Bruce (construction) and Jamie
(photomonitoring) have likewise been integral components. The Fort Selden staff,
especially Jose Guzman and Elva Melendez have provided help whenever it has
been needed. Thomas Moon developed the photodocumentation protocol.

Note: Technical specifications and plans for the construction of the test walls and
the spray system are available from P. G. McHenry, Jr.,
Albuquerque, N.M. 87107.
834
Griegos Rd. NW,
References

1. Agnew, N., "Adobe Preservation", A Report on a Three Month Research


Project at the Getty Conservation Institute. (1987), March, 105 p.

Figure 17. a) Close up ot geodrain


showing black plastic core and white
2. Agnew, N., Druzik, J. R, Caperton, T. and Taylor, M., "Adobe: The Earliest
Composite Material" : in Working Group 10. ICOM Committee for Conservation.
Preprints, v. 2, Sydney, Australia, (1987), 439-446.
geotextile fabric. Manufactured
under the name Cordrain, by Nylex
Corp., Australia. b) Photograph
showing geodrains in place and being
3. Agnew, N., Preusser, F. and Druzik, J. R,
"Strategies for Adobe Preservation" :
in 5th International Meeting of Experts on the Conservation of Earthen
Architecture. Rome, ICCROM, October 22-23, 1987, Guirimand, Grenoble, France,
buried by author.
(1988), 3-12.

4. Coffman, R L., Agnew, N. and Selwitz, c.,


''Modification of the Physical
Properties of Natural and Artificial Adobe by Chemical Consolidation" : in
Proceed Symposi
ings of um G: Materials Issues in Art and Archaeo logyII.
Materials
Research Society Spring Meeting, San Francisco, California, USA, April 16-21, 1990,
(InPress).

5. Hansen, E. F. and Agnew, N., "Consolidation with Moisture-Curable


Isocyanates: Polyureas and Polyurethanes" : in Proceedings of the 9th Triennial
Consolidation Studies 249

Meetin& of the Comm Conserv


ittee for ation. The international Council of
Museums. Dresden, German Democratic Republic, August 26-31, 1990, Press). (In
6.

473.
Bruno, A., Bultinck, G., Chiari, C. and Trossarelli, c., ill-IV,
"Contributions to the
Study of the Preservation of Mud Brick Structures", Mesopotamia. 1969, 443-

7. Third International Symposium on Mudbrick (Adobe) Preservation, 29


September - 4 October, 1980, Ank
ara, ICOM and ICOMOS, p. 273.

8. ibid., p. 281.

9.

10.
Torraca, Ci., ibid., p. VI.

Clifton, J. R., "Preservation of Historic Adobe Structures - A Status Report",


Report No. NBS TN-934 of National Bureau of Standards, Washington, D.C., 1977.

11. Ref. 1., p. 20-21.

12. Ref 1., p. 53.

13. Clifton, J. R. and Davis, F. L., "Protecting Adobe Walls from Ground Water" ,
National Bureau of Standards, Washington, D. c.,
PublicationNBSIR 79-1730, 1979.

14. Carter, T. H. and Pagliero,


(1969), 22, No. 1 and 2, 65-76.
R., "Notes on Mud-Brick Preservation", Sumer

Figure 18. Photographs showing


geotextile-covered and uncovered
walls (#43, 44, respectively) a) prior to
burial, and b) burial. after

geobars
cFoignutraeini1n9g. sStcrhuemacturtailcrdeiiangfroarcminogfetleesmt waentlss.
Adobe 90

ABS TRACT THE GETTY ADOBE RESEARCH PROJECT AT FORT SELDEN I I .


A Study o f the I nt e ract i o n o f Chem i c a l C o n s o l i dant s
Treatment o f adobe with with Adobe and Adobe Const ituent s
hexamethy lene di i s ocyanat e ­
de rived po lyme r s a n d s i l an e s
i n c r e a s e s compre s s ive Ri chard C o f fman , Char l e s S e l w i t z and Nevi l l e Agnew*
st rength and con s o l idat i on , Getty Conservat i o n I n st i tute
and enhanc e s r e s i st ance t o 4 5 0 3 Glencoe Ave . Marina D e l Rey , CA 9 0 2 9 2
di s aggregat i on by wat e r . To ( * ) Author t o whom corre spondence should b e addre s s ed
better unde r st and the
modi fi cat i o n s of the
propert i e s of natural adobe ,
the e f fects o f t re a t i n g
individual components o f
adobe ( c l a y , s i lt , a n d s and)
Background and Introduction
T h i s study was i n i t i ated by our interest in chemi c a l and phy s i c a l
methods o f pres ervat i on o f h i s t o r i c a n d archae o l o g i c a l e arthen
with chemi c a l c o n s o l idan t s s t ructure s , spec i f i c a l l y adobe . Many cultural s t ructures are
were examined . Several constructed o f adob e , wh i ch e a s i l y l o s e s s t ructural integrity when
d i f fe rent c l ays ( Na- and Ca­ wett e d . P re l iminary l aboratory work de alt with the inte ract i o n o f
montmo r i l l o n i t e s and k a o l i n ) various chem i c a l con s o l idants w i t h adobe [ 1 ] . After i s ocyanates
were t reated with i s o c yanate and s i l an e esters were shown t o be succe s s fu l i n con s o l i da t i n g
and s i l an e s a s we re mixtures adobe , w o r k f o c u s e d on the interact i o n o f those c o n s o l idant s w i t h
of c l ay and quart z s an d , p u r e c l ay s , s ince these are the component s most a f fe cted by
quart z s a n d a l on e , and wet t ing . T h i s pha s e o f rese arch invo lved study i ng the e f f e ct s o f
adobe . X - r ay di f fract i o n speci f i c chem i c a l con s o l idant s ( i s o cyanate [ hexamethy lene
( XRD ) ana l y s i s , s canning d i i s o c yanate ( HD I ) -de r i ved p o l yme r s ] and s i l ane s ) on pure c l ays .
e l ectron m i c r o s c opy ( SEM) , The r e su l t s showed that t r e atment of c l ay s with i s ocyanat e s and
grain s i z e ana l y s i s , water s i l a n e s markedly increased the c l ay ' s wat e r repe l l en cy but did not
s o rp t i o n ana l y s i s , and con s o l i date the c l ay [ 2 ) . Although chem i c a l t r e atment inhibit s
compre s s i o n t e s t i ng we re wat e r upt ake and p a rt i a l l y b i nds the c l ay part i c l e s t ogethe r , it
used t o evaluate the e f fe c t s does not prevent failure o f indivi dual c l ay p a rt i c l e s when
o f t h e chemi c a l s o n sub j e cted to deformat i onal stres s .
con s o l i dat i on , modi f i cat i on
o f compr e s s ive s t rengt h , and
r e s i s t ance t o d i s aggregat i o n
by wat e r in c l ay s , s an d - c l a y
Mate als Met odo ogy
ri and h l

mixture s , a n d adobe . The


r e s u l t s of the s e evaluat i o n s
i n d i c a t e t h a t the c l ay type
next s t ep was t o make an

c l ay .
artificial
Aft e r f a i l ing to c o n s o l idate pure c l ay s by chemi cal t r e atment the

mm)
adobe by comb i n i n g three part s
quart z s and (with a un i f orm part i c l e s i z e of 0 . 5 with one part
T h i s m i xture r e s emb l e s a " t yp i c a l " adobe and, when wet t e d ,
p l ay s a s i gn i f i cant role in resulted i n a workab l e , coherent , art i f i c i a l adobe mat e r i a l .
the prope rt i e s o f Three s e t s o f a rt i f i c i a l adobe were made us ing three d i f f e rent
i s ocyanat e - and s i l a n e ­ c l ays obt a i ned from the Clay Mine r a l s S o c i e t y , S ource C l ay
t reated adobe . Mine r a l s Rep o s i t o r y at the Depa rtment of Ge o l ogy , Unive r s i t y o f
M i s souri , Co lumb i a , Mi s s ouri . T h e c l a y s are :
A) SWy- 1 , Na-montmo r i l lonite , Crook Count y , Wyoming ;
KEYWORDS B ) SAz - 1 , Ca-montmo r i l lonite , Apache Count y , Ar i z on a ;
C) KGa- 2 , Kao l i n , Warren Count y , Georg i a .
Adob e , c l ays , compre s s ive
s t rength, c o n s o l idat i o n , T e s t p lugs were manuf actured by comb i n i n g the s and : c l a y
i s ocyanat e s , s i l a ne s , wat e r mixture with wat e r t o form a v i s cous s l urry t h a t was p o u r e d i n t o
repe l l ency , x-ray sma l l ( 2 2 m m x 4 0 m m ) cyl indr i c a l molds . P l ugs o f adobe were
d i f fract i o n anal y s i s manu factured in a s im i l a r fashion by d i s aggregat i n g adobe b l ocks
obt a i ned from Ft . S e l den , New Mex i c o , then reconst itut i n g the
mat e r i a l with wat e r into a v i s cous s l urry and recasting it in the
cy l indr i c a l molds . Pure quart z p l ugs were a l s o made by g r i nding
the qu art z s and t o obt a i n a var i ab l e s i z e range d i s t r ibut i on .
Of
TABLE I : Evaluation

C l ay type Mean Stress


Compression Tests
on Sand:Clay Plugs

& &
(Letter refers (Std. D e v i a t i o n )
Mean Modulus
(Standard
< 4 5 �m = 5 % , 4 5 �m to 7 5
�m = 8 0 % .
the molds .

T h i s was then s i eved and recombined in the f o l l owing proport i ons :
= 5 % , 7 5 �m to 1 8 0 �m = 1 0 % , and > 1 8 0
T h i s mat e r i a l was then mixed with water and poured into
After the molds were f i l led they were p l aced in an
to f i gures at Fai l ure) in Deviation) i n oven set at 5 0 0 C to dry . The r e s u l t i n g s and : c l a y , adobe , and
I and 2 ) MPa MPa quart z p lugs we re removed from the molds and give n t ime t o
equ i l ibrate with the amb i ent l aboratory condi t i on s ( 2 2 0 C , 4 0 - 5 0 %
KGa-2 ( A ) 0.099 (O.027l 1 1 . 1 24 (4.305)
RH) •
KGa-2
KGa-2 (C)
(6) 2.695 (0.360)
3.220 (0.9 1 6)
I 1 6.563 (35.335)
1 47.49 1 (64.594)

KGa-2 (E)
KGa-2 ( D )

SAz - I (A)
2. 1 96 (0. 1 60)
1 .627 (0.200)

0. 1 3 1 (O.037l
1 24.9 3 1 (42.099)
92. 1 69 ( 1 5.699)

1 0.776 (5.4 1 6)
Aft e r equ i l ibrat i on to a const ant weight , the p lugs were
t reated with e i ther an i s ocyanate o r a s i l an e . The i s o c yanate ( an
HO I - de r i ved prep o l yme r ) i s manufactured comme r c i a l l y by Mobay
C o rporat i o n under the name De smodur N- 3 3 9 0 � ( DN- 3 3 9 0 � ) and i s
SAz- I (B) 1 . 1 04 (0.2 1 2) 74. 1 67 ( 1 9.636)
produced a s a 9 0 % con cent ration i n a n aromat i c hydro c a rbon and n­
SAz- I (C)
SAz- I
SAz- I
(0E) 2.676 (0.306)
1 .400 (0 226)
0.93 1 (0. 1 40 )
1 6 1 .605 (66.047)
99.407 (27.434)
67.423 (25.340)
but y l acetate mixed s o lvent . Treatment requ i re d d i luting the DN-
3 3 9 0 � with a 2 : 1 xylene : methyl ethyl ketone ( ME K ) s o l vent mixture
t o f o rm s o l ut i o n s with concent rat i on s o f 5 % t o 2 5 % . The s i lanes
SWy- 1 (A) 0.0 1 5 (0.003) 1 .646 (0.6 1 9) were t et raethy l o rtho s i l i cate ( TEOS , ( C 2 H 5 0 ) 4 S i ) , with and without
SWy - 1 ( B ) 0.099 (0.04 1 ) 1 3.746 (6.042)
methy lt r i ethoxy s i l ane (MTEOS , CH 3 ( C 2 H 5 0 ) 3 S i ) , wh i ch is added t o
SWy - 1 (0)
SWy - 1 ( C )

SWy- 1 ( E l
0.297 (0.297)
0.456 (0.059)
0.365 (0 027)
20.353
45.50 I
27.22 1
(I
< 3 . 79 1 )
0.637l
(5 356)
achi eve wat e r repe l l ency . According to the manu facturer these
cont a i n 75% act i ve s o l i ds in an acetone -MEK s o l vent and are
prod�ced by P r o S o Co I n c . under the name Con s e rvare S t on e
TM
S t rengthener H ( S S - H , with MTEOS ) a n d Cons ervare S t o n e
St rengthener OH� ( S S -OH , without MTEO S ) . The s i l an e s do not
requ i re addit i on a l mixing with s o lvent s . The i s o c yanate
po l yme r i z e s i n s itu by react ing with wat e r inherent l y p r e s ent in
the c l ay s t o f o rm urea l inkages with r e l e a s e o f carbon d i o x i de .
The s i l a n e s po l yme r i z e by hydro l y s i s with atmosphe r i c and in s i t u
m o i sture , wh i ch creates a network S i - O p o l ymer that t i e s int o and
Consolidation Studies

c ro s s l i nks the substrate being t r e at e d .


EVALUATION OF COMPRESSION
TESTS ON SAND:CLA PLUGS
lEI�•SS-SS-OY H
DN-3390

Untreated
E ach plug was t re ated with 3 to 6 ml o f s o l u t i o n , then
covered to inhibit s o lvent evapo r at i on and p e rmit tho rough
penet r at i on of the cons o l i dant int o the p l ugs .
and curing, the plugs were t e s t e d for compre s s ive s t rength,
After t r e atment

r e s i s t ance to d i s aggregat i on i n wat e r , and p o r o s i t y . A set o f


unt reated plugs was a l s o t e s t e d t o deve l op b a s e l ine value s .
Comp r e s s ive st rength determinat ions were obt a i ned u s ing an I n s t ron
Value. In ( ) are
polymer loading Unive r s a l Testing I n s t rument , mode l 4 2 0 1 , with a 5 0 0 Kg cap a c i t y
In wL % compre s s ion l o a d ce l l . T h e s amp l e s w e r e t e s t e d u s ing a c r o s s head
speed of 5 mm/minute with a s amp l i ng rate of 10 p o i n t s per s e cond .
Wat e r repe l l ency and r e s i s t ance to d i s aggregat i o n by wat e r was
WQ.>­I- <� det ermined by subme rging the s and : c l a y , and adobe plugs in wat e r
f o r approximat e l y 4 8 hours . T h e p l u g s w e r e removed f r o m the wat e r

><..Jo en a n d weighed, p l aced in a n oven a t 5 0 0 C for a minimum o f 4 8 hours ,


rewe i ghed, and p l aced back int o wat e r . The p l ugs· went through ten
wet-dry c y c l e s o r unt i l they began d i s aggregat ing . Me rcury
i n j e c t i on p o r o s imetry was used f o r p o r o s i t y meas urements on sma l l
p i e c e s ( approxima t e l y one t o two cub i c cent imet e r s ) o f t re ated and
unt reated adobe s amp l e s .

The re l at i on ship o f i s o cyanate s o l ut i on concent rat i on and


number of app l i cations with depth o f penet rat i o n and con s o l i dation
was also exami ned . Three i s o cyanate s o lut i on s ( DN- 3 3 9 0 � MEK )
+
were mixed in concent rat i ons o f 2 5 % , 1 2 . 5 % , and 6 . 2 5 % . Each
s o l ut i on was brushed onto a 1 6 cm x 10 cm area of an adobe b l o c k ,

O. 1
MEAN STRESS AT FAILURE
IN MPa
3 4 from Ft . S e l de n , New Mex i c o , i n 1 , 2 , 4 , 8 , a n d 1 6 coat
app l i cat i on s . The t reated b l ocks were wrapped in p l a st i c , cured
for two wee k s , then uncovered and cut i n ha l f . One hal f was
archive d f o r future study wh i l e the other hal f was p l aced i n a
F i gure 1 . Re s u l t s o f l a rge cont a i n e r of wat e r t o wash away the uncon s o l i dated p o rt i on .
compre s s ion t e s t s showing
mean s t r e s s at failure for
unt reated and i s o cyanat e - or
s i l an e - t re at e d s and : c l a y
Experi nta Resu
me l lts
The r e s u l t s o f compre s s ion t e s t s f o r the s and : c l a y ( artificial
plugs .
artificial
adob e ) plugs a r e shown in t ab l e I and f i gure 1 .

KGa - 2 »
Unt reated
adobe plugs failed at l ow s t re s s l e ve l s ( 0 . 0 2 MPa t o
0 . 1 3 MP a ) i n t h e fo l l owing order o f decre a s ing s t rengt h , SAz - 1 >
SWy- 1 . Chemi c a l t r e atment marke d l y incre a s e d the
compre s s i ve s t rength o f the plugs and a l tered the s t rength
re l at i on s h i p s t o the f o l l owing , KGa - 2 > SAz - 1 » SWy- 1 . Both
i s o c yanat e - and s i l ane-treated p lugs exhibited between a s i x - to
t h i rt y-two - fo l d increase i n compre s s i ve s t rength , based on mean
s t re s s value s . P lugs made with c l ay SAz - 1 exhibited a twent y - two­
f o l d increase i n comp r e s s ive st rength when t reated with
i s ocyanat e , and a ten- f o l d increase when t re ated with the s i lane s .
Treatment with the s i l an e s resulted in a thi rty- f o l d increase i n
comp r e s s ive s t rength for tho s e plugs made with c l ay SWy- 1 , whe reas
TABLE II Evaluation of Compression Tests
on Adobe, and Quartz Pl ugs
i s o cyanate t reatment only resulted in a twent y - f o l d increase .
Both i s o cyanate and s i l anes caused comparab l e i n c re a s e s in the
comp r e s s ive s t rength o f plugs made with clay KGa-2 .
Plug Type
( L e t ter refers
Mean Stress & &
(Std. D e v i a t i o n
Mean Modu 1 us
(Standard
Table I I and f i gure 2 show i s o cyanate t r e atment o f the adobe
to figures
3 and 4) at Fai lure)
MPa
In Deviation) i n
MPa plugs r e s u l t e d i n a three- to twe lve- f o l d increase i n compre s s ive
s t rength ove r unt reated adobe . Treatment with e i ther s i l ane
Adobe (A) 0.975 (0. 1 70) 93. 1 99 (45.338) resulted i n a s i x - f o l d increase in the compr e s s ive s t rength
r e l at i ve t o unt reated adobe . For most o f the adobe s amp l e s t e sted
Adobe (8)(C) 3 .2 1 8 (0.558) 28 1 . 1 55 ( 1 1 8.578) the s i l ane-t reated plugs exhibited a greater compre s s ive s t rength

((0F))
Adobe 4. 1 57 (0.877) 274.9 1 5 (53 243)
than the i s o cyanate-treated plugs . However the s i l ane l o ading of
Adobe 4.679 ( 0 557) 338.6 5 1 (50 280)
Adobe ( E ) 8.070 (2.236) 399.022 (97.507)
4 . 3 % - 4 . 5 % was great e r than the l o ading on any of the i s ocyanat e ­
Adobe 1 2 082 <3. 342) 5 1 1 .326 (78 337) t reated s amp l e s ( s e e f i gure 2 ) . In spite o f t h i s , an i s o cyanate
l o ading of 3 . 2 % resulted in a great e r increase i n comp r e s s i ve
Adobe (G)
Adobe

Adobe
((HJ)) 5.862 (0.884)
5.667 (0 723)

1 .466 (0 43 1 )
47 1 .56 1
453.604 (88. 1 98)
( 1 5.782)

89.5 1 1 ( 1 4.2 1 9)
s t rength than was observed for the s i l ane-t reated s amp l e s .
t ogether . with the r e s u l t s i n t ab l e I and f i gure 1 , indicate that
at equ ival ent l oadings the i s ocyanat e s would be supe r i o r to the
s i l anes for enhan cing the compre s s ive st rength o f
Th i s ,

artificial
and
Adobe 2.305 ( 0 3 1 8) 1 73.542 ( 4 1 .45 1 )
Adobe ( K ) 2 . 5 5 8 (0.595) 1 99.973 ( 1 9. 1 50)
natural adobe .
Adobe (L) 2.482 (0.474) 1 87.2 1 3 ( 1 8.090)
T ab l e I I and f i gure 2 a l s o show that wet , i s ocyanat e - t re ated
Quartz (A)
Quartz
Quartz ( C )
Quartz (0)
(8) 0.05 1 (0.0 1 5)
3.469 ( 1 .40 1 )
1 .700 (0 39 1 )
0.5 1 7 (0. 1 34)
I
9 029 (6.807)
245.830 ( 1 56 099)
1 0.56 1 (58.729)
45.776 ( 1 5. 1 54)
adobe h a s a compr e s s ive st rength s i mi l a r t o dry , unt reated adobe .
I s o cyanat e - t reated adobe sub j e cted to mult i p l e wet - dry c y c l e s i s
weaker than a s im i l a r l y t reated adobe not expo sed t o wet - dry
c y c l i n g , a l though it st i l l rema ins st ronger than the init i a l
unt reated adobe . The compre s s ive s t rength o f untreated and
t re at e d pure quart z p lugs is a l s o presented i n t ab l e I I and f i gure
2. I s o cyanate t r e atment increased the compre s s ive s t rength o f the
qu art z plugs e i ght y - f o l d , whe reas s i l ane t r e atment r e s u l t e d i n a
t en- and t h i rty-t hre e - f o l d increase ( S S - OH and S S - H ,
re spe c t i ve l y ) .

F i gure 3 demon strates how chem i ca l l y t reated s and : c l a y . and


adobe p lugs withstood d i s aggregation i n wat e r r e l at i ve t o
unt reated p l ug s , and t o e ach othe r . Unt re at e d adobe a n d p lugs
made with quart z s and and c l ay SAz - 1 o r KGa - 2 ( C a - montmo r i l l o n i t e
Adobe 90

and ka o l i n , re spect ive l y ) d i s integrated as soon as they came i n


EVALUATION OF COMPRESSION cont a ct with wat e r . T h e plug made w i t h c l ay SWy - l (Na­
TESTS ON ADOBE,
AND QUARTZ PLUGS
montmo r i l l o n it e ) did not d i s aggregat e ; instead i t ab s o rbed wat e r
a n d swe l l e d into a ge l at inous ma s s . F i gure 3 a l s o shows that
!lEI�S�• S5-HO
DN-3390

Untreated
p lugs made with k a o l in were very r e s i s t ant t o d i s aggrega t i on by
wat e r once they had been chemi c a l l y t reated, regardl e s s of the
type of c o n s o l idant u s e d . I s o cyanate t r e atment provided exce l l eDt
re s i s t ance to det e r i orat i on by wat e r for the adobe p l u g and only
s l i ght l y l e s s e r protect i on for the s and : c l a y plug made with c l ay
Plugs containing 0.6%
DN-3390, which have SAz - 1 whi c h det e r i o r ated during the tenth wet - dry cyc l e . Howeve r ,
gone through wet-dry the i s o cyanate and S S - H had l imited succe s s i n prevent ing
cycling
d i s aggregat i o n o f the p lugs made with c l ay SWy - 1 wh i ch began

/ d i s a ggregat ing a ft e r four wet -dry cyc l e s . In cont r a s t , S S - OH


provided very l i t t l e protect i on against det e r i o rat i o n by wat e r for
the s and : c l a y p lug made with c l a y SWy- 1 . Ne i ther o f the two

GIf)..::.J:I s i l a n e s provi ded protect ion for plugs made with c l ay SAz - 1 .
Howeve r , both s i l anes were p a rt i a l l y succe s s ful in protecting the
adobe plug against rapid det e r i o r at i on by wat e r , with S S - H

WoIID.I. gglllill±illlilillill]
(2.4%)
(3.20/0)
outperforming S S -OH .

The eva luat i on of po l ymer concent ration and number o f


app l i cat i ons with depth o f penet rat i on reve a l e d that the de epe s t
penet rat i on was atta ined by app lying s i xteen coat s o f t h e most

<I: Values In ( ) are


polymer loading
di lute ( 6 . 2 5 % ) s o lut i on . F i gure 4 shows the d i f ference i n depth
of penet rat i on between natur a l adobe b l o c k s t re ated with the 6 . 2 5 %
s o lut i on a n d t h e 2 5 % s o l ut i on _ S i xteen coats o f the 6 . 2 5 %
In WI.
% s o lut i on con s o l i dated approximat e l y 9 2 0 c c o f adobe , wh i l e s im i l a r
app l i cat i o n s u s ing concent rations o f 1 2 . 5 % a n d 2 5 % r e s u l t e d i n

MEAN STRESS AT FAILURE


IN MPa
1 con s o l idat i on o f 6 8 0 cc a n d 4 7 0 c c o f adobe , respect ive l y .

P o r o s i t y me asurement s , u s ing me rcury i n j e ct i on p o r o s imet r y ,


we re made on a number o f ' t re ated and unt reated adobe p l ugs . The
F i gure 2 . Re s u l t s o f p o r o s i t y of unt reated s amp l e s var i ed , ranging from 1 3 % to 2 5 % _
comp r e s s i on t e s t s showing P o r o s it y me a s urements of the t reated plugs a l s o f a l l within t h i s
mean s t r e s s at f a i lure f o r range . Because o f thi s , no d i f ference in p o r o s i t y was
unt reated a n d i s o cyanat e ­ o r d i s t inguished between unt reated and t reated s amp l e s . However the
s i l ane-treated adobe , and sma l l volume of the test specimens is probab l y i n s u f f i c ient to
qu art z p l ug s . ensure s t at i s t i c a l homogene i t y , and a i r voids could e a s i l y a c count
for the w i de range o f values

N u m b e r of w e t - dry cy c l e s
n e e d e d to c a u s e f a i l u re
Discllssion
o 2 4 6 8 10 Tables I and I I and f i gures 1 and 2 , show that the comp re s s ive
s t rength of adobe is far greater than that of s and : c l a y p l ugs .
T h i s i s be l i eved to be a funct ion of the grain s i z e d i s t r ibut ion

Nt9('\JI and compo s i t i on . Natural adobe i s made of p a rt i c l e s with a wide


grain s i z e d i s t r ibut ion ranging from
to c l ay- s i z e .
p ebb l e - or granu l e - s i z e down
Therefore the part i c l e s are c l o s e l y p a cked with

Y:: abundant grain-to-grain contact s , minima l pore space , and good


cohe s ion . I n addit i on , much of the adobe mat e r i a l is comp o s e d o f
hard, r e s i st ant p a rt i c l e s such a s quart z o r f e l dsp ar . When t h i s

..«D0<l> «NI mat e r i a l i s t reated w i t h i s o cyanate or s i l ane e s t e r s , a p a rt i c l e ­


t o -part i c l e bond i s formed by the po lymer which g l u e s t h e ma s s
t o gethe r and enhances th8 cohe s i vene s s o f t h i s natu r a l l y st rong,
coherent mat e r i a l . I n cont rast the s and : c l ay p l ugs cont a i n

"-<l> (f) mat e r i a l with t w o d i s t inct grain s i z e s : s and- s i z e a n d c l a y- s i z e .


The s and- s i z e quart z part i c l e s are very s t rong and coherent

l-('\J>->-0. 3:>- I
SS-SS-OHH ��12]
I socyanate
part i c l e s that provide most o f the s t ructural st rength of the
plugs . Howeve r , the c l ay part i c l e s , wh i ch are inherent l y we ake r ,
act p r imar i l y a s b i nder o r f i l l e r between the quart z grains .
Compared with adobe , t h i s type of mixture has fewe r g r a i n - t o -grain

u (f) • cont a ct s , more pore space , and i s an inherent l y weaker mat e r i a l


that c a n fai l e a s i l y .
plugs i s shown in f i gure 5 .
The d i f f e rence between adobe and sand : c l ay
When s and : c l a y p lugs are t re at ed with
i s ocyanate o r s i l ane e s t e r s the p o l yme r bonds to the qu art z grains

..0«Do<l> and t o the out e r surfaces o f the c l ay part i c l e s . However p o l ymer


cannot penetrate between individu a l c l ay l a y e r s and p l ug f a i lure
occurs by de l aminat i on . o f the c l ay l ayer� when sub j e ct e d t o
de f o rmat i o n a l s t re s s .

The e f fe ct o f grain s i z e and comp o s i t i o n i s further


F i gure 3 _ Graph showing demonstrated by the enhanced compr e s s ive s t rength of c on s o l i dated
re l at i on ship between c l ay pure qua r t z p l ugs , a s shown i n f i gure 2 and t ab l e I I . The i n i t i a l
type or adobe , type o f st rength o f t h e p lugs , which i s s im i l ar t o that o f s and- c l ay
chem i c a l con s o l i dant , and p l ug s , i s due t o the quart z grains and a weak cohe s i o n between the
number o f wet - dry c y c l e s va r i ab l e - s i z ed p a rt i c l e s . Howeve r , the compre s s ive s t rength o f
needed t o init i at e t re at e d pure quart z p lugs more c l o s e l y app ro x i ma t e s the
d i s aggregat ion . comp re s s ive s t rength of t reated adobe p lugs . Thi s is probab l y
becau s e the i n i t i a l s t rength o f t h e quart z p a rt i c l e s i s enhanced
by the supporting s i l i ca-po l ymer network .

Th i s r e s e arch shows that the t r e atment o f s and : c l a y and adobe


plugs with i s ocyanate and / o r s i l anes r e s u l t s i n con s o l i da t i o n and
enhanc e s the ab i l i ty of the p lugs to re s i st rap i d de t e r i o rat ion by
Consolidation Studies 2 53

expo sure to wate r . In s and : c l a y plugs , the degree of r e s i s t ance

16 coats 8 coats
t o det e r i o r at i on appears t o be contro l l ed by the c l ay mine r a l ogy
and type of c o n s o l idant u s e d . I s ocyanate attaches i t s e l f t o p o l a r
s i t e s t h a t c o v e r t h e b a s a l s u r f a c e s o f montmo r i l l o n i t e s [ 3 , 4 ] .
P o l a r s it e s are a l s o pres ent on the surface o f quart z and kao l i n

Z nf
o e
o
part i c l e s in the form of a water f i l m . T h i s provides nume rous
atta chment s i t e s that enab l e the i s ocyanate t o f o rm s t rong bonds
between c l ay and qu art z part i c l e s . S i l an e s att a ch by l i gand
exchange with OH m o l e c u l e s that occur on the b a s a l surface o f
ppl ication
k a o l i n p a rt i c l e s and o n the edges o f montmo r i l l o n i t e s [ 4 ] .
Although l i gand bonding of the s i l an e s i s s t ronger than the van
der Waa l s and i on - dipo l e forces of the i s o c yanate bondin g , the
sheer number of p o l a r att achment s i t e s appears to a l l o w the
i s ocyanate to deve l op a st ronger ove r a l l bonding than the s i l an e s

I- 8,9 -l I- 8,9 -l
cm
cm
[4] . Because s i l an e s d o n o t deve l op a st rong enough b o n d br i dging
quart z and montmor i l l on i t e part i c l e s , p lugs made w i t h t h o s e
mat e r i a l s are e a s i l y det e r i o rated b y t h e phy s i c a l movement o f c l ay
part i c l e s as they hydrate and dehydrat e . T h i s exp l a i n s why
s i l an e s did not p e r f o rm we l l on montmo r i l l on it e -bearing p l ug s , but
did p e r f o rm we l l on k a o l i n-bearing plugs . In cont rast , i s o cyanate
deve l op s a bond s t rong enough t o prevent o r dimi n i s h the
det e r i o r at i o n of the s and : c l a y plugs by the hydrat i o n - dehydrat i on
act i o n . T h i s p e rmi t s the i s o cyanate t o p e r f o rm bet t e r at
protect i ng the plugs against det e r i o rat i o n by wat e r .

The l a c k of d i f f e rence in p o r o s i t y o f untreated and t re ated


s amp l e s i s probab l y because neither i s o c yanate n o r the s i lanes are
heav i l y l oaded into the s ampl e s . The s e p o l ymers f o rm t h i n f i lms
ove r the surfaces o f the c l ay and s and part i c l e s instead of
f i l l i ng the pore spaces with p o l yme r , wh i ch is commo n l y the c a s e

Si de videptew hofofadobepenetblratiockson showi


f o r : ng w i t h s o m e chemi cal con s o l i dant s . I f s u f f i c i ent p o l yme r was
app l i e d , a porosity d i s t inction would be recogn i z ed because the
pores wou l d become c l ogge d . Howeve r , t h i s could be a s o r more


25%6,25%s o l u t i on

sol ution
de s t ruct ive t o adobe than l e aving i t unt reat ed . I s o c yanate
inhib i t s det e r i orat i o n by wat e r but a l l ows water t o p a s s through
the adobe , thereby preventing wat e r f r om becoming t rapped i n the
adobe and caus ing internal damage . Th i s a l s o exp l a i n s why the
F i gure 4 . Cros s - s e c t i o n a l i s ocyanate works so we l l on adobe .
v i ews o f adobe b l o c k s
showing depth o f I ncreased penet r a t i o n and c o n s o l idat i o n as a funct i o n o f
con s o l idat i o n f o r a 6 . 2 5 % s o l ut i on concent r at i on and number o f app l i ca t i ons i s a l s o r e l ated
s o l u t i o n ( di agonal patt e r n ) t o the above . At h i gh concent r at i o n s the p o l yme r s o l ut i o n become s
a n d a 2 5 % s o l ut i o n t o o v i s co u s to permit adequate penet rat i o n int o the adobe . By
( h o r i z ont a l patt e r n ) o f app l y i ng mu lt i p l e coat s o f di lute s o lut i o n s the p o l ymer can
i s o cyanate u s i ng 8 coat s and penetrate deeper into the adobe substrat e . I f more concentrated
1 6 coat s . s o l ut i on s were used, the p o l ymer could become concentrated at or
n e a r the surface . T h i s would result in the s u r f ace becoming
s e a l ed against addit i on a l t re atment and could a l s o t rap wat e r
introduced through cap i l l ary r i s e f r o m t h e ground . Thi s , in turn,
Sand-clay Adobe could cause accel erated det e r i orat i o n o f the i nt e r i o r o f the adobe
and lead to premature c o l l ap s e of the adobe wa l l or s t ructure .

Summary
I s o cyanate and s i l an e s have been shown to enhance adobe ' s
r e s i stance to d i s aggregat i o n by wat e r , as we l l as incre a s ing
comp r e s s ive st rength . Howeve r , the adobe ' s compo s i t i on and grain
s i z e d i s t r ibut i o n are important fact o r s t o cons ider when deciding
whi c h c o n s o l i dant t o use . It appe a r s that adob e s cont a i n i ng
k a o l i n i t e are e f f e ct i ve l y c o n s o l idated with s i lane s , whe r e a s tho s e
adobes containing montmo r i l l on i t e , and / o r m i xed- l a y e r c l ays ,
respond better to i s o c yanat e .
Succe s s fu l t r e atment i s a l s o r e l ated t o the p o l ymer s o lut i on
concent rat i o n , and depth o f penetrat i o n . H i gh p o l ymer
concentrat i on s , o r inadequate penet r at i o n , could result i n the
F i gure 5 . S chemat i c deve l opment of a p o l ymer- r i ch outer s k i n . This could cause
repr e s e nt at i o n o f sand : c l a y acce l e r ated det e r i o rat i o n of the adobe by separat i o n of the
a n d adobe p l ugs showing t re at e d , outer l a y e r from the untreated inner core . T h i s type o f
di f fe rences i n g r a i n - t o ­ phenomenon has been obse rved o n several i s ocyanat e - t re at e d t e st
grain cont act s , packing and w a l l s at Ft . S e lden , New Mex i c o , where the c o n s o l idated outer
porosity . s u r f ace h a s det a ched from the rest of the wa l l . Th i s is b e l i eved
to be a result of the d i f ference i n the coe f f i c i ents of expan s i o n ,
upon wett i n g , between the unt reated and t reated p o r t i o n s o f the
adobe . Thi s type of behav i o r was not obse rved f o r s i l ane-t reated
t e st w a l l s , po s s ib l y because the s i l an e s permit more expan s i on o f
the adobe than does the i s ocyanate . T h i s indicates that t o
succe s s ful l y t re at adob e , o r e a rthen s t ructure s , i t i s n e ce s s ary
to insu re deep penet ration and prevent deve l opment o f a r i g i d ,
non- expandable outer s k i n .

References
I. oject
Resea N . H . Agnew, "Adobe P r e s e rvat i on " ,
rch P r at the Getty
Conser vatRepo tutea
rt on
ion I nsti
Three Month
, ( 1987) .
2 54

2. R . L . C o f fman , N . H . Agnew, M . Ge i s , and C . Selwit z , � The


E f fe c t s of Hexamethylene D i i s o cyanat e-Der ived P o l yme r s on the
Phy s i c a l P r op e rt i e s o f S e lected Natu r a l C l ays " , ( Unpub l i shed, i n
review) .

3. B .K . G .
Complexes , Then g , Formation and Properties of Clay-Polyme
( E l sevi e r , Amsterdam, 1979) 123-15 4 .
r

4. G. Spo s i t o , ( P r i vate commun i cat i o n , 1989) .


Consolidation Studies 2 55

ABSTRACT THE GETTY ADOBE RESEARCH PROJECT AT FORT SELDEN I I I :


AN EVALUATION O F THE APPLICATION O F CHEMICAL CONSOLIDANTS T O TEST WALLS

S ilane esters and polyisocyanates


were applied to adobe test wal l s *
Charles selwitz , Richard Coffman, and Neville Agnew .
in various concentrations and The Getty Conservation Inst itute
solvent mixtures by spraying, 4 503 Glencoe Avenue .
brushing, multiple coat ing, and Marina del Rey, Cali fornia 90292
bulk infiltration. After one USA
month ' s curing they were sUbj ected
to an accelerated weathering
Introduction
regime of two water spray cycles
per day for two months , and
T h i s study i s concerned with the evaluation o f chemical consol idants used to
subsequently they were allowed to
stab i l i ze o l d , weathered adobe of historical interest . Adobe test wal l s were
weather naturally . The wal l s were
treated with d i f ferent chemicals by a variety of application strategies and
evaluated visually according to a
after curing were subjected to a daily schedule of direct water spraying for
nume r i c a l rat ing s y s t em .
two months . This is very severe t reatment for adobe . It provided highly
I socyanates provide very ef fective
accelerated weathering and enabled us to evaluate materials and procedures in
stabi l izatio� by a combination of
a relatively short t ime period .
brushing and bulk infiltration,
but there must be a su itable
Test Wa l l Evaluation
network of accession holes for
deep del ivery of polymer, and the
I n the winter of 1988 , twenty-f ive adobe wal l s were constructed at Fort Selden,
amount of consol idant must be
New Mex ico . Bricks , 30 cent imeters ( 12 inche s ) square and 1 0 centimeters (4
opt imized. The s i lane ester
inches ) thick, were made from local clay-bearing soil which contained 9 1% sand
formu lation, Stone Strengthener H,
and 8% s i lt and clay . The c lay consisted of roughly equal amounts of kaolinite ,
is easier to apply and gives good
i l l ite and smect ite . These bricks were made into east-west running wal l s 150
c o n s o l i d a t i o n w i t h o u t
centimeters ( f ive feet ) long and tall and 3 0 centimeters thick, with a mortar
discoloration; but the treated
of the s ame composition . The wal l s stood on two layers of adobe brick below
walls show erosion not seen with
grade . On each side a surface "plaster" coating made from the same mortar was
isocyanate treated wal l s . Old,
aged adobe may require a
trowe l led over the r ight half of the wal l , i . e . , an area
in height and 75 centimeters (2.5 feet ) in widt h .
ISO centimeters ( 5 feet )
The walls were allowed to
r e hyd r a t ion b e f ore e ither
weather naturally for several months and in May were treated with a number of
consolidation c a n b e done .
conso1idant s . After treatment they were wrapped in polyethylene sheeting and
allowed to cure for one month and then subjected to water spraying for two
months . On each side a nozzle four feet f rom the wall del ivered 22 liters of
KEYWORDS
water over a 1 5 minute period twice a day as a dispersed spray . I n addit ion,
these outdoor walls experienced an unusually rainy New Mexico summer .
Photographs o f northern and southern exposures o f some o f the wa l l s treated with
Adob e , bru shing , bulk
DN 3 3 9 0 or s i l ane esters are provided in the left columns . They can be matched
i n f i l trat ion , c o n s o l i d at i o n ,
with descriptions o f treatments in Tables 3 and 4 .
dilut ion e f fect s , i socyanate s , mud
brick, rehydrat ion , s i l anes .
Chemicals and

Most of the studies


Appl i cat ion Procedures

focused on two types of consolidant s : polyisocyanates and


s i l ane esters . Most of the isocyanate studies were done with DN3 3 9 0 , is an
i socyanurate trimer of hexamethylene diisocyanate . It is supplied as a 90%
concentration in a mixture of n - butyl acetate and aromatic hydrocarbons .
Stone strengthener H ( SSH ) is a mixture of methyl tr iethoxys i l ane and
tetraethoxys i l ane ( ethyl s i l icate) in a ketone solvent , and it contains
dibutyltin dilaurate as catalyst . The active component in Stone Strengthener
OH ( SSOH ) is ethyl s i l icate . The i socyanates were given to u s by the Mobay
Chemical Company while SSH and SSOH were donated by ProSoCo and applied to
certain test wa l l s by their personne l .

Both types o f chemicals seem uniquely suited for adobe because they cure
by mechanisms that involve water inherently available in clay . The s il ane esters
hydrolyze to a s i l icate network that can crossl ink appropriate s i l iceous
functionality in the adobe while polyisocyanates readily l ink up and consol idate
in the presence of water through polyurea format ion .

Application was generally done by brushing, but spraying or bulk


inf iltration were used in a number of case s . Bulk infi ltration i s a procedure
in which holes are drilled into the adobe wall and funne l s are used to supply
the consolidant ; This insures that impregnation at a substantial depth is
obtained. Where b u l k infiltration is used, surface application of consol idant
by brushing is also employed .

Generally, no two wal l s were treated by exactly the s ame procedure.


Treatment of wal l 18 with DN3390 by a combination of bulk infi ltration and
brushing gave the best stabilization, and its descr iption may be i l lustra�ive .
Four evenly spaced 19 m i l l imeter diameter holes were drilled across the top to
a depth of 30 cent imeter s . Similarly four hole s , one meter above the bas e ,
evenly spaced horizontally o n the south s ide and one hole a t each end, were
drilled downward at 4 5 ° angles . Also a series of holes was made at 15 cent imeter
intervals around the base . A solut ion of 9 l iters of DN3390 in 27 l iters of
methyl ethyl ketone and 27 liters of toluene was prepared . Funnel s were used
to deliver 12 liters of this solution through the top hole s , 8 l iters through
the side holes and 28 l iters to the base ho les . Twelve liters were brushed on
the surface in f ive coat s , and an additional 3 l iters was applied around the
base . The treated adobe was then t ightly wrapped in polyethylene f i lm for four
week s .

* To whom correspondence shou ld b e addressed


Adobe 90

Evaluation of the ef fectiveness of consolidation was done as soon as the


spraying was concluded, then three months later, and finally after an additional
twelve months . Subject ive quality rating s , ranging from zero for a coll apsed
wall to ten for an unblemished wal l , were made separately for brick surfaces
( Table 1) and adobe mud coated surfaces which are descr ibed as "plaster covers"
( Table 2 ) . separate rating systems were needed because wal l s with plaster coats
tended to deteriorate by having the cover pull away from the surface while the
brick surfaces cracked and eroded . At the conclusion of spraying , one of us
10 110 lga.
a of det.rioratiOil of _! ldJWl ( CS ) ,
genera l ,
who was
overall
not involved in
survey of the test wal l s .
the treatment , created the rating
The evaluation was carried out by
system by a
Biag ... I.. 1 1 crack
an examiner (CS) who did not know the procedures that had been used on the
particular wal l being rated . The Getty Conservation Institute ' s overa l l program
MaJ.1 100b at
crack.
rong bud
, but - y at Fort Selden looked at a number of stabili zat ion approaches , including the use
of inserted rebars , protective geotext iles and reburial procedures , but this
Poo rer �1 presentat ion ' w i l l consider only the evaluation and comparison of chemical s used
for conso lidation.

bt ld.n. s.a.e .ro


...i.... crac g. .ion

bt 1ti.J:lg b.....
en.i.... cr.c , i.r .ro.ion
Resu lts with the I socyanate, DN3390
.a:t ial. ro-
.r fallen f _11

The results o f this study are summarized in Table 3 , which i s concerned with
the isocyanate DN3390 , in Table 4, which examines the consolidation of adobe
with s ilane ester s , and in Table S which compares d i f ferent types of isocyanates .
Results which descr ibe the ef fectiveness of DN3390 are developed by comparing
the ratings of five treated wal l s ( 1 6 through 2 0 ) with an untreated wal l . The
untreated wal l , wall 2 1 , had a l l but washed away by the time spraying had ended,
but the treated wal l s were relatively intact and standing ( Table 3 ) . There was
considerable variation in the quality of these consol idated structure s . Wal l
1 8 w a s essentially unchanged while wal l 19 was badly cracked and eroded . The
critical d i f ference between the wal l s was the use of bulk infiltration which
provided very deep penetrat ion. Wa l l 1 8 had access holes drilled across the top,
in the midsection and around the base and the largest amount of consol idant , 9
ItATIIfO VlU.UB DBSCIlInJOII liter s , was used . The second best wal l ,
across the top and along the base and employed much less consol idant ,
wal l 2 0 , had infiltration holes only
3 l iter s .
This wal l received a rating o f 2 8 ( out o f a possible maximum o f 40 ) versus a
... 1'.. crack., DO aiga. of ..p,.ratiOil
value of 37 for wal l 18 . The plaster l ayer on the south s ide of wall 20 had

.-p& ion - • aagge ioa


1Ior.
r.t
cr.ck. .t of pu l led away , not at the top but in the lower part of the wal l .
suggest that good results can be obtained with less consol idant but that broader
These findings

P art mab"pa i.Dg - .ocl••


4i.ac.
ial. ratioD, DOt .iaibl., but
le by tapp t
p l acement of infiltration holes is important . In the wa l l s where only brushing
cr.c ld.ng was used ,
were poorer .
and where there was no direct deep placement of consol idant ,
There was more cracking and breaking o f f of adobe .
results
Gener a l l y , the
Jtd:.eD, iDgpu.l.DOt ... ••• h_ru.ab
by tapp
.i....
,
ling
i.ibl.,
y cU..c.
"i.r
le
peeling away o f the plaster coat was a more critical and noticeable problem than
crac kiDg the deterioration of the brick surface .

rota •• 1 par.tioD i. ...i.ib1e Studies were made on the depth of penetration achieved by brushing . Samples
of adobe powder were obtained at measured d i stances from the surface by drill ing.
Pla.t.r ia rea to dy P.l1 off
In laboratory studie s , these were brought to constant weight at 1 S 0 ° C to drive
off water and then heated at 600°C in air to burn off the polyme r . weight and
ft.. ••
f.11_ off
p1 ter co co.p
....r haa 1et.1y concentration of
for wall 1 7 is shown in Figure 1 .
polymer were then determined by dif ference .
This wall had 4 liters o f DN3390 brushed over
A sample result

the top and s ide s , which is a relat ively large amount . Typically polymer
concentrations between one and two percent are found within the outer inch of
adobe . Very much less occurs between one to two inches from the surface , and
essentially no polymer can be found any deepe r . We did notice in other studies
that when large amounts o f solution are used on upper surface s , polymer i s found
deeper into the brick at lower levels of the wal l .

T1\BLE 3

COHPJ\RISON OF THE CONSOLID1\TION OF ADOBE WALLS WITH DN3390 USING DI FFERENT MOUNTS OF ISOCYANATE AND
OIFFERENT APPLICATION STRATEGIESA

WALL AMOUNT OF DN3 3 9 0 DELIVERED I N LITERS SU MMAR Y OF Rl\TINGS

Brushed on Applied INTERNAL BULJt. DELIVERX' NORTH SOUTH TOTAL


TOTAL Top ' sides To Base TOP MIDSECTION BASE BRICK PLASTER BRICIt PLASTER RATING

10 3.0 2.0 1. 0 22

'
17 4.S 4.0 O.S 23

18 •• 0 1.7 ••S 1.7 1.1 4 •• 10 37

1. 4.S loS 3.0 15

20 3.0 1.. O.S •• 6 O•• 28

21 none

Wall 2 1 North
A. All runs ( except 17) at 12 . 8 \ ON 3 3 9 0 in equal volumes o f methyl ethyl ketone and lD.ixed xylene ••
N o catalyst was used.

B. Material delivered into deeply drilled holes by using funne l s .

C. Approximately one-third o f t h e O N 3 3 9 0 was a t 6 . 4 \ concentration.


Consolidation Studies 2 57

F r G U It E
POLYMER
WT. % DN3CONTENT
90 vs DEPDETERMI
TH OF PENENTRATIONS
AT I O N
- -
2. 0 \ ALL '17.W

15ffi 1. 50
t-

;;S�:z: 1. 0 \
t;:�
a.
UJ'"'

� 0, 50 \ ------
t-:z:
0ua:
r-----....
Wall 2 1 South UJ
-'0
a.
���
0. 000. 0 1. 0 2. 0 PEl ET3.RA0 4. 0 5. 0 6. 0 DEPTH Of nON IN INCHES

f POLYHEA CONTENT fOA WALLS 17SW

The runs in Table 3 show that , in general , better results are obtained with
brushing when more consol idant is used . Our results also indicate that it is
better practice to apply more solution at the upper sides and top rather than
to drench the base . This is confirmed by comparing wal l s 17 and 1 9 . In each
case 4 . 5 liters o f DN3 390 were del ivered . In wall 1 7 , 4 . 0 l iters were put onto
the top and sides and 0 . 5 liters to the base . For wal l 1 9 , 3 . 0 liters were put
into the base by bulk infi ltration, and only 1.5 l iters were used on upper
surfaces . Wal l 19 showed the poorest consolidation of any of the wa l l s treated
with DN3 3 9 0 . Overall the studies show that the i socyanate can provide ef fective
Wall 19 North consolidation o f adobe , but there must be a suitable network of acces s ion holes
for bulk infiltration o f polymer and the amount of consolidant must be opt imized.

Results with S i lane Esters

Eight test wal l s were used to examine the ability of s i l ane esters to stabilize
adobe . Table 4 outlines the results where SSH was used on f ive wal l s , SSOH on
two wal l s , and SSH in a dif ferent solvent , mineral spirit s , on a single wall .
All of these systems used 0 . 5% dibutyltin d i l aurate as catalyst . In six of the
ser ies , applicat ion was by brushing or by a combination o f brushing and bulk
inf iltration . This was done in a manner to attempt to replicate the applicat ion
procedures used with DN339 0 . In two cases the s i l ane esters were del ivered to
the wall by spraying.

T7!..B LE 4

COMPARISON OF THE CONSOLIDATION OF ADOBE WALLS WITH SILANE ESTERS USING DIFFERENT AMOUNTS OF CONSOLI DANT
AND DIFF ERENT APPLICATION STRATEGI E S

Wall 19 South WALL S I LANE


A
METHOD OF AHOUNT DELIVERED I N LITERS RATINGS
ESTER DELIVERY TOTAL BY BRUSHING BULK INFILTRATION NORTH SOUTH TOTAL
'rap &: SIDES BASE TOP MIDDLE BASE BRICK PLASTER eRICK PLASTER

22 SSH BRUSHING ..
19 . 5 . 0.7 25

23 SSH SPRAYING 18.5 18 . 5 27


24 SSH BULK
a

BRUSHING
AND 45l 3 lO 14 2.
25 SSH BRUSHING 22 lO l 2 2.

2. SSH BULlt AND


BRUSHING
18 . 5 7.5 1.5 •. 5 3.0 23

2. SSOH SPRAYING 18 . 5 18 . 5 14

30 SSOH BULK AND


BRUSHING
10 . 5 11. 5 lO
c 0.7 33
3l SSH BRUSHING '.5 ••8

A. A l l runs ( except 3 1 ) i n polar, ketonic solvent. All runs containe<1 catalyst.

B. Bulk indicates material vas delivered into deeply drille<1 holes by usinq funne l s .

Wall 18 North
c. In minera� spirits
Adobe 90

I n general it was found that SSH was e f fective while SSOH was not . This
i s not unexpected since the major property dif ference provided by the two is
water repe l lency . The formu lation containing methyl tr iethoxys i lane gives a
surface that rejects water, while the formulation without the alkyl s i l ane , the
SSOH, does not , and this makes a critical d i f ference when attack is by spraying
water . There seemed to be little dif ference between wal l s that were treated by
spraying , brushing or the combination o f brushing and bulk infiltration . The
wal l s treated with SSH retained their original l ight tan color and showed no
pu l l ing away of plaster coat s . We attempted to determine why these layers peeled
away from some o f the i socyanate treated wal l s but did not , in any case, do so
when the consol idant was SSH . Thermal expansion studies done by thermomechanical
analyses showed no dif ferences between samples treated with DN3390 and SSH .
However, comparison of swelling coe f f ic ients by the same procedure suggests that
adobe treated with SSH is less rigid; this may explain the dif ference . We do
not have the technology to measure low polysiloxane concentrations in adobe and
thereby obtain the depth of SSH penetrat ion. The general correlation of non­
peeling with the amount of consol idant suggests that stabil ity is a l so a matter
of getting s u f f i c ient material deeply enough into the structure .

The SSH treated wal l s a l l rece ived a good total rating that appears to be
Wall 18 South
independent of the amount of consol idant used or the method of appl ication . This
i s in obvious contrast to results obtained with DN3390 . The low viscosity of
the s i l ane esters enables large amounts to be applied with ease and to achieve
good penetrat ion, which is an important advantage . The total quantity of the
consol idant itself applied in the s i l ane ester tests ran from 9 . 5 to 45 liters ,
a range that was totally above that of the 3 to 9 liters used in DN3390 runs .
This may explain why results with the isocyanate tests were more dependent on
the amount of consol idant and the method of appl ication .

The main problem with wal l s tre�ted with SSH and a major difference in the
results obtained with isocyanates was their tendency to show surface erosion.
Originally the adobe brick wal l s had a relatively smooth appearance . After
treatment with isocyanates and subsequent exposure to water, the surfaces
generally s t i l l retained a flat , even quality . They were darker and, depending
on the amount of DN3 3 9 0 applied, sometimes were cracked . The surfaces had become
hardened but were otherwise structurally unchanged . In contrast , the s i lane
ester treated surfaces were grainy and eroded and in some cases showed ruts and
cracks . This dif ference in behavior is probably generally true in stone
consolidation, i . e . , where s i l ane esters are compared to thermoset resins like
epox ies and isocyanates ' for the stab i l i zation of sandstone and l imestone, but
only in the regime o f water attack on treated adobe can this be di scerned so
clearly and so quickly .

Wall 29 North
The best result in this series was obtained with SSH in a mineral spirits
solvent , rather than the more conventional cosolvent , ( approximately 20% methyl
ethyl ketone 10% acetone ) . Mineral spirits is a petroleum distil late
consist ing o f al iphatic and aromat ic hydrocarbons . Wal l 3 1 , which was treated
with the s i l ane in this solvent , received a rating o f 33 while wall 2 2 , treated
identically except for the use o f the mixed ketone solvent , had a rating value
of 2 5 . We were able to smell res idual aromatic hydrocarbons in the test wa l l s
after spraying w a s stopped. It i s likely .that the better performance observed
in wal l 31 resulted from the presence of heavier hydrocarbons , which provided
additional water barrier and repellency propert ies .

TABLE 5
COMPARISON OF THE CONSOLIDATION OF ADOBE WALLS
USING DIFFERENT TYPES OF DIISOCYANATES AND DIFFERENT AMOUNTS OF CONSOLIDANT

1
WALL CONSOLIDANT POONDS OF PLACEMENT OF CONSOLIDANT RATINGS
CONSOLIDANT
USED IN
'"
TOP , SIDES BASE
BRICK
NORTH
PLASTER
SOUTH
BRICK PLASTER
TOTAL E

Wall 29 south 3 3 W Diphenylmethane


d i i socyanate

3 3 E Dicyc lohexylmethane
HJ\LF WALL

3.8

3.2
3

2
.•
••
0.'

0 • • 20
d i i socyanate and
C
catalyst

° 7.2 2.
HE DN 3390 7.2

3 4 W DN 3 3 9 0
°
1.8 1.8 12
3 5 . Dicyclohexylmetbane
d i isocyanate
O
1.8 1.8 12
3 5 E D i cyclobexylmetbane 7.2 7.2
C
d i i socyanate

A. Each quantity of consolidant was d i s solved in a mixture of 4.5 liters of methyl ethyl ketone and 4 . 5
I iters o f mixed xylenes.

B. App lication was by brushinq on o f 2 to 6 coats.

C. Approx imately 3 . 0 m i l l i l i ters o f dibutyl tin laurate added to solution.

o. Approximately 0 . 5 m i l l i l iters o f dibutyl tin laurate was added to solution.

E. Total i s calculated on a full w a l l bas i s .

Wall 25 North
Consolidation Studies 2 59

Results with Other Consolidants

other isocyanates examined in this program were diphenylmethane d i isocyanate


and dicyclohexylmethane d i isocyanate . Both are commercial products marketed by
Mobay . Results outl ined in Table 5 show that these worked about as well as
DN3 3 9 0 . This table also demonstrates that the use of opt imum amounts of
impregnant is needed for good consol idat ion . In one set the east half of wall
3 4 was treated with 7 . 2 pounds of DN3390 , while the west half received only a
quarter of that amount or 1 . 8 pound s . Respective summary ratings of 26 and 12
i l lustrate the benefits of u s ing a suff icient amount of consol idant . Almost
comparable results were obtained when dicyc10hexylmethane d i isocyanate was used
in a s imilar manner . These experiments are additional evidence of the
criticality of the amount of isocyanate employed and the lesser importance of
the type o f p�yfunctional i socyanate that i s used .

Two additional treatments should be mentioned . Wal l 32 was treated by


brushing and bulk infiltration with 4 . 4 pounds of Acryloid A-2 1 dis solved in 50

Wall 2S South
liters o f a mixed solvent consisting of methyl ethyl ketone,
acetate .
was treated with 3
During the two months of water spraying this wall collapsed .
l iters of D N 3 3 9 0 in much the same way that treatment was
xylenes and butyl
Wall 3 6

carried out on wal l 16 except that half of the methyl ethyl ketone in the
cosolvent with xylene was replaced with butyl acetat e , and 3 m i l l i l iters of
dibutyltin laurate , catalyst , was added . A summary rating of 2 6 for this wall
suggests that neither factor had a major impact on the quality of the treatment .

Reevaluations at Later Periods

Reevaluations of all o f the wa l l s was done three months and then f i fteen months
after the initial asses sment . A sample of the reevaluation of 10 wal l s is shown
in Table 6 . Wal l s treated with both i socyanates and s i l ane esters are included.
These data show that during the year and a quarter after the water spraying there
was no additional deteriorat ion. An exception to this occurred with wal l s where
the plaster coatings had pulled away . This behavior led to initial ratings o f
3 or 4. I n most o f these c a s e s this separated layer subsequent ly fell off to
mandate a drop in rating to a value of one . The rating s , each t ime , were done
without consider ing the treatment . The consistency of the ratings over time
provides validation of the evaluation procedure .

TABLE 6
REEVALUATION OF WALLS THREE MONTHS AND FIFTEEN MONTHS AFTER THE END OF SPR1\YING

WALL 1I.MOUNT OF CONSOLI D.ANT


TO TOP
BTYPE
, SIDES
SURFACE
I LITERS
BULK BRICK
NORTH FACE
PLASTER
SOUTH
BRICK
FACE
PLASTER
SUMMAR Y RATING

MONTHS ----_._--> 15 15 15 15 Q 15

, . ON3 3 9 0 2.3 8 22 25 2.

17 ON3 3 'l O 4 . 0B 2'" 20

, . ON) ) :I 0 1 . 7B • 2 • 10 10 10 10 10 37 3. 3.

19 DN3 3 9 0 1 . 58 15 17 15

20 ON3 3 9 0 LOB O • • 2. 30 31

2' SSH 18 . 58 27 2. 2.

,, 21

3,s5w"F:" s H"
2. 8S0H 1 1 . 58 10 . .

31 8 . 8B 31 25 30

DCHH 1 . 5B : 2 14 14

DCHH 6. as �. 24 2.

B. Applied by brushing

S. Applied by spraying

M. Disso lved in mineral s p i r i t s instead ot ketone solvents.

D. On wall 3SE 3SW,


and
full wall ba s i s .
the amount o t consolidant and the s umm a ry ratings were doubled to compare on a

More Recent Study of Brushing Procedures

A second set o f five test wal l s was treated with the i socyanate DN 3390 in May
198 9 . The purpose o f this test series was to determine i f the high quality of
consolidation that was achieved by bulk infi ltration in the previous test wa l l
program could b e obtained s imply b y brushing . The strategy was t o u s e more
dilute solutions of DN 3390 in order to obtain deeper penetration . The use of
lower concentrat ions meant that a larger number of brushings ( more than the
seven applications typically used in the earlier set ) would be required to supply
the quantity of isocyanate needed for a very good consol idat ion .
260 Adobe 90

Table 7 summarizes the results obtained in this phase and shows the amount
of DN 3 3 9 0 as l iters of i socyanate without solvent . These studies were done on
quarter and h a l f wal l segments in order to examine more variables but the
quant ities of consol idant
comparison with earlier studie s .
the f irst two wal l t reatments
are standardized to a fu l l wall
The number o f brushings was relatively low in
( 5 2 and 5 3 )
value

but rose to 11 on wall 5 4 and 1 8 on


t ; simplify

wa l l s 5 5 and 5 6 . Correspondingly the concentrations , which are not shown , range


from 12 . 8% to 2 % .

TABLE 7
CONSOLIDATION QUALITY RATINGS ON FIVE NEW TEST WALLS

Wall DN3 3 9 0 solvent Numb er ot Post catalyst RATING s umm ary

per full App lications wetting Brick Plaster Full Wall

wall l i ters Solvent Basis

52W
52E
3.0
3.0
IIII IIX No
No
30
22

53W
53E
1.5
1.5
IIII IIX Yes
Yes 24
0

54. 2.3 IIII 11 none No 14


0
54E

55WW
SSWE
5 5 EW
2.3

5.2
5.2
5.2
1111++""
IIII
,,
11

.
.
.
none

"
none
none
II
Yes

No
No
No
2.
2.
30
30
55EE

56WW
56WE
56EW
56EE
5.2

2.6
2.6
2.6
2.6
1111++""
IIII
,
1.
1.
.

.
. none
none
II
No

No
No
No
No
24
24
30
30

catalyst was dibutyl t i n laurate


IIX =
:: methyl ethyl
mixed xylenes
ketone

B = butyl acetate

The rating system applied to these wal l s after one month o f being wrapped
in plastic fol lowing treatment , two months of water spraying, and three months
o f natural exposure is also provided in Table 7 . Before discussing the numerical
rating s , it should be noted that these five wal l s were plagued with problems that
did not seem to beset the original wal l s . The wal l s , as del ivered b y the
contractor before t reatment was applied, were o f poor quality with much erosion
and cracking . Apparently some patching was done , but it was not sufficient .
The left half of wall 53 and the right hal f of wal l 54 rather quickly coll apsed
when the water spray program was employed . These mass ive failures made no sense
based on the nature of the treatment . They may have resulted from residual deep
cracks that permitted spray water to penetrate through the consolidation to the
core of the wal l with the obvious disastrous result s . These walls may also have
suf fered f rom an inadequate amount of consol idant around the base .

Nevertheles s , this phase s t i l l provided information on the ef fect ivenes s


o f multiple low concentration applications . Failed wal l s were those using the
smallest amount of consol idant , 1 . 5 and 2 . 3 liter s . Wa l l s treated with higher
levels o f consol idant , 2 . 6 to 5 . 2 liter s , . all remained intact . The use of a
very large number of brushings , 1 8 , gave treated wall ratings of 3 0 , even with
the application o f only 2 . 6 liters of DN 3390 , a relatively small amount . It
does appear that multiple coatings with dilute systems w i l l give better results
than fewer brushings at higher concentration , but the consolidation i s not as good
as the optimum use of deep bulk treatment .

The Consolidat ion of Aged Adobe

The purpose of this research was to develop proceedings and formu lations to
consolidate aged, weathered adobe of structures of historic importance . In the
course of time, adobe exposed to the atmosphere becomes weak and friable.
Presumably the clay platelet s , initially f locculated and al igned, become more
dispersed . Void spaces are enlarged which leads to a weakening of the weathered
surface . We tried several brushing experiments on smal l , nineteenth century adobe
remnants at Fort Selden. Both DN 3390 and SSH were used . To our dismay, the
aged adobe took up the consolidants but was not mechanically strengthened .
ACKNOWLEDGEMENTS
Working with both Fort Selden adobe and a tan adobe f rom Tel Dan in I srael we
have found that when the adobe is recast into mud bricks and dried, it can be

We

that
wish
Incorporated
company
to
and
thank
Frances
and Michael
ProSoCo
Gale of
Jakst i s
consol idated under treatment cond itions that fail with adobe structures that had
been in existence for some t ime . We theorize that a slow dehydration takes
that takes the clay to a form that prevents the water-based curing mechanism from
occurring. To test this hypothes i s , a chunk o f old Fort Selden adobe was soaked
p lace

and Dr . Stanley Siranovich of the


Mobay Chemical Company for the with 10% aqueous methyl ethyl ketone . This permitted a rehydration without

donation o f generous quant ities of physically altering the shape of the adobe . After the organic solvent
their products and their help in evaporated, the piece was t reated with 12% DN 3390 in methyl ethyl ketone and
the use of these mater ial s . This xylene . The product , after curing and drying, was mechanically strong and hard .
work was done in co l l aboration Thus we have qualitatively demonstrated that historic adobe can be consol idated
with Michael Taylor and the Museum with isocyanates and s ilane ester s , but the need for rehydration indicates that
of New Mex ico . We are considerable research and development rema ins before on-site procedures can be
part icularly grateful for the recommended .
assistance of Jose Guzman, Elva
Melandrez and Jaime Romero at Fort
Selden.
Consolidation Studies 261

ABSTRACT The Preservation of Prehistoric Mud Plaster at Mesa Verde National Park

Mud plaster is frequently found Allen S . Bohnert , Curator


in prehistoric sites within the Southeast Archeological Center
American Southwest . Mesa Verde National Park Service
National Park, a unit of the PO Box 2 4 1 6
United States National Park Tallahassee, Florida 32316
Service , contains over 3 , 9 3 0 U . S .A.
recorded archeological sites .
Approximately 590 of these are
c l i f f dwe l lings . More than 600 INTRODUCTION
years after abandonment , many of
the sandstone masonry walls in Mesa Verde National Park, estab l i shed in 1 9 0 6 , is the only National Park in the
these buildings retain at least United States National Park Service created expre s s ly to preserve cultural
remnants o f original mud resources . Mesa Verde i s noted for its wealth o f archeological resources ,
plaster . part icularly its c l i f f dwe l l ings . The resources are of international signif­
icance, and in 1978 the park was selected as a World Heritage Site .
This paper discusses the multi­
faceted plaster preservation Of the 3 , 9 3 2 archeological sites recorded within the park, nearly 600 can
program being formulated at the be c l a s s i f ied as c l i f f dwe l l ings [ 1 ] . The majority are " s ites of the Mesa
park. Initial steps included Verde Branch of a Formative Culture known as Anasaz i " [ 2 ] . The Anasazi resided
reviewing archival materials and throughout the four-corners region, an area where four state boundaries meet in
thoroughly surveying extant plas­ the American Southwest . They l ived in what is now Mesa Verde National Park
ter in e ighteen c l i f f dwe ll ings , from approximately A . D . 450 until the area was abandoned around A . D . 1300 .
containing over one thousand Vil lage locations shifted from the mesa tops to c l i f f alcoves toward the end of
architectural space s . Surveys of this period , and increased building activity in the alcoves began between A . D .
mud plaster on this scale are 1 2 3 0 and 1 2 60 . The resultant structures range i n s i ze from a single room to
unprecidented in the United more than 200 rooms [ 3 ] . The alcoves sheltered Anasazi architecture from the
States . Survey methodology, elements , and the semiarid environment o f southwestern Colorado further served
resu lt s , recommended management to preserve this spectacular resource ( see f i g . 1 ) .
action s , and future research
needs are discussed. Addition­ Much Anasazi architecture is finished with mud plaster, especially the
ally, the results o f prel iminary later sandstone and mud mortar c l i f f dwe l l ings . It should be noted , however ,
studies o f the mineralogical that plaster was used to some extent by the Anasazi throughout their occupation
characteristics o f the mud of Mesa Verde . Nusbaum , for examp l e , reported plaster in the earliest known
plaster from Mesa Verde are
presented .
Anasazi dwe l l ings , pithouses , excavated in 1 9 2 6 at Step House ( S ite #
[4] . Unt il recently the focus o f most plaster research has been o n highly
MV
128 5 )

decorated plaster, rather than its systematic documentation or its potent ial
relationship to cultural change and room function.
KEYWORDS
Nordenskiold, the first to scient i f ically document and record the archeo­
logical resources o f Mesa Verde , recorded decorated plaster in several c l i f f
Archeology , Anasaz i , c l i f f dwe l l ings during h i s 1 8 9 1 exploration [ 5 ] ( see f ig . 2 ) . Subsequent archeologi­
dwe l ling s , mud plaster , plaster cal work by Fewkes [ 6 ] , Cattanach [ 7 ] , Rohn [ 8 ] and others a l so provided some
conservation, Mesa Verde National
Park
plaster data. In the Long House report ( Site # MV 1200 ) Cattanach states :

Some wal l s were " p lastered" in places by smearing surplus mortar


extruded between blocks and across the face o f the surrounding
wal l . I n most cases , however , the overall plaster was prepared
and applied independent ly o f the wall construct ion . The material
was usually derived from the red and brown s o i l s and was probably
t i nted more often than we realized with pigments made from
hemat ite , kaolin and the like [ 9 ] .

Rohn similarly describes Mug House plaster ( Site #


consistent appearance o f the pink plaster in the rooms .
MV
122 9 ) , noting the
He hypothesizes that
" • • •red loess from the mesa top continued in fashion for plastering after
it had gone out o f style as mortar • . ••
" [ 10 ] . While discuss ing kiva
plaster , Rohn states that " • • •almost a l l the individual plaster consisted
of a very thin brown or tan adobe body that had probably been acquired from
Adobe Cave . Only in Kiva G was the reddish brown , mesa-top loess used
instead . [ 1 1 ] "

Mud plaster i s found in prehistoric sites throughout the American South­


west . It was not until Watson Smith ' s 1 9 5 2 pub lication [ 12 ] that a relatively
comprehensive survey o f pueblo wall paintings became availabl e . Smith ' s report
included brief descr iptions of plaster in notable sites throughout the South­
west , including Mesa Verde National Park. More important l y , the report
includes an excellent descr iption and analysis of decorative element s , materi­
als compri sing the mud plaster, and pigments used in prehistoric sites in the
Jeddito Valley , Arizona .

It was not until the 1980s that there was a renewed interest i n , and
increased focus on , mud plaster . Reports by Chiari [ 1 3 ] , Schwart zbaum [ 14 ] ,
and others describing successful conservation treatments performed on decorated
adobe surfaces became accessible to managers and resource specialist s . Addi­
tiona l l y , two reports by Constance S i lver appeared , one in 1980 [ 1 5 ] and the
other in 1982 [ 1 6 ] . Both publications used the extensive resources available
through the International Center for the Restoration and Preservation of
Cultural Property ( ICCROM ) . These reports , more than any other publications ,
highl ighted the lack of attention being given mud plaster and the need to
pursue both the analysis and conservation o f the prehistoric plasters in the
American Southwest . S ilver ' s reports graphically il lustrated the need to
develop comprehensive approaches to documenting and preserving both exist ing
plaster and that being uncovered during current excavat ions . These two
report s , combined with a certain amount of expanding professional interest in
mud plaster , served to help increase general awareness that the plaster was
indeed an integral component o f Anasazi architecture and that its preservation
was not ensured given present circumstances .

The pilot treatments described in the 1982 report involved the consol ida­
tion , reattachment , and cleaning o f in situ plaster [ 1 7 ] . The report describes
treatment methods and materials used in detai l , and they will not be reiterated
here . It should be noted , however , that this was the first t ime such treat­
ments were tested on prehistoric mud plaster in the U . S . Mesa Verde National
Park played an important role in the project since the treatments were per­
formed in Kiva C and on the free-standing west exterior wal l of room 2 8 in Mug
House . The pilot treatment project also alerted park staff to the fragile
nature o f the plaster resource.
MESA VERDE PLASTER SURVEY

Park Management , becoming keenly aware of the lack of an integrated plaster


data base and the continuing loss of this resource, obtained funding to begin a
thorough plaster survey . Contracted survey work has been conducted during 1 9 8 5
( Contract #-1490-4-000 6 ) a n d 1987 ( Contact #-1490-7-0002 ) .

Major objectives, common to both survey pro j ect s , were:

1. to record the location and general condition o f a l l extant plaster


in selected archeological sites ;

2. to compare general ly , where documentation exist s , the extant


resource to that which was present historical l y ; and

3. to provide recommendations for preservation treatment of the


resource , keeping in mind the int imate relationships existing between
the plaster and its masonry support .

A primary objective of the 1 9 8 5 survey included a substant ial archival


work component . Mesa Verde National Park has accumulated a considerable amount
of archival material since 190 6 . Most is located in distinct f i les in the
Archeology Museum or at the Research Center. One major series of records
pertains to ruin stab i l i z ation activities . The park ' s archeological ruins have
undergone periodic stab i l i zation since 1908 . A second major archival data base
i s the archeological site record f ile . Other plaster data are located in
historic photograph files and in the park ' s extensive archeological excavation
records . A l l of these resources include substant ial amounts of photographic
and written plaster-related dat a .

These dist inct data bases are not fully cross-referenced, nor are they
automated . The idiosyncrasies of the records provided certain chal lenges to
the contractor while integrating the plaster-related data. As mentioned previ­
ous l y , plaster has been addressed di fferently by past researchers . It may or
may not have been specifically mentioned or reported in detail when archeo­
logical sites were surveyed and when excavation or stab i l i zation work was docu­
mented . For example, older written stab i l i zation records or completed archeo­
logical s ite forms may not have discussed the presence of plastered wal l s .
However , photographs documenting the site o r the work may clearly i l lustrate
the presence of plaster . Also, when existing plaster was noted , the termin­
ology describing or classifying it was not consistent . Decorated plaster may
have been referenced using those terms or included under the grouping of
"pictograph s " as was done in the 1964 Wether i l l Mesa survey report [ 18 ] .

Nevertheless , the 1 9 8 5 plaster survey contract successfully compiled and


integrated much plaster data from the record groupings l isted above . E ighty­
five of the park ' s c l i f f dwel l ings [ e ighty-four are l i sted in the report ] were
reported as having documented prehistoric plaster [ 19 ] within as many as 1 , 86 1
architectural spaces . These data d o not indicate t h e number of plastered
wal l s , the extent of the plaster when documented , or the condition of the
plaster . But , even when using these l imited dat a , one can readily see that the
c l i f f dwel l ings of Mesa Verde contain extensive and important mud plaster
resources . How extensive, how important, and i n what condition can only be
determined through the examination of each architectural space in the eighty­
f ive sites . Complete and systemat ic documentation of such an extensive
resource is a t ime-consuming and expensive proposition . Obviously priorities
must be established and a plan of action carefully formulated .

With the inception of the 1 9 8 5 survey pro j ec t , the contractor and park
staff collaborated on the development of recording forms and format s . Each
architectural space within certain c l i f f dwe l l ings was to be fully documented,
both i n written and photographic form . !t was decided to use a two-part ,
hierarchical survey form. One part generally documented individual architec­
tural spaces . The second form was des igned to document each wal l containing
plaster . Examples of these forms are available from the author . Data were
recorded for each plaster ' s location, method of application, number of layers ,
presence of design element s , condit ion, color (Munse l l ) , and extent . During
the second survey minor revisions were made to this form, such a s prompting for
environmenta l and 10cational data pertinent to plaster preservation and for
specific data concerning previous documentation and noted plaster deterioration
[ 19 ] . However , data consistency was a primary object ive of both surveys .

The plaster recording system used existing site maps and room numbering
systems , and it was hierarchical by design. The system uniquely ident i f ies
each wal l forming the plaster ' s immediate structural support , which i s the
basic recording unit . For example , MY 640-55-N denotes plaster on the north ,
interior w a l l of room 5 5 , in s i t e #MY 640 . Similarly, M Y 640- 5 5 -N-EXT iden­
t i f ies the same wal l , but an exterior surface. The format and the system
generally l i nked documented plaster walls to their spatial and functional
context s , served to systemat ize documentation while considering future auto­
mation of these dat a , and ass isted in the analysis of structural concerns or
ambient causes of deterioration [ 2 0 ] .

Photographic documentation included at least one bl ack-and-white photo­


graph of each plastered surface . Additional photograph s , b lack-and-white ,
color, or infrared were taken as needed to i l lustrate certain design elements
and condition s . Each photograph was given a unique reference code on a photog­
rapher ' s log. Photographs also included the appropr iate Kodak Scale and were
referenced on plaster recordation forms . Drawings were made when needed to
further i l l u strate complex or very poorly preserved mot i f s .

The 1 9 8 5 survey established precedents for standardized plaster and


condition terminology. Fully standardized masonry and architectural nomen­
c l ature has not been adopted by park staf f . Howeve r , terms such as shaped or
unshaped stone, single or double coursed wal l s , tower, kiva, and room were
recorded during the plaster survey as they were used during previous archeo­
logical work in the park. The descriptive terminology for plaster and condi­
tion was kept to a useful minimum and was kept general . Further, each term was
i l lustrated in final reports to facil itate their use and interpretation by
future researchers . The glossary of terms used to document plaster, for
examp l e , inc luded : aura incised design, painted design, incised and painted
Consolidation Studies

design, floor band, bichrome design, wainscot , colored wash, dado, de facto
plaster , undecorated plaster, and absence of plaster [ 2 1 ] .

The condition of the plaster was one factor evaluated when recommendat ions
for further action were made by the surveyors . Terminology for spec i f ic
examples of deterioration types and overa l l condition were established for both
the surfaces and the plaster substrate . Surface deterioration caused by
accretion ( i . e . , bird excrement , smoke blackening, nearby stabilization work ,
muddy runof f ) , and conditions such as powdering, f l aking, separation from
substrate and weathering losses were i l lustrated and defined . Along with the
above surface condition nomenclature, several types of plaster deterioration
were i l lustrated . These included lacuna, separation from wal l s , cracks ,
interlayer cleavage , and f a i ling supports [ 2 2 ] .

FIELD SURVEY RESULTS

The surveys documented plaster i n eighteen sites containing j ust over one
thousand architectural spaces . The c l i f f dwe l l ings chosen for initial survey
were the largest and best preserved . prel iminary data suggested that substan­
tial amounts of plaster, both decorated and undecorated, were s t i l l present in
all sites selected for documentation . Another factor influencing site selec­
tion was the interpretive importance of the site. Nearly a l l sites chosen for
survey were open to public vis itation or visible from public overlooks .

Over 1 , 900 square meters ( 20 , 500 sq. ft . ) of plaster on 8 6 9 surfaces were


documented during the surveys [ 2 3 , 2 4 ] . Approximately 75 percent of the plaster
and the surface are in fair to poor condition. These figures certainly i l lus­
trate an extensive resource in need of attention . Numerous preservation­
related recommendations resulted from the surveys . The more prevalent recom­
mendations were :

1. to apply s i l i cone beading to the ceil ing of an alcove to form false


driplines which divert water away from the buildings ;

2. to stab i l ize architectural supports - often the base of a wal l ;

3. t o closely monitor certain plasters ;

4. and for a conservator to evaluate selected plaster for treatment .

It is not practical , nor is it feasible, to attempt to preserve a l l


recorded plaster . A multicomponent action plan must b e implemented i f the
plaster resource i s to be documented and a significant sample preserved . A
suggested action plan includes the fol lowing six major sections:

1. Active stabilization: Stabilize wal l s , control moisture to the extent


possible , control rodents or birds, and el iminate insects - - continue
developing a comprehensive approach to site preservat ion .

2. Monitoring : Develop a methodology and schedule to monitor systemat ically


selected plaster based upon criteria such as overa l l condition, presence of
design element s , uniqueness , representativeness of more mundane finishes ( i . e . ,
washe s , monochrome or bichrome designs, auras , floor bands ) , interpretive
importance, feas ibi l ity of preservat ion, and previous treatment .

3. Evaluation for conservation treatment : Backf i l l selected spaces , solicit


conservation treatment proposals and contract for treatment , closely evaluate
the e f fectiveness of the pilot treatments performed at Mug Hou s e .

4. Training: Further sensitize park s t a f f ( i . e . those responsible for


management , interpretat ion, stab i l ization, and protect ion ) to the importance
and fragil ity of the plaster. Train the park ' s Stabili zation Spec ialist and
Ruins Stab i l ization Crew so that plaster monitoring, documentation, and
preservation can be fully integrated into both the present stab i l izat ion
program and future excavat ions , and ensure that newly recorded data are
compat ible with previous survey data.

5. continue field surveys : Survey remaining c l i f f dwe l l ings in priority


order, again ensuring compat ibil ity of new data with previous survey dat a ,
and improve survey methodology .

6. Expand researc h : Continue to support and forma l i ze research projects


related to plaster composition ( p igment s , binders , and mineral s ) , material
sources , plaster fabrication, determination of salt s , and conservation
treatment ( c leaning, consolidation, and reattachment ) .

PLASTER ANALYS I S

Mud plaster from Anasazi sites h a s not undergone scientific analysis until
relatively recent ly . Smith l 2 5 ] reports �he composition of kiva plaster from
two s ites in the Jeddito Valley as 90% sand and 10% clay. The report also
l i st s thirteen colors and possible pigment sources , noting that it i s unclear
i f pigments were added or if the clay and sandstone were previous l y stained .
All pigment s , except black, were apparently derived from inorganic material s .
smith also hypothesizes that a n organic binding medium could have been mixed
with pigments [ 2 6 ] .

Recent studies by S ilver [ 2 7 ] produced strikingly different and thought­


provoking result s . x-ray dif fraction analysis shows that prehistoric plaster
from Anasazi ruins in the four-corners area and a po s s ible mud source i n Mesa
Verde National Park contain no mineralogical c l ay s . Plaster from Lowry Ruin
contains 75% quart z , 5 - 1 0% feldspar, 5 - 10% mic a , and traces of additional
materia l . The raw mud from Mesa Verde contains 50% quartz and 50% anorthite .
The Lowry sample was also tested using stains spe c i f ic to organic compounds .
The plaster tested negat ively, while a paint layer from Lowry tested positively
for carbohydrates and proteins . The raw mud sample was not tested for organic
materia l s . As suggested by S i lver, the lack of mineralogical clay in her
samples i s remarkabl e , and the presence of organic binding media may part ially
account for the cohesiveness and durability of the paint layers and plaster .
Adobe 90

Mesa Verde plaster samples were recently examined by Dr. Mary Griffitts
[ 2 8 , 29 ) . X-ray d i ffract ion of kiva and room plaster samples from step House
and Mug House demonstrate a departure from both Smith ' s and s ilver ' s data.
Table I i l lustrates comparative sample data from S i lver ' s and Griffitt s '
report s . O f part icular note i s the relatively low percentage o f quartz and
high percentage of calcite in the white plaster from both Mesa Verde sites
analyzed by Griffitt s . Additionally, measurable percentages of mineralogical
clay are present in the Mesa Verde samples .

Table I . Comparative Results of X-Ray D i f fraction ( After S ilver 1 9 8 7 and


Griffitts 1989 , 19 9 0 ) - - Note : Totals may not equal 100% due to
sample variab i l it y .

S ite/S ampl e Quartz Calcite Feldspar Kaolinite Gyps um Mica

Lowry Ruin 75% 5-10% 5 - 10%

River House 80% Tr 15-20%


Ruin

MVNP - Raw Mud 50% 50%

Mesa Verde National Park Plaster Sample Data Follows :

step House - 93-95% 0-10% 10-15% 2-3% 1-2% 1%


Kiva A - dark

step House - 37% 15% 17% 2%


Kiva A - white

Mug House - Rm 43% 30% 8% 1% 2% 1%


2 8 , ext-white

Mug House - Rm 72% 4% 8 3% 2% 2%


2 8 , ext-pink

Mug House - 90% 15-20% 5%


Kiva C

While exam�n�ng thin sections from Mesa Verde plaster Griffitts ident i f ied
serec ite , a very finely divided mica resulting from the weathering of feldspar .
Serecite is not found in any of the geologic formations studied by Gr i f f itt s .
It is present , however , in the lower zonee of the park ' s loess soi l s , j ust
above an underlying layer of cal iche . The X-ray d i f fraction data and occur­
rence of serecite suggests that loess soi l s may have been combined with cal iche
( calcite) to create certain plaster s . Caliche evidently i s a major component
of the white plaster samples subjected to the above analytical test s . An
empirical experiment by Gri f f itts produced a good plaster ( 5 0% caliche and 50%
loess ) of similar color and appearance to some original Cliff Palace plaster .
The Mesa Verde plasters have not been analyzed for the presence of organic
binders . This should be done in the future . With such tests one must be
certain the samples have not been contaminated by previous conservation
treatments or altered during burial . It would also be interesting to test
hypothesized source material for the presence of organic material as a control ,
t o ensure the raw materials do not contain substances which might skew the
analysis .

CONCLUSION

Mud plaster in archeological sites of the American Southwest has not received
the attention it warrants or need s . The plaster survey program at Mesa Verde
National Park is unprecedented in scope , with over one thousand architectural
spaces inventoried and over 1 , 900 square meters ( 2 0 , 500 sq. ft . ) of plaster
documented . Surveying, locating, and recording plaster and documenting its
condition are necessarily the first steps in a comprehensive plaster resource
preservation program . The Mesa Verde survey project i l lustrates the importance
of such inventory data and the extent of the plaster resource in that park. It
also i l l u strates the need to establish preservation priorities and a multi­
faceted action plan since it i s not feasible or practical to conserve a l l
plaster . A suggested action plan includes stabil ization, monitoring, evalua­
tion of treatment needs , training, continued survey, and expanded research pro­
grams . Such programs are not only appl icable to Mesa Verde , but also to a l l
other areas with t h e potent ial f o r prehistoric plaster among their resource s .

Silver ' s 1 9 8 7 study discussing the presence o f organic binding media and
the absence of mineralogical clay in certain Anasazi plaster is intriguing.
Similarly, the high calcite content , presence of mineralogical clays, and the
presence of serecite in samples of Anasazi plaster from Mesa Verde i s also
interesting . Only a few samples comprise the present data base and more
samples must be analyzed. Additional research in the areas of plaster compo­
sition , pigment s , conservation treatment s , and plaster replication is needed .
This basic research must be performed if we are to reconc ile apparently
discordant data and to continue making progress with actual conservation
treatments and ensure the long-term preservation of Anasazi plaster .

With further refinements in our ability to ident ify the composit ion of
prehistoric plaster and material source s , plaster may prove to be a useful
cultural and temporal indicator . Anasazi masonry styles have been used to
i l lustrate certain attributes characteri z ing deve lopmental phases by which
sites are c l a s s i f ied [ 30 ) . Regional variations - in masonry style have also been
used to ident ify sites attributed to d i fferent branches of the Anasazi [ 3 1 , 3 2 ) .
The architectural finishes ( i . e . , plaster and paint s ) applied to the masonry
potent ially have similar capab i l ities .
Consolidation Studies

NOTES

1. Joan Pace, "Mesa Verde National Park C l i f f Ruins Tot a l s " ( Unpublished paper
on f i le Mesa Verde National Park, 1 9 80 ) , 1 .

2. National Park Service, Mesa Verde National Park - Resource Management Plan
( Unpubl ished draft on file Mesa Verde National Park, 1990 ) , 2 8 .

3. Ibid . , 31.

A. Jesse L . Nusbaum, The 1 9 2 6 Re-excavation o f step House Cave , Mesa Verde


Research Serie s , no . 1 ( Mesa Verde National Park:
ation, 1 9 8 1 ) , 2 0 .
Mesa Verde Museum Associ­

5. G . Nordenskiold, Ruiner of Kl
A . Norstedt & ippb
Soners , 1893 ) , 1 2 7 - 1 2 8 .
o ningar I Mesa Verde ' s Canons , ( Forlag : P .

6. Jesse Walter Fewke s , Antiquities of the Mesa Verde National Park. Spruce
Tree House, Bureau of American Ethnology , Bulletin No . 41 ( Washington, D . C . :
Governement Printing Office, 1909 ) , 10 .

7. George Cattanach, Long Hou se , Archeological Research Series , No . 7-H


(Washington, D . C . : National Park Service, 1 9 80 ) .

8. Arthur H . Rohn , Mug Hous e , Archeological Research Serie s , No . 7-0


(Washington, D . C . : National Park Service, 1 9 7 1 ) .

9. Cattanach, Long Hou se , 14.

10 Rohn, Mug Hou s e , 48.

11. Ibid . , 70.

12 . Watson Smith, Kiva Mural Decorations at Awatovi and Kawaika-a , Papers of


the Peabody Museum of American Archaeology and Ethnology , Vol . XXXV I I
( Cambridge : Harvard University , 1 9 5 2 ) .

Sympo
13 . Giacamo Chiari, "Treatment of Adobe Friezes in Peru , " in Third Internation
sium on Mudbrick Preservat ion, ed . E . Madran ( Ankara : ICOM and ICOMO S ,
1 9 80 ) , 3 9 - 4 6 .

14. Paul M . Schwartzbaum, C . s . Silver, and Christopher Wheat ley , "The

E . Madran ( Ankar a :
Sympo
Conservation of a Chalcolithic Mural Painting on Mud Brick from the Site of
Teleilat , Jordan , " in Third Internation sium on Mudbrick Preservat ion, ed .
ICOM and ICOMOS ) , 1 7 7 -200 .

15. Constance S . Silver, " State of Preservation of Pueblo Indian Mural


Paintings in the American Southwest " ( Paper on file ICCROM, Rome , 1980 ) .

16. Constance S . Silver, " 1 9 8 1 Report on the Development of Methods for the
Conservation of Pueblo Indian Mural Paintings in the American Southwest"
( National Museum Act Report on file Smithsonian Institution, Washington, D . C . ,
1982 )

17 . Ibid . , 1 7 -2 2 .

18 . Alden C . Hayes , The Archeological Survey of Wetheri l l Mesa , Archeological


Research Serie s , No . 7-A (Washington, D . C . , National Park Service, 1 9 64 ) , 1 2 0 .

19 . Constance S . S i lver , " Summary of the Results o f the 1 9 8 5 Project : Survey


of the Plaster and Rock Art at Mesa Verde National Park" ( National Park Service
Contract #1490-84-04 Report on f i le Mesa Verde National Park, Colorado, 1 9 8 5 ) ,
Vol . 1 , 1 2 -2 8 .

20. J . Fetterman and L . Honeycut t , "The 1 9 8 7 Mesa Verde Plaster Recordation


Project" ( National Park Service Contract #1490-7-0002 Report on f i l e Mesa Verde
National Park, Colorado, 1 9 89 ) , Vol . 1 , 3 .

21. S i lver, " Summary of the Results of the 1 9 8 5 Proj ect : Survey of the
Plaster and Rock art at Mesa Verde National Park , " glossary.

22 . Ibid . , glossary .

23. Ibid . , 55-59 .

24. Fetterman , "The 1 9 8 7 Mesa Verde Plaster Recordation Project , " 9 - 5 1 .

25. smit h , Kiva Mural Decorations at Awatovi and Kawaika-a, 18.

26. Ibid . , 18-3 0 .

27. Silver, " Summary of th� Results of the 1 9 8 5 Project : Survey of the
Plaster and Rock Art at Mesa Verde National Park , " 7 5 - 10 5 .

28. Mary L . Gri f f itt s , letter t o author, 2 2 June 1989 .

29 . Mary L . Grif fitt s , letter to author, 27 March 1990 .

30. Haye s , The Archeological Survey of Wetherill Mesa, 39-4 1 .

31. Robert P . Powers , W . B . Gil lespie, and s . H . Lekson, The Outlier Surv ey.A
Regional View of Settlement in the San Juan Basin , Reports of the Chaco Center ,
No . 3 ( Albuquerque, National Park Service, 1983 ) , 307-32 7 .

32 . Alden C . Hayes , D . M . Brugge , and W . J . Judge , Archeological Surveys of


Chaco Canyo n , Publications in Archeology, No . 1 8 - A ( Washington, D . C . , National
Park Service, 1 9 8 1 ) , 48-69 .
266

ACKNOWLEDGEMENTS

Financial support for this project was supplied by the Horace Albright Employee
Development Fund. The research was conducted while I was Regional Curator ,
Rocky Mountain Regio n , and it could not have been concluded without the support
of the Division of Cultural Resources . Special thanks are due to Dr . Mary
Gr i f f itts for her timely research and gracious ass istance . The staff of the
Mesa Verde National Park Research Center was extremely helpful , providing
photograph s , reports , and work space . Thanks are also due to Margo Surovik
Bohnert for her support and patience and to Dr . Ann Johnson for her editorial
assistance . I remain responsible for a l l errors and omissions .

Figure 1 . 1 8 9 1 photograph o f Square Tower House .

Figure 2 . 1 8 9 1 photograph o f Square Tower House, by G . Nordenskiold.


Consolidation Studies

ABSTRACT CHEMICAL S URFACE T REATMENTS AND CAPPING TECHNI Q U E S OF EARTHEN


STRUCTURE S : A LONG-TERM EVALUATION .

Among the v a r i o u s techn i ­


que s o f e a rthen a r c h i t e c ­ Giacomo Chiari
tu r e p r e s e rv a t i o n w h i c h Dipartimento d i Scienze della Terra
have b e e n app l i ed i n t h e Via Valperga Caluso 3 7 . 1 - 1 0 1 2 5 Torino .
past , t h e cons o l idation o f Italy
vertical surfaces b y chemi­
cal agents and the capp ing
of t h e top p a rt o f wa l l s Introduction
p l a y a n i mp o r t a n t r o l e .
The i d e a l character i st i c s The idea of solving the problem of adobe preservat i on by coating
o f a good c o ns o l i dant a r e the surface with some perfect conso l idant shoul d b e d i sm i ssed .
outl ined . The consol idation Each preservative shows advantages and disadvantages ; the perfect
me c h a n i s m s o f s y n t h e t i c treatment has not yet been d iscovered and probably never w i l l be .
r e s i n s and ethyl s i l i c a t e Adobe is a weak material that has always been used with the idea
ar e d i s c u s s e d , t o g e t h e r o f constant maintenance and repa i r . I n most cases the wal l s were
with the i r advantag e s and origina l l y protected by roofs , whi ch in archaeol ogical excava­
d isadvantages . The perform­ t i ons are m i s s ing . One cannot expect to stop the natural evolu­
a n c e o f f i e l d t re atments tion and modi f ication o f the mater i a l . A l l we can hope to do i s
with ethyl s i l icate and of to reduce the speed o f deterioration .
cap p i n g t e c h n i q u e s a ft e r
twenty years o f app l i cation Chemical surface treatments
are e v a l u at e d f o r a c a s e
st u d y i n I r a q . A w a y t o An ideal consol idant for adobe should have the fol l owing charac­
rega i n g o o d adh e s i o n o n teristics : 1 ) confer water res istance but not water rep e l lency in
det e r i o rated surface s , ord e r to a l l ow wat e r migration both in l iquid and vapour phas e ;
based o n the app l i ca t i on o f 2 ) l eave pores and cap i l laries ope n , and al l ow for other impreg­
r i c e paper ( mo i stened and nat i o n s , e v e n w i t h d i f f e re n t p r o d u ct s ; 3 ) c o n f e r m e c h a n i ca l
pressured) is described . strength and abrasive resistance both in dry and wet conditions ;
4 ) have good penetration , i . e . , low v iscosity ; 5 ) should not form
f i l m s on the s u r f ac e , n o r show an abrupt p l anar boundary w ith
KEYWORDS respect to the untreated core ; 6) have a thermal expansion coef­
f i c i ent s im i l a r to that of adob e ; 7 ) should not change the co­
l ou r , or cause gloss ; 8 ) be resiBtant to stresses caused by salt
Adobe , Earthen Architec­ crysta l l i zation , capil lary rise o f ground water , and freeze-thaw
ture , C o n s o l i d a t i o n , Ethyl cyc l es ; 9 ) be durab l e , i . e . , resistant to water , and photo-oxida­
S il icate , Syntheti c resins , t ion ; 1 0 ) be easily app l i cable , possibly also in damp conditions ,
Capp ing , Long-term evalua­ and cheap ; 1 1 ) should not be harmful to the operators ; 1 2 ) should
tion . be revers ibl e , if possible ( se e also [ 1 ] ) .

It is my op inion that a product ful f i l l ing a l l the s e charac­


teristics does not exist .

The c o n s o l i dants most u s e d on adobe a r e synth et i c r e s i n s ,


u s ua l l y the rmop l a s t i c , and ethoxy s i l an e s . I t i s important to
und e r l i n e that g e n e r a l statements on t h e b e h a v i ou r o f who l e
c l asses o f compounds have almost no meaning . Products that have
the same nominal compo s it i on may vary greatly f rom one producer
to anoth e r ; a l s o the app l ication techn i qu e could i n fluence the
f i n a l r e s u l t s . Each i n d i v i du a l p roduct shou l d , there f o r e , b e
tested on t h e spe c i f i c material , possibly with accelerated aging
t e s t s , w e t - d r y c yc l e s , and s a l t s c ry s ta l l i z a t i on . E v e n t h i s
·
cannot assure that a consol idant which has given good results in
the l aboratory woul d behave equa l ly wel l on the l ong-term f i e l d
appl ication . Large f i e l d comparative tests a r e presently carried
out on s p e c i a l l y bu i l t wa l l s at Fort S e l de n ( New M e x i c o State
Monument s ) and in Grenob l e ( CRATerre ) for various products and
capp ings . The results of these tests w i l l b e extreme l y valuab l e
t o ident i fy suitab l e consol idants . Keeping thi s i n mind , one can
try to d e s c r ib e the c o n s o l i d a t i o n m e c h a n i s m and to e v a l u a t e
advantages and disadvantages o f both types o f consol idants .

synthet i c res ins

Syntheti c resins are l ong chains o f organic polymers derived from


268 Adobe 90

a vast range o f monomers . The most commonly used are the polyvi­
nyl acetates , acry l ics ( among them Acriloid B7 2 and Primal AC3 3 )
and polyisocyanates [ 1 ] . They can be used in solut i on i n organic
s o l v e n t s or in water emu l s i o n s , or the p o l ymer i z at i o n can be
obt a ined in s it u by the u s e of c a t a l y s t s or by react i o n with
atmospheric moisture .

Solut i ons are best suited for surface consol idation , because
the products are pure , and they show good aging propert i e s and
penetration .

W a t e r emu l s i o n s are o b t a i n e d by add i t i o n o f s u r f a c t a n t s


( usual ly soap-l ike substance s ) . These additives may increase the
speed of deteriorat ion of the res in , by ox idat ion , breaking o f
the polymer chain , and cross-l inking between cha ins . The results
are change in c o l our , brittlene s s , and break down of mechan ical
propert i e s . These mechan i sms requ ire the action of l ight ( espe­
c i a l l y UV-rad i a t ion ) and oxygen . In the case of emu l s i o n s the
po l ymer g l ob u l e s s u spended i n water are re l a t iv e l y l arge , the
l i quid has a high viscos ity , and penetrat ion i s l ow . Water is not
a good carrier in the case o f adobe , s ince it causes swel l ing of
F igure 1 . Scanning E lectron
the clay part i c l e s and decreases the mechanical propert i e s w i th
Microscope ( S EM) image of a
the risk of material detachment during the treatment . Emu l s ions
no n t r e a t e d s a m p l e f r o m
Hatra . shou l d t h e re f o re b e a pp l i e d as adhe s iv e s o n l y , b y i n j e c t i o n
ins ide the wal l s , and never on the surface .

synthetic resins act as consol idants by penetrating ins ide


the p o r e s and c o a t i n g the l o o s e part ic l e s . Chem i c a l rea c t i on
normally does not take place between the polymer and the materi­
al . The strengthening is obtained by the setting of the resin at
the moment i n which the s olvent evaporates . I n many cases , re­
verse m i g r a t i o n o f the p o l ymer to the s u r f a c e , as the s o lvent
( especially i f highly volat i l e ) evaporates , causes the formation
of a thin f i l m . I f the coating is not porous , which is the case
for most synthet ic resins , and does not a l l ow for water trans­
port , both i n the l i quid and vapour pha s e , the water that can
gain access beneath the protective l ayer causes stress and de­
tachmen t . Most synthet i c r e s i n s have high therm a l exp a n s i o n
coe ffic ient s , o f one order of magnitude larger than adobe . S ince
the surface tends to be warmer than the ins ide during the day ,
and colder during the n ight , stress . is devel oped at the inter­
face , with pos s ib l e detachment . Among the advantages o f synthetic
re s i n s one can quo t e : good m e ch a n i c a l p r o p er t i e s , a c e r t a i n
degree o f revers ibil ity with non-crossl inking types , and the fact
F igu r e 2 . S EM image o f a that they may act as adhes ives as wel l .
samp l e taken from the wa l l
treated with ethyl s i l icate Technical description of the ethyl s i l icate reaction with earthen
in Hatra . Although there is material
very l it t l e d i f ference be­
tween the two " l andscapes " , Ethyl s i l icate ( tetraethoxysilane) is partly inorganic and partly
the c o n s o l i da t e d p a rt i s organ i c , but a fter complete curing ( which may take a l ong t ime )
water resistant .
it l eaves a purely inorganic residue . There are several kinds o f
commer c i a l products ( se e t h e Mater i a l s s e c t i on ) . T h e monomer
consi st s o f a s i l icon atom to which four ethoxy groups are bond­
ed , si ( OCH2CH3 ) 4 ) . When a water molecu l e reacts with an alcoholic
residue , hydrolys i s takes p l ace : one ethyl a l cohol molecu l e i s
formed ( which evaporates ) and a n acidic residue remains attached
to the s i l icon atom to form S i -OH .

The hydrolysi s reaction is the fol l owing :

CH 3 CH 2 -O 0-CH 2 CH 3 CH 3 CH 2 -0 0-CH 2 CH 3
", � "' �
si -- -- Si
CH 3 CH 2 -O
/' "- i
° i CH 2 CH 3 CH 3 CH 2 -O
/' "-
0 + H-0-CH 2 CH 3
I
H .!. O-H
I
�--
I
H
Consolidation Studies

The hydrolysis of the four groups can take place at d i fferent


t ime s . I f the four alcoho l i c residues are a l l hydrolyzed , s i l ic i c
a c i d i s f o rme d : S i ( OH ) 4 . When t w o a c i d i c g r o u p s b e l on g i ng t o
d i f f e re n t m o l e c u l e s interac t , c ondens a t i o n or polymeri z at i on
takes place . One water molecule i s released ( and i s aga in avai l ­
abl e f o r t h e hydrolys i s react ion) a n d a strong s i- o - s i bond i s
formed .

The condensation reaction is the fol l owing :

F igu re 3 . App l i c a t i o n o f
mo i s tened r i c e paper : the
pressure a l lows adhes ion to
be reestabl ished .

+ H -o-
H

A sma l l amount of acid acts as a catalyst for hydrolysis . In


the app l i cation to adobe , there is no need to increase the speed
o f the react ion , and better results are obtained without addition
o f acid [ 2 - 4 ] . with polymeri z ation , a three -d imens ional network
o f s i l ica tetrahedra sharing one vertex is formed . S ome of them
usua l l y s t i l l have ethoxy groups attached to the s i atom . The
Figure 4 . Detachment o f the
clay particles abundant in adobe have a l arge number o f hydroxyl
paper w ithout any l o s s o f
groups ( -OH) l ocated on the i r surface . Cond e n s a t i o n may occur
surface material .
between the acidic group o f the s i l ica framework and the hydrox­
y l s of c l ay part ic l e s . Of cours e , an extreme ly l a rge number o f
such bonds a r e formed , and , i f the clay particles a r e sufficient­
ly close to one anothe r , the s i l ica framework helps to keep them
together .

One of the maj or causes of deterio r ation of adobe is water,


which separates the clay particles constituting the binding agent
of the b�icks . On excess ive wetting eventual l y the c l ay is d i s ­
persed i n a water suspen s ion . The ethyl s i l icate treatment , by
add i ng s trong bonds between the c l a y , prevents c l a y p l a t e l et s
from be ing separated b y water , a n d therefore g ives the material
the necessary water resistance . For the first period after treat­
ment , s i l ica gel is formed ins ide the pores and the total porosi­
F igure 5 . Overa l l v i ew o f ty is reduced . Whi l e the polymerizat ion continues , the s i l ica gel
the capp ing with one l ayer contracts and the pores reopen . with t ime , very l it t l e material
o f s trengthened b r i c�s at remains ins ide the adobe , even the micro-pores are a lmost com­
Te l l ' Um a r , s h o w i n g t h e pletely open , but the c l ay part i c l e s are s t i l l bonded together .
plant overgrowth . The overal l e f fect of the treatment is to confer water · res istance
to the material but not water rep e l l ency , ne ither at vapor nor
l iquid l evel [ 5 ] .

F igure 1 shows a scann ing e l ectron m icroscope ( SEM) p icture


o f an untreated samp l e from Hatra , whi l e Figure 2 shows a treated
one . It c a n be s e e n that the changes due to the treatment are
hardly detectable .

The t r e atment i s c omp l e t e l y i r r ev e r s i b l e , v i o l a t i n g the


p r i n c i p l e t h a t every i n t e rv e n t i o n i n c o n s e rv a t i on s h o u l d b e
tota l l y revers ibl e . The f a c t that n o t o n l y t h e surface a n d ap­
pearance o f the material , but also its intimate structure , under­
goes so l ittle modi fication may in part j ustify the irreversibi­
l ity . Other conso l idant s , even total ly d i f ferent i n nature , can
F igure 6 . A section of the
eas i l y be appl ied , s ince the porosity and polarity o f the materi­
cover i ng with one l ayer o f
al are pract ically unchanged .
strengthened bricks , s t i l l
in perfect condition . (Tell
' Umar) . The app l ication by spraying makes i t s use very easy and has
Adobe 90
the advantage of obtaining a l a rger penetration in those parts
which are more porous than others , thus l eaving a very irregular
separation between the treated and untreated parts . This strongly
red u c e s the c h a n c e of d e t a chment of the s t rengthened l a yer .
Cl eaning of the surface and extraction o f soluble s a l t s can be
done a ft e r c o n s o l i da t i on . Th i s is n o t p o s s i b l e i f synth e t i c
res ins are used .

Among the disadvantages , bes ide its irreversibi l ity , one can
recal l that the treatment cannot be app l ied to wet surfaces . In
thi s case , the excess o f water causes the hydrolysis reaction to
take place at much h igher speed than polymer i z at i on . A g l os sy ,
frag i l e crust is formed , which crumb l e s without making the de­
F igure 7 . On the north s i ­ s i red connections with the clay particles . synthetic res ins also
de , t h e u n p r o t e c t e d wa l l present appl ication problems to wet surfaces .
ha s s u f f e r e d a n e n o rm o u s
loss of materi a l , due to
Another d i sadvantage i s that ethyl s i l icate i s not an adhe­
the ra in . ( T e l l ' Umar) .
s ive , but only a consol idant . If a gap already exists between two
b l ocks o f adobe , they w i l l be ind ividua l l y consol idated but not
"glued" together . To achieve adhe s ion between a l ready separated
parts , one should make use o f other strategies : the most obvious
is to intervene as soon as poss ib l e , idea l ly during excavat ion ,
in order to have surfaces that are not yet damaged by weathering .
I f the surface is already damaged it is possible to regain adhe­
s ion by remodel ing the surface us ing rice paper , water and pres­
sure . The paper a l l ows the cru s t to b e susta ined wh i l e i t i s
carefully moistened . The material becomes s l ightly plastic , and ,
by exerci s ing pressure with a hard sponge , one can reestabl i sh an
acceptable degree of adhes ion . When almost dry , the paper can be
detached without any l o s s of mater i a l , including p igments ( See
fig . 3 , 4 ) . Both o f these strategies were app l ied with success in
the c o n se rv a t i o n of a pa inted frieze i n Carda l , Peru , i n 1 9 8 7
( unpub l i shed ) . I f the detached crust i s thicker than a few m i l l i ­
meters , the risk o f i t fal l ing during the app l ication of the rice
paper is too high . Inj ections of synthetic resins ( water emu l ­
F igure 8 . The canal i zation
s ions ) in the interior o f the wal l are recommended i n this case .
of ra i n w a t e r o v e r T e l l
' Um a r . I n s p i t e o f s o m e
plant growth and some minor Preservat ion o f s ites in Irag: a case study
damage , this dra inage sys­
tem worked properly . In 1 9 6 8 a prel iminary campaign for the conservation and preserva­
t ion of archaeological finds in unbaked earth was carried out in
Iraq , with the a im of documenting the problem o f mud-brick dete­
rioration [ 6 , 7 ] . Vari ous l aborator� tests of surface protect ion
were p e r fo rmed , u s i ng most of the products w i d e l y used at the
time . Ethyl s i l i cate seemed to give the best prel iminary results
and was s e lected for maj or field tests , done in the Sel eucia area
and in Hatra in 1 9 6 9 . Synthetic resins ( p o l yv inyl acetates and
acryl i c s ) were also used on a minor sca l e , by inj ection .

One l iter o f S i l ester Z N S , three l iters o f ethyl a l coho l


( 9 6 ° ) , and 1 ml o f hydrochloric acid as a catalyst was used to
treat 1 m2 . The penetration was 2 -3 cm in depth , and a fter a day
or two the wal l reassumed its previous colour . No changes in ap­
pearance were obs erved , but the water r e s i stance was great ly
enhanced . After a month o f spraying water on sel ected spots three
times a day , there was no evidence of eros ion , whi l e on a nearby
untreated part a hole was formed after the first three sprayings .

C app
i ng o f the top part o f wal l s

The yearly average rainfal l in the area is about 3 0 0 rom , concen­


Figure 9 . A natural cana l i ­ trated in a few torrential storms . Furthermore , in some sections .
za t i o n o f water , l e ft peop l e had to wa lk on top of the wal l s . For these reasons , it was
unprotected , to be compared decided that the chemical surface protection a lone would not be
to F i g . 8 ( Te l l ' Umar) . suffic ient , and two capping techniques were tried . The first one ,
used in the "Archives" of Seleuc ia , consisted of a l ayer of a few
cm formed with a m i x o f earth and sand , with the addition o f a
minimum amount ( 5 - 8 % ) of portl and cement , to avo i d an excess ive
Consolidation Studies

hardne s s . This capping was app l ied in two layers on the moistened
surface . After good compres s ion and drying of the first l ayer ,
the few unavoidable cracks were sealed by a second , thinner l ayer
made with a mixture less rich in water . For a few days straw mats
s l ight l y s p r i n k l e d w ith water covered the capp ing to avo i d an
exces s ivelY quick drying process . with thi s procedure the crack­
ing of the surface was pract ica l ly avoided .

Trials of capping with addition of asphalt were a l s o attempt­


ed , but immediately proved to be d is astrou s , perhaps because o f
the poor homogeneity o f the s lurry , which was mixed b y hand .
F igu re 1 0 . A v i ew of the
"Archives " in Seleucia . The A second type of capping was used to cover a large section o f
top of the wal l s was capped a wa l l , at Te l l ' Umar . I t cons i sted o f a s in g l e l ayer o f new
with s o i l -cement , which did stab i l i z ed bricks , made with the same s o i l -cement mix with the
not prove to be e f fect ive , addition o f straw . This a l l owed for a proper drainage system for
s ince there was format i o n the water that ran down from the artificial h il l . A l arge quanti ­
o f water poo l s a t the bases
t y o f bricks w a s manufactured us ing wooden molds . They were wel l
o f the wal l s .
dried in the sun f o r over a month , covered with wet straw mats
for the first week and turned over every three days . The result­
ing bricks were perfectly sol id , without any cracks .

Evaluation of the results

All the work i n I raq was done mainly in 1 9 69 . In the spring o f


1 9 7 1 , the ethyl s i l icate treatments were s t i l l in perfect condi­
tion , and it was easy to notice the difference between the treat­
ed and untreated parts , which had already severely suffered from
rain . The capp ing with s o i l cement showed some cracks , and in a
few parts water i n f i l tration had eroded preferent i a l channe l s .
The new l ayer o f bricks was perfectly pres e rved w ith the excep­
tion o f the vert i c a l wa l l s at the end o f the drainage system
which , being immersed in wate r , had col lapsed . At that time , some
rep a i r s were made . A fter th i s , no m a i ntenance work at a l l wa s
done to the s ite , which was abandoned .

The present s ituation . •a fter twenty years


F igure 1 1 . Another v i ew of
a s ma l l room i n the " A r ­ In May 1 9 8 9 a c r i t i c a l eva l u a t i o n o f the work wa s d o n e . The
chive s " . capp i n g w i th o n e l ay e r of s t ra ightened b r i c k s g a v e the b e s t
resul t s ( see figs . 5 , 6 ) . The canal that w a s dev is e d to d i sperse
the water proved to be e f fective ( see fig s . 7 , 8 , 9 ) . At one point
the water found a d i f ferent path , and a lot of damage occurred .
It wou l d have been e a s i l y avoided i f the s ite had been m a i n ­
tained .

The necess ity of maintenance can never be stressed enough .

Some of the vertical wal l s were broken by the pressure of the


ro o t s a n d t r u n k s o f tre e s . On the north s id e , where the n ew
bricks were not put in p l a ce , e r o s i on due to water c a u s e d the
loss o f large mas s o f earth ( see f ig . 7 ) .

Overa l l , ' the technique of putting one l ayer of new bricks on


top of the wal l s can be j udged in a very positive way . Of course ,
the original material is no longer visibl e , but a l l archaeologi­
ca l and a r c h i t e ctura l i n f o rma t i on is s t i l l r e t r i ev ab l e f rom
d i rect inspection . Thi s sacr i fi c i a l l ayer can eas i l y be removed
in case more excavat ion is needed . This techni que should not be
'
confused , by any means , with large reconstruction o f wal l s that
F igu re 1 2 . D e t a i l of the are quite o ften carried out even us ing baked bricks set on top o f
s o i l - cement capp ing , par­ sma l l rema ins o f adobe wal l s . T h i s type o f intervent ion should
t i a l l y detached and crack­
never be done .
ed . B e s i d e s damage due to
temp e r a t u r e c h a n g e s a n d
wa t e r i n f i 1 t r a t i o n , it The general condition of the Archives was disappointing ( see
shou l d b e remembered that figs . 1 0 , 1 1 ) . Most o f the smal l rooms were f i l l ed with earth , in
peo p l e wa lked on the cap­ part f a l len from the wal l s , in part carried in by the wind . In
ping . some parts the capping had resisted , whi l e in other parts it was
Adobe 90

cracked , a l l owing water infi ltrations ( S ee f i g . 1 2 ) . This con­


firms the op inion [ 8 , 9 ] , that in the case o f sma l l rooms l ocated
below the field level , without dra inage and with the poss ibil ity
of water pool format ion at the base of the wal l s , the only poss i ­
ble intervention is complete , immediate backf i l l ing .

To c h e c k i f the c a p p i ng reduced in any way t h e s p e e d o f


deteriorat ion , a comparison was done with the nearby excavation
of "Via Porticat a " , which had undergone the same abandonment for
twenty years . Most of the wal l s compl etely disappeared , with the
exc e p t i o n o f two rooms , capped w i th s o i l cement ( s e e f ig s .
1 3 , 14 ) •

It can be concluded that the endurance o f wa l l s can be en­


hanced , t o a c e r t a i n extent , by s o i l -cement capp i ng , provided
that the bases of the wal l s are not in direct contact with wate r .
It should b e noted , however , that constant maintenance is needed ,
which s e ldom can be ensured . Furthermore , s ince the weakest point
is the connect ion between the capping and the original vertical
surface , capping the top part of a wal l has some meaning only if
F igure 1 3 . Rooms a� " V i a the vert i c a l s u r f a c e s are consol idated a s we l l . I n any other
Porticata " , in the Seleucia cas e , the capping i s almost ine ffective .
area , without any conserva­
tion work .
The ethyl s i l icate treatments done at the Arch ives were not
visibl e , s ince most o f the wal l s were covered with debri s . After
excavation the consol idated part was not relocated . The mo isture
and s a l t content were extreme ly high , and the very next day the
ent i r e surface o f the excavat ion wa s covered w i t h white s a l t
efflorescence s . The water tab l e l evel was l e s s than a meter deep .
Unfortunately there is not enough documentation to estab l i sh when
and why the t reatment fa i l e d . What c a n be s a i d i s that under
dra s t i c cond itions - e . g . , when a wa l l i s impregnated with s a l t
water f o r twenty years - the consol idation with ethyl s i l icate is
not e f fective enough to protect it .

In Hatra environmental cond it ions are d i fferent from S e l eu­


cia . The rainfa l l i s more or l e s s the same , but the water table
is much l owe r . The treated wa l l had a stone base and did not
collaps e . The e ffect o f ra in al one on treated and untreated parts
can be seen . One row of bricks at the top was l e ft untreated for
comparison . Figures 15 and 16 show. that clay from the top row was
washed down and covered the treated bricks . This clay encrusta­
tion was e a s i ly removed by the u s e o f a rough b ru s h , wi thout
e f fecting the consol idated surface . Even very fine deta i l s ( s ee
F igu r e 1 4 . T w o r o o m s a t
"Via Port i cata " wh ich were figs . 1 7 , 1 8 ) were perfectly preserved .
pr o t e c t e d w i t h the s o i l ­
ce m e n t c a p p i n g . I n t h i s I t c a n be c o n c l ud e d t h a t in t h i s c a s e ( a s i n many o t h e r
case the capp ing performed s imilar ones , f o r exampl e in Chan Chan , Peru ) . t h e ethyl s i l icate
bett e r , p r o b a b l y b e c a u s e treatment did confer enough strength to the surface to counteract
peop l e d i d n o t wa l k over the e ffect of rain for twenty years .
it .
General conclusions

Of the various conservation measures undertaken in I raq twenty


years ago , some endured this l ong period of abandonment remark­
a bly wel l , and some did not .

Among the positive interventions one can note : a ) the cover­


ing of the top of wa l l s with a sacr i f ic i a l l ayer of one row o f
new bricks ; b ) the repairs of a l ready damaged wal l s b y the use o f
the same kind o f bricks , when wel l anchored to the original part ;
c) the water d i sposal obtained by the cana l i z a t i on o f ra infa l l
us ing a d hoc des igned paths made with the strengthened bricks ; d )
t h e surface treatment o f vert i c a l wa l l s w i t h e t h y l s i l icate ,
provided that the bases of the wal l s are not damaged by wate r .

Even sma l l faults in t h e execut ion o f t h i s k i nd o f work can


Consolidation Studies 273

result in seriou s damage . It seems advisabl e , there fore , to re­


evaluate the interventions a fter a period of time , correcting the
p o s s i b l e m i stake s . It shou l d be stressed that a regul a r mainte­
nance program is o f paramount importance .

Among the f a i l ures i s the capping done with a thin l ayer o f


strengthened s o i l d i rectly on top o f wal l s . Although i t produced
some r e su l t s , it w a s not s u f f i c ient to p r e s e rve the wa l l s f o r
such a long period , especiallY under the very harsh conditions at
S e l eu c i a Archive s . The formation of water p o o l s at the bases o f
wal l s rema ins the biggest prob l em .
F igu r e 1 5 . Wa l l t r e a t e d
wi t h e t h y l s i l i c a t e i n The ethyl s i l icate treatment on surfaces of wa l l s severely
Hatra , a s i t a p p e a r e d i n a f f e ct e d by s a l t water a l s o s ee m s not t o be a s t ro n g e nough
19 6 9 . The t o p l a y e r o f protect ion .
bricks was not consol idat­
ed .
References

1. N . Agnew , F . Preusser and J . R . Dru z i k , " strategies for adobe


p r e s e rv a t i o n . The Getty C o n s e rv a t i on I n s t i t u t e r e s e a rch
program" , 5th Intern . Meeting o f experts on the conservation
of Earthen Architecture . ICCROM-CRATerre ( 19 8 7 ) : 3 - 1 1 .

2. G . Chiari , " Conservac ibn d e l o s monumentos arqueologicos en


adob e : Peru " , UNES CO , RLA/ 0 4 7/ 7 2 - FMR/ SHC/OPS/ 2 4 3 ( UNDP)
( 19 7 5 ) .

3. G . Chiari , "Treatment of adobe friezes in Peru " , I I I Interna­


t i onal Sympos ium on mud-brick ( adobe ) preservation . Ankara ,
Turkey . ICOM-ICOMOS ( 19 8 0 ) : 3 9 -4 5 .
Figure 1 6 . View of the same
wal l in 1 9 8 9 . The bricks on 4. R . Rossi Manaresi and G . Chiari , " E f fectiveness o f conserva­
th e t o p a r e w a s h e d a w a y t i on treatments of a vol canic tuff very s im i l ar to adobe " ,
fr o m t h e r a i n , a n d t h e I I I International sympos i um on mud-brick ( adob e ) preserva­
rna t e r i a l ran over the t i o n . Ankara , Turkey . ICOM-ICOMOS ( 19 8 0 ) : 2 9 -3 8 .
l ower , consolidated part .
5. G . Ch i a r i , " Co n s o l i d a t i o n o f a d o b e w i th e t h y l s i l i c a t e :
control of l ong term e f fects us ing SEM" , 5 th International
Meeting o f experts on the conservation o f Earthen Architec­
ture . ICCROM-CRATerre ( 19 8 7 ) : 2 5 -3 2 .

6. A . Bruno , G . Chiari , C . Trossare l l i and G . Bult inck , " Contri­


bution to the study of the preservation o f mud-brick struc­
tures " , Mesopotamia III-IV , ( 19 6 8 - 69 ) : 4 4 3 -4 7 9 .

7. G . Torraca , G . Chiari and G . Gul l ini , "Report on mud-brick


preservation" , Mesopotamia VI I , ( 19 7 2 ) : 2 5 9 - 2 8 7 .

8. A . A l v a a n d G . Ch i a r i , " Protect i o n a n d c o n s e rvat i o n o f


excavated structures o f mudbrick" , i n Conservation o n Archae­
o l ogical excavat ions . ed . N . P . Stanley Price ( Rome : ICCROM ,
F igu re 1 7 . D e t a i l of the
1 9 8 4 ) : 1 0 9 - 12 0 .
same wall in 1 9 6 9 .

9. G . Ch i ar i , " Characte r i z ation o f adobe a s bui l d ing material .


Preservation techniques " , International symposium and train­
ing workshop on the conservation o f adobe . Lima-Cuzco ( Peru)
UNDP/UNESCO/ICCROM ( 19 8 5 ) : 3 1-4 0 .

Materia l s

S I LESTER ZNS : ethyl s i l icate partially condensed ( about 1 0 mole­


cu l e s of monomer ) , produced by MONSANTO ( US A ) . Dea l e r : P i etro
carin i , Via S . Marta 2 3 , Milano ( Italy) Tel . 0 6 - 8 7 4 4 7 7 .

TEOS : tetraethyl-ortho s i l icate , produced by Union Carbide Corpo­


ration . 2 7 0 Park Avenue . New York 1 0 0 1 7 (USA) .
F igu r e 1 8 . S a m e d e t a i l
after parti a l c leaning with Wacker Strenathener OH : Ethyl s i l i c ate m ixed with s o l vent ( to­
a h a r s h brush ( s e e e s p e ­ luene) and catalyst , produced by Wacker-Chemie GmbH , Prinz regen­
c i a l l y t h e wh i t e g y p s um tenstrasse 2 2 , Munchen ( FRG) .
mortar ) . Even sma l l deta i l s
are w e l l preserve d .
274 Adobe 90

ABSTRACT SYNTHETIC LATEX-SOIL SLURRY , A NEW ADOBE PRESERVAT ION TECHN IQUE

The probl ems of unnatural Richard L. Ferm


appearance and poor adhe­ The International Foundation for Earth Construction
s ion usua l ly encountered 3 2 8 2 Theresa Lane
when synthetic l atexes or Lafayette , CA 9 4 5 4 9
l atex p l asters are app l i ed U . S . A.
to adobe are overcome by
use of a l atex- s o i l s l urry
app l ied as a thin coating . Background
The key property of these
coatings is the abi l ity to A wide variety of polymers and po lymer i z ab l e monomers have been
prevent the passage of eva luated in the preservation and restoration of earthen or stone
l iquid water whi l e structures and artifacts [ 1 ] . Most of these are appl ied as pene­
a l l owing the escape of trating so lutions or sprays . Synthetic l atexes ( aqueous po lymer
vapor water from the emul s ions or dispers ions ) have been used for such work only to a
interior of the structure . l i mited extent , mainly as binders for pl asters or mortars .
Due to the high f luidity Typica l examples are Fenn ' s work at Fort Bowie [ 2 ] and Bent ' s O l d
of the s l urry , appl ication Fort [ 3 ] .
can be made a s imp l e brush­
ing operation . The author has investigated the app l ication of l atexes to
soi l s for a number of years in both l aboratory and outdoor
exposure tests that have c l early indicated the benef i ts and
KEYWORDS prob l ems presented by these materia l s . Latexes , l ike most aqueous
treatments , penetrate very s l owly and darken the surfac e ;
moreover heavy app l i cations give a n unnatural g l ossy appearance .
Adobe , l atex , s o i l - s lurry , On the other hand , l atexes are easy to work with and c l ean up ,
stabi l i zation , eros ion pro­ and some have exce l l ent mechanical properties and resistance to
tection . degradation . These properties depend on the structure of the
po lymer and a l so on the type of emu l s i fier used in making the
l atex . The emul s i f ier greatly affects the degree of wetting and
spreading of the po lymer spheres in the l atex on the soi l
parti c l es . The re l ationship between the e l ectrica l charge of
emu l sion partic les and that of minera l surfaces referred to by
C l i fton [ 4 ] has been found to be too simp l i stic to a l l ow a narrow
definition of the best emu l s i f iers . Many of the l atex emu l s i f iers
are nonionics .

The author has evaluated a l arge number of synthetic


polymer l atex commerci a l products and f ound only three or four so
far that are suitab l e for adobe preservation/restoration work .
These are Airf lex ( R ) 5 0 0 and 5 1 0 , an ethy l ene-vinyl acetate
copo lymer ; UCAR ( R ) 3 6 5 , a viny l -acry l ic copo lymer ; and pos sibly
Rhop l ex (R) E- 3 3 0 , an acry l ic res in emu l s ion , which was promising
in some initial testing . The f i e l d testing was a l l done with
Airf l ex 5 1 0 or UCAR 3 6 5 .

Latex used as a binder for p l aster does not usua l ly cause a


co l or change prob l em i f soi l identica l or very s i mi l ar to the
substrate is used . Sometimes s o i l eroded from the structure to be
treated can be used . Latex- soi l p l asters tend to have a short
useful l i fe due to adhes ion f a i l ure . This is because most
pl asters are fairly thick , 0 . 5 cm or more , and are usua l ly mixed
with l atex contents of 5 % or more based on the weight of dry soi l
used . Note that most l atexes contain 4 5 % to 5 5 % by weight of
active po lymer . The combination of thickness and high po lymer
content can make a moisture barrier that wi l l a l l ow the
accumu l ation of water at the interface between the wa l l and the
p l aster . Both additional damage to the interior and l oss of the
p l aster wi l l result in time . This prob lem has been described in
deta i l by C l i fton [ 5 ] who points out that restoration work on
adobe by indiscr iminate app l ication of preservation mater i a l s can
cause more damage than benefit .

A New Approach : Latex-So i l S l urry

In an attempt to ba l ance the benefits of po lymer-conta ining


coatings and plas ters against the prob l ems they may create , the
use of thin coats of latex- s o i l mixtures of s lurry consistency
was investigated . F i e l d tests on adobe test wa l l s did not deve l op
any of the prob l ems described above and were very effective in
preventing erosion for four years of San Francisco Bay area
weather . These coatings function by shutting out l iquid water ,
but a l l owing vapor water to escape from the interior , and thus
prevent interior condensation . Key features are the l iquidity
provided by adding suff icient water to the coating m ixture to
give a s l urry and thus a l l ow a thin appl ication , and a reduced
amount of l atex , 0 . 5 % to 3 . 0 % ( as supp l ied ) based on the dry
weight of the soi l in the mix . The cured coatings were about 1 . 5
to 3 mm thick .
Consolidation Studies 275

App
l ication Technique

The cons istency of the l atex- soi l s l urry is regu l a ted by the
water content and does not appear to be critica l . The amount o f
water t o b e used is determined by making t r i a l mixes with the
test soi l and genera l ly is about 1 7 % to 2 0 % of the weight of dry
soi l . The optimum amount of l atex wi l l probab ly be about 1 . 0 % to
1 . 5 % of the dry s o i l weight , but shou l d be confirmed by making
several sma l l sca l e app l ications with s lurries containing a range
of l atex concentrations . The test s l urries are appl ied to a
suitab l e substrate , a l l owed to cure for several days of dry
weather , and then evaluated for eros ion resi stance . Rubbing the
wet test surface with the f i nger should not resu l t in a l oss o f
soi l after moistening with a water spray . The minimum amount o f
l atex that gives satisfactory eros ion resi stance shou l d b e used .
The l atex shou l d be f i rst mixed with the total amount of water to
be used to ensure a uni form coating . An ordinary paint brush
works we l l as a s l urry app l icator . A gunite-type sprayer could
probab ly be used for a l arge pro j ect . The s lurry should be mixed
occasiona l ly , but settl ing does not appear to be a prob l e m . The
s l urry can a l so be stored in a c l osed container and remixed again
before use .

The substrate should be in sound condition prior to s l urry


app l ication . Loose mate r i a l should be removed by brushing i f not
too deep or extensive . Reinforcement of the surface prior to
s l urry appl ication can be accomp l i shed by a l ight treatment with
a penetrating resin solution . The app l ication rate must not be so
heavy that it wi l l form a water vapor barrier . Acryl oid F - 1 0 ( R ) ,
a 4 0 % solution of butyl acryl ate in an aromatic naphtha , appl ied
as a 1 0 % solution in xyl ene works we l l . Acry l oid F - 1 0 has
exce l l ent weathering properties , but darkens the soi l surface to
an extent that it would probably not give an acceptab l e appear­
ance without a latex- s o i l s l urry overcoat .

The app l ication of l atex- soi l s lurry wi l l not prevent the


wicking up of ground moi s ture into the l ower l ayers of the an
adobe structure . In some cases the inj ection of a chemical grout
or insertion of a meta l barrier pl ate at ground l eve l , as
suggested by C l i fton and Davis [ 6 ] , may be considered .

F i e l d Eva luations

A l l tests were made on wa l l s constructed of unstabi l i zed adobe


bricks and p lain soi l mortar obtained f rom the Hans Sumpf Adobe
Company of Madera , Ca l i fornia . Thi s adobe soi l is considered an
idea l s o i l f or brick making . The composition is shown in Tab l e I .
These unstab i l i z ed bricks were made in a spec i a l p l ant run during
which the usua l aspha l t emu l sion stabi l i zer was withhe l d .

The test wa l l s were constructed and coated i n the fa l l of


1 9 8 2 in the San Francisco Bay area . Four of the wa l l s were about
1 . 5 m wide and 1 m high . The f i f th wa l l was 3 m l ong and 1 m
high . Soi l s l urry composi tions ranged f rom 1 . 5 % to 3 % l atex ( a s
supp l ied ) based o n the d r y soi l weight . O n l y the Airf l ex 5 1 0 ( R )
and UCAR 3 6 5 ( R ) l atexes were used . App l ication o f the s lurries
was made with an ordinary paint brush . The coatings had a
thickness of about 1 . 5 to 3 mm . P l aster of the same composition
was trowe l ed on one of the short wa l l s and on the l ong wa l l . Some
of the wa l l areas were given a l ight spray of a 1 0 % solution of
Ac ryloid F - 1 0 ( R ) in xy l ene one ' day prior to the s lurry appp l ica­
tion .

Inspections were made at interva l s unt i l January 1 9 8 6 when


short ly thereafter the four short wa l l s were demo l i shed to make
room for another pro j ect . The f i fth l onger wa l l sti l l stands
today and continues to be used for new exper imenta l work . At the
last inspection in 1 9 8 6 a heavy rain was f a l l ing making i t
possib l e t o evaluate c o l or and surface hardness under wet
conditions . A l l app l ications , except the heavier p l aster
appl ication , were in exce l l ent condition . The thick p l aster had
serious adhesion fai lures . Otherwise no d i fferences were observed
between the two l atexes or their concentrations . A l l surfaces
when wet had the usual dark c o l or of wet soi l ; they a l l returned
to the normal s o i l color when dry . When the surfaces were pressed
with a b l unt rod , the areas sprayed with Acry loid F - 1 0 (R) prior
to s lurry app l ication had greater resistance to indentation than
the corresponding untreated areas .
Adobe 90
The longer wa l l which had been coated with a heavy latex ­
soil plaster developed serious adhesion prob l ems and damage to
the interior core . I t has s ince been repaired and is being used
to study other restoration methods .

Conclusions

Appl ication of latex- soil coatings as a fl uid s l urry is a


promising method for preserving and restoring adobe structures .
However , since a l l field testing has been conducted with only a
s ingle soil type and o utdoor exposures have been l imited to only
several years , the method should be regarded as experimental
unti l tested more extens ively . Thin appl ication and use of
minimum l atex content in order to avoid formation of a water
vapor barrier prevent wind and water erosion without damaging the
core soil structure . The slurry is easy to apply and shou l d be
lower in cost than many preservation techniques . The method
should a l so f ind appl ication for earth housing improvement in
developing countries as we l l as for the preservation of
historic a l structures .

Tab l e I

Compos i tion of Hans Sumpf Company Adobe S o i l

Particle S i z e Analys is
By wet sieve :
Grain s i z e , mm % F iner , by we ight
2 . 30 99.9
1 . 00 95
0 . 59 83
0 . 30 64
0 . 20 56
0 . 15 52
0 . 08 45
By Coul ter Counter ( R ) :
On fraction below 0 . 0 8 mm grain s i z e
Mean grain diameter : 7 . 3 pm
5 % by volume is greater than 2 7 �m
9 5 % by vo lume is greater than 2 . 0 J.L m
Clay Type
s Kao l inite and c l orite : 1 5 % - 2 0 %

Materials List
Bentonite : Nil

Airf l ex 5 1 0 ( R ) , Aqueous emul sion of ethy l ene and vinyl acetate


copolymer , 5 5 % solids , Air Products & Chemica ls , Inc . , Box 5 3 5 ,
All entown , PA 1 8 1 0 5 . Telephone : ( 8 0 0 ) 3 4 5 - 3 1 4 8 .

UCAR 3 6 5 ( R ) , Aqueous emu l s ion of v inyl -acry l ic copo lymer , 5 5 %


s o l ids , Union Carbide Corporation , 3 9 Old Ridgebury Road ,
Danbury , CT 0 6 8 1 7 - 0 0 0 1 . Telephone : ( 2 0 3 ) 7 9 4 - 6 3 0 0 .

Acry l oid F- 1 0 , Butyl methacryl ate po lymer , 4 0 % s o l ution in VMP


naphtha , Rohm and Hass Company , Independence Ma l l W . ,
Philade lphia , PA 1 9 1 0 5 . Te l ephone : ( 2 1 5 ) 5 9 2 - 3 0 0 0 .

Notes

1 . Frank J. Bockhof f and Esther Bockhoff , " Synthetic Po lymers for


I mpregnative Cons o l idation of Cul tural Artifacts , " Po lymer News
7 , No . 2 ( 1 9 8 0 ) : 5 7- 6 7 .

2 . Dennis B . Fenn , " Initiatory Report , Adobe Preservation , Fort


Bowie National His toric S i t e " ( I nterna l report , Western
Archeo l ogica l Center , 1 9 7 6 ) .

3 . Dennis B . Fenn . " I nitial Report , Mud P l a s ter Preservation


Research , Bent ' s Old Fort Nationa l Historic S i t e " ( I nterna l
report , Western Archeological Center , 1 9 7 7 ) .

4 . James R . C l i f ton , " Preservation of Historic Adobe Structures ,


a Status Report " NBS Technica l Note 9 3 4 , U . S . Department of
Commerce , 1 9 7 7 , 1 0 .

5 . Ibid . , IS, 19 , 2 3 .

6 . James R . C l ifton and Frankie Davi s , " Protecting Adobe Wa l l s


from Ground Water " NBS I R 7 9 - 1 7 3 0 , U . S . Department of Commerce ,
1979 .
Consolidation Studies 277

ABSTRACT DETER I ORAT I ON AND CONSERVATION OF SOME MUD BRI CK IN EGYPT

Mud b r ick f rom two archaelo­ F ATMA M . HELMI


g i c a l sites at Abu- S ir and C airo Univer s i t y , Facu l t y o f Archaeology ,
Mat aria in Egypt was studied Cons ervat ion Department ,
by X - ray d i f f ract io n , atomi c G iza , Egypt
absorp t ion , t h i n sect ion
ana l y s i s , and scan n ing 1. I NTRODUCTION
e l e c t ron mi croscope . X-ray
d i f f ract ion data showed t hat Mud br ick has been widely used in anc ient Egypt s i nce the
it con s i s t s of t he f o l low ing predynas t i c p er io d in Naqqad a , Upper Egypt . It was common l y used
minerals : qu�rt z , p l agio� i n construc t i n g tombs i n t he 1 st and 2nd Dynast i e s at Saqqara an d
c l ase and pot ash f e l d spars , Aby-dous . C l ay was abundant a l l ov er Egyp t , so houses were b u i l t
mica , an d variab le amount of f rom mud b r i cks wh ich were s u i t a b l e t o t he dry c l imate o f Egypt .
c l ay mineral s . Deter iora­ The s i z e o f the an c i ent Egypt ian mud br ick var i e d , some had t he
t ion phenomena are due to same d imensions as the recent ones , wh i l e ot her s were very l arge
chem i c a l we at her i ng by wat e r size . I n the Egyp t ian museum , there are two an c i ent mud�b r i ck w i t h
a n d e f f ect o f salt . Consolida­ dimen sions 9 6 . 5 x 5 3 . 3 x 3 0 . 5 cm . As stone became known , tombs
t ion of mud br ick was carried and t emp l e s were const ruct e d from it , whereas houses and phar ao n i c
out by apply ing t e t raetho_. pal � w e r e s t i l l bu i l t f rom mud b r i c k both for p o o r p e o p l e and
xys i l ane , met hy l t r imet hoxy for noble s . This is the reason that most of an c ient Egyp t i an
s i l ane , and me t hy lme t hac rY ­ house s and palaces ; ha,vel vanishedl, be cause mud br ick is l e s s durab le
lat e-but y l acry l at e copolyme r . than stone used for tombs and temp l e s [ 1 ] .
Resu l t s were examined by
scan n ing e l ect ron mi cro scop � . Mud br ick is composed of s and , s i lt , c l ay , and f ibrous organic �
mat er i a l s such as s t r aw may a l so be adde d . Sand was added to
increase comp r e s s ive strength and m i n im i z e c racking when the adobe
KEYWORDS
dried . Non - c l ay m i n e r a l s act as internal b i nder . They reduce
cont ract ion and prevent crack ing [ 2 ] .
MUD BRI CK , DETERI ORAT I ON ,
CONSOL IDAT I ON , WEATHER I NG ,
At excavat ion s i t e s in Abu - S i r an d Matar i a , it was found t hat
X-RAY D I FFRACT I ON ,
mud b r i ck s t ructure s were very f r i ab l e and exten s ive l y de t e r i o",, __
PETROGRAPHY .
r at e d . The aim o f the present work is t o study the deter iorat ion
factors o f adobe . in these two s i t e s and to f i nd out t he suitable
conso l idan t s for the i r conservat ion .

2. EXPERI MENTAL

2.1. Mud Br ick Samp l e s

Two mud b r ick samp l e s from Abu- S i r ( Ol d K l n gdom 1 s t Dynast y ) and


Mat ar ia ( Roman Pe riod ) were studie d : The samp l e s were very
f r iab l e and had pale grey co lour .

2.2. X-Ray D i f f ract ion An a l y s i s

T h e samp l e s were ground in an agate mo r t ar to a f ine powder ,


pressed in the spec imen ho lder , and then moun ted in a Ph i l ip s X-ray
d i f f r actome ter .

speed �
The operat ing con d i t io n s were

0 . 1 , char t � 5 , Range
silt : 0 . 1 .

� � Cu �
Gene rator
radiat ion ( 1 . 5 4 1 8 A O ) w i t h Ni f i l t e r , 40 Kv , 2 0 rnA cur rent tube ,
1 x 1 0 3 , t ime cons t a n t � 1 , and

2.3. Chem i cal and Atomic Ab sorp t i on Analyses

Comp l e t e chem i cal an a l y s i s of mud br ick samp l e s was carried out .


Also t he samp l e s were imme rsed in deion i z e d wat e r for 2 4 hour s ,

ions and groups ,�


and the wa shing wat er was an alysed for det ermi nat ion of 't he f o l lowing
Na , K , C a , Mg , C I , 80 4 , C0 3 .

2.4. Thin Sect ion Ana l y s i s

Mud br ick samp l e s were s e c t i o ne d , moun t e d on mi croscop ic s l ides .


D i f f erent minerals of each samp l e were ident i f i ed us ing a Le i t z
po l a r i z i n g mi croscope .

2.5. Conso l idat ion

Three cub e s 5 cm3 were cut from each mud br i ck samp l e , then treated
w i t h t e t raethoxy s i lane [ TEOS ] , t r imethylmethox y s i l ane ( MTM08 ) , and
methylmethacrylate-but y l acry l a t e ( MMAB A ) copo l ymer separ a t e l y b y
penet rat ion through cap i l l ary r is e . The f i r s t two consol idan t s
were u s e d wit hout di lut ion , b u t t he l a s t copolymer w a s d i lut e d i n
1 : 1 to luene an d xy l ene . After o n e month t he samp l e s were treat e d
once aga i n , t hen a l l owed to s t and for one mon t h before examinat ion
with scan n i ng e l e c t ron mi cro scope .
Adobe 90

2.6. Scann ing Elec tron M i croscope ( SEM ) Examinat ion

The t reated samp l e s were sput t er - coa t e d with go l d ( 1 0 nm th ickne ss)


then were examined b y SEM to comp are t he act ion o f the conso l idan"us
and whe t he r t hey f i l l e d pores , and t he shape o f the po l ymer l inks
between grains a f t e r the polyme r i zat ion proce ss .

3. RESULTS

3.1. Abu- S i r Mud Br i ck

X-ray d i f f ract ion data of t h i s samp le F i g . 1 ( a ) indicated that it


con s i s t s of oc - quart z oc 8 i 0 2 ( 5 - 0 4 9 0 ) , Anor tho c l a se ( Na , K )
A I S i 3 0 8 ( 9 - 4 7 8 ) , a l b i t e Na A I S i 3 0 8 ( 1 0 - 3 9 3 ) , and trace amoun t s of
b io t i t e K ( Fe , M g ) 3 A I S i 3 0 3 0 1 0 ( OH ) 2 ( 2 - 0 0 4 5 ) , kao l in i te AI2Si295 (OH4
and montmo r i l l o n i t e NaO . 3 ( AI , Mg ) 2 S i 4 0 1 0 ( OH ) 2 ' n H 2 0 ( 2 9 - 1 4 9 8 ) .
Thin sec t ion exam i n at ion F i g . 2 ( a ) showed quar t z cry s t a l s ( wh i t e ) ,
and large grains o f orthocla se ( s imp l e twinn ing ) i n a mat rix o f
f i ne gra i n e d s i l t and c l ay mineral s , and organ ic mat e r i al s . Most
of t he const i t u t i n g grains were angular . Chemical an a l y s i s showed
t hat i t con s i s t s o f 8 4 . 8 % ( by ."tt . ) insolub l e i n HCI ( s i l i ca and
s i l i cate minera l s ) , 8 . 2 2 % A 1 2 0 3 ' 1 . 3 0 % Fe 2 0 3 , 0 . 6 7 % Na 2 0 , 3 . 2 1 %
K 2 0 , 1 . 2 0 % CaO , and 1 . 2 5 % MgO . Atomic absorpt ion an a l y s i s o f the
samp l e washing wat e r showed t hat i t cont a in s 0 . 0 4 2 Na+ , 0 . 0 1 7 K+ ,
0 . 0 1 5 Ca 2 + , 0 . 0 0 3 Mg 2 + , where vo lumet r i c chem ical ana l y s i s
con f i rmed t h e pre sence o f 0 . 0 1 5 C I - , 0 . 0 0 4 s0 4 2 - a n d 0 . 0 3 C0 2 - .
3

SEM micrographs of t h i s samp le a f t e r t r eatme n t w i t h TEOS ,


MTMOS , and MMABA are shown in F i g . 3 ( a , b , c ) respect i v e l y . In
t he case o f TEOS t he polymer network was formed o n t h e grains and
const i t ut e l i nks w i t h i n pore s . MTMOS forms l e ss a cont inuous
layer on t he grains and a l so did not succeed i n creat i n g good
l inks between grain s . The copolymer shows a spongy f orm o f the
unhomogenous r e s i n l inks .

3.2. Mat ar i a Mud B r i c k


X-ray d i f fract ion data o f t h i s adobe showed that i t con s i s t s o f
t h e f o l lowing mineral s : -quartz GC - S i0 2 ( 5- 0 4 9 0 ) , a l b i t e
NaA I S i 3 0 8 ( 1 0 - 3 9 3 ) , anortho c l ase ( Na , K ) A I S i 3 0 8 ' ( 9 - 4 7 8 ) , a n d
t race amount s o f b iot i t e K ( Fe , Mg ) 3 A I S i 3 0 1 0 ( OH ) 2 ( 2 - 0 0 4 5 ) , and
montmor i l l o n i t e Na n ( A I , Mg ) 2 S i 4 0 1 0 ( OH ) 2 ' n H 2 0 ( 1 2 - 2 1 9 ) . Thin
se ct ion anal y s i s of t he samp le demonstrated quar t z grai n s , some
sma l l and others large , p l agiocl ase f e l dspar ( Lame l l ar twinn ing )
in a mat r ix of f ine grained s i l t and c l ay miner a l s . Chem i ca l and
atom ic ab sorp t ion analyses of t h i s samp l e showed that it con t a i n s
7 8 . 0 % ( by wt . ) insolub l e in HCI ( S i l i ca and s i l icate minera l s ) ,
9 . 5 8 % A 1 2 0 3 , 3 . 5 0 % Fe 2 0 3 , 0 . 6 8 % Na 2 0 , 5 . 0 4 % K 20 , 1 . 4 0 % CaO , and
2 . 4 8 % MgO . I t s washing wat er contain s 0 . 0 3 8 Na+ , 0 . 0 1 7 K + , 0 . 0 0 8
Ca+ , 0 . 0 0 2 Mg 2 + , 0 . 0 1 2 C I - , 0 . 0 0 3 s04 2 - , and 0 . 0 3 C0 3 2 - .

MTMOS , and MMABA are given in Fig . 3


SEM micrographs of t he samp le after tre atment with TEOS ,
( d , e , f ) respect ive l y .
i s clear t hat TEOS was p r e c ip itated a s nodules and penetrated
t hrough pores and around grai n s . MTMOS Fig . 3 e showed that
It

l inks were formed a l so be tween grains but there are st i l l large


areas where l i t t l e p r e c ip i tat ion o f the polymer occure d . MMABA
d i d not succeed in forming ne twork l inks of the po l ymer .

4. D I SCU S S I ON

The prope r t i e s of mud br i ck and i t s durab i l i t y to we athe r ing


depend to a great extent on the ir con s t i tuent s and the int eractions
be tween them . A l s o , on the local envi ronmen t a l con d i t i ons . The
amount of sand wh i ch i s most l y the maj or component , s i l t o r c l ay
minerals wh ich act as a b inder , and the exist ence of organ i c
mat t er , l ime s t one , o r f ir e d br ick , a l l p l ay a n impo r t an t role in
the deter iorat ion proce s s .

I n the present work , re sul t s showed that anc i ent Egypt ian mud
b r i c k samp l e s con s i st e s s en t i a l l y of quar t z , p l agioc lase and
pot ash f e l dspar s i n a loose packing mat r ix o f s i l t and c l ay
mineral s , and or gan i c mater i al s . The author t h i nks that t he main
internal caus e s o f deter ior at ion of the examined mud b r i c k were
the loose packing and t he angular i l l - sorted cons t it ut i n g grains
o f d i f f erent s i zes . The cement mat e r i a l was not d i s t r ibut e d
regular l y be tween quar t z grains . There a r e l arge a r e a s o f f ine
s i l t and c l ay part i c l e s , where other areas have concent r at ion o f
l o o se angular quar t z grains a s cou l d b e seen from thin sect ion
analy s i s and SEM m i crographs ( F igs . 2 , 3 ) . . A l so , the percentage
o f s i l t and c l ay i s much more than in the ideal adobe stated by
ConsoI'dation Studies
1 27 9

Q
(a)

����j�lM�;����'���:J
An
)'tl"). N."/ywI.K BVM..
68 I 160 512 , 44) , 40, 3I6 , I 28
I I
20 i'2 4
Q
(b)

60. 52 44 36 28 4
Fig . 1. .
X-Ray. d i f f r ac t l' 0 n patt ern s ° f Abu-S lr ( . ) , . n d M.t .r i.
( b ) �ud b r i ck . Q a r t z ( Q ) , A l b i t e ( A ) " Anorthoc '.se ( An ) ,
':l
b iot l t e ( B ) , Kao l ln i t e ( K ) , an d Montmo r l l lo n i t e ( M ) .
280 Adobe 90

Cl ifton [ 3 , 4 ] . The adobe based on d i f f e re n t c l ay mineral s may


react to increasing wat e r co n t en t . Montmor i l lo n i t e adobe w i l l be
much more respon s ib l e to comp r e s s ive deformat ion than a kao l i n i t e
o n e f o r t he same amount of ab sorbed water [ 5 ] .

Water is a ser ious factor for adobe deter iorat ion . Absorp­
t i on of water cause s swe l l ing of c l ay m i n e r a l s and evaporat ion
g ive rise to shr inkage , cracking , and breaking . The locat io n of
Abu- S i r excavat ion i s n.ear cult ivat e d l an d ; t he adobe s t ructure
wal l s were found immersed in water due to the h i gh wat er t ab l e
level in thi s are a . So , e f f ect of water on adobe i s eros ion of
sur f ace along c r acks and f i s sure s , leaching of c l ay and s i l t
ma t r ix , a n d di ssolut ion o f solub l e salt [ 6 , 7 ] . Migrat ion o f
solub l e s a l t occurred towards t h e surface by evaporat ion , an d
recry s t a l l izat ion took p l ace at the surface l e d to adobe
det e r i orat ion .

Temperature f l uctuation betwedn day and n i ght in both excava­


t ion s i t e s is another p o s s i b l e f actor for phys ical weathermng ,of
adobe i n Egyp t . Swe l l ing by hum i d i t y i n the early morn ing , and
shri nkage at mi d-day cycles l i kew ise , may b e imp l i cated . Al so ,
t hermal s t re s s and unequal expan s i on of the d i f f erent con s t t t ue n t s
t ake s p l ace g i v i n g r i s e to adobe d i s int egrat ion .

W i n d borne sand and deur itus mat e r i a l s mechan i c a l l y a t t acked


adobe in both areas cau s i ng it s abr as ion . R i e derer [ 8 ] , has noted
t he intense mechan ical a t t ack of w i n d on s t one in t rop i ca l
count r i e s .

The app l icat ion of consol idan t s is very important for adobe
conservat ion , wh i c h t ransform in s i t u into a po lymer confer adobe
new proper t i e s of durab i l i t y and weather r e s i s t ance . I n t he
present work it is fouQd from t he obt ained dat a t hat ,TEOS is t he
most suitable con sol idant for t he s t udied Egyp t i an adobe . I t has
low v i scos ity and penetrat e s we l l t hrough adobe st ructure .
Hydro l y s i s t ake s p l ace by mo i s t ur e in a i r and in adobe i t s e l f
forms a network S i -O polymer i n i t s s t ruct ure a n d ethanol i s
evaporat e d . T h i s agrees w i t h t he accept a n ce of F i e l den [ 9 ] ,
Gamarra [ 1 0 ] , and Lewin [ 1 1 ] t hat TEOS w i t h or without MTMOS and
MEOS resp e ct i v e l y as a sur face t reatmen t f o r adobe . The author ' s
current r e search i s the app l icat ion of MTMOS as water rep e l lent
mat e r i a l after treatment of adobe with TEOS . . MMABA i s excluded
because i t may cause cracking b y t ime and e f f e c t of U . V . radiat ion
i n s it u , t h i s i n addit i on t o the weak f ormed l i nks in adobe
structur e .

5. CONCLUS I ONS

Deter lorat ion o f mud b r i ck in Egypt at Abu- S i r and Mat ar ia


excavat ions is due to the usua l causes ( gr ound wa ter , t hermal
cy c l ing , wind abras ion ) but is acce lerated by the adobe t exture .
I l l -sorted an d loose packing of t he d i f f erent s i ze cons t ituent
grains . TEOS was found to be t he b e s t conso l i dant for Egypt ian
adobe st ructure .

6• REFERENCES

1.

2.
A . Lucas , Ancient E gyp
t i an Mat e r i a l s and I ndus t r ies ,
Edward Arnold Lt d . , London , 1 9 4 0 ) , 8 7 - 9 0 .
( 3 r d ed . ,

G . Torraca , Porous Bui l d i ng Mater i a l s -Mat e r i a l s sc ience for


Arch i t e ctural Con servat ion , ( 2 nd e d . , I CCROM , 1 9 8 2 , 9 5 - 1 0 4 ) .

3. J . R . C l i f t on , P . W . Brown , and C . R . Robb i n s , Methods for


charact er i z ing Adobe b u i l d ing Mat er i a l s , Report No . NBS TN- 9 7 7 o f
t he Nat ional Bureau o� S t andards , Washington D . C . , ( 1 9 7 8 ) .

4. J . R . C l i fton , Preservation of H i s t o r i c Adobe Structures , A


St atus Repor t , Report No . NBS TN- 9 3 4 of t he Nat ional Bureau of
S t andards Washingt o n , D . C . ( 1 9 7 7 ) .

5. T . W . Lambe , and R . V . Whitman , S o i l Mechan i cs , ( S I Ver s ion , J .


W i l ey , New Yor k , 1 9 7 9 ) , 3 5 .

6. G . Torraca , " A n I nt er n at i onal Pro j e ct for t h e S t udy o f Mud


b r i c k Pre servat ion" , in Cons e r vat i on of Stone and Wooden Obje c t s ,
( I I C , New York Conf erence , 1 9 7 0 ) , 4 7 - 5 7 .

7. C . R . St een , " S ome Recent Exper ime n t s i n S t ab i l i z i ng Adobe and


Stone " , I b i d . , 5 9 - 6 4 .
Consolidation Studies 281

8. J.
R i ederer , " ' Protect ion f rom Weat hering of Bu i l d i n g Stone in
Tropi cal Coun tr i e s " , ( I I C Prep r in t s of the Bologna Congr e s s
Conservat ion o f Stone and W a l l Paint ings , 1 9 8 6 ) , 1 5 1 .

9. B . M . Fei l den , Conservat i o n o f Histor i c Bui l d i ngs ( Butterwor t h ,


London , 1 9 8 2 ) , 7 5 .

10.enen
EJ:
R . M . Gamar ra , " Con servat ion of S t ructures and Adobe Decorative
ts in Chan Chan " , in Adobe I nt ernat ional Sympo s ium and
rraining Wo rkshop on the Con servat ion o f Adobe , L ima Cusco ( Peru ) ,
(1 9 8 3 ) , 6 9 .

11. S . Z . Lewi n , and P . M . S chwartzbaum , " I nve s t i gat ion o f t h e Lo ng­


Term Effe ct iven ess o f an Ethyl S i l icate Based Conso l i dant on Mud
br i ck" , in I b i d , 7 7 .

Fig . 2. Thin sect ion photographs of Abu- S i r ( a ) , and Matar i a ( b )


mud-br ick showed quart z , potash and p l ag i o c l a s e fe l dspars
i n a mat r ix of s i l t and c l ay miner a l s . x n i col s , 6 . 3 x .
Adobe 90

b e

c f

Fig . 3. S EM mic rographs of Abu- S i r mud b r i c k ( a , b , c ) and


Mat a r i a mud br i ck ( d , e , f ) , aft er t r e atment wi t h
TEOS , M'FMOS , and MMABA r e spect ively , 4 0 0 X .
Consolidation Studies

ABSlRA CT THE WEATHERING CH ARA CTERISTICS OF THE ROCKS OF THE KE Z IER GROTTOES
AND RESEARCH INTO THE IR CONSERVATION

THE KEZIER GROTIOES, in the


Xinjiang Autonomous Region of
Huang Ke zhong, Jiang Huaiying, Cai Run , Feng L i j u an
Scient i f i c and Techn i cal Inst itute for Conse rvat ion of Cultural
Re l ics
China, were carved in mudstones
People ' s Republ i c o f China

and sandstones. These natural


sediments are poorjly consolidated
and behave like earth, disintegrating
Introduction
The Ke z ie r Grot toes ( F ig . l ) are located in the Xinj iang Autonomous
rapidly when immersed in water. Reg ion of north,west e rn China . They were carved in Tert iary
mudstones and sandstones which are so poorly con s o l idat ed and
Extensive research tests are reported cemented that they d i s integrate immediately upon b e ing wetted . In
on the use of potassium silicate of t h i s respect they are s imilar to man-made earthen mat e r i a l s .

high molar ratio Si02 : K20

(typically 3.8) together with


magnesium fluorosilicate and, in
some instances, silanes such as
methyltriethoxysilane as
consolidants for the sediments.
These yielded water-resistant
products. ·Freeze-thaw, salt­
resistance, accelerated aging, and
water-absorption tests are reported.
In field tests, walls were wrapped in
gauze, wetted with glycerine and
covered with plastic to prevent rapid
drying and the formation of K2C03 .

KE YW
ORDS
F ig . l Ke z ier Grottoe s ; view near Cave 1 8 on the western s ide .
Cave art, China, Consolidation of
sediments, Grottoes, Magnesium In 1 9 8 6 t he grottoes were surveyed, and it was noted that
fluorosilicate, Potassium silicate rap id det e r iorat ion was occurring desp ite the ext reme ly sparse
rainfall in the region . What l i t t le p recipitat ion occurs is
focus sed into run-off chann e l s that d i s s e ct surfaces and
per iodica l ly cause col lapse o f large section s . Within the
grottoes alkal ine salts also cau se f laking o f the wall paintings .

In order to understand the characte r i s t i c s of d i s integrat ion


and deve lop strategie s for consol ida t i on and protect ion o f the
grot t oe s , a re search program was undertaken . Thi s addr e s s ed the
mineralogy, chemical composit ion , soluble s a lt s , and extens ive
laboratory and in-situ t e s t s o f consol idant s . Acce lerated aging
t e s t s are present ly underway .

Rockthological
Li
Characteristics and Their
characteristics : Physical and Mechanical Properties
The geological strata of the grottoes region is t he upper unit o f
t h e Tert iary p e r iod . The sediments are lacu s t r ine in origin and
are comp r i s ed of mudston e , sandston e , as we l l as l e s s e r amount s of
coarse sandstone , and grave l mudstone . The sandstone is grey t o
greyi sh-green, and it s mineral component s a r e mainly quart z ,
f e l dspar, a s we l l a s a l e s s e r amount o f a b lack mineral .
Typ ically the t hickne s s of s ingle bed was 3 - 4 m . The cementing
agent s are calcareous or calcareous-mud, and the binding
propert i e s are poor, as shown by rapid dis integrat ion o f the rock
in rainwater .

The rock wal l s in the grot toes have f i ve to s ix layers o f


mudstone , t h e t hickne s s o f each be ing about 2 m . They are brown
or gray i s h brown in color and contain solub l e s a l t s with loca l � y
f ine gypsum layers , which weather eas i ly .

Microscopy:
Thi s work was carried out by the Geological Institute o f the
Academy of Sciences of China . Comprehens ive chemical analys i s , X­
ray d i f f raction ana l y s i s , di f ferent i a l thermal ana lys i s and
scann ing e lect ron microscope examinat ion has been done . The
Adobe 90

sandstone was poorly cemented and cont ained a great de a l o f


carbonate ( 3 6 . 5 % o r s o ) , and other s o luble s a l t s such a s NaC I ,
MgC I 2 , and C a S04 0 2H 2 0 . Aft e r being fu l l y submerged i n water the
rock lost i t s st rength and comp l e t e l y di saggregat e d ( F i g s . 2 , 3 ) .
The mudstone contained 5 - 6 % montmo r i l lonit e .

Fig . 2 P h o t o graph showing s amp l e F ig . 3 Photograph showing the


J f the rock imme d i a t e l y a ft e r s ame s amp l e ( fi g . 2 ) 1 0 mi nut e s
b e i n g p l aced i n wat e r . late r .

Table I Physical properties:

Sandstone:
P hys cal properties and Mecha ical strength :
i n

The phy s i c a l propert i e s are presented in tab le I . In view of the


bulk specific gravity = 1.86g/ cm3 fact that the sediments were s o poorly compacted and frag i le , a
po int loading inst rument was used for t e s t s of mechan i c a l
specific gravity = 2.72 st rength . T h e advantage o f this method is that it avo ids the
porosity = 20.9% problem of i rregular samp l e s , wh i le a l lowing for the examination
pore volume = 0.113 cm3/g o f seriously weathered sample surface s . Typ i c a l averaged va lues
mean pore radius = 27.7 1 0-5 cm
x o f mechan ica l propert i e s a re pre sented in table I I .
water content

Mudstone:
= 1 .55 - 7.3%
Consolida Testing
nt

Select of mate
ion r ials :
bulk specific gravity = 2.01 g/ cm3
specific gravity = 2.74 Chemical con s o l i dation shou ld improve the physical and chemi cal
void ratio = 0.267 propert i e s o f the rock and their water-re s i s t ance , thereby
water content = 2.74% stopping t he i r det eri oration due to rain . The protect ive mea sures
we cons idered u s ing were t reat ing the rock with weather-re s i st ing
chemica l s .
Table II Mechanical properties:
According to the propert i e s of the rocks and c l imate of the
Ke z ie r Grott o e s , we decided first that the prerequ i s i t e s for the
Semi-weathered rock: se l e ct ion of consol idants were improved wat e r-res i stance and
St = 0.86 kg/ cm2 compress ive st rength . I n view o f the fact that the rock i s poorly
Sc = 18.23 kg/ cm2 compact e d , the con s o l idants shou ld have good penet rat i o n , show
res i s t ance to freez ing and weathe r ing . Because the chemical
mat e r i a l s were not used directly on the mu ra l paintings within the
Weathered sandstone:
grottoe s , a l it t le di fference in color was t o lerab l e .
St = 0.66 kg/ cm2
Sc = 13.93 kg/ cm2 From the f i rst , we intended to u s e pot a s s ium s i l i cate for
t reat ing the rocks and invited our col leagues o f the
Administ rat ive Unit of Bingl ingshi Grott oes , Gan su Province t o
Semi-weathered mudstone:
j o i n i n the t e s t s . Two pre l iminary i n - s i t u test s e r i e s , and many
St = 23.58 kg/ cm2
laborat ory test s , we re carried out i n October 1 9 8 6 and May 1 9 8 7 .
Sc = 497.8 kg/ cm2 St rength increased bec�use of the use of pot a s s ium s i l icat e , but
init i a l l y depth of penet ration and wat e r-res i stance we re not
where St = tensil strength, s u f f i cient . Thus pot a s s ium s i l icate was not used for large - s c a l e
t reatment a t that t ime . I n orde r t o furthe r t h e pro j e ct , o u r
and Sc = compressive strength i n s t itute t h e n resea rched a n d deve loped seve ral new chemi cal
weathe r-res ist ing mat e r i a l s based on part icular grade s o f
pot a s s ium s i l icate t ogether with other addit ive s .

Test principles :
The rock of the Ke z i e r Grott o e s con s i s t s of quart z , f e l dspar and
carbonat e s ; it is fragi l e , flakes on rubbing, and dis int egrat e s at
once i n wat e r . Because the main component of the rock is s i l icon
dioxide , a chemical con s o l i dant which can form an inert s i l i c on
dioxide or s imi l a r compound was s e l ected . Potass ium s i l icate i s
weathe r - r e s i s tant but i s s t rongly alka l ine ( pH 1 4 ) , t h u s acidic
magnes ium fluor o s i l icate wa s added t o neutra l i z e the alkali and
Consolidation Studies

generate inert s i l i co n diox ide and other compounds that wou ld


con s o l idate the rock and l eave an inert sub strat e :

where n is t he mo l a r rat i o o f S i 0 2

In addit i o n , organic s i l i con compounds such as me thylt r imethoxy­


s i l oxane o r met hylt r i ethoxy s i loxane can form high-mo lecular
s i l i cone under the action of a weak ba s i c catalyt ic agent :

-O-S i-O-

CH 3
I CH 3 0

I I
R 0 Si - O

I
OH -
R +H 2 0 -� II Si- O - Si - 0

0
CH 3
I
+ R 0 H

OR
n II -O- S i -O -
n

where R i s ethyl ( C 2 HS ) o r methyl ( CH 3 )

This react i on i s fast . In order to meet pract i cal usage , it i s


necessary t o regulate t h e amount o f water a n d hardening agent a s
we l l a s increase the amount o f a l coho l i c s o l vent t o s l ow the
react i on . The alcoho l serves a l s o to s low t he increase in
v i s c o s i t y as the po l ymeri zat ion proceeds and thus fac i l itates
penet rat ion int o the rock .

Tests to dete ine opt mula


rm the imum for

A great deal of work on the select i on of the best fo rmu las


:

suitable for consol idat ion o f the Ke z i e r Grottoes was n e ce s s a ry .


Because o f s e r i ou s we athe ring o f the rock , s amp l e s which cou l d be
used for t e s t s were l imited and it was impos s ible to obtain many
samp l e s for t e s t s . P o t a s s ium s i l i cate of mo lar ratio of S i02 : K20
over 3 . 4 has been u s ed abroad; in our count ry pot a s s ium s i l icate
of highest rat i o 4 . 1 in l iquid st ate can be produced . In brief,
t h e formu l a developed was based on a n aqueous so lut ion o f
pot a s s ium s i l icate -- whose value was " f rom 3 . 4 t o 4 . 1 b u t ma i n l y
3 . 8 , diluted four- f o l d w i t h wat e r - - and was a l s o reacted with
magne s ium f louoro s i l icate .

Add ing ma gnesium fluoro s i l icate to pot a s s ium s i l i cate


solution precipitat e s cat k in - l ike s i l i con dioxide immediat e l y .
The test called for brushing d i l uted potass ium s i l icate s o lut ion
ont o the opp o s i t e two s ide s o f the sample unt i l it was permeated
complet e l y . Aft e r seven days o f drying it was brushed with 3 %
magne s ium fluor o s i l icate s o lut ion unt i l comp l e t e l y permeated .
Aft e r a further seven days of drying it was brushed with 5 %
magne s ium f lu or o s i l i c at e s o lut io n .

Samp l e s for t e s t ing depth of penetrat ion we re brushed on one


s ide. only unt i l permeated comp l et e l y . The s i z e of the s amp l e was
5 x 5 x 5 cm . The ma in advantage of organ i c s i l i cones such as
methyl t r iethoxy s i l ioxane was its supe r i o r wat e r-r e s i stance .
Howeve r , the h i gh cost of this mat e r i a l prevented i t s u s e in the
Ke z i e r Grottoe s . In these experiments a group of t e s t s o f
comp o s i t e mat e r i a l s o f inorgan ic mat e r i a l s a n d t h e orga n i c
s i l i c ones w a s invo lve d .

In the course o f t
h e experiments i t wa s found that i f o n l y
s imple brushing w a s done , without a n y wrapp ing o f t h e s amp l es ,
then poor penet rat i on occurred on the second appl icat ion , and the
sur face was suffused with white . The reason is t hat a f t e r
brushing w i t h potass ium s i l icate s o lut ion t h e surfaces d r i e d too
qu i ckly and crys t a l s o f pot a s s ium s i l icate aggregated o n the
sur face . This converted into s i l icon dioxide and pot a s s ium
carbonate after contact with the a i r . A thin impermeab le layer
often formed on the sur face , prevent ing the second and t h i rd
brushings from penetrat ing . Howeve r , if after brushing the
samp l e s were wrapped with gau z e wetted with glycerine , the
mo i s ture evaporated s lowly and the pot a s s ium s i l i cate rema ined in
the inner part o f the samp l e s , such that wh ite pot a s s ium carbonate
did not form on the surface . Subsequent brushing with 3 %
magnesium f luoros i l icate s o lut ion , o r 5 % for third t ime , a l lowed
easy penet rat ion . After brushing with orga n i c s i lanes the samples
286

a l s o were covered with gau ze wetted with glyce r in e , which reduced


the volat i l i z at i o n and prevented organic s i l icone resin from
forming on the surface and creat ing a g l o s s y , re f lect ive laye r .

Judging from indoor t e s t s , pot a s s ium s i l i cate of low mo lar


rat i o had a low st rength and uneven penetrat ion wh i l e the samples
con s o l idated with pot a s s ium s i l i cate o f rat i o 3 . 8 had a high
st rength and even penetrat i on of about 5 cm . For samp l e s treated
with pot a s s i um s i l icate of rat i o above 3 . 8 the st rength did not
increase .

Tests in-situ:
area o f about I
We s e l e cted seven rock . wa l l s of Ke z i e r Grottoes ( F ig . 4 ) , each with
m2 • Aft er physical tests in-s itu t o determine t he
rebound s t rengt h , samp l e s were sent to Be i j i ng for test ing depth
of penet rat i on and water-res i stance . Ana lys i s of test resu lt s
showed that the pure inorganic mat e r i a l s gave deep penetration and
high st rengt h , but water-re s i stance was poor . On the other hand,
organ i c s i l i cones gave good water-re s i stance , but the st rength was
poor and the penetrat ion was uneven . We conc luded that us ing
compo s i t e inorgan ic and organ ic mat e r i a l s for con s o l idat ing the
rocks o f Ke z ier wou l d , ide a l l y , provide the best protect ion .

Fig . 4 S i t e test ing .

Systematic Tests:
In order to ve r i fy the test data further, we s e l e cted t wo of the
better formu l a s and divided the tesf s amples into two group s .
Group A was the comp o s i t e mat e r i a l and Group B the inorgan ic
mat eria l .

Formu l a A : pot a s s ium s i l icate o f mo lar rat i o 3.8


3 % magne s ium f luoro s i l icate
5 % magnes ium f luoro s i l icate
met hyltr i ethoxys i l oxane 100
ethanol 70
hardening agent SCP 0.8

Formula B : pot a s s ium s i l icate o f mo lar rat i o 3.8


3 % magnes ium fluoro s i l icate
5 % magne s ium fluoro s i l icate

Examination
Prepa samp of
ration of
Consolidation Materials
les :

In order t o make the data from l aborat ory t e s t s and those from the
in-s itu t e s t s comparab le , we made it a ru l e that the preparat ion
o f the samp l e s i n the laboratory must be ident ica l with the
con s o l idat ion t reatment in-situ . Thus , in the proce s s of
con s o l i dat ion of the samp l e s only spraying o r brushing was u s e d ;
subme rging o f the samples w a s not a l lowed, a n d that guarant eed the
re l i ab i l i t y of test data .

Den sity and porosity :

Mercury i n j ection poros imetry was used to measure den s it y and


porosit y . It was found that the dens ity of Group A was somewhat
Consolidation Studies

higher than that of Group B : Both A and B were sprayed, and both
A and B had much higher den s it y than the o r ig i n a l rock s .

The c a l cu l at i on of the porosity wa s by conversion between


den s it y and pore volume . The r e s u l t s showed that the poro s it y o f
group A s amp l e s w a s much l ower than that of group B a n d o r i g i n a l
rocks .

Pore volume and mean pore radius:


The pore vo l ume and the mean pore radi u s we re mea sured for group A .
and group B . The pore vo lume of group A wa s 0 . 0 1 6 6 4 cm3 / g , and was l
lower than that of 8 5 % of the o r iginal rocks whos e value wa s
0 . 1 1 2 6 7 cm3 / g . At the same t ime it obvi o u s l y wa s lower t han that
o f group B . The mean pore radius of group A wa s higher than that
o f the o r iginal rocks but lower than that o f group B .

Strength:
Two methods t o measure st rength i n the laboratory and i n - s itu were
used . In the l aborat ory , samp l e s ( 5 x 5 x 5 cm) were sprayed on
opp o s ite s ide s . Aft e r drying, the sprayed s ide s were vert i ca l l y
mea sured by the method of point loading . The t en s i l e s t rength o f
S t = 2 . 2 6 kg/cm2 a n d compre s s ive st rength o f S c = 4 7 . 5 1 kg / cm2 were
3 . 4 t imes great e r than those of the weathered rocks and 2 . 6 t ime s
greater t han those of semi -weathered rock s . They were a l s o h i gher
than those of group B .

With regard t o mea s u rement i n - s itu , because the concent rat ion
o f consol idant de creased with depth into the samp l e , the use of
the po int loading method for mea s urement was impo s s ib l e , and we
used the rebound method to measure the surface st rength of rock .
The surface s t r ength of the original rock wa s t o o low to calcu late
i t s real strength; thus o n l y the change i n rebound va lue before
and after t reatment is used t o des cribe the i r change o f s t rength .

Al l the rebound va lues of the o r iginal rocks were l e s s than


1 0 , and indent ations occurred on the rock surfaces after the
me a s u rement . The mean rebound value o f Group A was about 1 5 ,
whi le that o f Group B was about 1 3 . O n l y a t race o f indentation
occu rred on the t reated rock surfaces after the rebound
mea surement .

Tests for dept of penetration:


h

Mea s u rement of penet rat i on was based on. the d i s integrat i on o f the
o r i g in a l rocks after subme r s ion i n wat e r . In laborat ory t e s t ing
o n l y one s i de of the samp l e o f the 5 cm cube was sprayed wh i c h ,
after drying, w a s fu l ly submerged in water . The uncon s o l idated
part o f the samp l e d i s integrat e d , and t h i s method showed that the
depth o f penet rat ion o f consol idat ion was more than 5 cm . Tests
f o r penet rat i on depth in-situ with t h e same b a s i c method proved
that the depth was 4 - 5 cm .

Tests wate sta


for r-resi nce :

T e s t samp l e s f o r wat e r-re s i s t ance were divided in two . One was


submerged for a l ong t ime . The samp l e s , 5 cm cube were sprayed on
a l l s ide s and f u l l y permeated and cured, then submerged in wat e r .
From 6th Augu st 1 9 8 7 to the pres ent , none have di s integrat e d ,
whe reas t h e o r iginal r o c k s dis integrated in l e s s than 1 0 minute s .
The other t e s t was t o dry and wet the samp l e s a lt e rnat e l y by
submers ion in wat e r for 24 hours and then by plac ing them in an
oven at 6 0 ' C for 72 hours to dry . Some 1 6 cycles have been
completed so far without g i s integrat ion .

Water absorption tests:


The s e determined the amount of wat er moving through the samp l e
surface p e r u n i t area with t ime . The purpose wa s t o compare the
change o f cap i l larity and the wat er-re s i stance o f surfaces o f the
rock before and after spraying . The s amp l e s were sprayed on one
s ide then prepared into rect angu lar c o l umns o f 2 x3 x 5 cm, dried and
weighed and then the sprayed surface of 2x3 cm wa s brought into
contact with wat e r . Lat e r it wa s weighed and the amount o f wat e r
moving through t h e sample surface per u n i t a rea was ca l c u l at e d . A
p lot of water uptake with t ime wa s constructed . Both l aboratory &
i n - s i t u r e s u l t s were determined . For group A the wate r -absorbing
capacity was 0 . 0 0 8 g / cm2 in 2 hours i n - s i t u , and 0 . 3 4 g / cm2
indo o r s , and the capi l lary r i s e of wat e r wa s 2 cm in 2 hours . The
wat er-abs o rbing capac it y o f the original rocks could reach 1 . 1 5
g / cm2 , and cap i l l ary he ight could reach 5 cm . in 1 0 minut e s . For
group B the wat er-ab s o rbing capa city was 0 . 8 9 g / cm2 in 9 0 minut e s ,
and cap i l l a ry he ight of 5 cm in 3 5 minut e s . It was thu s c l e a r
t h a t t h e wat e r - r e s i st ance o f qroup A has qreat l v increa sed .
288

Freeze-thaw tests :

The 5 cm cube s amp l e s were fu l l y sprayed, oven-dr ied and we ighed,

±
then submerged in a vacuum chamber
f re e z e r at - 2 0 ' C
for 1 hour , and p l a ced in a
2 ' C for 4 hours be fore being t aken out and
thawed at room temperature . This proce s s comp leted a cycle . F rom
the r e s u l t s of n ine cycle s , the freez ing-re s i s t ance of group A wa s
bet t e r than that of group B , as we l l as the origina l rock .

Tests resista
for salt nce:

The prepa rat i on of the samp l e s was the same as for the freeze -thaw
tests . Samp l e s were placed in s at urated sodium s u l fate s o lut ion
for 20 h ours , then dried at 6 0 ' C for 7 2 hours . After f i ve cycles
the s t ab i l it y of the samp l e s o f group A wa s clearly better than
that o f group B .

Tests for aging :

The protect ive mat e r i a l s were sprayed on thin samp l e s 4 x 5 0 x l O O cm,


and then put into aging apparatus of WE- SUN t ype , the t emperature
was cont r o l led at 4 0 ' C , and sprayed with wat e r every 3 hou r s . The
apparatus has run for 1 0 0 0 hours so fa r . Cracks occurred on the
samp l e s because of their t hinne s s , and part of the samp l e s o f
group A weakened a t t h e corners ; most o f the samp l e s o f group B
l o o sened and di s integrated . T e s t s are cont inu ing .

Morphological observation:
The rock samp l e s before and after t reat ing were observed with the
s canning e l ectron micros cope . It t u rned out that the compo s i t e
mat e r i a l s e ither had s e a l e d t h e sample surfaces o f group A, o r had
f i l led the pore s , but the pores had not been sealed comple t e l y .
F l a k e - l i k e inorgan ic mat e r i a l s a l s o occurred on the rock s u r face
( F i g . 5 ) and t o a depth o f 3 cm ; f l ake - l ike o r cat k i n - l ike
inorgan ic mat e r i a l a l s o covered the rock s u r faces o f samp l e s of
group B, but pores were f i l led in to a l e s s e r ext ent , and s ome
mineral grains were not covered with inorgan ic con s o l idant . Under
the pola r i z ing micros cope we obs e rved the compo s i t e mat e r i a l s
part ly f i l led i n t h e pores t o a depth o f 1 c m i n group A . O n the
sur faces of the samp l e s the compos ite mat e r i a l s covered the
min e r a l s thicker than 1 cm . T e s t s in s it u 2 were ob ser ved under
crossed p o l a r i z ed l i ght , i t could be s e en that the pore s on the
roc ks ' s u r faces increased, and many minera l s could be r e s o lved .
The compo s it e mat e r i a l s showed sma l l granu lar st ructure , and the
inorgan ic mat e r i a l s showed cat k i n - l i k e s i l i c on di oxide .

F ig . 5 Scanning e lect r on micrograph of the coating mat e r i a l


cove r ing t h e minera l s i n t he rock sample .

Conclusion
Pract i cable methods o f protect ing the s it e against e r o s ion by
wat e r we re deve loped . The se methods u s e d spray ing , or brushing o f
pot a s s ium s i l icate s o lut ions o f high molar rat i o o f S i02 : K20
( t ypica l l y 3 . 8 ) , with othe r inorganic a�di t i ve s , on the surfaces
o f the rocks . In s ome t e s t s the s e procedure s were comb ined with
a l k y l a lkoxys i l a ne s , but the high cost o f these mat e r i a l s prevented
extens ive f i e ld u s e . The problems of K2 C03 e f f l ores cence we re
large l y e l iminated through u s e of the approp r i at e rat i o o f
s i l icate and by cove ring t h e sur faces with glycerine- impregnated
gau ze t o retard drying .
Consolidation Studies

ABSTRACT PRESERVING THE E I GHTH CENTURY B . C . MUD BRICK ARCHI TECTURE AT


GORD ION , TURKEY : APPROACHES TO CONSERVAT ION

Conservation of the e i ghth S tephen P . Koob * , Mark H . Roge r s , G . Kenneth S ams


century B . C . mud brick Freer G a l l e ry of Art and Arthur M. Sackler G a l l e ry
archite c ture a t Gordion , Smith sonian I n s t i tution
Turkey , has been long Wash ington , D . C . , 2 0 5 6 0
overdue . P a r t i a l burning U . S .A.
he lped pres erve some o f
the mud brick , b u t l i tt l e I ntroduc tion
h a s b e e n done t o protect
it s i nce excavation ove r There are many individual factors that contr ibute to the decay o f
thirty years ago . Two excavation s i te s , but . by far the mo s t des truc tive and frus trating
methods of intervention i s the combination of neg l e c t and s low decompo s i t ion through t ime .
have been t e s ted : conso­ I t is inexcusable that the great r e l i c s o f our p a s t d i s s o lve away
l idation with acrylic unattended whi l e new trenche s are dug nearby . The primary concern
r e s i n s and protection at this t ime should not be the d i s c overy o f new c iv i l i z ati ons , but
with a coat ing of mud the preservation of those a l ready uncovered .
plaster . Both methods
are viewed· a s temporary The ancient Phrygian c i ty of Gordion s e rves as a perfect
treatments that can be example o f a s i te where such action i s neces sary . Wh i l e this
replaced when a more paper wi l l concentrate only on the needs o f the s i te ' s mud brick
permanent protective building s , Megaron 1 and Megaron 4 , it i s by no means intended
treatment is found . to obviate the attent ion needed by Gordion ' s other monument s .

KEYWORDS A B r i e f S i te H i s tory and D e s c r ipt ion

Conservation , architecture , Megaron 1 and Megaron 4 be long to the Early Phrygian c i tadel that
mud br ick , consol idation , is the dominant feature o f Gordion today . Excavations conduc ted
acryl id , mud p l a s te r . by Rodney S . Young between 1 9 5 0 and 1 9 7 3 and by Mary Vo igt in
1 9 8 8 - 1 9 8 9 , under the auspices o f the Unive r s ity o f Pennsy lvania ,
have uncovered numerous occupation leve l s within the 3 0 0 x 5 0 0 m
" C i ty Mound" s i tuated on a f l a t plain a l ong the S akarya River
about 1 0 0 km southwe s t o f Ankara , Turkey ( S e e f ig . 1 ) . Settle­
m e n t s r u n t h e gamut from Early Bronze A g e s trata in t h e t e l l ' s
lowe s t l ayers to an early Roman Emp ire encampme nt in the topmo s t
r e g i o n s o f t h e mound . T h e expo s ed s e c t ions o f Early Phryg ian
Gordion lie in an enormous 1 5 0 x 2 0 0 m trench c le ared from
the e a s tern hal f of the C i ty Mound dur ing Young ' s excavation s .
L i t t l e was known o f the Phrygians unt i l Young ' s inte n s ive s e arch
made it clear that by the end o f the e i ghth ce ntury B . C . Phrygia ,
with Gordion s e rv i ng as i t s capital , was a prominant civi l i z at i on
exerting cons iderable control over central Anato l i a . I t i s also
apparent that Gordion reached t h e he ight o f i t s prosperity under
the rule o f the legendary King Midas and that Phryg i a me t i t s
dem i s e a s a dominant power under h i s leadership when Gordion
was burned and des troyed by the Kimmerians , a group o f nomad i c
invaders , in c a . 7 0 0 B . C .

The Early Phrygian c i tade l , c a l led the " De s truction Leve l "
because o f the s igni fi cant fire damage and c o l l apse o f many o f i t s
s tructure s , i s div ided i n t o three ma in prec inc t s : a p a l a c e a r e a to
the northwe s t , a high terrace to the southwe s t support ing two long
service s tructure s , and a mu ltiroomed bui lding that bordered the

F i gure 1 . Gordion , ove r a l l CFW


p l an o f destruction leve l .
�.

100,,"

* Author to whom correspondence should be addre s s ed


c i ty ' s northwe s t fort i f icat ion wa l l . To the southeast i s a
ma s s ive gate comp lex that s e rved once as a monumental entrance
to the c i ty but was unde rgoing a rebu i lding pro j e c t at the t ime
of the K imme rian con f l agration . The burned rema ins of the two
mud brick megara s tand in the palac e are a , which c o n s i s t s of two
courtyards separated by a he avy enc l o sure wa l l running northeast­
southwe s t and each containing group s o f megara . Megaron 1 is
located in the southwe s t corner o f the " enc losed courtyard "
nearest the gate complex ( so c a l l ed becau s e o f a l i ght retaining
wa l l that separates i t from the gate s tructure s ) , and Megaron 4 is
s i tuated on the northwe s t s ide o f the " open square " , the courtyard
northwe s t of the he avy enc l o s ure wal l .

Megaron 1

Excavated in 1 9 5 6 and dated to the ni nth or e i ghth century B . C . ,


Megaron 1 is one of the e a r l i e s t bui ldings of the D e s truc tion
Leve l . The s tructure measures 9 . 5 x 1 7 . 5 m and cons i s t s o f a
porch and inner room , each with a central hearth ( S ee fig . 2 ) .
Entry wa s gained through two central doorways p l aced symme tri c a l ly
a long the longitudinal ax i s , one leading from the enc losed court­
yard to the front porch and the other pierc ing the c ro s s wa l l
between the porch and the inner room . Except for a stone soc l e
t h e megaron ' s wa l l s we r e b u i l t entirely o f mud brick a n d wood and
then covered with a thick coat of mud p l a s te r . The e x i s t ing w a l l s
a r e three b r i c k s t h i c k with a l ternat ing n i c h e s and p i e r s on the
inner and outer face s . The niches were l e f t by wood posts s e t
i n t o t h e wa l l faces on t h e e x t e r i o r a n d i n t e r i o r o f t h e bui lding
and are wide enough to sugge s t pairs o f po s t s . The rema ins o f
t h e wa l l face s s tand t o a uni form height o f about 1 m , wh i le the
central row of bricks r i se s somewhat higher and i s p ie rced at
regu lar interva l s by hol e s . A hor i z ontal beam topped o f f the
vertical posts and bricks of e i ther wa l l face a� th i s leve l . The
voids through the inner brick core were for short c r o s s p i e c e s that
t i ed the framework o f each s i de together . The level above th i s
f i r s t horiz ontal beam wou ld have had i t s own po s t s and another
wood leve l i ng course above them .

During the fire , the roo f - -made of beams covered with a layer
of reeds coated in turn by c l ay-- c o l lapsed a l ong with all the
wa l l s above the first level ing cour s e . The copious amounts o f
wood used i n the timber framework and roof beams in conj unction
with the highly f l ammable reeds mus t have made the blaze q u i te
inten s e . This wa s made apparent by the vitri fied s tate of in
s i tu mud p l a s te r that melted in the fire and began running down
the wa l l s . Many o f the individual mud bricks , particularly tho s e
l o c a t e d i n t h e niches l e ft behind by burned out posts , w e r e baked
hard and b l ackened because o f the i r proximity to the f l aming
timber s .

Megaron 4

Megaron 4 was comp leted at the end o f the e ighth century j u s t


be fore the Kimme rian f i r e . I t r e s ts about 2 m above the other
buildings o f the open square upon an extens ion of the c i tade l ' s

F i gure 2 . 1 9 5 6 - Megaron 1
j u s t a f ter excavation .
Consolidation Studies

F i gure 3 . 1 9 6 1 - Megaron 4
j us t a fter excavat ion .

southwe s t terrace . The s tructure has exterior d imens ions o f


1 2 . 3 x 2 2 m ( S e e f ig . 3 ) . T h e interio r arrangement con s i s t s o f
a very shal low front porch and a deep inner room conne cted by
a central doorway in the c r o s s wa l l between the two rooms that
is matched by a doorway o f the same width between the facade ' s
re turn s . Unl ike Megaron 1 , the wa l l s were made o f s o l id mud

thickne s s . There are no vertical niches

hor i z ontal s t r ing cour s e s .


ill
brick without a wood frame and range from 1 . 3 2 � 1 . 4 5 m in
the inner and outer
face s and no bricks preserved at a uni form level to sugge s t
Charred wood and reed found on the
floors indicate that Megaron 4 had a roof s imi lar to Megaron 1 ,
and a pattern o f post ho l e s in the main room sugg e s t s that there
wa s once a three- s i ded gal l e ry leve l in the rear chamber a long the
back wa l l and flank s . As with other. examp l e s of this gal lery
s y s tem found in many bui ldings of the Des truc t i on Leve l , Megaron
4 ' s gal l e ry prov ided its own support without pierc ing the actual
wa l l fabr i c .

Wh i le Megaron 4 was r a z e d a lmo s t to i t s foundat ions , it s e ems


that the absence o f a h a l f - t imbering sys tem did not fue l a s severe
a fire a s that encountered by Megaron 1 . The roof and gal l e ry no
doubt burned and c o l lapsed fo l l owed by the subsequent fall of mo s t
o f t h e supe r s truc ture . However , t h e in s i tu s e c t ions o f mud
brick , wh ich vary in height from 0 . 2 5 mete r s to 0 . 7 5 mete r s , do
not exhib i t the same degree of burning and d i s c o loration as those
in Megaron 1 .

Decay of the Megarons

C l imatic conditions in the Gordion region are inordinately har sh .


The ma j or i ty o f the average annua l 3 5 0 mm of precipi tation fal l s
dur i ng the winter months and i s accompanied b y freez ing tempera­
ture s ( - 1 5 ° C ) . During the s umme r , when excavations are ongo ing ,
the area is quite dry and windy ' with only l imi ted rainfa l l ,
exceedingly high temperature s ( 4 0 ° C- 4 5 ° C ) , and correspondingly
low humidity ( 1 0 % - 2 0 % ) . T h i s is an ide a l envi ronment for the
promotion of mechan i c a l weathe r ing on the exc avation s i te ,
particularly for unprotec ted surfaces that are broken down by
repe ated frost-wedging and wet-dry cyc l e s and washed or b lown
loose by rain and wind . •

Of a l l the eros ion at Gordion the mo s t notab l e post-excava­


t i on deterioration h a s taken p l ace on Megaron 1 and Megaron 4 .
Unt i l the 1 9 8 9 season , no inte rvention had been taken to c ir cum­
vent the s lumping of the ea rthen masonry that mak e s up the
rema in ing wa l l s of the s e monuments exc ept for an app l i c a t ion
o f a c ry l i c re s i n to a sma l l s e c t ion o f Megaron 1 in 1 9 8 2 ( s e e
" P revious Tre atment " be low ) .

Megaron 1 s e ems to be better preserved than Megaron 4 ,


a l though i t was excavated f ive years e a r l i e r and h a s been exposed
t o the environment for a longer period o f time . T h i s i s more than
l i kely due to the more complete baking that the mud bricks o f
Megaron 1 underwent from the intense temperatures generated b y the
ignited timbers of the build ing ' s wood framework . Another factor
F i gure 4 . Ove r a l l deteriora­
tion o f Megaron 1 .

may be that Megaron 4 i s in a more vulne rab l e pos ition . It is


e l evated above the level o f the other bui ldings in the palace area
and does not bene f i t from any shade dur ing the day or any sort of
protection from wind and rain . Megaron 1 , on the other hand , i s
bordered b y terracing wa l l s bu i l t a t l ater s tages o f the c i tadel
that extend a long its southeast f l ank and rear wa l l . These o f fe r
a t l e a s t some shade a n d certainly act a s s c reens from some o f the
e l emen t s .

D e s p i te the d i screpan c i e s in states of pres ervation both


me garons are in terrible cond i t ion . For the f i r s t several years
a f t e r it was unearthed , Megaron l ' s individual bricks and their
cour s e s were c l e a r ly d i s ce rnible and uninterrupted except for the
niches l e ft by timber pos t s . Now , after thirty- four years o f
cont inuous contact with snow , rain a n d wind , many sect ions o f
the wa l l s have d i s s o lved comp l e te ly , whi l e the outer surfaces o f
others have " me l ted " s o that the wal l s ' original forms a r e barely
recogn i z ab l e ( S e e fig . 4 ) . In megaron 4 , the bulk o f the wa l l s
have e i ther d i s appeared o r have been covered with s lump that was
once original brick sur face . It i s only at the r e turns o f the
c r o s s wa l l and facade that one can s t i l l make out the s e ams that
demarcate d i f ferent bricks .

Previous Treatment

Prior to 1 9 8 9 , the only treatment undertaken on the mud brick


architecture was the app l i c a t ion in - 1 9 8 2 o f a n acrylic resin
( Acryloid A - 2 1 ) on a sma l l s e c tion o f Megaron 1 [ 1 , 2 ] . I t is
unc l e a r whether the app l i cat ion h a s b e e n e f fec t ive in preventing
further deterioration , particularly a s the treatment was not
c a r e f u l l y documented or photographed . Furthermore , i t i s doubt ful
that the high mo lecular weight ( and corre sponding viscosity) o f
the r e s i n permitted deep penetrat ion .

On- s i te Examination and Treatment Te s t s

I n 1 9 8 9 , w e made a detailed examination o f the mud b r i c k architec­


ture o f Megara 1 and 4 and d i s c u s sed the deterioration o f the mud
brick , the previous tre atment , the samp l i ng of burned and unburned
s e c t i on s , propo s a l s for tre atment t e s t s , and future r e s e arch . We
reques ted and rece ived perm i s s ion from the Turk i sh Archaeological
S ervice to take samp l e s o f the mud brick . The ana l y s i s o f s amp l e s
w i l l determine t h e condition and s tructural s trength o f the burned
and unburned mud brick .

Megaron 1 was chosen as a s i te to t e s t two treatment methods


for the protection of the mud brick architecture . The f i r s t
propo s a l w a s to cover a sma l l mud b r i c k s e c t i on / p ier w i t h a modern
mud p l a s t e r ( or mud s t uc c o ) as a protect ive coating , wh ich ,
a l though obscuring deta i l , was not unlike the prote c t ive mud
p l a ster that had originally covered the surface . The s e cond
propo s a l was to consol idate another sma l l section with an acrylic
c o l l oidal d i spersion (Acrysol WS - 2 4 ) , both a s a comparison to the
mud p l a s t e r app l i c a t ion and to the previous acry l i c tre atment
( Ac ryloid A-2 1 ) app l i ed in 1 9 8 2 .

Toward the end o f the season we c arried out both of the s e


t e s t s , after f i r s t photograph ing , c leaning , a n d samp l i ng d i f ferent
s e c t ions o f the mud brick . One o f the local workmen , who had been
Consolidation Studies 29 3

rep l a s te r ing the mud brick in the Gordion excavation compound


( house and wa l l s ) , mixed up a thick mud p l a s te r and care fu l l y
app l ied i t d i r e c t l y ove r o n e o f t h e mud brick s e c t i on s / p i e r s .
T h i s was a l lowed to dry and touch-up was done in any cracks
that developed . Detai led photographs were taken to rec ord the
tre atment s o that any weathe r ing and change over the winter can
be compared in 1 9 9 0 ( s e e f i g . 5 ) .

P r ior to the on- s i te con s o l idat ion t e s t with Acrysol WS - 2 4


w e experimented o n a large fragment o f mud brick found lying in
the middle o f Megaron 1 . This fragment was taken back to the
excavation compound and consol idation tests were begun u s ing
d i f ferent concentrations o f WS - 2 4 and u s ing two d i f ferent methods
F i gure 5 . S e c t i o n / p i e r o f of app l i c a t ion .
Megaron 1 a ft e r mud p l a s te r
app l ic a t ion . Acry s o l WS- 2 4 was chosen as the consol idant because of i t s
exc e l lent phy s ical and chemi c a l prope r t i e s [ 3 ] . I t i s a very
fine part i c l e - s i z e acry l i c polymer o f h igh mo lecular weight
d i s p e r s e d in a water medium , and its extreme ly low v i s c o s ity
ensures deep penetrat ion . The hardne s s and durab i l ity of the set
re s i n ( a fter water evaporation ) should provide more than adequate
s t r ength and protection against adverse c l imatic condition s .
After s e t t ing , the r e s in i s removable with s o lvents ( e . g . ,
f l u shing with acetone ) , but in practical t e rm s this would be
very d i f f i c u l t .

The f i r s t method chosen was by s imp l e we t t ing : dr ipping and


pour ing various concentrations o f WS - 2 4 direc'tly onto the mud
brick . Th i s was done o n ly to one end ( the sma l l end ) and i t was
found that concentrations up to 2 0 % r e s in were readi ly absorbed
by the ��rous brick , except in the areas where the brick fabr ic
had been s intered or fused by the des truction fire . A 4 % s o lu t i on
o f WS - 2 4 was chosen for the c omp l e te con s o l idat ion , and a ft e r
thorough wetting ( at w h i c h p o i n t a bui ld-up o f consol idant
oc curred on the s u r face ) , the brick was l e ft in the shade to dry .

I n order to a s s e s s the penetration o f the c on s o l idan t , the


brick was broken in hal f a f t e r forty- e i ght hours when i t f e l t
d r y to t h e touch . I t was easy t o s e e where t h e consol idant had
penetrated because i t had s l ightly darkened the subs trate and ,
in fac t , had not comp l et e ly dried in the cente r . Comp l e te
penetration was achieved except on the one s ide where the brick
was burned .

The s econd app l i c a t ion t e s t was carried out by plac ing the
rema i ning brick fragment in a c losed sys tem and a l l owing a 4 %
s o lution o f WS- 2 4 t o percolate o r " w i c k " up from the bottom .
The brick was p l aced in a doub l e - l ined p l a s t i c bag with enough
s o lution at the bottom to cover one-quarter of the brick , and
the p l a s t i c bag was then s e a led . The perco lati on r i s e of the
consol idant was monitored over the next several day s , and once
the s o lution had approached h a l fway. up the brick ( i n thirty - s ix
hours ) , more solution was added to replace that wh ich had been
F i gure 6 . Con s o l idation te s t absorbed ( s e e f i g . 6 ) . After another forty-eight hour s , the
u s ing 4 % WS- 2 4 on mud brick . consol idant had reached about three-quarters o f the way up the
brick , but hardly moved a fter that . The brick was there fore
removed the next day , and an unsuc c e s s fu l attempt was made to
break it ( again to a s s e s s the penetrat ion of the consol idant ) .
The unconso l idated end was immersed in 4 % WS- 2 4 , and when dry ,
the fragments were returned to the f ind spot on s i te to weather
the winte r . The brick was photographed and a c areful exami nation
w i l l be done in 1 9 9 0 to a s s e s s d t s cond i t ion and the e f fe c t ivene s s
o f the consol idation .

On s i te , one o f the sma l l e s t and most deteriora ted s e c t i o n s /


p i e r s in t h e S W c o r n e r o f Megaron 1 wa s consol idated u s ing WS - 2 4 .
I n this c a s e , a 4 % s o lution was both sprayed on with an atom i z e r
a n d dr ipped from t h e t o p i n t o o n e o f t h e mo s t expo s ed are a s . The
s e c t ion , a l though only four cour s e s high and two wide , l i t e r a l l y
drank up t h e consol idant , a n d over 2 l i t e r s o f s o lution were
app l ied . As the upper bricks became saturated , an exc e s s of r e s i n
began to build up in o n e area , a n d i t w a s dec ided t h a t t h e end
point had been reached . The exc e s s was l ightly bru shed o f f u s ing
deionized water . Photographs were taken throughout a l l s tages. o f
the treatment in order t o record and compare any changes the next
summer ( s e e fig . 7 ) .

F i gure 7 . S e c t i o n / p i e r o f Addit ional samp l e s o f mud brick were taken from Megaron 4 .
Megaron 1 a fter 4 % WS - 2 4 Ana l y s e s o f the mud brick samp l e s w i l l be undertaken to determine
app l i c at i on . the ir compos ition and cohe s ive s trength , and , together with the
analyses of the anc ient mud p l a s t e r , w i l l aid in s e lec ting a
future treatment .
29 4

Conc lus ions

The condition of the burned Gordion mud brick mak e s i t an


extremely interest ing and problematic c a s e s tudy in the preserva­
t ion o f in s i tu earthen architecture . Two methods have recently
been t e s ted for the protection o f the mud brick . The f i r s t
treatment te s t involved the c omp lete cove r i ng o f a mud brick
sec tion with a mud p l a s te r o r stucco . The sec ond method t e s ted
re turned to the use of an acry l i c c on s o l idan t , but in a f i ne ly
d i spersed aqueous med ium . Both treatments have advantages and
l imitation s , when one considers the nature o f the intervent ion
a s " exterio r " o r " i nterior " .

The mud p l a s te r s tucco approach i s more in keeping with what


the bui lding had as an original fac ing , but now that the bui lding
is exposed to the e l ement s , it w i l l require regular maintenanc e .
I n add i t ion , the e f fec t iven e s s and protect ion has to be weighed
against the somewhat di sturbing aesthetic appearanc e . Even s o ,
the covering preserves the general appearanc e , and the interven­
tion i s minimal as it doe s not introduce any new and unknown
mate r i a l s into the mud brick .

The introdu c t ion o f the acrylic con s o l idant WS - 2 4 , though


apparently penetrat ing deeply , presents many unknown s , regarding
i t s cohe s ive s t r ength , loc a l i z ed concentrat ion , and interior
penetrat ion . As a temporary conservation me asure , it a l lows
the current cond i t ion of the s tructure to remain v i s ible whi l e
prote c t i ng t h e mud brick from further decay .

Until a s u i table and e f fe c t ive consol idant can be found to


comp l e t e l y prote c t the expo s ed mud brick architecture at Gordion ,
the dec i s ion w i l l have to be made as to what temporary measure i s
mo s t acceptab l e . Further t e s t s and r e s earch a r e needed t o a s s e s s
both t h e phy s i c a l requirements o f the mud brick and methods for
c omp l e te impregnat ion . The expo s ed architecture at Gordion o f fe r s
an important opportunity a n d s i te f o r future s tudy a n d immediate
attent ion .

NOTES

1. , Gordion Notebook 1 7 3 , Gordion Archives at the


Unive r s i ty Mu s e um , Phi lade lphia , 1 1 1 .

2. M . H . Roge r s , " S i te Conservation at Phrygian Gordio n " ( Honors


E s s ay , Unive r s i ty of North Carol ina at Chapel H i l l , 1 9 8 9 ) .

3. S . P . Koob , " Consol idat ion with Acry l i c Col loidal D i sp e r s i on s �


Preprints , A I C Annual Me eting , 1 9 8 1 , 8 6 - 9 4 .

MATE RIALS

Acryloid B - 7 2 , Acrys o l WS - 2 4 : Acry l i c r e s ins , manu factured b y


Rohm & Haa s , Phi ladelphia , PA ; supp l ied by Conservation Mate r i a l s
Ltd , Spark s , NV , U . S . A .
Consolidation Studies 29 5

ABS 1RA CT CONSOLIDATION OF A NEOL ITHI TIIE WITIl


C EAR N SITE POTASSIUM SILICATE

Li Zuixiong,
This paper presents the results of an Preservation Research Institute,
Dunhuang Academy, Gansu Province,
experimental study of consolidation
The People's Republic of China
with potassium silicate of a neolothic
earthen site at Dadiwan, Gansu
Province, China. Details of the
preparation of the potassium silicate
In troduction
Weathering is the most serious cause of deterioration of earthen structures, and a solution to the

solution, its properties, and the


mechanism of its action with
problem is urgently req uired. By investigation and analyses of the ruins of a neolithic earthen site

at Dadiwan, Qinan County, Gansu Province in northwestern China, it was concluded that

montmorillonite - containing clay are


described. The physical and
montmorillonite contained in the clay is one of the main factors contributing to weathering. The

purpose of the present research was to find ways to reinforce this type of earth by enhancing the

chemical properties of the potassium degree of consolidation by chemical means and thus prevent weathering.

silicate-clay composite were


determined by testing and
Ruinsof house of Yangshao Period in Dadiwan.Oinan. G ansu[1J
instrumental analysis. Calcium At the beginning of the 1980s, two rare, large ruins of houses, designated F405 and F90 1 , built

flourosilicate and aluminium silicate during the neolithic period of late Yangshao, were excavated at a site at Dadiwan. A unique

were used with the potassium aspect of the site is that the floors (in area about 300 and 1 3 1 m2 respectively) are made of burnt

silicate solution. Of crucial Liaojiang stone (the depositional calcium carbonte in loess), which is quite similar to contemporary

importance is the molar ratio of Si02 man-made concrete type ceramics. These floor materials are believed to be the earliest known

to K20. The optimum ratio is 3.8 - cement and man-made concrete filler.

4.0. Higher concentrations of K20

give poor results through eventual


formation of K2C03 by atmospheric
carbonation.

KE YW
ORDS

Consolidation, Potassium silicate,


Montmorillonite

Fig. 1 - The F901 ruins of the house of the Yangshao period in Dadiwan.

SelectionofConsolidants
The clay in the structures at Dadiwan contain some montmorillonite. As is well known,

montmorillonite can absorb much water into its layered structure and swell considerably. Upon

drying the clay contracts. Such an alternation of expansion and contraction accelerates weathering

damage. It was the intent of the research to reinforce the earthen structures by permeating them

with inorganic cementing materials which are similar to and can react with the clay, thereby

improving the cementation and make the earth more weatherproof and stable. This, in theory, is an

ideal way to reinforce and prevent ellrthen ruins from weathering.

Initially, tests on the material from Dadiwan were conducted with polyvinyl alcohol,

polyvinyl-acetate emulsions, and sodium and potassium silicate solutions of different molar ratios

of metal oxide to silicon dioxide. Less than ideal results were obtained. Subsequently it was

found that using potassium silicate with a high molar ratio of Si02 : K20, together with appropriate

solidifying agent, cross-linker and surface active agent to reinforce the structure by permeation is

an ideal method for consolidating.

At first, it is necessary to determine the optimum molar ratio of SiOz : K20 for the aqueous

solution of potassium silicate. In the case of too Iow a ratio, after the potassium silicate and clay
Adobe 9 0

have reacted with each other, too much K20 will remain which will form K2C03 under the action
of the atmospheric C� and moisture:

If the potassium carbonate that results is not treated in time, or is removed, the stability of
the cemented mass will be impaired. When the ratio is too high, it has a weakened cementation,
which causes the strength of the solidified layer to decrease and, at the same time, tends to form a
coagulated layer preventing the mass from being permeated to any depth. In such cases, spraying
is difficult to carry out and does not produce good results. How to determine the optimum
modulus for the potassium silicate? Through cementing an amount of clay (C) containing some
montmorillonite [2] with an aqueous solution of potassium silicate (PS) at the same concentration,
but with different molar ratios of Si02 to K20, a PS-C composite mass is made. Then test
properties such as CO2 - resistance, weatherability, water resistance and thermostability may be
determined. By these tests the optimum ratio of the potassium silicate solution was determined to
be 3 .8 - 4.0.

Solidifying agents and cross-linkers are also important factors affecting the stability and
applicability of the PS-C material. Tests have been carried out on more than 10 types of
solidifying agents such as aluminium trichlorid� , iron trichloride, sodium fluorosilicate, calcium
fluorosilicate and potassium aluminium sulphate. Judged by the solidifying speed and the stability

of the PS-C-cemented mass, calcium fluorosilicate as the solidifying agent was found to be best.
Through testing, powdery aluminium silicate was selected as the cross-linker.

Pre.parationof Silicate Correct Ratio


Potassium of Molar

An industrial potassium silicate (molar ratio about 2.6) was diluted with water to a specific gravity
of 1 .3 and then caused
bath to 100·C and s tirredto
react in a 1 5 L stainless steel autoclave. It was heated by jacket-water
vigorously. At the same time, powdered silicon dioxide (80 mesh),
mixed with warm water into a paste, was added gradually into the autoclave. The operating
pressure rose to 21 kglcm3• About two hours later, samples for determining the ratio were taken
from the autoclave. If uired
the req ratio was obtained, the PS preparation was considered
complete. The PS obtained with an optimum ratio is a somewhat yellowish, thick colloid. It
would be substantially colorless if purer starting material was used. The measured surface tension,
density, viscosity and pH are given in Table 1.

Table 1 . Surface tension, density, viscosity and PH values of potassium silicate (P.S.) solutions at
given temperatures

P. S. solution Temperature Surface Tension Density Viscosity PH

Original
(0C)

8.0 +/- 0. 1
25.0 +/- 0. 1
(dyne /

76.67
em)
75. 17
(glcm3)

1 .2529
1 .2488
(CP)

1 1 8.89
1 8.729
15 12

Diluted* 8.0 +/- 0. 1 74. 3 1 1 .0478 1 .7789


25.0 +/- 0. 1 72. 8 1 1 .0449 1 . 1425
15
%
* - Potassium silicate solution diluted with water by 4X, + 1 CaSiF6
10 - 1 1

The PS diluted four times with water is colorless and transparent. The procedure for preparation
of the working solution is as follows:

Determine the dilution according to the requirements for application and measure the water
required 6
. Weight out respectively the Ca S i F and A 1 2 (S i �h required and add them to the
water. Stir the mixture and then pour it into the PS solution. Finally, add about l Oppm of NNO
as diffusion agent and stir uniformly. It is then ready for' spraying.

Preliminruy Analysis of Mechanism of Action Between the Clay(C) and Aqueous Solution of
High Molar Ratio of Potassium Silicate (PS)'
Consolidation Studies 29 7

X-ray Diffraction Analysis

Fig.2 X-ray diffraction spectrum of clay containing montmorillonite.

Fig.3 X-ray diffraction spectrum of PS-c.

Fig.2 is the spectrum of the clay containing montmorillonite; Fig.3 is the spectrum of PS-C. The
spectra show that the clay minerals' crystallinity was destroyed by treatment with the PS. The clay
minerals have turned substantially into an amorphous gel. Even if montmorillonite still exists in
the clay, the clay treated with the PS solution, which is mainly potassium silicate, will have a
reduced space between layers, i.e., the C-axial length will be reduced, thus resulting in reduced
expansion by absorbed water due to exchangeability of the montmorillonite with the potassium
ions.

fraredSpectral nalysis
In A

FigA is the infrared spectrum of clay containing montmorillonite; Fig.5 is the infrared spectrum of
PS-C. The infrared spectrum shows that the OH peak intensity near 3500 cm -1 of the mass has
dropped considerably, and the OH shifted by 10 wave numbers. This is another reason why the
PS-C cemented mass has higher water resistance.

,.

FigA Infrared spectrum of clay contamed montmorillonite.

Di fferentialThermal nalyis
A (PTA)

differential thermal curve of the PS-C. The differe


Fig.6 is the differential thermal curve of the clay containing montmorillonite; Fig.7 is the
ntial thermal curves show that the clay has the
first endothermic peak at 84'C, the second at 56I'C and the third at 754'C, while the three
endothermic sygnals of the PS-C are of 1 1 3'C, 684'C and 837'C respectively. That is to say,
after the clay was treated with PS, its water-loss temperature rises. It is possible that part of the
clay-absorbed water is converted into structural water on the clay resulting in an increase of the
strength and water resistance.
Adobe 90

· zo

8.<: I
o Q(]O ilQOiI
Fig.S Infrared spectrum of PS-C.
2000 ' 1600 . /200 ()OO 'lor-
Fig.6 Differential theITIlal curve of
contained montmorillonite.
Scaon lectron icros
in& E M cope(SEM) Examination

Fig.7 Differential theITIlal curve


of PS-c.
Fig.8. SEM photograph of clay containing montmorillnite, 9000 x.

Fig.8 is a SEM photograph of clay contained montmorillonite; Fig.9 is a SEM photograph of PS­
C. The SEM photograph shows that an interwoven fibrous network of crystals, similar in
structure to that found in hydrated gypsum, has fOITIled. This is an important factor in imparting to
the PS-C composite good water resistance and mechanical strength.

Fig 9 SEM photograph of PS-C, 9000 x.


Consolidation Studies 299

Testin�Resistanhysicalce Chemical Pro.perties


Water
Main P and of PS-C

Test pieces of PS-C of different molar ratio were immersed in water until they disintegrated.
Among test pieces which have been immersed for 16 months, the PS-C of high ratio (3.8-4.0)
have not disintegrated, while those of lower ratio (below 3.5) disintegrated in a much shorter time.

'COz Resistance
-
Resistance to C(h is one of the important indexes in examining the stability of the PS-C cemented
mass. In the case of too Iow a ratio the potassium silicate will retain too much K20. This K20
will gradually create K2 COl under the action of C02 in
effloresce in moist air. air and cause the PS-C cemented mass to
This can be established by the following test: Small pieces of PS-C of
different molar ratio are placed in a vacuum desiccator with water at the bottom and filled with
COz. It is found that PS-C test pieces made with potassium silicate of low ratio went white and
viscous on their surface within a few days. Samples of high ratio showed no change after months.

Weatherability
Small test pieces were placed outside and exposed to sunlight for six months, then tests were
carried out for water and C(h resistance. It was found that test pieces of high molar ratio were
stable and did not disintegrate when subsequently placed in water. Also no deterioration took place
under C(h saturated with water vapour.

Basicityofthe PS-C Cemented Mass


Because the PS solution is a strong-base, it might be considered that clay treated with the PS
solution (that is the PS-C cemented mass) will be unstable and effloresce in air. This is not so: the
silicon and aluminium in the expansible clay can cement with water and calcium only in a highly
basic environment and increase the mechanical strength and stability of the PS-C cemented mass to
a great extent. High basicity has become a simple and effective way to improve the engineering
performance of expansible clay and has been widely used in civil engineering at home and abroad.

The original PS solution we used had a pH = 12 which dropped to about pH = 10 after


being diluted. The PS-C cemented mass is quite stable which has been proven by the tests outlined
above and in site applications.

CONSOLIDATION WITIi THE


PS OF
PERIOD HOUSE AT DADIWAN.
F901 REMNANTS OF THE LATE YANGSHAO

In August 1984 and in October 1985, site consolidation was undertaken on the weathered remains
of the Yangshao house by spraying with PS solution. A 13-meter wall, one cooking stove and
two post holes have been consolidated. Up to now, five years later, no abnormal changes have
taken place on the surface of the structures.
1 1 ).
In our view an ideal effect was achieved (Figs. 10,

In the site testing the weathered remnants of the house were sprayed once, twice and three times
respectively for the purpose of comparison. From observations made over more than one year it is
recommended to spray three times. Structures which have been sprayed three times still have good
permeability, that is to say, it is possible to do fourth or fifth treatments if necessary.

CONCLUSION

Testing for consolidating the weathered archaeological earthen structures at Dadiwan, and on-site
application, have shown that the PS solution reported here can achieve an ideal result in terms of
strengthening the substrate while retaining good permeability. As seen by physical and chemical
properties, testing, X-ray diffraction, inf rared , scanning electron microscopy and differential
thermal analyses, the PS-C is a kind of stable weatherproof inorganic composite which does not
undergo marked changes in appearance. The tests have achieved the aim of changing the cemented
state of clay containing montmorillonite through reinforcing by means of permeation with PS.
There is, however, a problem to be solved through future research, that is, when applied to wet
earth the PS shows poor permeability and a low rate of solidification. Furthermore, Si02
3 00 Adobe 90

coagulates on the surface under such conditions and this results in a reduction
and strengthening effect.
in the consolidation

Fig. 10. The wall and skeleton pillar on adherent wall of F90J house-ruins after consolidation.
(Photo, March 1990.)

Fig. 1 I. The Pillar hole of propping beam wooden pillar of F90J house-ruins after consolidation.
(Photo, March 1990.)

Acknow led�ements;
We thank Mr. Wang Hengtong and Mr. Wang Qi for their assistance in the experiments.
Consolidation Studies 3 01

ReferencLi es:
1 Zuixiong, 'Conservation of Cave Temples in China', Bulletin Society for Scientific
Studies Cultural roperty
on P No. 13, ( 1987) 1-7, (in Japanese).

2 Stambolov, T., and van Asperen de Boer, J.R.J., 'The Deterioration and Conservation of
Porous Building Materials in Monuments' , I

1979) 45-58.
nternationalCentre Conservation,
for Rome

3 Li Relics&Museum,
Zuixiong, 'Deterioration ofBingling Temple, North Cave Temple and Maiji Mountain' ,
Cultural No.3 ( 1985) 66-75, (in Chinese).

4 Li,
Nishiura, T., and Z., 'Experimental Study on the Consolidation of Fragile Porous
Stone with Potassium Silicate for the Conservation of Cave Temple in China' ,

Conservation Far TIC


108- 1 12.
of Eastern Art. 12th Internation Congress, September 1988, Kyoto,

[1] The late Yangshao period of Dadiwan i s 4,900-5,500 years before present.

[2] Clay (C) containing montmorillonite for testing was taken from the North Cave Temple,
Qinyang, Gansu Province.
3 02 Adobe 90

ABSTRACT GENERAL CONSIDERATIONS ON THE PRESERVATION OF EARTHEN


ARCHITECTURE IN MINAS GERAIS , BRAZ I L ; A PROPOSAL FOR REINFORCE­
There is a considerable MENT OF A BRAYED MUD WALL STRUCTURE
historical and artistic
patrimony o f earthen ar­ Sergio Jose Fagundes de Sousa Lima , architect
chitecture in Bra z i l dat­ SPHAN - Technical O f f ice of Sao Joao del Rei
ing from the colonial Rua Marecha1 Deodoro , 12 - Centro
period . A systematic 3 6 . 3 0 0 - Sao Joao del Rei
study of the consolida­ Minas Gerai s , Bra z i l
tion of this architecture
is essential . This paper S ilvia Puccion i , engineer
considers intervention S PHAN - Technical Consul tory
methods for earthen Av . Rio Branc o , 4 6 , 5 · andar - Centro
structures as wel l as an 2 0 . 09 0 - Rio de Janeiro
analysis of their con­ Rio de Janeiro , Bra z i l
tents .

To i l lustrate the type of Brief Historical Overview of the Use of Earth in Architecture
proj ects that have been
undertaken in Bra z i l , a The use of earth as a construction material dates to prehistoric
case study o"f the conser­ times , to ancient Mesopotamia and Egypt . In Pre-Columbian
vation of the Bas i l ica of America , its usage was widespread among the Aztecs and other
Our Lady of P i l l ar is North American peoples .
discussed . This proj ect
involved the restoration In S outh American countries such as Bra z i l , the Spanish and
of a severely deteriorat­ Portuguese introduced European techniques of earthen architecture
ed edifice , which re­ during the early stages of colonization . As a result , the early
quired special analysis urban architecture of these countries is characterized by fea­
and structural reinforce­ tures such as wide window s i l l s , the predominance o f wal l s over
ment of the masonry of a empty spaces , and spec i f ic wal l thic knesses . Every type of
brayed mud wal l which had architecture in the "mineira" ( 1 ) history of the eighteenth and
incurred static damage . nineteenth centuries--civil , Official , or religious--was in­
The goal of the project fluenced . Earth became the most viable building material for
was to maintain the old structures and dividing wal l s of edifices because of the abun­
material whi l e g iv ing the dance of the material in the region and its environs and the
masonry a new sustaining mastery of the technique by artisans .
capacity , taking into
account the original con­ The construction systems that use this techni que offer a variety
structive system . of applications as seen in the edifices of our historical cen­
ters . Examples of earthen techniqeus are adobe and formigao
( norma l ly used for structural purposes , at l east on the outside
KEYWORDS wal l s ) and the pau-a-pigue , taipa-de-sebe or taipa-de-mao for
dividing wal l s and internal seal ing .
Conservation , restora­
tion , brayed mud wal l , Less frequently we find another material cal led moledo , equa l ly
pau-a-pique , adobe , mole­ characteristic of that region , made of decomposed c l ay , pebbles ,
do , reinforcement , con­ and rocks . Because of its resistance "( load-bearing capacity ) , it
sol idation . is a l s o used in the structural masonry and dividing wal l s , in
some cases exposed to open air . ( See f ig . 1 . )

As an a lternative to the use of earth , we find stone used as a


material especi a l l y in edifices of monumental proportions . The
use of stone inevitably increased the construction cost because
of the d i f ficulty of handling and transportation .
In a few cases rock is used for the entire structural system of
the bui lding , though in the maj ority of cases it forms only part
of the structure , the foundations and corners , for example .
Another advantage o f stone i s its decorative use .

General Considerations about the Preservation of Earthen Ar­


chitecture

A. Systems conservation

Over the l ast six years , the National Artistic and Historical
Patrimony Secretary ( S PHAN) has exercised greater influence over
the historical centers of the region Campos das Vertentes with
the objective of repairing the damage occuring as a result of the
social and economic decline in Minas Gerais . Some old edifices
were in critical condition . Others required intervention at
speci f i c damaged points , which were eas ily repaired . In other
cases the stabil ity of the building was compromised , requiring
more extensive intervention . In the case of structures whose
inhabitants kept up comprehensive maintenance--from the roof to
water and e lectrical instal l ation and sewage system--the physical
condition remained sound , reducing conservation costs .

B. Urgent work staff

In order for S PHAN to be e ffective in the historical centers , the


F i g . 1 : Wal l with "mole­ Federal Government establ ished an "urgent work staff" to restore
do" base and adobe mason­
ry . Tiradentes , Minas
Gerais . Photo : Lima . 1. The style representative of Minas Gerais ( a state in Bra­
zil) .
Consolidation Studies 3 03

Description o f the Con­ and maintain the architectural patrimony as wel l as furnishings .
structive Systems This staff is not sufficient to undertake a l l of the needed
interventions and can only addres s the "most urgent" cases .
Brayed mud wal l : Thi s
system i s one in which Through thi s staff the Bra z i l ian Government is abl e to offer to
the wal l s are of mass ive the people of the historical centers , who are general ly in need
monolithic mud , which is of f inancial resources , ass istance in preserving their bui ldings .
formed in wood moulds We o ffer , free of charge , the manual labor of special i z ed sta f f ;
held in position by tra­ private patrimony owners contribute the necessary materials when
verses and wood sticks . resources are ava ilabl e ; SPHAN supp l ies the technical orienta­
t i on .
Adobe : Thi s material con­
s ists o f mud bricks with C. criteria and intervention methods
dimens ions o f 2 0 x 2 0 x
4 0 cm . It contains vege­ The f irst step in the restoration and conservation of a given
table fibers or manure . damaged edifice is to examine the existing damage and to prepare
It is put to dry in shad­ a condition report . After verification of the financi a l condi­
ow and , after a few days , tion of the res ident , the subsidy is determined which the S PHAN
in the sun . It di ffers w i l l provide for the execution o f the work .
from conventional bricks
in that adob e is not I f a dec i s i on is made to undertake the intervention , it i s
baked in an oven . necessary to study the possible treatment approaches according t o
a t l east three criteria :
Formigao : This variation
o f the mud wal l is made - The persistence in the use of the techniques , traditional or
by not s i fting the mud contemporary , which have been tested and which sat i s fy the
but rather mixing it with purposes of the conservation proj ect .
stony earth to make a
conglomerate-l ike con­ - The use of native material which is appropriate to the material
crete . being treated and which reduces costs .

pau-a-pique : Thi s is a - The h istoric authenticity of the obj ect as wel l as its contem­
kind of seal ing which poraneity . Interventions should be indicated in distinct and
consists of p ieces o f transparent ways .
wood s e t perpendicular
between the "baldrames " We can now del ineate a chart of practical examples of s o lutions
( the l ower support ) and adopted for the treatment o f certain material s .
the " frecha i s " ( the upper
support ) , attached by D. Conservation o f the wal l
holes or nai l s . Normal ­
l y , other thinner p ieces , 1. Pau-a-pique
sticks , are attached on
both s ides . After the We f ind the pau-a-pigue system in several edi f ices , which show
texture is made the mud external deterioration due to contact with moisture caused by
i s thrown and pressed on poor roofing or by s o i l saturation . In these cases , the solution
by hand . is to e l iminate the humidity or the s4rcharge and to investigate
the rupture , d i sintegrat i on , deformations , etc . that have oc­
Moledo : Thi s is a con­ curred . From thi s point on we promote the repair of the masonry
glomerate made of clays , o r , as an alternative , in s ituations of extreme disaggregation
" sa ibro" ( f ine gravel ) where repair is not possibl e , the option is reconstruction , using
and decomposed rock , re­ comtemporary techniques such as conventional fired brick , either
moved in blocks from the solid or " emptied" ones or " s ical" block ( ce llular concrete ) ,
mines , in the shape of etc .
cobble stones of varying
dimen sions . Because this 1 . 1 . Repair o f the elements
material i s very res is­
tant , it i s applied " in -Paus-a-piques : those that present a useful section are main­
natura . " tained and conserved and rectified when necessary .

-sticks : when comp l etely deteriorated , new sticks are sub­


stituted .

-Mud wal l : i f l osses are registered a fter the disaggregation o f


the plaster , the mud i s often repaired using a mixture o f mud­
cement ( - 1 : 2 0 ) , compatible with the old system in level o f
coefficient o f d i l atation . Eventual ly a layer of rough cast and
two l ayers o f f inal cast are appl ied .

In the most serious cases , when a considerable l o s s of mud i s


aggravated b y the instabil ity o f the wood texture , a texture o f
galvani zed wire i s used o n both s ides t o consol idate the whol e
system . ( See f i g . 2 . )

2. Brayed mud wal l

Monol ithic ( big block) masonry structure o f l arge dimensions


presents problems when it is directly exposed to open a i r . After
the rough cast is dislocated , humidity ( especially rain) disag­
gregates the structure . Also , structural problems arise due to
12-W-GALYVAOfNIZTEIDYNWIG RE SECTION f i s sures in the l ower areas of the foundation , deterioration o f
the beams , etc . Usua l l y , in the case of one- or two-story struc­
ture s , the e ffects of the problem cease with the treatment of the
prob l em--that i s , if the roof is repaired or the foundation is
Fig . 2 : Pau-a-pique wal l . reinforced , only the work of recomposing the damaged spaces
I l lustration by Lima .
Adobe 90

rema ins . At most , it is possible to introduce an " atirantamento"


( steel and wood wires to hold the wal l s together) on the masonry .
In cases of excessive horizontal stress , this intervention
consol idates the masonry with rigid structure .

3. Adobe and moledo ( 2 )

Because o f the pecul iarities o f these material s , they "behave"


very wel l statical ly ( rigidly ) , except in s ituations such as
those that occurred with the brayed mud wal l . Presently , there is
no other method of intervention in these systems of construction .

The dimensions of these materials are considerable : adobe almost


invariably measures 20 x 20 x 40 cm ; moledo is quite variab l e ,
but its s iz e is at l east equal to the adobe brick . Usua l l y they
are used double in masonry , resulting in a minimum width of 4 0
cm , with the right enclosure s .

4. General technical solutions

In any particular edifice there are different systems of earthen


construction--structural masonry , both external and internal , and
internal dividing wal l s . I n certain s ituations there i s a lack
of cohesion in the j unction of two systems which causes f i ssures
on the corners . A galvanized iron texture can help to connect
the masonry .

Every time we undertake an intervention on the roof of an edifice


with structural systems of earthen construction , it is advisable
to take advantage of the moment to execute a crowning belt , to
connect the system by stab i l i z ing the skeleton of the buil ding .
The e f fects of weather must be avoided , s ince the areas closest
to the eaves suffer the greatest from accumul ated ra in water .

In order to a l l eviate the e ffects of ascending and l ateral


pressure of water on the masonry , it is adv i sabl e in certain
cases to construct a ventilation ditch , a system whose purpose is
to el iminate the direct contact between the masonry and the soil ,
by forming an intermediate "mattres s of a ir . "

A orooosal for reinforcing a structure of a brayed mud wal l .


Specific case study: Bas i l ica Cathedral of Our Lady of P i l l ar

A. The brayed mud wal l

The brayed mud wall was largely used a s a system o f construction


in the " Paul ista Plateau" and in the state of Goia s . In the
northeast , the use of this technique was not often uti l i zed , but
there is record of the use of this system of construction in 1 6 6 0
i n the city o f Salvador , Bahia .
In Minas Gerai s , the use of the mud wal l was less frequently
used . Other materials (wood , stone , etc . ) were p l enti ful , and
furthermore the region is quite uneven , a fact which favors the
use of other techniques .

Reports of travelers from the n ineteenth century ( e . g . , Richard


Burton , 18 6 8 ) have been found which prove the existence of civil
and religious bui l dings constructed in " taipa" in several regions
of the State ; such is the case of the Bas i l ica of Our Lady o f
P i l lar in S a o Joao d e l Rei .

B. Bas i l ica Cathedral o f Our Lady o f P i l l ar

This is an edifice made of brayed mud wal l , dating from the early
eighteenth century . The high altar contains the main chapel arid
altar p iece , which was compl eted three decades later .

The construction includes an area of 1 . 0 5 6 m2 with an average of


5 . 50 m height . The church square covers an area of 129 m2 ,
reached by a fl ight of sta irs made of "cantaria" ( carved stone ) .
The construction consists of two parts , a posterior one which
includes the high chapel and sacristry and another , anterior ,
which includes the main nave and the lateral corridors extending
into the main fayade , which was rebuilt during the middle of the
nineteenth century in stone masonry . ( See fig . 3 . )

ymm
The main fayade has a central door in the lower part , f lanked by
four sma ller doors s e trically set in' relation to the central
axis under f ive windows of the balcony finished by a triangu l ar
" frontao" topped by a stone cross . The fayade is f l anked by two
towers provided with pyramid dome s . Al l the deta i l s of the
Fig . 3 : Fayade of the fayade are made of carved stone . Inside , the temp l e is one of
Basil ica Cathedral of Our
Lady of P i l lar . Photo by
Lima . 2. Material made of decomposed clay , pebbles , and rock .
Consolidation Studies

the most express ive of the "mineiro" baroque for its profusion o f
g o l d carving , mainly in the high chapel where proportion and
harmony and grace predominate . The nave has seven l ateral altars
in the same style as the high chapel .

C. Reinforcement proposal

The bui l d ing is settled in the southeast part o f the town of Sao
Joao del Re i , in the footh i l l s o f Rosario . ( See f i g . 4 . )

In the time s ince the building ' s construction , the streets o f the
town have been paved , including those streets that begin behind
the church and rise directly up the h i l l . This paving keeps the
surface water from f i l tering down into the s o i l , as it did in
former times . The resulting accumulation of water on the poster­
ior part o f the church causes erosion o f the s o i l that holds the
foundations at the right s ide o f the bui lding .

In addition , there is also the poor condition of the drainage


system meant to col lect rain water and conduct it to the s ewage
system .

These facts caused the " l is ionamento" and the deformation of the
right wal l .

The deformation o f the wal l also caused the rupture o f the eaves
and d i s l ocation of the t i l e s , a l l owing the infiltration of ra in
Fig . 4 : Aerial view o f
water on the mud wal l s over the course o f many years . This
part o f the historical
caused the disaggregation o f the earthen masonry , reducing its
center o f Sao Joao del
enduranc e , increasing the intrinsic tension of the material on
Rei . In the center is
the critical points of the structure , for instance in the "vaza­
the Basil ica Cathedral .
dos" ( empty spaces ) o f the doors , windows , niches , etc .
Photo by Lima .
The altars at the right wal l of the central nave , l ocated at the
openings that pass through the masonry , were in a severly damaged
state . One could clearly see the lateral " flarnbagem" ( as a
result of pressure) of the carved wood columns that form the
l ateral altars .

As a first step , we removed the artistic elements , and we provid­


ed for the support of the "vao" ( empty space ) .
(-- - The conservation of the roof had already been comp l eted in 1 9 8 6 .
An investigation into the stab i l ity o f the building was elaborat­
e d , beginning with the investigation and geological study of the
s o i l , in order to evaluate the support conditions , inspection and
charting of the damage , with the setting of "testemunhos " ( an
instrument used to register the deformation) on the wal l s . These
observations help to show the extent of the movement of the
structure .

At the same time , we planned and implemented drainage on the


posterior part of the building in order to conduct the surface
water to the sewage system . Repair o f the system for draining
rain water from the roof was also proposed and implemented . ( See
fig . 5 . )

After thi s work was completed , we began to observe the damage by


periodic inspection , to verify the condition of the deformations .

These steps were necessary to reduce the financial costs of the


intervention , by first el iminating the causes of damage and then
intervening to repair the consequences .

At present , we are e l aborating the structural reinforcement o f


the mud wal l and intend to prevent the reinforcement o f the
damaged masonry and in this way avoid its SUbstitution .

T� T T T r T T T TTTTTT T T T f
POINTS OF
STRHT
MO,5UREMEI'IT OF THE FOUNOA.TION
We know a l l the patterns of reinforcement are based on the
behaviour of the supporting structure . In the case of the mud
wal l s , the supporting structures have the fol l owing characteris­
tics : They are considered cyclopean structures , for the height­

�..;;;.-+- �P. ARENT


I
POINTS
DRILLED

DRAIN
UNDERGROUND
o I

[)\AIN AND

OF GUTTERS OOF
2 3

GUTTER
section relation of their pieces is l ow and so is its capacity to
absorb horizontal stresses . The spring coefficient and the
ducti l ity both have l ow values . Another element to be cons idered
is the l ow capacity of deformation in plastic regime , which means

'" CONDUCTORS

MAN
DAMAGED BRAYED MUD WALL
ON THE R

SORY
a material has only a l imited capacity to d i s sipate energy
without art iculating . As mud is a very fragi l e material , after
being damaged it is d i ff icult to analyze the static behaviour :
Fig . 5 : Ground floor plan each new e l ement presents d i fferentiated dimensions and condi ­
o f Our Lady o f P i l l ar . t ions o f support , producing secondary e fforts o n the critical
I l lustration by Lima . points .
3 06 Adobe 90

Thus , we consider that any proposal for reinforcement should take


into consideration the original state of tens ions , considering
the e l ements working togethe r . Also the finenes s o f the pieces
should be cons idered . This i s a basic fact for the qual i fication
of the "portante" ( load-bearing) capacity .

ymm
Some points of the proposed reinforcement are to be considered :
The continuity of forms and volume s , the s e trical d i stribution
of the elements , improving resistance to tensions , the density
and uniformity of the material s , the rigidness and duct i l ity o f
the sections , the capacity o f res istance compatible with the
e fforts and the l imited capacity of deformation .

We then proposed reinforcing and consol iding the wal l by f i l l ing


the empty spaces and areas o f loss with a mixture of s o i l -beton­
ite-l ime and reinforcing the empty spaces in order to support the
surchages without modifying the original structural system of the
wal l s .

The reinforcement involves instal l ation o f a parabol ic arch o f


discharge over the empty space , in both s ides o f the wal l and
additional arches d i sposed upon the former in order to red i stri­
bute the strength l ines near its original proceeding .

These arches w i l l be made of baked mud bricks with mortar with a


REINFORCEMENT
l - BRICK ARCH F OR BRAYED MUD W ALJ..
2-WOOD CF lJClUt fOR
STR E COM�ESSEO BRAYED MUD WAL.L mixture o f s o i l -cement .
3-AREA INJECTIOfiI '" ntE INJURIES

5-
I-W
STAUCT IJPPOfn'C# OF CONCR
.-REINfORC£M£NT

S
OOO
URE
ETE LWDER SWPOAT
REINFORCEMENT IN WOOO tiN THf:
a:SHAP£n£ "SOS9ORS"PCfiCH)
Clf" A The necess ity o f reinforcing the straps of the open space and
improving the conditions of support of the " impostas" of the arch
1-1- SF\IICEOFOF
"SCISSORS"
EMPTY
THE R OOf'
THE HIGH ALTAR
w i l l be evaluated on s ite .
- Vlf. Of AN I'Wt..IlMr
I
---INJURY ON THE on£R FACE
A moulding o f the empty space w i l l be executed in wood with a
rectangular c losed porch .
Fig . 6 : Brayed mud wal l
reinforcement proposal .
For the vertical sections that were crushed , a belt w i l l be used
I l lustration by Lima .
with wood p ieces on both s i des , properly l inked by transverse
pieces . ( See fig . 6 . )

Conclusion

As one can observe from the text above , the task o f preserving
the earthen architecture in Braz i l is sti l l at a very rudimentary
stage , without systematic studies showing the technical and
financi a l d i fficulties found in thi s area where the Government is
not tota l ly aware o f the need for a l arger investment in research
and l abor . In recent times technical - scientific studies have
begun to appear that support restoration work � thi s is developing
s l owly . Despite these problems , the work developed up to now has
been rewarding . Quite satis factory results can be obtained by
using s impl e technique s , making it possible to save the maj ority
of Bra z i l ian patrimony from ruin .
Consolidation Studies

Bib l iography

1 . F . M . Lea l , Restauracao e Conservacao de Monumentos Bra s il eir­


o s . Rec i fe : UFPe , 1 9 7 7 .

2 . Ministerio da Saude , Casas de Terra . RJ: SES P , 1 9 5 8 .

3 . S . de Vasconcelos , Arguitetura no Brasi l : S istemas Con­


strutivo s . BH : UFMG , 1 9 7 9 .

4 . L . Saia , Morada Paul ista . SP : Ed . Perspectiva , 1972 .

RJ: HAN
5 . A . A . M . Franc o , Desenvolvimento da Civil i z acao Material no
Brax i l . SP No . 1 1 , 1 9 4 4 .

6 . C . N . do Peru , Reports o f the International Committee on


Conservation o f Mud Bricks . Lima : Anai s do Congre ss o , october
1 9 8 0 e October 1 9 8 2 .

7 . G . crot i , Progetta z i one Struttural e e Consol idamento del l e


Constru z ioni . C6pia xerografica sem referencia .

8 . S . Mastrodicasa , Dissesti statice del l e Struture Edil ivie .


Milano : Editore U lrico Hoepl i , 1 9 7 8 .

9 . C . A . C . Lemos , Arguitetura Bra s i leira . Sao Pau lo , Edi90es Mel ­


horamentos , USP , 1 9 7 9 .
.'
Seismic Mitigation
"
Seismic Mitigation 3I
ABSTRACT COMPORTAMI ENTO ESTRUCTURAL DE LA QUINCHA

Quincha ( or reed b inding)


is a type o f construction
u s ed i n Peru in colonial
times and up to the f i rst
J . Bario l a , M . T inman , R . Ort i z , B . Alberc a ,
Pont i f ic i a Univers idad C a t 6 1 ica del Peru
APTO . 1 7 6 1 , L ima 1 0 0
J. Varga s

decades of thi s century . Peru


It i s made of c l ay ,
reed s , and timb e r ,
mate r i a l s with which high I n t roducc i6n
qual ity housing can b e
achieved and which have La quincha es un s i s tema cons truct ivo empleado durante l a colonia y
demonstrated good h a s t a las primeras decadas de e s t e s i g l o en el Peru . Los materi a l e s
res istance to us a d o s e n l a quincha s o n t ierra , madera y cana , c o n los cua l e s s e
earthquake s . The present logran e s t ructuras l iviana s . E n z o n a s urbana s , la quincha se u s 6
research , which comp r i s e s mayori t a r i amen t e como segundo p i s o , sobre u n primer pi so de adobe .
experimental and E s t e t ipo de e d i f i c aci6n brinda innumerables ven t a j a s como hab i t a t
analytical stud i e s , was e n l a c o s t a d e l Peru , p � r e l agradable ambiente que proporc iona , p�r
d i rected at expanding la seguridad que o frece durante t erremotos , y por su ba j o costo .
b a s i c knowledge on its Luego del s i smo de 1 7 4 6 , la construcc i6n de quincha se adopt6
structural behav i o r . The mas ivamente por el buen comportamiento observado . La Univers idad
conclu s ions of this study C a t 6 l i c a de l Peru y l a Techn i c a l Unive r s i t y of Nova Scot i a , Canada ,
have a l l owed for the con apoyo del I nt erna t ional Development Research Centre of Canada ,
propo s a l of s impl e rul e s inic iaron en 1 9 8 8 una inve s t igac i6n interd i s c i p l inaria con e l
o f d e s ign . These rul e s o b j et ivo d e promover e l u s o d e edificac iones d e quincha en zonas
can b e u s ed f o r the margina l e s de L ima . La inve s t igac i6n abarc6 desde e studios
conservation o f experimenta le s hasta e l di seno y construcc i6n de di ferentes obras .
h i storical monuments . E s t e traba j o presenta mayormente los aspectos basicos de ingenier1a
e s t ructural que sean de interes para un publ i c o amp l i o interesado en
la construcc i6n y conservac i6n de edi ficac iones de quincha .
KEYWORDS
Carac t e r 1 s t i c a s de la Quincha
Quincha ( reed b inding) ,
structure materia ls , S e describe a cont inuac i6n el t ipo de quincha usado en el proyec to
experiments , design , de inve s t igac i6n de l a Univers idad C a t 6 l i c a , hac iendo alus i6n a
s e i smic behav ior . ot ros t ipos de quincha t radiciona l .

La c iment ac i6n , usada en d i f erent es obras construidas como


parte del proyec t o , con s i s t e en una l o s a de concreto armada con cana
( Fig. 1). En las construcc iones t radic ionales de quincha se
encuentran c imentaciones de mamposter1a de ladri l lo 0 de piedra
a s ent ada con c a l .

Las paredes son formadas por paneles prefabricados de madera de


dimens iones 0 . 8 0 m de ancho y a l tura s de 2 a 2 . 4 0 m , los que se
c l avan entre s 1 . Los pane l e s con s i st en en un marco de madera con

co loca la
c o n barro .
estera,
travesafios como mue s t ra l a F i gura 1 . Una vez unidos los paneles se
t e j ido de canas chancada s , que luego se tarra j e a
L o s pan e l e s t ienen secc iones de madera d e 2 . 5 x 7 . 5 cm .
El espesor t erminado de los muros es 1 5 cm . Se usan pane l e s de dos
t ipos : pane les convencionales y paneles s 1 smicos ( F i gura 2 ) .

carrizo El
t ipo de cana que
( Arundo Donax ) , el
muy comun e l uso de bambu
y carrizo entero arqueado
se usa para l a s e s t eras y techo e s e l
cua l crece en l o s bordes d e l o s r 1 o s . Es
part ido y chancado ( Guadua Augus t i f o l i a )
sobre travesafios de madera .

El techo es de vigas de madera , y cobertura de c ana con mortero


de cemento . Se us an vigas de madera de torn i l l o espaciadas a 0 . 8m .
Las canas proporcionan un a i s lamiento adecuado , a pesar de su
reduc ido espesor . El mortero de cementa se apl ica en espesores del
orden de 2 cm , con 10 cua l se logra una cobertura l iviana , durab l e
y resi stente a la l luvia . La quincha tradic ional hace uso de l a
" torta d e barro " , una capa d e barro d e 5 a 1 0 cm d e espesor . La
F i gura 3 mues t ra una perspect iva de la e s t ruc tura .

Mat e r i a l e s

Como se ha mencionado , los m a t e r i a l e s princ ipa l e s d e l a quincha son


la t ierra , madera y cafia . Para la presente inve s t i gac i6n se us6 la
madera denominada " Torn i l l o " ( Cedre l inga Catenaeformi s ) , proveniente
del Oriente del Peru .

Se e s tudiaron di ferentes especies de cana y bambu (1),


obteniendose los va lores de la re s i s tenc ia a la t racc i6n . Por
e j emp lo para l a cana brava ( Ginerium Sagit t atum ) se observ6 una
resi st enc ia de l orden de 2 8 0 0 Kg/cm2 , seme j ante a la de a l gunos
aceros .
3 12 Adobe 90

Morteto-----7��
M�o--!----n�
Estera -------

Tar ajeo de barto----"""'I'!-:


Tar ajeo 2a. ca.,a------;--
Sobrecimiento------t'''' "-"',-<::--- Losa
F i gura 1 De t a l l e de Panel y C imiento

F i gura 2 T ipos de Panel e s


F i gura 3 E s t ructura de Quincha
Seismic Mitigation 313

La " pa j a " puede s er cua lqu i er f ibra vege t a l res i s tente y


duradera , p�r e j emp lo el ichu ( St ipa Ichu que ) crece en la s ierra del
Peru . Tamb ien es p o s i b l e usar cesped , pa j a de arroz , e t c . Es
conven iente que l a pa j a tenga f ibras de aproximadamente 10 cm de
l argo .

La t ierra es usada para el t arra j eo de l a s paredes en


comb inac i6n con arena grue sa y pa j a . El tarra j eo se a p l i c a en dos
capa s , cuyo espesor debe ser en 10 p o s i b l e 1 . 5 y 3 / 4 de cm . La capa
de l gada , l a cua l puede hacerse con arena gruesa 0 arena f ina ,
perm i t e cubrir l a s f i suras e irregul a ridades de la primera y darle
un acabado f ina l . Normalmente , l a mezc l a 6pt ima ( arena : t ierra )
e s t a entre 1 : 1 y 1 : 2 p�r volumen para suelos de mediana p l a s t ic idad
que son los mas corrientes . Se afiade una cant idad de pa j a
equiva lente a l 1 % del peso t o t a l seco . Cuando s e observa exce s iva
f i surac i6n , es necesario rea l i zar pruebas para lograr una mez c l a
adecuada . El porcenta j e que se afiada dependera de cuan arc i l losa
sea l a t ierra . La forma mas pra c t i c a de determinar e l porcenta j e de
arena gruesa adecuado es fabricar una serie de mez c l a s con
propo r c i ones en volumen , arena : t ierra , de 1 : 3 , 1 : 2 . 5 , 1 : 2 , 1 : 1 . 5 ,
1 : 1 , con 1% del peso seco en pa j a , t a rra j eandose paneles de prueba .
Debe observarse la traba j a b i l idad de la mez c l a y la f i suraci6n a l
secar , escogiendose la mez c l a con adecuada t raba j a b i l idad y que
perm i t a controlar la f i suraci6n ( entendida como s o l o la apar ic i6n de
f i suras f inas ) , con la menor cant idad de arena gruesa p o s i b l e . El
exceso d e arena grues a hace que l a mez c l a s e a muy deb i l y f a c i lmente
ero s i onabl e . Las Referenc i a s 2 - 6 , cont i enen d i f e rentes es tudios
sobre las e s t ructuras de t ierra . La Referen c i a 4 trata e l tema de
la durabi l idad de t arra j eo s de barro .

Para el techo se us6 (1) un recubrimiento de mortero


cemento : arena grues a en proporc i ones p�r volumen 1 : 6 y de 2 . 0 cm de
espesor .

Ensayos de Laborat o r i o

S e rea l i z aron di ferente s ensayos p a r a e s tudiar e l comportamiento de


l a s e s t ructuras de quincha ( 1 ) . A cont inua c i 6n se presenta un
resumen en e l que no se inc luyen det a l les y resultados tecnico s .

La c ompo s i c i6n de los t arra j eo s de l o s e spec imenes fue


mantenida en propo r c i ones aproximadas arena : t ierra de 1 : 2 . En
a l gunos c a s o s debido a l a variab i l i�ad de la p l a s t i c idad de la
t ierra , no fue p o s i b l e mantener l a mi sma proporc i6n . La pa j a usada
fue el ichu , en una cant idad i gual al 1% de l peso seco del materia l .
La cant idad de agua fue def inida de acuerdo a una t raba j a b i l idad
estandar medida con l a agu j a de Vicat ( 3 ) . Las paredes de quincha
fueron s imi lares a las descri tas anteriormen t e , excepto que el
espesor fue de 5 cm en lugar de 7 . 5 cm .

Se rea l i z aron di ferentes ensayos de pan e l e s para estudiar las


propiedades mecanicas b a s i c a s , t a les como su capac idad para res i s t ir
cargas ver t i c a les , l a s que son produc idas p�r e j emplo p�r e l peso
del techo . Las resi s tenc i a s obtenidas fueron bastante superiores a
l a s que se dan en construcc iones de 1 y 2 p i s o s , p�r 1 0 cua l e s t a
s o l i c i t a c i6n no se con s i dera c r i t ic a . D i ferentes acciones , p�r
e j emp lo el vient o , s o l ic itan a los muros en d irecc i6n perpendicular
a su p lano . Con l a f in a l idad de e s tudiar l a respue s t a de las
paredes de quincha ante este t ipo de cargas se rea l i z aron
exper imentos observandose que el espec imen era capaz de sufrir
deformac iones bast ante a l t a s , s in darse dafios importantes y s in
Carga perder res i s tenc i a . Adi c i ona lmente , se rea l i z aron tamb ien ensayos
p1
dinamicos con la mi sma f inal idad ; ver i f icandose nuevamente un buen
comportamiento frente a cargas perpendiculares al p lano de las
paredes .

Ensayos de Paredes bajo Carga s Latera l e s

E s t e ensayo s imul a la acc i6n d e los s i smos sobre las paredes


mediante l a ap l icac i6n lenta de c argas laterales . Se ensayaron 3
paredes de la forma mos t rada en la F i gura 4 , a las cua les se les
apl i c6 ademas de la c arga latera l , una c arga ver t i c a l de 2 0 0 0 N p�r
metro de ancho para s imular e l peso del techo . E l o b j e t ivo del

p3 �
�------�
Carga
ensayo fue e s tudiar e l comportamiento ante cargas latera l e s , y en
par t i cular determinar l a i n f l uenc ia de l a relaci6n ancho / a l tura de
l a s parede s . Los mec a n i smos de f a l l a observados se indican en l a
Tabla 1 . En todos l o s mecanismos de f a l l a se pudo aprec iar c omo
parte del fen6meno la separa c i 6n del t arra j eo y la es tera . Como se
aprec i a , l a resi stenc i a p�r metro de ancho varia en menos de un 2 0 %
para l o s tres espec imenes , p�r 1 0 cual puede cons iderarse para f ines
pra c t icos que permanece constante .
F i gura 4 Ensayo de Paredes
3 14 Adobe 90

Tabla 1 R e s i s tencia de Paredes ante Cargas Laterales

Espec imen Res is tenc ia Mecani smo de Fa l la


p�r m de
ancho
( N/m )

Pl 4 , 600 Volteo d e l a pared y separac i6n


estera- t arra j eo
P2 4 , 100 F i sura d i agonal y separaci6n estera-
tarra j eo
P3 3 , 90 0 F i sura horizontal y separa c i6n
e stera- tarra j eo

Ensayo de un M6dulo de Quincha

Se ensay6 un m6dulo de quincha ( F ig . 5 ) de 4 x 4 en p lanta y a l turas


de 2 . 40 y 2 . 80 . Todos l o s pane les fueron de 2 . 4 0 m del a l tura , p�r
10 que fue necesario un e l emen t o reticular de madera para lograr la
pendiente del techo . El techo del m6dulo tuvo un peso de
aproximadamente 80 Kg/m2 . El espesor de t odas l a s paredes fue de
aproximadamente 0 . 10 m . Los o b j e t ivos del ensayo eran (1)
desarro l la r model o s que permitan e s t imar la respue s t a , (2)
invest igar s i e l techo t raba j a c omo un diafragma r igido y ( 3 )
e s tudiar e l comportamiento global y de det a l le s de l a c on s t rucc i6n .
El e spec imen fue somet ido a tres corridas con di ferente s
a c e leraciones maximas d e l s i smo . El s i smo usado fue e l regist rado
en L ima en 1 9 7 0 .

La f a l l a ocur r i 6 cuando el despl a z amiento maximo ( n ivel del


t echo ) a lcanz6 un va lor equiva lente a l 0 . 5 % de l a a ltura del m6dulo .
La f a l l a se debi6 a l desprendimiento del barre de la e s tera , 1 0 cual
consti tuye un punto c r i t ico del s i s tema . La f a l la fue fragi l ,
reduc iendose instantaneamente la r e s i s tenc ia a un 5 0 % de la in ic ia l .
No hub ieron dafios de importancia en l a s paredes t ransvers a l e s a l
movimiento , e n parte porque e l techo c o n s t ituy6 una restricc i6n
e f e c t iva al desplazamiento de l a s paredes t ransver s a l e s . El
amo r t i guamiento medido del materi a l fue de 8 % , valor alto en
comparaci6n con o t ros mater i a l e s , 10 cua l e s una ven t a j a debido la
d i s ipac i6n de energia que e s to permite durante un s i smo .

F i gura 5 Ensayo del M6dulo

Dens idad de Paredes

E s deseable que los resultados experiment a l e s se t raduzcan en reglas


s impl e s de ana l i s i s . Dado que la r e s i s tenc ia de la e s t ructura es
proporc ional al area de l a secc i6n t ransve r s a l de l a s paredes
( espesor x ancho ) , e s posible expres a r los requerimientos de
res i s tenc ia en e s t o s t ermino s . E l anal i s i s s i smico se rea l i z a
corrien t emen t e e n d i recc iones horizonta l e s . La F i gura 6 mues tra
como e j empl o l a p lanta del m6dulo ensayado y defin iendose dos
d i re c c i one s , longi t udina l y t ransver s a l .

Con e l t ermino densidad de paredes


paredes dividida p�r e l area techada .
nos referiremos a l area de
Podemos c a lcular , entonces
una densidad de paredes en l a direcc i6n longitudinal y otra en la
Seismic Mitigation

d irecc i 6n transver s a l .

Por e j emp l o , para el m6dulo ensayado ( F ig . 6 ) , considerando la


direcc i6n longi t ud ina l , tenemos dos paredes de 4 m cada una . El

1-1'---( ·1 4 1'1 area de paredes e s 2 x 4m x 0 . 1 0m = 0 . 8 0 m2 en e s t a direcc i 6n .


S i endo el area t echada 1 6 m , 2 la densidad de paredes es 0 . 8 0 m2 / 1 6
m 2 = 0 . 0 5 6 el 5% . Con e s t a densidad de paredes , el m6dulo

4r1'1
T res i s t i 6 s i smos muy severos cuya proba b i l idad de ocurrenc ia es muy
<II>III
cdl- 1 . 6 1'1
L
remota .

En base a l o s resultados experimenta l e s y ana l i s i s pos teriores


I- Longitudinal ( 1 ) se pudo conc luir que una dens idad del orden de solo un 2 . 5 % , en
cada direcc i6n transvers a l y longitudina l , seria sufic iente para

1 0.10 1'1 J
T
res i s t ir s i smos con una ace lera c i 6n pico del 2 0 % de la gravedad ,

1 . 61
1'1 a c e l erac i6n de d i sefio usada en el Peru . Se determin6 ( 1 ) que de
e s t e valor , una cuarta parte deberia cons i s t i r en paneles s ismicos ,
de manera de garant i z ar la seguri dad en c a s o de un s i smo muy severo .
E s t o s valores son a p l icables s i empre y cuando l o s pesos unitarios de
los componentes sean seme j antes a l o s usados en este es tudio , es

ecciones para
Figura 6 eDil rAnaIisis dec ir , t echos de 80 Kg/m2 y parede s de 1 0 0 Kg/m2 •

Resumen y Conclus iones

E l e studio resumido en este documento fue orientado a inves t i gar


experiment a l y ana l i t icamente la respue s t a s i smica de la
cons trucc i6n de quincha . Los ensayos inc luyeron l a s imulac i6n dE
d i f erentes fen6menos t a l e s como l a acc i6n de l a s cargas de gravedad
y los s i smos . La es tructura de l a quincha permi te el traba j o
comp lementario d e t ierra y madera . E l tarra j eo de t i erra res i s t e
las fuerza s laterales producidas por fen6menos t a l e s como l o s
terremotos , mientras la madera res i s t e e l peso d e l a e d i f icac i6n .
Desafortunadamente , la cafia teniendo una gran res i s tencia no cumple
una func i6n es tructura l . Los resultados experiment a les demostraron
que para edif icac iones de 1 y 2 p i s o s , las cargas de gravedad no son
norma lmente cri t icas . Los ensayos dinam i c o s del m6dulo y los
ensayos de parede s c on cargas laterales perm i t ieron aprec iar que el
punto critico del s i s t ema es la adherenc ia barro - e s t era ,
produc iendose una f a l l a fra g i l en la cual se reduce bruscamente l a
res i s tenc ia . Por este mot ivo se ha propues t o el uso de pane l e s
s i smicos como u n a medida adic ional d e segur idad . La forma de
ver i f icar la segur idad s i smica de la construc c i 6n propues t a es en
terminos de la densidad de paredes , e s t a b lec ida como un 2 . 5 % del
area techada en cada d irecc i6n de ana l i s i s .

Referenc i a s

1. J . Bario l a , "Quincha Construct ion , " F i n a l Report to the


I nterna t ional Deve lopment Research Centre of Canada , Departamento
de I ngenieria , Pont i f i c i a Unive r s i dad C a t 6 1 i c a del Peru , 2 4 0 pp . ,
1990 .

2. J . Bario la , " Dynamic S t ab i l i t y of Adobe Wa l ls , " Doc toral


D i s sertat ion , Department o f C i v i l Engineering , Univer s i t y of
I l l in o i s a t Urbana-Champa ign , 1 9 8 6 .

3. J . Varga s , J . Bario la , M . B l ondet , P . K . Mehta , " S e i smic


S t rength o f Adobe Masonry , " R I LEM , I nterna t ional Union of Test ing
and Research Labora tories for Mater i a l s and S t ructures , Vo l . 1 9 ,
No . 1 1 2 , Pari s , France .

4. E . Heredia , J . Bari o l a , J . Varga s , P . K . Mehta , " Improving the


Mo i s ture Resi stance o f Adobe S t ructures , " R I LEM , Interna t iona l
Union of Test ing and Research Laboratories for Mater i a l s and
S t ructures , Vol . 2 1 , pp . 2 1 3 - 2 2 1 , Paris , France , 1 9 88 .

5. J . Bari o la , J . Varga s , D . Torrealva , G . O t t a z z i , " Earthquake -


Res i s tant Prov i s ions for Adobe Construc t ion in Peru , " 9 th World
Conference on Earthquake Engineering , Tokyo -Kyo t o , Japan , Augu s t
1 98 8 .

6. J . Bario l a , M . A . S ozen , " S e i smic T e s t s of Adobe Wa l l s " ,


EARTHQUAKE SPECTRA , Earthquake Engineering Research I n s t i tute ,
Berke ley , C a l iforn i a , 1 9 9 0 .
Adobe 90

ABSTRACT

This paper synthesizes


LA ARQU ITECTURA DE QUI TO FRENTE A LOS S I SMOS .

Ines del P ino* , Patrick De Sutter y Mar i o Moran P .


research on seismic inter­ Inst ituto Nacional de Patrimonio Cultural
ventions carried out on Cuenca 9 1 8 y Me j ia
rel igious monument s in the Quito - Ecuador
h i storic center of Qu ito
over the last two years .
This study addresses struc­
tural diagno s i s , ana lys is
and evaluat ion of estab­
Obietivo
S intet izar las exper ienc ias obtenidas en e l manej o de tecnicas
l ished methods for re in­ de conservac i6n y reforzamiento en los monumentos co loniales
forcement and consol idation re l igiosos de Quito afectados por s i smos .
of structure s . I t briefly
describes two case stud­
ies .
Marco referencial para e1 estudio
1. Aspectos geotect6nicos referentes a la s i smicidad regional :
The structural inter­
vent ion was aimed at res­
toring the rubble-work ,
wh ich had deteriorated due
0)
El Ecuador es un pa i s andino ubicado en e l continente Sud Ameri­
cano ( Quito , 0 . 180 S - 78 . 500 cuyo tectonismo esta definido
per los mov imientos de las placas mar i t ima de Nazca y continental
Sud Americana que , al chocar , forman una zona de subducc i6n in­
to earthquake s , by using terplaca ( Ver graf . 1 , 2 ) . La geomorfo logia andina y particular­
materials and techno logy mente la de Quito , esta caracteri zada por un volcani smo activo y
that were consistent with e l tectonismo de las formac iones de l Holoceno const ituidas per
the or iginal construc t ion . dep6s itos aluvial , glac ial , lagunal , cenizas volcanicas y secuen­
c ias sedimentarias de l P le istoceno ( Ver graf . 3 , 4 ) [ � ] .

KEYW ORDS La zona de subducci6n interplac a , e l tectonismo andino y e l


volcan i smo const ituyen las principales areas fuentes d e l o s s i s­
mos ( Ver graf . 5 , 6 ) [ 2 ] [ 3 ] .

" STRUCTURA
" H I STORIC CENTER OF QU I TO " ,
L I NTERVENT I ON " ,
" EARTHQUAKE " , " SE I SM I C IN­
TERVENTION " , " ADOBE " .
TECTON/SMOSUDAMERI
DE CSUBDUCCI
ANO ON MAPA GEOLOGICOEO PIOECHINLACHAPROVINCIA

•• I'1'.AU.ALU e.o'_T�
_ SOtANOAH()INAJtlJ/GARE1U.

S> PLACA NAZCA


TECTON/SMO DE SUBDUCCION
FUENTES DE GRAF I COS

Graf . 1 y Graf . 2 : compen­ 2. Rie sgo s i smico para monumentos h i st6rico s : E l rie sgo s i smi­
diD de varias publ icaciones co e s medido en func i6n de la maxima intensidad s i smica ( Mercal l i
sobre Tectonismo Andino . Modificada-MM ) que e l monumento h a resist ido durante su vida ut i l
sin l legar a l colapso [ 4 ] ; e sta intens idad const ituye e l s i smo de
Graf . 3 y Graf . 4 : Mapa comprobaci6n y so lamente sera aplicable para cada monumento cuyo
geo16gico de la Prov inc ia periodo de observac i6n sea mayor de 300 afios .
de P ichincha . Ministerio de
Energia y Minas , 1985

Graf . 5 : E laboraci6n en
MAPIlPROVIOE NCIFALLAMI
A OE PIECNTOS
HINCHAOE 'LA SISTEMATIZACION OE LOS FALLAMIENTOS I
base al " Mapa s i smotect6ni­
co de l Ecuado r " , INECEL y
al " Mapa Hidrogoe 16gico de
la Provinc ia de Pich incha " ,
Mini sterio de Energia y
Minas , 1985 .

Graf . 6: Estudio micro s i s­


mico de l val le internadino
entre Latacunga y Guay l la­
bamba . Minard Hal l , Escue la
Po l itecnica Nacional , 1978 .

* Autor a qu ien debe ser dirig ida la correspondenc ia .


Seismic Mitigation

La de finic i6n del s i smo de comparaci6n en terminos de la


acelerac i6n permite di senar e l reforzamiento l imite adecuado para
e l monumento danado p�r este .

3. Microzonificac i6n s i smica de Quito : Las condiciones dinami­


cas de l sue lo de fundac i6n determinan otro parametro , pue s una
re spue sta inadecuada de la interacc i6n sue lo-e structura produce
un riesgo que debera ser moderado con una intervenc i6n e structu­
ral aprop iada .

El proyecto de microzonificaci6n s i smica para e l Centro


Hist6rico de Quito permite conocer parametros dinamicos de l sue lo
local ( Ver graf . 7 ) .

PROYECTO OC MI C ROZONI F ICACI O


PARA EL CENTRO HISTORICO DE QUITO N SI S MI C A

ZONA 2.2.55 0.0.Tz46


Para Consfrucciones Historicos 0=2

T,
ESPECTRO DE RES PUESTA GRAF. 7
4. Valor arquitect6nico del centro hist6rico de Quito : Existen
dos razones que determinan el valor de Centro H i st6rico :

4. 1 En el periodo co lonial ( 1534- 1810 ) , Quito fue la c iudad de


la regi6n Norte de l pai s menDs afectada por e l impacto s i smico
ya que , en el mi smo periodo , c iudades vecinas debieron ser re­
construidas 0 reub icadas .

4.2 La unidad e spac ial y formal de l Centro H i st6rico de Quito


se mant iene a 10 largo de 400 anos y constituye una de las pocas
capitales de Latinoamerica que posee un centro de tal magnitud y
caracterist icas f i s icas .

5. Vulnerab i l idad s i smica de los monumentos hist6ricos cons­


truidos en t ierra : En general , los conventos e iglesias de Qu ito .
se construyeron or ig inalmente en adobe . Paulat inamente e st o s
monumentos se reconstruyeron en ladr i l lo . En monaster iDS de
monjas se dieron los mismos cambios pero manteniendo e l adobe en
los c l austros y cerramientos perimetrale s .

Las construcc iones h i st6ricas de adobe presentan alturas max1mas


de 10 metros , con muro s cuyos espesores promedio son de 1 . 20 mt �
en planta baj a y 1 . 00 mt . en planta alta .

Las t ipologias de las construcc iones de adobe , que han demo strado
su resistenc ia y estab i l idad , se caracterizan por tener valores
baj o s , inferiores a 5 para la relac i 6n altura- e spesor . Cuando
esta relac i6n t iene valores superiores a 5 , su estab i l idad frente
a los s i smos d i sminuye en ausenc ia de s i stemas de arriostramiento
lateral .

Lo expuesto demue stra que formas esbeltas como torre s , muros s in


arrio stramiento y s i stemas abovedados no han resist ido los s i smos
de intensidades mayores que V ( MM ) , 10 que j ustifica la existen-
Adobe 90

c i a de torres de ladr i l l o y reforzamiento de los muros con ladri-


1 10 , piedra y madera para mej orar su resistenc i a .

Reconocimiento
sismica de las tecnologias tradiciQnales de preyenci6n
1. Tipo logias arquitectonicas constructivas de l centro hi stori­
co de Quito , procesos construct ivos y materiale s : El examen en
muros de 17 monumentos hi storicos coloniales a 10 largo de 22
B l ames de importanc ia ha dado como resultado las s iguientes con­
c lusione s :

1.1 La total idad de los monumentos re l igiosos conservan aun


muros de adobe , muros mixtos adobe- ladr i l lo y solo en dos casos
muros de bahareque loc a l i zados en planta alta .

1.2 En todos los casos , las cubiertas son de madera y tej a con
predominio de una e structura t ipo rey y de l t ipo dos aguas adosa­
das .

1.3 La relac ion de los muros , en alto y ancho calcul ado por
pisos , es la siguiente : en planta baj a , la re lacion es de 3 . 35 y
en planta alta es de 4 . 27 , con excepc ion de l monasterio de Santa
Catal ina que t iene una re lac ion promedio de 6 . 0 .

1.4 Las dimens iones mas general izadas para e l adobe co lonial van
desde 44 a 46 cm . de largo por 22 y 23 cm de ancho y entre 12 y
13 cm de espesor . Los adobes produc idos en el presente siglo
t i enen una dimensi6n genera l i z ada de 40 x 20 x 1 0 cm . Mientras
mas ant iguo e s el adobe , mayores son las dimensione s . La perdida
de componentes organicos que son la paj a y e l estiercol e s pau la­
t ina con el t iempo . Los adobes actua l e s produc idos en la zona
c i rcundante a Quito e l iminan los ingredientes organico s , reducen
dimensiones y desmej oran · su cal idad .

1.5 Todos los muros estudiados son muros soportantes de adobe ;


se encontraron ej emplos de arcos de descarga y de b6vedas primi­
t ivas unicamente en el Hospital San Juan de Dios e Ve r graf . 8 , 9 ) .

$1!lTCItlA DE au.

J:
Graf . 8 : Propuesta de con­
finamiento para el Hospital
t.J.. t .J..
" San Juan de Dios " de Qu i­ SISTEMAS TRADICIONALES SISMORESISTENTES MODELO
ANAL/SIS ESTRUCTURAL PARA EL
DE LA ESTABILIDAD.
to , e l aborada por Ing .
Mario Moran . GRAFICO No.8
Patologias presentesestructural
de comportamiento en las tipologias constructiyastradlciooal
en la arauitectura V sistemas
Los s i smos ocurr idos en la co lonia dieron paso a la ampl iac ion de
igles ias y conventos y a su reconstrucc i6n parc ial 0 tota l , razon
par la que en la actua l idad se encuentran muros mixtos con rel le­
nos de ladr i l lo y piedra , que se unen con morteros de barre
cal -arena , perdiendo su homogene idad .
de 0
Los dafios s i smicos , corre sponden a una interacc ion entre las
partes del muro y los s istemas de entrepiso y l a estructura de
cubierta .

Criterios
solicitadasdeporanalisis
sismosestructural para construcciones en tierra
1. Determinac ion del s i smo de comprobaci6n para e l ana l i s i s de
la edificac ion : Este tema constituye la principal preocupac ion
para la determinac ion de l nive l de intervencion e structura l .
Seismic Mitigation 3 19

En los estudios real izados en Quito se ha ut i l izado l a s i guiente


metodo l o g i a .

a. Clasificac i6n de las intensidades s i smicas registradas que


han afectado al monumento .

b. Determinac i6n del area fuente de los s i smos que han causado
danos a la edi ficac i6n .

c. Calculo de l a intensidad s i smica en e l monumento :

I = 8 . 1871 + 0 . 938 M - 1 . 759 Ln ( R+40 ) [5]


d.
M
R == Magnitud ( MS K )
Distanc i a hipocentral (Km)
Estudio crit ico comparat ivo de las intensidades con los
danos causados en el monumento , para e stablecer el s i smo de
comparaci6n que servira para los ana l i s i s de estab i l idad .

2. Mode laj e de los empuj e s s i smicos y cargas gravitac ionales :

EL EVA CION 2 . 1 . Empuj e s s i smico s : La aceleraci6n se calcula mediante l a


siguiente ecuaci6n d e atenuac i6n para Quito :

Ln a = 295 . 2 14 + 0 . 6479 M - 1 . 2 1 5 1 Ln ( R+40 )

E l periodo fundamental del monumento e n t ierra , por su gran


[6]
vo lumen y baj a altura , se 10 podr ia considerar simi lar al periodo
predominante de l sue lo , ya que conjuntamente con el muro cons­
t ituirian un cuerpo continuo .
--------,
-- ���
--
---
:
J
I
I
El empuje s i smico , como fue rza de inerc i a , se ha calculado
ut i l iz ando las formas espectrales propue stas en e l proyecto de
microzonificaci6n s i smica para e l Centro H i et6rico y la ace­ [7],
_ ______

_ __
__ __ ....
--.J I..I ..
I
leraci6n determinada para el s i smo de comprobac i6n .

2.2 Cargas grav itaciona les : Const ituyen e l peso propio de los
I materiales que conforman el muro y las cargas de ut i l i zaci6n como
I

-l-jI I s istemas de entrepisos , tab i quer ia y cargas vivas , transmit idas


c omo carga uniforme d i stribuida en toda la longitud del muro .

3. Determinac i6n de los si stemas resistentes y definic i6n de l


I modele estructural de ana l is i s : De l ana l i s i s de las geometr ias
en planta y e levaci6n de l monumento y de los l ineamientos de las
fisuraciones produc idas por los s i smos , se determina en cada caso
part icular , los s i stemas estab i l i z antes y s i smorres i stentes de la
edificac i6n los cua les son modelados eetructuralmente para e l
ana l i s i s .

: .
CORTE
£SCALA 2 _ 402
___
GRAFICO No. 9
4.
zos :
Metodos de ana l i s i s estructural y determinaci6n de esfuer­
Las caracterist icas mecanicas· part iculares de los mampues­
tos y de los muros , conj untamente con los problemas de comporta­
miento de l a f i surac i6n y micro fisuraci6n condic ionan la neceei­
Hospital " San Juan de Dios " dad de recurrir a metodoe de ana l i s i e consistentes . Se ha ut i l i­
de Quito . I ntervenci6n zado genera lmente metodos estat icos de calculo , apoyados en l a
e structural en uno de los res istenc ia d e materiales , ap l icados sobre secciones unitarias de
muros de adobe . los e lementos .

5. D iagn6stico y formulac i6n de la teoria de l comportamiento


estructural de l monumento : La comparaci6n cuant itat iva y cua­
l itat iva entre la resistencia d i spon ible del muro y los esfuerzos
mayorado s , calculados por efecto de las cargas gravitac ionales y
s i smicas , permiten formu lar la teoria del comportamiento estruc­
tural sobre la cual se bas a el proyecto de reforzamiento .

Desarrol
i. lo de tecnicas de mejoramiento
Tecnicas de Reforzamiento :
estructural Las tecnicas ut i l izadas son de
dos t ipos :

1.1. Reforzamiento de la mamposter ia mediante forramientos su­


perficiales de mal la metalica recubierta con mortero y anc l ada a
la mamposteria con micropi lotes corto e .

1.2. Reut i l izac i6n d e las tecnicas tradic ionales de arriostra­


miento talee como entrepisoe , l laves y so lerae de madera , ayuda­
dos por t i rantes de acero co locados a nive l de entrepiso y cu­
bierta .

2. Tecnicas de Coneo l idaci6n : Lae tecnicae con fina l i dad de


consol idaci6n ut i l izadas en los monumentos consisten en l a rest i­
tuci6n de su cont inuidad perdida por los agrietamientos 0 inter-
3 20 Adobe 90

venciones provocadas p�r el cambio de uso , mediante l a co locaci6n


de microp i lotes inyectados en la mamposter ia .

3. Tecnicas de Mantenimiento :
tales como l a humedad , erosion , sobreut i l izaci6n 0
Los muros danados per agentes
intervenc iones
inadecuadas fueron reparadas e l iminando l a causa de los danos y
l iberando los materiales incompat ibies con la naturaleza de l muro
de t ierra para luego recuperarlo mediante tecnicas trad i c ionales
de protecc i6n y reposic ion puntual de fa ltantes , en muchos casos
a l igerando las cargas mue rtas innecesarias y planificando un uso
compat ible con la res i stenc i a y estab i l i dad d i spon ibles en la
obra murar i a .

Casos de Intervenci6n
Hospital San Juan de Dios : Esta intervenc i6n se centra en e l
muro Sur de l hospital ( Ver graf . 9 ) , destruido por efectos de
enve j e c imiento , sobreut i l i zac i6n y perdida de los mecani smos
e stab i l izantes s i smorre s i stentes .

Este muro de adobe data de 1564 , es posiblemente e l mas


ant i guo de Quito y tinico en su forma , por los nichos abovedados
en planta baj a .

Las condiciones pato logicas del muro determinaron e l d iseno


de una intervenc i6n estructural cuyo obj et ivo es el " conf inamien­
to armado " de las masas de adobe agrietadas y fisuradas para que
por medio de inyecc iones de morteros de terrocemento recuperen l a
cont inuidad y la resistenc i a adecuada para obtener una estab i l i­
dad con una seguridad consistente con la c ruj i a de la cual e l
muro e s parte .

El di seno de intervenc i6n se complet6 con propue stas de


confinamiento de l sue lo de fundac ion y l a repo s i c i 6n de los meca­
n i smos de arriostramiento s i smorresi stentes origina le s .

Cap i l l a d e l Robo : En e ste monumento , cobra importanc ia l a con­


servaci6n de la e spadana de l front ispicio de la Capi l l a , por sus
condiciones de gran esbe ltez y la construcc i6n mixta con dos
materiales : ladr i l l o- adobe . { Ver graf . 10 )

Graf . 10 : estado actual y


propue sta de reforzamiento
y consol idacion estructural
de la Cap i l la del Robo .

Levantamiento rea l i zado por


e l Instituto Nacional de
F/lCHAOA CAPI;'LLA DEL ROBO
Patrimonio Cultural .

La masa sobre saliente de l a e spadana , per su esbe l t e z , e s un


e l emento vulnerable a l s i smo ; las vibraciones y desplazamientos
laterales produc idos afectaron a l muro inferior de adobe , agrie­
tando lo y f i surando l o con l a consecuente perdida de resi stenc i a .

E l estudio e structural demo stro l a neces idad d e amort iguar y


arr iostrar la e spadana a fin de contro lar la re spuesta s i smica de
este e l emento . Para e l lo se d iseno un si stema de tensores ver-
Seismic Mitigation 3 21

t icales de arriostramiento anc lados a dados de hormig6n ubicados


en el piso , que cumplen doble func i6n : por una parte , controlar
el vo lcamiento , y por otra , amortiguar las vibracione s .

La resistenc i a del muro de adobe fue mej orada mediante e l


uso d e micropi lotes d e hierro , inyectados e n l a mamposter ia y e l
revestimiento de esta con una armadura de mal la metal ica adher i da
a los micropi lote s .

Conclusi6n
El obj et ivo de la intervenc i6n estructural en los edificios his­
t6ricos de Quito es la de garant izar su estab i l idad , sin preten­
der un reforzamiento mas a l l a del resist ido por el monumento a 10
largo de l t i empo .

La estab i l idad s i smica de los monumentos co loniales cons­


truidos en t ierra depende de los mecani smos de arrio stramiento en
madera proporc ionados por los sistemas de entrepiso y cubierta .

Los test imonios actuales ponen en evidenc ia que formas es­


t ructurales complementarias a los muro s , tales como arcos , b6ve­
das y cupulas no fueron concebidas en t ierra sino en otros mate­
r iales como ladr i l lo y p6mez . Estas formas fueron importadas
indistintamente desde zonas no s i smicas . Las torres han s ido , a
10 largo del t i empo , los e lementos arquitectonicos mas afectados
por los mov imientos s i smicos .

Los e studios de geofisica para e l centro historico de Qu ito


han revelado parametros dinamicos no compatibles con este t ipo
de edi ficacione s , 10 que obl iga a conservar los mecan i smos tradi­
c ionales de protecc ion s i smi c a .

L a recuperacion d e la resi stenc ia perdida por causas s i smi­


cas , de enve j e c imiento , sobreut i l izac i6n y falta de mantenimiento
de los muro s de adobe se ha practicado con la apl icac ion de tec­
nicas de reforzamiento para otros materiales que se han adapt ado
para las obras de restaurac ion en adobe .

El presente estudio pos ib i l ita e l desarro llo de futuras


invest igaciones relac ionadas con e l riesgo s i smico y e l comporta­
miento estructural de las edif icac iones h i storicas .

Notas bibl iograficas :

1. Direcc ion General de Geo logia y Minas de l Ministerio de

Investigacion
,Instrumentacjonde losSismica
Recursos Naturales de la Provin­
Energia y Minas del Ecuador , " Mapa Geo 16gico de la Prov inc ia de

cia de0.Pjchincha
P ichincha , "
( 1985 ) .

2.
ESPOL ,
Lara ,
1985 ) , 7 - 16.
del Ecuador ( Guayaqu i l :

3. M . Hal l , P . Ramon , " Estudio M i cros i smico de l Va l l e Interan­


dina entre Latacunga y Guay l labamba" ( Estudio Realizado para la
Direccion Nacional de Defensa Civ i l , Quito , julio , 1978 ) , 1 - 27 .

Estatjco
4.
v de Como Reglamentar
Proteccj6n
A . Giuffre ,

cjos de Inter9s HjstOrjcoSismica las


de Intervenc
Centros jones
Urbanos Vdel
de Restauro
Edjfj­
( Buenos Aire s : Comite Latinoamericano
de Estructuras , Comis ion CLAES- S I SMO , 1987 ) .

5. J . Palac io , " Determinacion de los N ive les Esperados de Ace­


leracion en el Pai s " ( Trabaj o presentado en el I I Encuentro Na­
c ional de I ngenieria Estructural , Cuenca-Ecuador , mayo 1 4 , 1987 ) .

6. Ibid . , 4.

7. M . Moran , " Propuesta de Microzonificacion S i smica para e l


Centro H i stor ico de Qu ito " ( Var ios e studios puntuale s , Quito ;
octubre , 1989 ) .
322 Adobe 90

ABSTRACT LE BAT I ANC I EN DANS LES ZONES A R I SQUES

Patr i c k de M a i so n n e u v e
Research on anti seismic
construction is
increasing and finding
concrete appl ications .
It befits us to look with
8718 000
ar c h i t e c t e -p r o fesseur
av . Sa i n t

F l-· a n c e
L i moges
Sur i n

attention and modesty at


La � e90n du p ass�
the works bequeathed to
us by architects of
Avant d ' avo i r un certa i n regard sur Ie b�ti anc i en dans
antiquity . They
l ' an t i qu i t � , i l est i mp o r t an t que nous d i s t i ng u i ons l es
respected earthquake­
d i ff�r e n t s b � t i s .11 e x i st e des fam i l I e s de b�t i qui ont des
resistant rules of
c arac t�r i s t i ques p ar t i c u l i �res se l o n l es m a t �r i au x qui
construction . S ince that
compo sent l es murs et par c ons�quent qui r�ag i ssent
time , we have invented
d i ff�remme n t . Pour cette r a i so n , i l faut que nous
l ittle and forgotten
appr�hend i ons , l es b�t is en fonc t i ons de l eur fam i l l e : l e b�ti
much .
e n pans de bois et Ie b�t i e n p i erre par e x emp l e , r�ag i ssent
d i ff�remment , v i e i l l i ssent d i ff�remmen t , so n t �ventue l l ement
sUjets � des ma l ad i es d i ff�rentes et l ' on vo i t d�j� lA une
r a i son pour sp�c i f i er l es i n terven t i ons sur un b�ti se l o n sa
nature . I l faut t o u j o u r s avo i r p r �sent A l ' espr i t c es
d i ff�r entes fam i l i es de b�t i .
b�ti en p i er res de tai l l e
b�t i en pans de bois
b�t i en terre c r ue
b�t i en terre cui te
b�t i c o mp o s i te

Chacune de ces fam i l i es se l o n qu ' e l le est c o ns t i tu�e de


m a i sons i so l �es ou en b l oc s aur a des r�ac t i o ns qu ' i l nous
faut essayer de c c.mprend r e .

Avo i r u n r eg ar d sur les t ec h n i q ues ant i ques de c o n s t r uc t i o n


en � t ud i an t l eu r s ves t i g e s , pour un no n a r c h �o l ogu� , c ' es t lui
perme t t r e de m i eu x c o mp r end r e l es tec h n i q ues du b�t i anc i en
sur l eque l il travai l l e . Les d i ff�rentes r e c h er c h e s no us
perme t te n t d ' aff i r mel- , p eu t -� t l- e avec pl-�somp t i on ma i s avec
une �v i de n t e s i nc �r i t� , que les t�mo i g nages du b�ti s i mp l e
t r ad i t i o nn e l de l ' an t i q u i t � � t a i en t d ' un n i veau
tec h no l og i q ue auss i avanc� que Ie nStre p a r fo i s pens� de
man i �r e p l us i n t e l l i gen t e . Nous savons par la t r ad i t i o n
�c r i t e , l e souc i des anc i ens de c o n st r u i r e en respec t a n t l es
r �g l es s i sm i ques . Les t r avau >: du p r o fesseur Bruno Hel l y d i se n t
ma i n t e s et ma i n t es fo i s " D i eu l es ava i en t pun i car i ls
n ' av a i e n t pas respec t � l es r �g l es . " M a i s q ue l l es so nt e l l es ?
Pour r�pondre � cette ques t i o n , i l nous f a u t traverser l es
s i tes et l es regarder avec un oe i l d i ff�ren t .
A Pompe i , nous d�couvrons gr�ce au tr avaux de Jean P i er r e Adam
l es i n terven t i ons de r e s t aur a t i ons des Roma i ns �·p res Ie
t remb l ement de terre d e 51 ap J.C. et avant Ie recouvrement
en 79 ap J.C.
Dans Ie s i te de D� l o s en Gr�c e , l es ves t i ges des
soubassements no us � t o nnen t . Par e>:einp l e , l es m a i sons qui se
t r o uven t
r emarquab l e
de mur s
p i er res

fo r t
la
POLl-l
proche
ses
p�r i ph�r i e
mo sal ques ) nous
� t o nnantes . Te l
i mp o r t a n t s avec , en t r e
mur
de

est
eux
la " ma i so n
montrent
c o mpos�
, de
de
des
des masques" (
c o mp o s i t i o ns
b l oc s
p e t i t e s p i er res q u i
de

par l eur pr�senc e , l eur nomb r e et l eur p o l i ssage ont de toute


�v i dence un rS le i mp o r t an t . Ce syst �me a n t i s i sm i que est
c omparab l e � celui des i l es vo i s i nes c o mme l ' ile de
Ser i ph c. s , l es ma i sons d u moyen-age de 1 433 , d e m � m e que l es
ma i sons p l us r � c entes de 1 885 sont c o mpos�es avec ce m@me
p r o c �d � , avec cependant une d i ff�rence ,� savo i r que l es b l oc s
sont souvent p l us pet i t s et mo i ns so i gn�s . Ma i s de toute
� v i denc e ,si u n e ma i so n de 1 433 n o u s e s t p arvenue , cette
m a i so n a pu sup p o r t e r b i en d e s t remb l emen t s d e terre . 5 i nous
r e s t o ns � D�los no us pouvons vo i r la enc o r e des c h o ses tr�s
I nt �ressantes . Te l mur cans l eq u e l l es p i er r e s s ' i mb r i quen t
l es unes dans l es a u t r es c o mme si e l l es se souda i en t e n t r e­
e l l es . Ce t t e c o mp o s i t i o n de mur no us la r e t r o uvons de l a me-me
man i �r � dans Ie s i te anc i en de la viei l le t h �r a , dans l ' ile de
Santor i n . Nous t r o uvons auss i dans Ie s i te d e D� l os des
b l oc s de p i er r e de g r andes l ongueurs . I ls l i a i sonnement deux
mur s ou ang l es de murs entre eux ou deux· m a i sons entre e l l es .
Ce t te t ec h n i que , nous la r e t r o uvons d ans l es s i tes des
c c. m p t c. i r s ven i t i ens , ma i s auss i dans l es . ma i sons
t r ad i t i o nne l l es de l ' ile d ' An d r o s en Gr�ce . Enf i n au Th�� t r e
de D�los , l es murs qui sou t i ennent les g r ad i ns so�t p e n s � s de
man i ere t r �s � t onnan t e . Un s i te c o mme D� l os est r i che de
remarques qu ' i l faudr a i t a p p r o fond i r et r a p p r o c h er de b i en
d ' au t r e s s i tes . I L e x i s te une fou l e de t ec h n i ques , t o u tes ces
tech n i ques
c o mmen90ns A
ne sont
e n t r evo i r
q u ' une p e t i te p ar t i e de ce que nous
Seismic Mitigation 3 23

Les b � t i s en pans de bois : c ' es t dans l es s i tes


arch�o l og i ques d ' Her c u l anum ou de P o mp e i que nous t r o uvons
des e x emp l es .
B i en que V i t r uve , dans son l i vre , n ' ac c o r d e que peu d ' i n t�r.t
aux ma i sons en p a n s de bo i s , au regard des � t udes q u ' o n peut
mener SLtr l es s i tes , de nombreLt >: � l �ments nous permet tent
d ' af f i rmer que l es pans de bo i s � t a i ent beaucoup p l us u t i l i s�
qu ' on ne Ie pense . A Her c u l anum , l es pans de bois � ta i en t tr�s
u t i l i s�s PO Ltl- l es � tages . No tr e r egard se por tera sur une
ma i so n que Ma i u r i Amadeo a fo u i l l � l a " Casa Gr a t i c i o " . Ce t t e
ma i so n ent i �rement e n pans de bois n e do i t p a s � t o nner dans
un tel s i te an t i que dont l es R . CH . so n t en p i er r e . Nous a l l o ns
essayer de c o m p r endre c e t te m a i so n d ans sa p r �sen t a t i on
ac tue l l e et dans l es d i ff�rentes r�hab i l i ta t i ons par l es
a r c h �o l ogues . Ma l g r� l es tec h n i q ues emp l o y�es ( Ie b�ton arm�
en p l ancher . . ) nous pouvons enc o r e a U j ourd ' h u i tr�s b i en
c o m p r end r e la techn i c i t� du pans de bois � 1 ' �poque r o ma i ne
et la r e trouver dans de nombreux b�t i s anc i ens e x i stants
a U j o u r d ' hu i . Ce qu i nous permet de m i eu x comprendre l es
p i �ces du p u z z l es manquan t e s . Les pans de bois de cette ma i so n
� t a i en t pens�s e t mont�s c o mme ceux que mo n t a i e n t l es
c h ar p e n t i er s du moyen-age et des �pc'ques su i vantes . Ces
c h ar p en t i er s qui c o nna i ss a i en t la fabr i c a t i o n des nav i r es
, sava i en t c o mment p e r me t t r e � une m a i so n en bois de se
s t ab i l i ser entre deu x m a i sons en p i el-res . En essayant de
c o mp r en d r e la m a i so n avec Va l er i o Papacc i o ar c h i tec te du s i te
ant i q ue nous pensons que la Casa Gr a t i c c i o qui est c o ns t r u i te
entre deux ma i sons , samn i tes , u t i l i se l ' espace du J ar d i n ou des
d�pendances d ' une de c es deux ma i sons , et a �t� constru i te
par un des deux propr i � t a i res so i t pour .tre l o u�e so i t
, c omme o n I e pense � la su i te d ' un t r emb l ement de terre
. Ce t t e ma i so n ne c o mp o r t e pas de murs p i gnons en p ans de bois
. Tous l es � l �ments p o r teurs ( p l ancher s ) reposen t sur des
p i l i er s en b r i ques . • Ceux-c i sont d�so l i d ar i s�s des murs des
m a i sons vo i s i nes . Nous ver r o ns d ans la t r o i s i �me p ar t i e de ce
d o c umen t qu"� la su i te d ' un t r emb l ement de terre il ne faut
j ama i s l a i sser un espace v i de entre deux ma i sons qu i ont
souffer t . Tr � s souvent l o r sque u n e ma i so n s ' es t
effondr�e , nous , l es ar c h i tec tes , ve i l l o ns � c o nso l i der l es
m a i sons vo i s i nes en sou tenant l es murs p i gnons par des
p o t eaux et des p o u t r es en bois . lc i , d ans Ie cas de la Casa
Grat i cc i o . l a m.me t e c h n i que a �t� emp l o y� . S i mp l ement l ' espace
v i de est devenu une c o n s t r uc t i o n . C ' es t Ie ph�nom�ne c l ass i que
de l ' u t i l i sa t i o n par l es p r o p r i � t a i res , d es espaces v i des .
En c o mp a r a n t cette ma i so n avec des ma i sons en pans de bois de
L i mo g e s ou de T o u l ouse • . . . piOUS c o m p r enons que l es Roma i ns
p o ss�d a i en t tr�s b i en la tech n i que du b�t i en pans de bo i s .

Les � t udes , l es o b serva t i ons et t o us l es ense i g nements que


no us tr ansmet c o nf i l-ment la c o nna i ssance des ma i sons en p ans
de bois que nous avons pu r �a l i ser dans Ie c ad r e de la
c o l l ec t i o n des 4 5 ouvrages d e la c o nna i ssance d u Ba t i Anc i en
d'E.D.F. Cela nous a i dent � c o mp r e n d r e l ' i mp o r tance du p ans
de bois dans l es s i tes � hauts r i sques s i sm i ques c o mme Pomp e i
et Her c u l anum .

Nous venons de vo i r � t r avers deux s i tes t e l l ement c o nnus de


l ' an t i q u i t � que l es b a t i sseurs ant i ques ava i en t une
c o nna i ssance des t e c h n i ques ant i s i sm i ques et qu ' i l sava i en t
l es u t i l i ser .
3 24 Adobe 90

II p ar t i e

Nous d i sp o sons d e t ec h n i q ues ant i s i sm i ques

Dans I e c a d r e des zones s i sm i ques ces memes c arac t �r i s t i ques


appar a i ssent sur I e bat i anc i en t r ad i t i onne l que no us
h ab i t o ns aU j o u r d ' hu i et que l ' on peut s i tuer d u moyen age �
nos j our s . Ce bat i t r ad i t i onne l , se l o n sa nature ( en p i er r e ou
autres ) a des c o ns t antes q u ' i l no us faut s a n s c esse r e t r ouver
et a p p r o fond i l- . Ces t ec h n i ques ne sont pas h i er a c h i s�es ma i s
tr�s l i �es a u x types d e b at i .

1 ) Le bat i en p i er r e
Comme no us l ' avons �voqu� p l us haut , l es ma i sons gr�cques o n t
l eurs murs c o m p o s�s d e m� l anges d e g r osses p i er r es e t de
p e t i tes p i er r es a v e c un savant agencemen t , l es c l o i sons so nt
en p ans de b o i s et Ie p l afond a la p ar t i c u l ar i t� d ' @ t r e fa i t
d e r o nd i ns d e b o i s sur l esque l s peuvent bouger des p i er r es
p l a t es . Le tout est recouver t d ' une i mp o r t a n t e � p a i sseur de
terre . Lo r s d ' un t r emb l ement de terre , Ie remp l acement d e ces
t ec h n i q ues par une d a l l e b�ton frag i l i se Ie b a t i . Nous
pouvons vo i r 15 � 20 ans apres un c er ta i n no mbre de
f i ssures , l es d a l l e s se r e l evant au x e x t r �m i t� s . Cec i nous
montre qu ' i l ne faut pas ap p l i q uer n ' i mp o r t e q ue l l e t e c h n i que
m a i s qu ' i l faut res ter p r �s des t ec h n i q ues d u b a t i anc i en et
des c o nna i ssance des ma90ns l oc a u x qui non seu l ement savent
l es me t t r e en oeuvre ma i s auss i savent l es e n t r e t en i r .
Tout I e monde c o nna i t les t ec h n i q ues p r �sentes dans d e
nombr euses r�g i ons d u monde , q u i consi stent � renfo r c er l es
soubassements des murs c o mme p ar e x emp l e � Annecy ou en
Corse . . . I l e x i ste auss i des t ec h n i ques p l us sub t i l e s e t qui
c ac h e une permanence c o n s t a n t e de r�hab i l i t a t i o n : Ce l l e d ' un
des prem i ers p r o j e t s � t ud i �s l ors du t r emb l emen t de t e r r e en
I ta l i e en no v . 1 980 dans la r�g i o n de la Campan i e .

Cette tec h n i que c o ns i st e e n une superpo s i t i o n de c o n t refor t s


et une u t i l i sa t i o n de l ' espace e n t r e c e u x -c i . E l le est
emp l o y�e dans sa p l us g r ande e x ub�rance dans cer t a i nes i l es
i t a l i ennes tel le que l ' ile de P r o c i de a ou l ' on c o n s t a t e en
e x am i na n t de p r es q u ' i l y a une success i o n d e c o n t r efo r t s q u i
avancent sur la mer . Ces m@mes r�f�rences s e r e t r o uvent en
Gr�ce dans l ' i le de Santor i n , emp l o y�es d e fa90n mo i ns
i mpor t a n t e ma i s on tl-ouve aussi d ' i mmenses c o n t r efor t s q u i
so u t i ennent enc o r e des ma i sons du moyen ag e .
Enf i n d e man i er e p l us sub t i l e , c ' es t dans l ' i l e d e Paros que
l ' o n u t i l i se la t ec h n i que d e l ' ar c � l ' i n t�r i eur de la
m a i so n . Lorsqu ' o n essaye d e comp rend r e I e po s i t i onnement de
cet arc , qu i s e s i tue souvent au c e n t r e !;I e la p i ece , o n
constate qu ' i l p r end appu i sur un a u t r e arc e x t�r i eur � la
m a i so n q u i sert d e passage c o uver t . Ce l u i -c i p r enant appu i
l u i -me-me SU1- l ' ar c d ' une au t r e m a i so n . C ' es t a i ns i qu ' i l
e x i s te un i mmense m a i l l ag e d ' ar c s � l ' i n t�r i eur comme
� 1 ' ex t er i eul- des ma i sons . Le non e n t r e t i en d e ces arcs
d i m i nue toute l eur eff i c ac i t e . Ce que no us a d m i r o ns avec
beaucoup de p l a i s i r l orsque nous nous p r omenons dans l es i l es
g recques sc.us l es passages c ouver t s n ' es t que l ' i n l assab l e
vo l on t e des h a b i t a n t s d e c r �er ce ma i l l age af i n d e se
pro t�ger . Ces � l �men t s que l ' o n c ro i t comme des e l �ments d e
decor sont d ' une i mp o r t ance a n t i s i sm i que c ap i t a l e .

Enf i n l e s ma i sons d e Santor i n o u l es r ares ves t i ges d e la


r � g i o n Ama l f i t a i ne en I ta l i e sont d e v�r i t ab l es a r c s en­
e l l es-m@me . To u t e la t o i t u r e c o ns t i tue un arc . On tr ouve auss i
de nombreuses p e t i tes t e c h n i ques p o u r l e s ouver tures ou les
fene- t r es , en Yougo s l av i e o u en g r �c e .

Avant d ' abor der l e s ma i sons e n p ans d e b o i s e t comme � l �ment


de t r an s i t i o n regardons l es ma i sons en p i er r es o u a u t r es
, dont la p ar t i e h a u t e e s t en pans de bois , t e l l es qu ' on l es
t r ouve dans Ie nc. r d de la Gr �ce ou en Yc.ug o s l av i e . A c h aque
n i veau , l es murs c o mp o r t e n t tous l es m e t r es des sor tes
d ' �c h e l l es en bo i s pos�es � p l at sur Ie mur . A i ns i la
ma i so n est s t r i �e sur t o u t e s a hau teur par du bois qui
ser t de c h a inage o u d e l i a i sonnemen t . Ce t t e t e c h n i que se
r e t r ouve en Turqu i e c c.mme l ' i nd i que Haroun Tass i e f d ans son
ouvrage

p r o p ag ande .
c o mme c o ns t i tuant
p r �c o n i s�e e t f i gu r a i t � ce
un mode l e d e ma i so n r u r a l e
t i tr e sur des aff i c hes d e
Seismic Mitigation 32 5

Pour Ie pans d e b o i s c ' es t en A l sace ( Fr a nc e ) que l ' on


r emarque l ' emp l o i de la t ec h n i q u e d u bois l ong qu i monte sur
p l us i eurs n i veaux . I l e x i ste dans c e t te r �g i o n une vo l on t � de
s�parer c h aque n i veau , une t r i angu l a t i o n d e la t o i t ur e , e t une
p i ece de bo i s " Ie Man " qu i �querre l es sab l i �res avec l es
p o teaux c o r n i �r es . La p ar t i c u l ar i t� de c e t t e p i ec e de b o i s est
c o ns t i tu�e par Ie fa i t q u ' e l l e se c h e v i l l e par l ' e x t�r i eur
. L ' o b serva t i o n d e s m a i sc,"s en pans d e b o i s nous permet de
c o n s t ater e x ac tement l a m@me t ec h n i que en Turqu i e et en
Gr�ce . Dans l ' 1le d e Lefkas on r emarque c e t t e t ec h n i que
s i m i l a i r e , l es restaurateurs d u d er n i er t r emb l ement de t er r e
ont p r �f�r� p o s i t i onner d e s esc a l i er s b � t o ns devan t ce type
d e m a i so n , i g no r an t comp l e t ement Ie fa i t que l es esc a l i er s
� t a i en t c o ns t r u i t s en b o i s . La p r o p r i �t � d u b o i s est de
pouvo i r s e d�former .

Enf i n l es ma i sons en t er r e , l es � t udes a ce j our ne no us


perme t t ent pas d ' aff i rmer enc o r e que nous d i sposons des
des t ec h n i ques a p p r o p r i �es . Par c o n t r e nous savons que la
p ar t i cu l ar i t� d e ces ma i sons est d ' avo i r c o mme o n Ie dit en
Daup h i n� " De b o nnes b o t tes et un b o n Chapeau " c ' es t a d i re
que la c ha r p e n t e d e la t o i ture avec s o n c h a 1 nage e t son po i d s
l i a i sonne l ' ensemb l e de la ma i so n sur d e s bases sa i nes . Toute
frag i l i t� des C h a i nages d ' ang l e s e n t r a 1 ne p l us souvent la
c h u t e des murs .
L ' �t ude d e s t e c h n i que anc i ennes par type de b�t i a pour but
n o n seu l ement d ' appor ter l es me i l l eurs r�ponses t ec h n i q ues e n
u t i l i sa n t auss i b i en l es t e c h n i ques anc i ennes que l es
tech n i q ues modernes . Ma i � sur t o u t d e c o mp r endre Ie
c ompor tement de ce bat i et r e t r o uver I e savo i r des g�n�r a t i o n
d ' u t i l i sa t eurs . Ma i s il nous faut enc o r e beaucoup c h er c her et
sur t o u t ana l yser .

III p ar t i e
Le b a t i a n t i s i sm i q ue : d es r�g l �s , une i mp l i c a t i on . . .

Par m i l es p r o b l emes que nous c o ns ta tons avant et apres un


t r emb l ement de terre o n r e l eve un c er ta i n nombre de
c o ns t antes c a u s e d e danger .
l ) le p r o b l eme du non e n tr e t i en et des c o ns t r u c t i ons abus i ves
2 ) le p r o b l eme de l a m�c o nna i ssance des f i ssUres , de
l ' i nterpr�t a t i on de la l ec ture de c e l l es-c i pour en �va l uer
Ie danger

En
des
zones a
mur s , que l que so i t
r i sques
h ab i t a t i o ns . Le
Ie
, I e danger
non-e n t r e t i en
type de
v i en t
de
b a t i , rend
souvent
la t o i ture
du

vu l n�rab l e
et
non e n t r e t i en
des
aux
secousses non seu l ement l ' h ab i ta t i o n m a i s auss i l es ma i sons
vo i s i nes . L o r s des
b�t i s ayant
p o u t r e ma l
observa t i ons c ' es t
souffer t
ancr�e • • • par
a
t r emb l ements d e

tel le
cause
e xemp l e . Par � i
c h em i nee
terre
d ' un � l �ment

qUi peut
, on a pu observer
defec tueu x , une
l es p r e m i eres
@tre en danger
des

, te l l e tui le ma l po s i t i onnee , te l l i n teau no n c o nfo r t� qui


peut c o ns t i tuer u n danger .
De m@lI!e l a c o n s t r uc t i o n sauvage est un p r o b l eme t r e s g r ave e t
qui c ependant s ' amp l i f i e . Un p r op r i �t a i r e d e d e r n i er etage qu i
c onstru i t une sur � l �va t i o n sans t en i r compte de son vo i s i nage
est un ree l danger . Dans Ie cas de Pc. l l a en I ta l i e c ' es t une
c o ns t r uc t i o n
que c e l u i -c i
c o n t r efo r t s .
Un a u t r e p r o b l eme
abus i ve
ava i t
qui
for t

appar a 1 t
a d � s t ab i l i s�
b i en t enu

, c ' es t
j usq u ' a
un

I e c h angement
i l ot
p r �sent
e n t i er , a l o r s

de
g r �c e

n a t ure
a ses

des
mat�r i au x . Lorsque un p r o p r i e t a i r e d ' un b�t i en p i e r r e pense
c o nso l i d er sa m a i so n en remp l a9ant une poutre b o i s par une
poutre me t a l l i que , ou un p l an c h er b o i s par un p l ancher
b � t o n , l es mat�r i au>: r � ag i r o n t d e man i er e d i fferente et
peuvent c reer des desor d r es . Une des p r em i er e s ma i sons que
j ' ai abor dee apres un t r emb l ement de terre ava i t un r e z -de­
c h aussee en p i e r r e d e tai l le , Ie prem i er �tage en b e t o n , Ie
deux i eme � tage en b r i que . Le f i ls ava i t c o n s tr u i t sur la
m a i so n d e so n pere e t ava i t r ea l i se une h ab i ta t i o n pour son
propre f i l s en b r i que au d eu x i eme � tage . L o r s d u t r emb l ement
de terre , l ' � t a g e en b � t o n s ' es t dep l ac � de 10 a 1 5c m , quant
a l ' � t a g e en b r i que , i l fut t o t a l ement effond r � .
De la m@me man i er e ,on a vu sur ma i sons ou la to i ture ava i t
� t � remp l ac�e par une t o i ture e n b�ton : des d a l l es ( to i t ur e )
ava i en t bouge de p l u s i eurs c e n t i me t res sur l es murs de p i er r e
Adobe 90

La l ec ture
ALl n i veau
comprendre c e q U i
du
d ' un
b�t i
ilot , la
A t r avers

s ' es t
l ec t ure
passe
ses f i ssures
des f i ssures
sur
nous
I e b �t i . No u s devons auss i
permet de

savo i r si l e s f i ssures sont danger euses ou nc. n .

Dans I e cadre des ref l e x i ons que no us pouvons f a i r e sur Ie


b�t i anc i en , l e p r em i er p r o b l eme est l a c o nna i ssance d e c e
b �t i . Fo r t l o ng t emps i g noree . c e t t e c o n n a i ssance n ' es t pas
enc o r e d i ffusee dans Ie c ad r e des eco l e s d ' a rc h i t e c t ur e .
C ' es t un ree l p r o b l eme c a r o n peut en vo u l an t rehab i l i t er
fa i r e p l us d e ma l
l es s i tes
acc r o i t r e
b�t i est
la
que d e b i en . D.n
archeo l o g i ques o u
d e s t r uc t i o n
i mpor t an t e car
des
l es

c ' es t
l ' a souvent
restaur a t i ons
s i tes . Ma i s
d a n s ces
la
z o nes
c c. ns t a t e dans
ont c o n t r i bue
c o n n a i ssance
que v i vent
du
les
A
f a m i l I es l es p l us pauvres , l es hommes t r ava i l l an t d ans
d ' au t res r eg i ons ou pays . Ap p o r ter des t ec h n i ques c i t ad i nes
peut .tre t o t a l ement c o n t r ad i c t o i r e p o u r Ie b�t i ma i s aus s i
l es
est
en oeuvre
est
v i eux
donc

e t o nnant
ma90ns
i mp o r t a n t
qu i

l eurs p r o p r es
d e constater
s o n t s u r p l ac e ne
d ' a i der ceux

q u ' une
qui
tech n i q ues q u i
so n t
l es c o nna i ssent
sur
respec t e n t
t e c h n i que que
p l ac e
Ie
l ' on
A pas . I I
me t t r e
b�t i . I I
t r o uve en
Yc.ug o s l av i e se r e t r ouve en g r ec e et dans d ' au tres pays . Nous
devons donc avant de f a i r e t o u t e c o nc l us i o n approfond i r no s
c o nna i ssances par des recherches et des echanges .
Un des p r i nc i pa u x p r o b l emes du b�ti res i d e dans l ' en t r e t i en
de c e l u i -c i . To u t b�t i e n t r e tenu a de me i l l eurs c h ances de
b i en reag i r .

Comme nous
t ec h n i ques
l ' avons vu
spec i f i ques d ans
p r ecedemment
l es
• c ' es t
z o nes s i sm"iques
Ia r ec h el-c h e d e
qu ' i l no us
faut
peres
a p p r o fond i r
de
souvent
l ' an t i q u i t e A nos
en e tu d i an t
j o urs •
l es tech n i q ues d e nos

b i b l i og r a p h i e
B . He l l y , J . P . Adam , H . Ta ss i ef , A . Ma i ur i ,

o
Seismic Mitigation

ABSTRACT :
On October 1 7 1 9 8 3 an earth­
quake me asuring 7 . 1 on the
DAMAGEPRIETA
LOMA TO HISTORIC
EARTHQUAKEADOBESANTABUILpINGS NEAR THECA EPICENTER QF THE 1989
CRUZ COUNTY
Richter scale with an epi­
center in Santa Cruz County Edna E . Kimbro
shook northern C a l i fornia Historical : Arch itectural : Conservation Research
cau s ing ma s s ive property
jamage and some l o s s o f At 5 : 04 P . M .. on October 1 7 , 1 9 8 9 , an earthquake measuring 7 . 1 on the
l i fe . Among t h e s t ructures Richter scale occurred centered in N i sene Marks State Park in Santa
a f fected were the l a s t four Cruz Count y . For 15 seconds it shook Northern Cal i fornia caus ing
H i spanic Era h i s t o r i c adobe mass ive property damage and some l o s s of l i fe . Over the ensuing
bu i l dings in Santa Cruz months thousands o f aftershocks fol l owed, 9 0 measuring over 3 on the
Count y . Two o f the four had Richter scale, 2 0 above 4 , and 3 over 5 . Santa Cruz County l i e s for
previou s l y been st rengthened the most part on the P ac i f i c plat e , west o f the San Andreas Fault
to varying extent s to with­ dividing the P ac i f i c and North Ameri can p l ate s . The P a c i f i c plate is
stand s e i smic event s . the gradua l l y moving north .
Santa Cruz Mi s s i on Adobe and
the two story Castro Adob e . Only four H i span i c era historic adobe structures remain in
The damage to each o f the Santa Cruz County where the twelfth Franciscan Mi s s ion was founded
four structures is de ­ by Spain i n 1 7 9 1 : the Branci forte Adobe ( Ca . 1 8 0 6 - 1 8 1 8 , the Bplco f f
scribed, eva luated and ma j o r Adobe ( C a . 1 8 3 9 - 1 8 4 4 ) , the Santa Cruz Mi s s ion Adobe ( 1 8 2 2 - 1 8 2 4 ) , and
factors contr ibut ing t o the the Castro Adobe ( C a . 1 8 4 6 - 1 8 5 0 ) . Although the se structures are the
damage are i dent i f i e d . oldest historic bu i l dings located i n this part of z one 4 , only two
o f the four had previously been st rengthened t o help withstand
KEYWORD S : sei smic event s . All four were presumably a f fected by the 1 9 0 6 San
Adob e , earthen archite cture , Franci sco earthquake j udging from old photographs and physical
seismi c , e a rthquake , s e i smic evidence o f damage .
mit igat i on

The Branciforte Adobe


A one - story, two-room building, the Branciforte Adobe ( C a . 1 8 0 6 -
1 8 1 8 ) , fared relat ive l y wel l in the Lorna P rieta earthquake o f 1 9 8 9 .
Origina l l y it was const ructed as a one - and onerhal f story, t i le
roofed one-room structure with adobe gable ends and a
loft above . Currently the Branc iforte Adobe has woodframed
addi t i ons to the west and south, a woodframed repl acement s outh end
tapanco
or

short wal l , and woodframed repl acement gable over the north end
adobe wal l . The west long adobe wa l l is no longer loadbearing as
the shingled roof now rests on the woodframed west wal l of an
Rgure 1. The Branciforte Adobe, 1987 addition and on the east adobe wa l l .

The Branci forte Adobe is oriented with the long wal l s running
north-south and faces east as was typical of early Cali fornia adobe
const ruct ion . The wal l s are 60 cm thick, with a height to thickness
rat i o o f 4 . 5 for the l ong wal l s . The adobes measure 3 0 cm in width
by 7 . 5 in depth by 60 cm i n length and are laid in Flemish bond . A
30 cm interior part i t i on wal l added to the structure a 1 8 4 8
features vert ical wood members with specially cast adobe i n f i l l
between, representing a seismically supe rior bui lding techn i que . A
simi lar adobe part it ion wa l l was added to the Santa Cruz Mi s s ion
Adobe circa 1 8 4 8 a s wel l .

Beside s the usual diagonal cracking about the door and window
open ings , through the wal l vert ical cracks occurred at the
previously undamaged northeast and northwest corners of the adobe .
A ve rt ical crack widened near the northwest corner where an opening
had been cut and part i a l ly f i l led by a cupboard .

Previous to the seismic event of October 1 7 , 1 9 8 9 the


Branci forte Adobe had been rehab i l i t ated between 1 9 7 6 - 1 9 8 6 , was we l l
maintained and i n a good state o f preservat i on . Moisture problems

long adobe wal l .


corredor
had been a l leviated by the installat i on of a French drain at the
north gable end and restorat ion o f the sheltering the east
These fact ors , together with the low height t o
thickne s s rat i o and prior l o s s o f the adobe gab l e , apparent l y
contributed t o t h e structure ' s pe rformance under eart hquake loads
Rgure
Adobe.
2.1990
North end wall, Branciforte despite its proximity t o the epicenter and lack o f s e i smic
strengthen ing .

The Bolcoff Adobe


Bolcoff Adobe , a one -story, two-room structure dating from ca 1 8 3 9
t o 1 8 4 4 , was the least obviously damaged b y the earthquake although
it is i l l -maintained and in a poor state of repa i r .

The long walls o f the Bolcoff adobe are o r iented north-south


with the facade facing east l ike the Branci fort e Adobe . The wa l l s
are constructed o f adobes measuring 2 0 c m in width, 7 . 5 c m i n
height , and 3 5 c m in length laid in Engl i s h bond . They are 7 0 cm
thick and the ratio of height to thiokne s s is 3 . 6 , we l l within the
Rgure 3. The Bo/coffAdobe, 1990 current code requi rement of 5 in s e i smic z one 4 .
Adobe 90

The north port i on of the west long wal l has been rep l aced with
formed concrete and reinforced concrete encapsulates the base of the
southern wall segment extending about a meter up the wa l l . The
south end short wa l l has been replaced with woodframing which
support s the heavy barrel t i led roof at that end . The roof p l at e

res t s o n concrete corner but t re s s e s at the n r th e n d and at t h e
midp o int o f t h e east l ong wa l l ; t h e adobe wal l s d o not carry the
load .

Mudst one f oundat i on s extend over a meter deep into the a l luvi a l
s o i l o f t h e former stream b e d o f W i l de r Creek . Ground leve l inside
and out of the s t ructure has been l owered over t ime about 30 cm
exp o s ing portions o f the foot ings . Above the stone foundat i on s
exce s.s i ve moi �ture has resulted in cons iderable erosion o f t h e wa l l

Adobe, 1990
Rgure 4. West elevation, The Bolcoff face , undermining the east wal l near its base .

E f fects of the October 1 7 , 1 9 8 9 earthquake - on the B o l c o f f Adobe


were not devastating as might have been expected given the
bu i l ding ' s condit i on . A pre -exist ing ve rtical crack widened in the
east wal l and diagonal cracking was obse rved in the n o rth end wal l .
There , high in the gable portion o f the wal l , an unframed opening
suffered the l o s s of s ome adobes near the apex and the opening
itse l f was displaced t o the west . The most dramatic e f fects
occurred i n places whe re portions o f the deteri orated adobe wa l l
surfaces f e l l away exp o s i ng the darker interior o f the mud brick s .

The moderate nature o f the damage to the Bolcoff Adobe i s


att r ibuted t o s t ab i l it y result ing from the l ow height t o thi ckne s s
rat i o o f the wal l s , the sma l l numbe r o f openings and the fact that
the we i ght of the roof is not borne by the wa l l s . The l ocation o f
the B o l c o f f Adobe farthest from the epicenter may have been a fact or
in favor o f i t s cont i nued survival .

The Santa Cruz Mission Adobe


The Santa C ru z M i s s i on Adobe , a one - and one-ha l f story seven - room
structure was completed in 1 8 2 4 for Indian neophyte quart e r s at the
mis s i on . On October 1 7 , 1 9 8 9 the bui lding was undergoing
restorat i on and had recent l y been retrofitted with a hor i z ontal
concrete bond or ring beam connect ing the hi storical outer wa l l s
with the newly reconst ructed adobe cross wa l l s . Al s o , a new p l ywood
sheathed r o o f we l l anchored to the bond beam had been const ructed
await ing i n st a l l at i o n of roof t i l e s .

The S anta Cruz M i s s i on Adobe i s atyp i c a l l y o riented with its


long wal l s running east -west and the reconstructed short wa l l s
Adobe, 1990
Rgure 5. The Santa Cruz Mission
al igned north-south . The wa l l s are const ructed o f adobes 2 0 cm wide
by 7 . 5 cm and 4 2 : 5 0 cm in length l a i d in Engl i s h bond . They are
8 2 . 5 cm thick with a he i ght t o thickne s s rat i o o f 2 . 6 6 for the l ong
wa l l s . The reconstructed short wal l s suf fered the greater damage ,
cracking d i agon a l l y with cracks forming "X" conf igura t i on s in the
wal l expan s e s and cracking diagonal ly at the top o f openings . What
init i a l l y appeared to be out - o f -pl ane displ acement turned out to be
poor workmanship in the construction o f the wal l s , the product of
fai l ure t o a l ign the courses vert i ca l ly .

At the we s t end or short wal l , a pre-exi s t i ng vert i c a l crack


which had been repaired with Port l and cement stucco in 1 9 6 6 widened
and deepened, extending through the wa l l at the corne r . At the s ame
corner a wide ve rtical crack opened in the south long wa l l , l e aving
an e s sent i a l l y f re e st anding co lumn o f adobe which toppled in an
afte rshock , cracking the bond beam on the west s i de . The upper
port i on of the west short end wa l l moved outward hor i z ont a l l y to the
west beneath the bond beam leaning more than 10 cm from vert ic a l .
The l ower port i on of the wa l l suffe red some out -o f -p l ane damage with
loss o f stones f r om the foundat ion .

Fact ors contribut ing to the instabi l ity o f the west end
inc lude : 1 ) f a i lure t o prope rly repair the e x i st i n g vert i ca l crack
at the corne r ; 2 ) a later do o r cut into the west wa l l west face
c l o s e t o the corne r ; 3) unstable foundat i on condit i on s not
completely remedied; 4 ) recent archaeological excavat ions very near
the wa l l ; and 5) lack of vertical anchorage o f the bond beam t o the
wa l l .

Rgure 6. Southwest corner, the Santa


Cruz Mission Adobe, 1990 The gab l e s confined between the ring beam and roof framing
suffered l it t l e damage . Some original m i s s i on period mud p l aster
with Native Ame r i can G r a f f i t i l ocated high on a short gable wa l l
survived more o r l e s s intact . Also intact i s a mud p l a st e red
surface decorated by I ndian neophyt e s on the interior o f the long
south adobe wal l . Thi s surface was painstakingly re-adhered t o the
adobe wa l l by Constance S i lver and Nan Rosenthal in 1 9 8 7 .
Seismic Mitigation 3 29

The l ong wa l l s a l igned perpendicu l a r to the primary north-south


direct i on o f the e arthquake mot i on were comparatively l i t t l e
a f fected by the e arthquake . Predictab l y , they suffered narrow
diagonal shear c racks at both exist ing and recent l y f i l led in door
and window openings . Narrow , near vert ical cracks appeared on the
exterior o f the wal l s marking the i ntersections with newly
reconstructed interior crosswa l l s . The o r i ginal perimeter wa l l s
exhibit noticeably bet ter workmanship i n the i r construct i on with
deep foot ings extending we l l over a meter below grade and mudstone
rubb l e ma s onry reaching high above grade as we l l . They have been
kept dry and roo fed, and have been l it t l e af fected by basal e r o s i on .

The response of the Santa Cru z Mi s s i on Adobe to the e a rthquake


may have been improved by the recent removal of a unoriginal second
story . That factor t ogether with the ins t a l l at i on o f a ring beam,
the l ow height to thickne s s rat i o of the wal l s , the recon s t ruct i on
of mu l t i p l e cro s s wal l s and the rigid roof di aphragm, contributed t o
the bu i lding ' s survival with only moderate damage .

The Castro Adobe


The Castro Adobe , ca 1 8 4 6 - 1 8 5 0 , the only two - story h i s t o r i c adobe
structure in the count y , is l ocated l e s s than 5 m i l e s from the
epicente r o f the Lorna Priet a earthquake . The wa l l s o f both s t o r i e s
a r e a uniform 7 0 c m t h i c k a n d the he ight - t o -thi ckne s s rat i o o f the
long wa l l s i s j u s t over 3, c alculat ing each story separate l y . The
adobe s , 3 5 cm wide by 7 . 5 in height by 70 cm in lengt h , are l a i d in
heading bond on sha l l o w mud-mortared cobble foundat ions about a
third of a meter deep . They are made of dark sub s o i l with s t raw,
whi l e the mortar is l i ght t op s o i l with considerable st raw admi xed .

Figure 7. The Castro Adobe, 1990 The bui l ding is oriented typica l l y with l ong wal l s running
north-south penetrated by nume rous opening s . The doors and windows
are symmet r i c a l l y disposed and a l i gned one above the other . The
large columns of masonry between openings cracked diagona l l y forming
large "X" shape s . There are but two adobe cro s s wa l l s on the l ower
f l o o r and none above . The second s t ory i s e s sent i a l l y one l arge
room 9 by 24 m .

North o f the two story block, a one -st ory adobe cocina,
kitchen, was con s t ru cted against the north wa l l with an ove r s i z ed
or

the cocina
opening in the e ast long wal l . Within the l ast 30 years the roof o f
w a s a ltered in s e i smically uns ound ways .
roof rested direct l y on the east long wa l l but was rai sed about 1 0
cm above the we s t long wa l l bearing on narrow adobe i n f i l l between
That i s , the

short studs and was not anchored to e ither s ide wa l l . During the
earthquake , the roo f ' s mas s i ve l aminated ridge beam rammed a hole
into the midpoint o f the two - story north wal l and cracked i t
vert i c a l l y through t h e wal l . The roof moved a s a uni t about 1 5 cm
north thru s t ing out the west l ong wa l l and producing a vert i c a l

(
\
wal l of the cocina
crack at t h e northwest corner over 1 0 c m wide .

from t h e base .
The north gab l e end
s u f fered ma j o r out -of -plane damage about a met e r
F i na l l y , a t iny pre-ex i s t ing vert i c a l crack through
the north gable end wa l l , where a corbel was inserted i n the 1 9 5 0 s ,
widened about 1 0 cm . With a ftershock s , the northwest corner and
part of the north end wal l became a freestanding column of adobe

\ leaning outward .

Figure 8.
Oamage from cocina ridge
beam, the Castro Adobe, 1990 F o l l owing the 1 9 0 6 earthquake , the Castro Adobe had been f i t ted
with wal l ancho r s at the f i r s t floor plate leve l ; the s e performed
we l l in 1 9 8 9 preventing the col lapse of the second f l o or . Also, a
wooden t i e beam had been embedded in the north wa l l penetrat ing the
east and west long wa l l s at the midpoint o f the second f l o o r . In
the 1 9 5 0 s , t h e h i s t o r i c roof ing system wa s rep l aced by trus s e s and
the spongy second f l o o r su spended from the bot t om chords by steel
rods through the f l oor j o i s t s .

In 1 9 8 7 , evidence of settlement of the south end of the

vert ical at the apex .


in situ,
bu i l ding p rompted inst a l l at i on of a concrete grade beam to st abi l i ze
the south gable end wa l l leaning more than 10 cm from
Wide welded steel straps were added ext e nding
about hal f way around the st ructure at the t op of the f i rst and
second st o r i e s j ust below the p l ate , to anchor l ong wal l s . On
October 1 7 , 1 9 8 9 , the upper port i on of the south gab l e end wal l
col l apsed outward, tak ing the steel strap down with i t . A wide
vert ical c rack opened in the south wa l l at the southwest corner and
a hairl ine crack opened at the southeast corne r . A d i agonal crack
(which had been repai red with expanded met a l l ath on the exter i o r )
widened i n the w e s t l o n g w a l l a t a doorwa y ; other crack s , s im i l a r l y
repaired on t h e int e r i o r , re- opened and widene d .
33 0 Adobe 90

At the north end of the two-s t ory b l o c k , the long wal l s cracked
at the corners t o beyond the wooden t ie beam, which proved o f
dubious va lue . Both the l on g and short wal l s c racked diago n a l l y
around d o o r s a n d windows . A ven t , which h a d b e e n opened through the
wall without a heade r , exacerbate d severe through- the - wa l l diagonal
cracks .

Much of the damage to the Castro Adobe wa s predictable and

cocina,
part i a l l y the result of misgu i ded intervent i on s : "new" ope n i ng s
without heade r s , a poorly de s i gned roof over t h e an i l l ­
conceived t i e beam, and a n end wa l l " s t ab i l i z ed" out o f vert i c a l .
A s i gn i f i cant r o l e was p l ayed by f l aw s inherent in the bui lding ' s
init i a l construct i on : sha l l ow foundat i on s bui lt on f i l l , a p aucity
of cross wa l l s , the u s e o f heading bond, incompat ible mort a r , one ­
and two- s t ory ad j o in i n g s e c t i on s , l arge number of ope n ings , the i r
al ignment , i n s e rt e d l o ad bearing c o rbe l s , etc . , etc . Cont r ibut ing
t o the bui lding ' s cont inued exi stence are the uni form wal l thickn e s s
o f both s t o r i es , t h e l ow height- to- thickn e s s rat i o , t h e t ying o f the
roof t ru s s e s to the f i rst f l o o r j o i s t s , the wa l l ancho rs , the north­
south orientation o f the l ong wal l s , the abs ence o f any prob l ems
a s s o c i at e d with moi sture , and the over a l l high level of ma i ntenance .

Con clus ions

The Lorna P r i e t a earthquake of 1 9 8 9 repre s ents o n l y the l a t e s t


demonst rat i o n o f t h e destruct ivenes s o f s e i smic event s t o
Ca l i forn i a ' s Spani sh C o l o n i a l and Mex i can earthen archit ectural
he r i t age . I n recent years several o'f the state ' s rare adobe
monument s have been devastated by e arthquakes . San F ernando Mi s s i on
church was who l l y de s t royed in 1 9 7 1 ; Mi s s i on S an Gabr i e l and the P i o
P i c o Adobe ( a St ate H i storic P ark) were s evere l y damaged i n 1 9 8 7
nece s s i t at ing the i r c l o sure ; and i n 1 9 8 9 , the Castro Adobe j o ined
the i r rank s .

None of the s e h i s t o r i c adobe s t ructures had been e ffect ive l y


strengthened t o withst and s e i smic event s o f appre c i ab l e magnitude ,
large l y becau s e of the high c o s t to retrofit an otherwise s ound
bui l ding to the s t andards o f C a l i fornia ' s H i s t o r i c Bu i lding Code .
The c o s t s are great e n ough that tho s e respon s ib l e for s tewardship o f
historical adobe resources gamb l e with geology . They pray t h e c o s t
of earthquake repa i r s a n d n e c e s sary st rengthe n i ng a ft e r a s e i smic
event will not exceed that of pre-e arthquake retro f i t t i n g . Thi s i s
a gambl e impo s s i b l e t o win i n the l ong run and one that p l a c e s
Ca l i forni a ' s irreplaceable h i s t o r i c a l he r i t age at r i s k .

Peru , Mex i c o and other Lat i n Amer i c an nations s i m i l a r l y


a f fected by "mother" e arth ' s moveme nt s , a n d even more r i c h l y endowed
with Span i sh C o l o n i a l earthen architectural ma s t e rp i e ce s , have
as sumed the l eadership r o l e in deve l oping s e i smic s t rengthen ing
techniques for adobe s t ructure s . The s e t e chniques have not been
wide l y t e s t e d or accepted, much l e s s we l l unde r s t ood, i n C a l i fo rn i a .
Engi neers unfami l i a r with them r i s k potent i a l l o s se s of h i s t o r i c
resources i n the i r m i s app l i cat i on . C l e a r l y , e x i sting s e i smic
improvement t e chniques suitable for historic adobe bu i ldings must be
introduced i n C a l i fornia and further research c ommenced t o explore
new methodo l og ie s .

I f the status quo cont inues and a lt ernat i ve effective and


affo rdable means to achieve s e i smic s t ab i l i z a t i o n are not
iden t i fi e d , t e s t e d , and adopted for use i n the state, C a l i f ornia ' s
proud mi s s i on past wi l l cont inue t o be l o st one p i e ce at a t ime : an
end wal l t oday , a gabl e t omorrow, a dome , b e l lt ower o r vault the
next day .
Seismic Mitigation 33 1

ABSTRACT
GUATE MALA
PREVENCION SISMICA EN LAS CONSTRUCCIONES DE ADOBE ,
DESPUES DE LOS TERREMOTOS DE 1 9 1 7- 1 9 1 8
EN LA CIUDAD DE

Thi s study focuses on the


ditt erent applied building Marla Elena Molina
systems used by engineers , 1 a . Calle Oriente No . 8
architect s , and builders ot adobe Antigua Guatemala
s t ructures built immediately
at ter the earthquakes that
at tec ted Guatemala in December Introducci6n
1 9 1 7 and January 1 9 1 8 . These
systems show the eftort in Guatemala es un pais que es constantemente at ectado por diversos terremotos .
improving the resis tance and E s tos han devastado su terri torio causando grandes perdidas humanas y
safety ot these s t ructures , ma teriale s . Entre estas ultima s , las construcciones en ado be , han sutrido
consider ing the constant threat severos danos .
ot turther seismic activity.
Sin embargo , actualmente encontramos numerosos ejemplos de
Many trips were made to Guatemala
City to explore the city ,
selected
case s ,
r ive
having
representat ive
descri bing each one ' s
intensa act ividad sismica. Por 10
edificaciones en adobe que han logrado resistir satistactoriamente e s ta
tanto , se hace necesario investigar y
conocer sus carac terist icas construct ivas y tomarlas como experiencia para
aprovecharla en la busqueda de alternativas de restauraci6n, consolidaci6n y
character�stics as well as the reestructuraci6n de edit icaciones antiguas , 0
bien , para programas de
consequent ettects that the vivienda nueva de bajo cos t o . Ademas , evaluar esta experiencia construct iva ,
s t ructures sut t e red due to the nos permi tira argumentar en relaci6n a la idea er r6nea y generali zada que el
1 9 7 6 earthquake s . The conclusion adobe es un material que debe ser desechado en areas de riesgo sismico .
is that it has not been adobe , as
material , that tailed but the El presente es un tra bajo preliminar que nos inicia en el conocimiento
lack ot adequate technology and y comprensi6n de los sistemas constructlvos utilizados en la arquitectura de
maintenanc e . adobe reali zada en la Ciudad de Guatemala despues de los terremotos de 1 , 9 1 7
y 1 ,918. (1)
Based o n this conclus ion, our
purpose is to promote �nterest Estos sistemas construct ivos cons tituyen un testimonio hist6rico del
and knowledge ot the expe rience avance tecno16gico alcanzado en esa epoca , en las construcciones que preveen
of this s tudy and that it will be el riesgo sismico , cuya et icac ia , actualmente puede ser evaluada mediante
useful in our work as
preservationists ot old buildings
and planners of new housing .
las respuestas de
(2)
compor tam�ento est ructural
editicac�one s , despues del terremoto de 1 9 7 6 .
observadas

No existe investigaci6n previa que haya tratado este tema . Por


en es tas

tanto10
este e s tudio ha sido ela borado con limltada lnformaci6n hist6rica y a traves
KEYWORDS de o bservaci6n de campo por la ciudad de Guatemala. En esta se ldent�ticaron
y o bservaron aquell�s construcc�ones que mue s tran un sistema cons truc tivo a

GUATE MALA
ADOBE STRUCTURES , SEISMIC D
, 1 9 1 7 - 1 9 18
AMA GE ,
base de ado b e ,
regis trando ,
constructlvas .
escogiendo cinco de los casos mas representativos y
mediante fotografias, Sus prlncipales caracteris t lcas

La epoca de const rucci6n de estas edit lcacione s , asi como los danos
sufridos a causa del terremo to de 1 9 7 6 , fueron esta blecidos a traves de
entrevis tas con los propietarios 0
usuarios de estos inmue ble s .

Casos o bservados :

Caso 1

En este caso se pudieron o bservar las siguientes caracte ris ticas :

Muros de adobe de 0 . 4 7 m . de espesor , que funclonan como muros


portantes de carga , detiniendo las cruj las principales de la vlvienda ( 3 ) .

Los muros entre ambientes son ta biques construidos con adobe colocado
de canto , de 0 . 1 5 m . d e espe s o r , reforzados con regl a s d e madera y alambre

X
es pigado de hierro galvanizado , colocado en torma de retlcula diagonal al
muro , con intervalos de 0 . 30 0 . 30 m. en ambas caras del mi smo .

Se utilizan varillas de acero de 1" de d�ame t ro , uniendo los muros de


carga , ancladas por medio de pernos sobre placas de metal ( ver fo togratia No .
1). Estos tlrantes coinciden con los tabique s , quedando empot rados en el
repello de una de sus paredes .

Con respec to a la volume tria ,


X X
las crujias prlncipales se
conformadas por ambientes cubicos de 4 . 5 4 .5 4.5 m.

La cubierta a dos agua s , con pendientes d e 45%


encuentran

aproximadamente , con
e s tructura de madera cubierta con lamina de zinc .

La tipoiogia es carac teris tica de las casas senoriales de la epoca


hist6 rica en es tudio , con una planta en torma de E con dos patios interiore s ,
uno inmediato a l area de zaguan ( area d e vesti bulo inmediata al ingreso
principal ) , y el o t ro pa t i o , ubicado en la parte pos terior de la vivienda
inmediato al area del comedor ( e spacio que s i rve de divisi6n entre ambos

cuartos de servidumbre ,

p6rtico 0 la cocina y la pila 0


patios ) , donde se encuentra ubicada el area de servic io , contormada por los

transici6n entre los patios y los ambientes interiores ,


lavadero . Como area de
se encuentra el
peristilo que sirve de area de ci rculaci6n y vestl bulo entre tales
espacios .
33 2

Condicion de la es tructura despues del terremot o de 1 9 7 6

Se puede observar que l a vivienda resistio aceptablemente el terremoto de


1 9 7 6 . Los dallos suiridos , tales como grietas y desprendimientos de material
en las cabezas de los muros y en los acabados , fueron reparados
inmediatamente . Actualment e , sufre de terioro por falta de adecuado
mantenimiento .

Caso 2 :

En e s te caso se pudieron o bservar las s i �uientes caracteristica s :

Muros d e adobe e n el exterior d e l a vivienda , reforzados con alambre de


h1erro galvanizado ,
intervalos de 0 . 25 X colocados en forma de reticula diagonal al mur o , a
0 . 25 m. Es tos muros de ado be , se encuentran apoyados
so bre un muro de ladrillos de mayor espesor , que se levanta a una altura
aproximada de 0 . 7 0 m. sobre el nivel exterior del suelo ( ver f o togratia No .
2).

Los tab1ques interiores , son muros livianos , const ruidos con un sistema
especial , consistente en una est ructura de parales de madera colocados cada
0 . 30 m . Es tos se encuentran cubiertos utilizando dos sistemas : a) machiembre
de madera; b) malla metalica cubi e rta con mezcla , repello y blanqueado ,
de jando vaci0 el inter10r del muro .

4 X La volume tria se encuentra conto rmada a base de espacios cubicos de 4


4 m. , con una cubierta de dos aguas con pendientes de 45% , y lamina de
zinc sobre es tructura de madera.
X
La tipologia en este caso , varia con respecto a la acos tumbrada en
e s t e periodo post-te rremotos 1 9 1 7 - 1 B , ya que presenta asime tria en la
volume tria . Ademas , cuenta con los dos patios interiores circundados por el
portico 0 peris tilo y posee un jardin exterior , que sirve de transicion entre
la vivienda y la banque ta y calle publica.

Condicion de la e s t ructura despues del terremoto de 1 9 7 6

Se pudo apreciar que l o s muros n o presentan grietas , sino unicamente


desprendimientos de material de acabados . El pro blema actual es una absoluta
talta de mantenim1ent o .

Caso 3

La cons t ruccion de este caso , es una vivienda popular ubicada en una zona
peri ferica de la ciudad y presenta las siguientes carac terist icas :

Muros exteriores de adobe , reforzados con reg las de madera en sentido


horizontal y ve rt1cal . En sentido horizontal presentan una separacion de
0 . 30 m. aproximadamente , en sentido vert1cal se encuentran colocadas a cada
0 . 90 m. aproximadament e .

L a cubierta d e lamina d e zinc sobre e s t ructura d e madera con una


pend1ente de 30% a una sola agua .

La altura es relativamente baja , 2 . 7 5 m . , con una volume t r ia en forma


rectangular exteriorment e . No fue posi ble o bservar su conformacion interna.

E l sistema construct ivo utilizado en este caso es similar al utili zado


en la arquitec tura vernacular de las poblaciones rurales del terri torio
guatemalteco . Este cons iste en tabiques que se construyen de palos
ent retej idos y barro ( 4 ) .

Condicion que presenta la e s tructura despues del terremot o de 1 9 7 6

'Carecemos de datos exac tos per o , e n base a l a observacion d e campo , pudimos


apreciar la ausencia de grietas . Los danos que presenta son en la cubierta ,
la cual · pareciera que se ha colocado provisionalmente . Ademas se o bse rvaron
. la perdida de los acabado s en la parte inf e rior de los muros , asi como e l
deterioro d e las piezas de madera debido a l a acci6n d e l o s insectos
( t ermitas ) , (ver f o tograiia No . 3 ) .

Caso 4

El sis tema empleado en este caso corresponde a inmue bles modestos , propiedad
de per sonas de pocos recursos economicos , localizados en zonas perifericas de
la ciudad de Guatemala .

Entre sus principales carac teri s ticas podemos mencionar las siguientes :

Los muros exteriores son de adobe colocado de canto , repe l lados


rusticamente , reforzados con madera y alambre e spigado , dispuesto sobre el
muro en torma horizontal a cada 0 . 30 m . Este muro de adobe se as ienta sobre
un pequeno muro de ladrillo que s e levanta aproximadamente 0 . 40 m . sobre e l
nivel d e piso exterior , ( ve r f otografia N o . 4 ) .
Seismic Mitigation 333

Una variedad de este tipo de muro la podemos apreciar en la t o tografia


No . 5 .

Estas viviendas tienen una volume tria exterior , en to·rma rec tangular 0
cuadrada , con ambientes interiores de pequefias dimens ione s . La altura es de
aproximadamente 2 . 7 5 m . y se encuentra cubierta con lamina de zinc sobre
e s t ructura de made ra , a una agua , con una pendiente de 25% aproximadament e .

Cond1ci6n despues del terremoto d e 1 9 7 6

En es tos casos n o se observaron grieta s , sino unicamente deterioro en


acabados exteriores y en las piezas de madera de refuerzo .

Caso 5

Durante el recorrido exploratorio , se pudieron o bservar varios e j emplos de


viviendas que utilizan en sus muros el ladri110 y el adobe alternativamente ,
por e j emplo :

tres hiladas de adobe y una hilada de ladr illo ,


una hilada de adobe y dos hi1adas de ladrillo .

Los muros son gruesos de aproximadamente 0 . 4 5 m .

L a cubierta es d e lamina d e zinc sobre e s t ructura d e madera , a dos


aguas , con pend1ences de 45% .

Su volume tria es tipica de las cons trucciones post-te rremotos 1 9 1 7 -


1 9 1 8 , donde predominan las plantas arquitect6nicas en forma de E y de F , con
sus dos pa tios 1nter10res , el p6rtico 0 perist i10 , el zaguan , etc .

Cond1ci6n de estas e s tructuras despues del terremo to de 1 9 7 6

En realidad carecemos d e datos suf icientes que nos permitan establecer con
prec1s16n los danos que su!rieron estos inmue bles a causa de dicho te rremo t o .
S i n embargo , los casos observados e n el centro d e l a ciudad d e Guatemala ,
mostra ban sus muros en buena condici6n , con excepci6n del deterioro en la
parte in! er10r de sus muros , debido a la humedad .

CONCLUS IONES :

Las construcciones que despues de 1 9 1 7 - 1 9 1 8 uti1izaron el adobe redujeron la


masividad de las edi ! i caciones anteriores mediante la sustituci6n de las
cubiertas de teja por cubiertas de lamina de zinc , asi como los muros
inter10res gruesos que tueron sustituidos por tabiques de lgados y livianos .

Para construir estas vivienda s , se uti1izaron los adelantos


tecno16gicos de la epoca (mal las , placas , pernos , alambres galvanizados y
t i rantes de metal) que fueron aplicados para crear nuevos sis temas
cons truc t ivos que perm1 tieran mejorar la resis tencia del adobe a los
es!uerzos de tensi6n a causa de los s i smos .

La expe r1encia const ructiva , que a grandes rasgos se encuentra


concen1da en este primer e s tudio , nos mue s t ra que las const rucciones de adobe
que !ueron reforzadas utilizando estos sistemas antisismicos , res istieron
aceptablemente el ter remoto de 1 9 7 6 , y , que las edificaciones de adobe que
!allaron, en su mayo r1a , carec1an de tecnicas constructivas anti s 1 smicas
complementarias 0 estuvieron sujetas a un deterioro previo por falta de
mantenim1ento .

En los casos es tudiados , se puede apreciar el uso t recuente de


materiales como la madera , el ado be , el ladrillo , y el metal . Es tos , en
diferentes combinac iones , con!ormaron los sistemas constructivos antisismicos
de la epoca.

E s necesario continuar la investigaci6n de este tema de la tecnolog1a


construct iva antis1smica uCil izada en el pasado , la cual , ignorad� ,
cons t i tuye una valiosa fuente del conocimiento para el uso y la conservaci6n
del adobe en zonas de riesgo s1smico .
334 Adobe 90

,
tf
FOLograt1a No. 1
Seismic Mitigation 335

Fotogratia No . 3

BIBLIOGRAFIA

eme rgenc�a post-te rremoto

1979). 1976"
Hall Hi bbi t s , J . E . , J . A. Flore s , "Medidas preventivas , intervenciones de
(Ponencia en S emina rio sobre P rotecci6n de
Monumentos en areas sfsmicas , Antigua Guatemala , UNESCO , ICOMOS , OEA, CNPAG ,
Nov .

Fotogratfa No . 4
Paniagua , Jose Ram6n , Vocabularios Basico de la Arquitectura ,
Catedra , 2a. Edici6n , COlecci6n Arte Catedra , Madrid 112. 1980,
Saenz Calde r6n , Manrique , "Es tudio de la Vivienda de Bajareque e historia
Ediciones

s i smica de Huehuetenango" . Tesis de grado para conferirle el ti tulo de

1987.
arquitecto , Fac . de Arquitectura , USAC ,

T6rtola , J . Ro berto , "La Vivienda de Bajareque en la Sub-Regi6n Altiplano


Occidental de Guatemala , Estudio de una Tecnologia Olvidada " . Tesis de grado
para cont erirle el ti tulo de arquitecto , Fac . de Arquitectura , USAC , 1986.
NOTAS

(1) desconociendo el grado de la magnitud que tuvieron


intensidad en grados de la escala Mercal l i .
y
La informaci6n hi st6r ica existente sobre estos te rremotos es escasa ,

10 estimando la

(2)
intensidad de

la muer te de
125,0000
grados en la escala de Me rcalli y y
S e calcula que este ter remoto alcanz6 e n l a ciudad d e Guatemala ,

grados en la escala de Richte r , calificado como devastador


personas .
una
una magni tud de 7 . 5
provo cando

1976"
J . E . Hall Hi b bi t s , J . A . Flore s , "Medidas preventivas intervenciones de
eme rgencia post-terremoto
( Ponencia presentada e n el Seminario sobre Protecci6n d e Monumentos en

1979).
areas sfsmicas , Antigua Guatemala , UNESCO , ICOMOS , OEA, CNPAG , Nov .

(3) La palabra crujia ,


carga .
planta del edit icio .
signif ica :
Cada una de las partes princip9les 0 1980,
Espacio comprendido entre dos muros de
naves en que se divide la
Tomado de J . R . Paniagua , Vocabularios Basico
la Arqui tectura , Ed . Catedra , 2 a . Ed . , Madrid 112.
----de --
--
Fotogratfa No . 5
(4) J.R. T6rtola , "la
Arquitec tura , USAC , 1986,
Vivienda
3).
de Bajareque en • • • • •" (Tes i s , Fac .
ABSTRACT
This paper presents the
REPAIR OF CRACKED ADOBE WALLS BY INJECTION OF MODIFIED MUD
conclus ions from a Nels Roselund
demonstration proj ect Kariotis & Associates
designed to develop a 7 1 1 Miss ion street , suite D
procedure for stab i l i z ing South Pasadena , CA 9 1 0 3 0
the wa l l s of the pio Pi co
Mansion Adobe of whittier
Cal i fornia damaged by Location and History of the Pio pico Mansion Adobe
earthquake .
The Pio pico Adobe is in the city of Whittier , Cal i fornia , about
Adobe wal l s of the building 6 . 4 km ( four miles ) from the epicenter of the Whittier Narrows
were cracked by the Earthquake of october 1 , 1 9 8 7 . It was original ly constructed in
whittier Narrows earthquake about 1 8 5 0 by Pio Pico , Governor of Cal i fornia , as headquarters
of 1 9 8 7 , which l e ft for his "El Ranchito . " It is near the San Gabriel rive r , and was
portions of the wal l s l oose apparently damaged by flooding in 1 8 6 7 -68 and again in 1 8 8 4 -8 5 .
and l ikely to become I n 1 9 0 5 the building was intended for demol ition for materials
unstable in further
earthquake shaking . The
procedure cal l s for f i l l ing
the cracks with mud that
for lan d
society was
fill for construction of a nearby bridge .
formed to save it ;
A historical
the society then went on to
conduct a restoration , 1907 to 1909 . The State of Cal i fornia
took ownership in 1 9 1 7 .
has been modi f ied with l ime
and fly ash for strength It was poorly maintained unt i l 1 9 4 4 when a second
and hardness . The mud is restoration was started during which wal l s were repaired , fl oors
mixed to a fluid and wal l s were replaced , and steel beams and tie rods were
consistency and pumped into embedded in wal l s . This work continued unt i l 1 9 4 8 . A third
the cracks under pressure . restoration was begun in 1 9 6 7 during which wal l s and foundations
The mud hardens to produce were strengthened .
a material that has
hardness , strength , and The 1 9 8 7 earthquake l e ft the wa l l s standing but extensively
permeabil ity cracked . The Department of Parks and Recreation of the state of
characteristics similar to Cal i fornia is preparing for stabil ization and restoration o f the
the adobe and that bonds to adobe .
the adobe . The material ,
tightly f i l l ing the cracks , Descript ion of the Bu ilding
keys together the irregular
surfaces on each s ide of It is a U-shaped building divided by adobe partitions into
the crack . This restores thirteen rooms . It has a part ial second story constructed o f
the interlocked e ffect o f boards o n sawn lumber j o ists supported i n pockets i n the wal l and
the original assembly of a lumber- framed roof supported on the tops of the wal l s . The
adobe bricks , stab i l i z ing wal l s and partitions are general ly about 60 cm ( two feet ) thick ;
the wal l . there is one 3 0 cm ( one foot ) thick. partition . The wal l s are
about 3 0 0 cm ( ten feet ) high in the f irst story . The gable end
KEYWORDS at the second story rises to about 2 4 0 cm ( eight fee t ) above the
second floor .

The 1 9 8 7 earthquake damage consisted principa l ly of cracks


Repa i r , cracks , seismic , in the adobe wal l s . Wal l s in which steel beams and tie rods were
stab i l i z ation , found ( probably from the 1 9 4 4 - 4 8 restorat ion) were cracked , but
restoration , inj ection , the most seriously cracked wal l s appeared not to have been
earthen architecture . reinforced . There are s igns of uneven settl ement of the building
probably related to floodings . The cracking patterns are
genera l ly vertical , indicated that , in general , the wal l s
responded t o earthquake shaking by rocking o n hori zontal p l anes ,
causing approximately vertical shear-related cracks near wal l
intersections where adj acent wal l s tended t o restrain rocking ,
and approximately vertical flexural cracks at some distance from
the intersections . Cracks extend through the wal l s and outl ine
blocks of l oose assembl ies of adobe bricks and mud mortar . The
rocking caused material to be eroded from the shear-related
cracks a s parts of the wal l s moved relative to one another . The
cracks with hori zontal orientation ( on which the wal l s rocked )
are generally closed by gravity . The genera lly vertical cracks
vary in width from f ine to as much as 2 . 5 cm ( one inch) or more .
There are many cracks in the 0 . 6 cm to 2 . 0 cm ( one-quarter to
three-quarter inch) width range . Some vertical cracks are at old
repairs containing brick , packed soil pieces of adobe occurred ,
and other material s . Some vertical cracks are partly the result
of separation due to fai lure of wal l s to return to their original
position a fter rocking , but most appear to be due to l o s s o f
material eroded by the relative movement each s ide of the crack .

Ant icipated Response to Earthquake Shaking

The response of the wal l s to earthquake shaking of the intensity


of past earthquakes is basically stable . The wal l s rock safely
at their base , l ightly restrained by the roof and f l oor
diaphragms . Assembl ies o f adobe bl ocks and mortar , loosened by
past shaking could be further l oosened and shaken from the wal l s
by future earthquake . If the loose assembl ies were stab i l i zed
into the wal l s , reoccurrence of the shear-related cracks and
fl exural cracks could be expected during future earthquakes , but
overa l l stabi l ity of the wal l s is anticipated .
Seismic Mitigation 337

The Demonstration Project

The Cal ifornia Department of Parks and Recreation sponsored a


proj ect to develop a means of stab i l i zation of the wal l s . The
demonstration proj ect was conducted by Kariotis and Associates o f
South Pasadena , Cal i fornia . It was proposed to stabil i z e the
cracked wal l s by inj ection of a fluid material into the cracks
that would harden to a soil-l ike f i l l . The material would have
to have propert ies of hardness , strength , and permeabil ity
s imilar to adobe to be compat ible . It should have insign i f icant
shr inkage after placement , and it should bond to adobe .
Samples of mud grout with various additional ingredients
were mixed in sma l l batches of about 9 kg ( twenty pounds ) . Molds
for casting mud samples were made by hand-compacting mo ist soil
around a piece of 5 cm ( two- inch) diameter pipe held at the
center of an approximately 15 cm ( s ix-inch ) diameter cyl indrical
metal can from which the top and bottom had been removed . When
the pipe was removed , a cyl indrical mold of s o i l was formed
measuring 5 . 7 cm ( two-one-quarter- inch) diameter by approximately
15 cm ( s ix- inch-long ) . The cyl indrical samples were cast by
pouring the fluid mud into the molds . Accompanying puddl e samples
were also made by pouring mud onto a moist soil or visqueen sheet
base . The samples were a l l owed to cure about four weeks . Then
the cyl indrical samples were sawn in cross section through the
can , mud sample , and mol d .

In order t o have a means of control l ing water content and


evaluate cons istency of fluid grout a simple flow-diameter test
was used : grout is poured from a 5 cm ( two inch) diameter by 1 0
cm ( four inch ) cylinder onto a flat , smooth , imper i ous surface to
make a roughly c i rcular puddle . The puddle diameter is measured ;
the average of two puddle widths measured perpendicular to each
other , one of which is the maximum width , is cal led the flow
diameter .

The sawn samples were observed for shrinkage , hardness ,


abrasion resistance , and permeabil ity . The puddle samples were
observed for breaking strength . Shrinkage was evaluated visually
by noting whether a crack formed at the perimeter between the
sample and the mold . Hardness , breaking strength , and abrasion
resistance were evaluated qua l itatively by hand , comparing the
samples with the soil mold material and with adobe in the wal l s
of the building . Permeab il ity was observed by setting samples in
a pan with sha l l ow water , and the time required for saturation o f
the sample was compared with time for saturation o f the adj acent
soil mold .

After the sma l l samples were evaluated , and a mix for


further testing was selected , larger samples were prepared for a
test of the inj ection procedure . To be able to evaluate the f l ow
of the fluid mud against adobe in a crack in the wal l , a test
frame was constructed to mount a panel of clear acrylic plastic
against the surface of an adobe wal l . Because of the irregular
surface of the adobe wa l l , the space between the plastic panel
and the adobe surface varied from zero to about 5 cm ( two
inches ) . The edges of the panel were sealed with wet newspaper ,
and mud was inj ected into the space between the plastic panel and
the wal l . The flow of the mud into the void was observed through
the plastic .
After the sample between the plastic panel and the wal l had
cured , the panel was removed and the sample observed for
hardness , breaking strength , abrasion res l stance , thickness of
the l eading edge of flow into thin z ones of the void , and bond to
the adobe .

The Injection Process

Mixed mud was transferred by buckets from a drum mixer to a


holding tank with an agitator . The holding tank fed a Moina pump
that forced the mud into a hose that del ivered the mud to a 1 . 3
cm ( one-ha l f - inch) diameter noz z l e that was used to inj ect the
mud into holes in the plastic panel . The pump was capable o f
devel oping 2 0 7 kPa ( 3 0 p s i ) pressure .

Results of the Demonstration Project

In general the samples modified with various amounts of Portl and


cement and/or fly ash and l ime had improved characteristics
compared with the unmod if ied soil samples . Samples containing
Portland cement tended to be harde r , stronge r , and less abradable
than the adobe of the building . Samples modi fied with fly ash
and l ime were made that had strength and hardness comparable to
adobe , but were somewhat more easily abraded than the adobe ,
especially at surfaces exposed to the a i r during curing . Twenty
grit s i l ica sand was added to some of the samples to try to
improve abrasion res i stance ; s i gn i ficant improvement was not
noted .

The unmodi fied soil samples had excessive shrinkage : cracks


formed around the perimeter of the sample s , and they became l oose
in the i r molds . The modi fied mud samples in general showed no
shrinkage ; except one sample developed a crack of about . 0 0 5 cm
( . 0 0 2 - inch ) width . A crack in the cross section o f one sample
seemed to be due to the e ffects of sawing .

I n itial shrinkage of samples in the molds was observed


within the first hour after placement . In most cases the
shrinkage produced a depress ion of about 1 . 3 cm ( one-hal f inch)
in the center o f the top surface of the sample over the 15 cm
( s ix- inch ) depth of sample below . However , initial shrinkage o f
about 3 . 8 t o 5 c m ( one-an-one-ha l f t o two inches ) occurred in
three samples . Most of the samples were mixed in smal l batches
in a s l owly rotating drum , but the samples exhibiting a large
amount of shrinkage were mixed at high speed . S ince excess ive
initial shrinkage could be detrimental to f i l l ing cracks , s l ow
mixing seems to be important , perhaps to avoid incorporation of
excess ive a i r into the mix . Too much water in the mix could also
contribute to excessive shrinkage .

Mud inj ected into the void between a plastic panel and the
adobe wal l flowed into spaces as thin as about 0 . 1 6 cm ( one­
s ixteenth inch) . There seemed to be l ittle resistance to
hori zontal f l ow of the grout into spaces wider than about 0 . 3 2 cm
( one-eighth inch ) . A head of about 2 5 cm ( ten inches ) was
developed above the inj ection noz z l e .

After curing for about four weeks , the inj ected mud had a
hardness comparable to the adj acent adobe and bonded to the adobe
wel l enough that when peeled from the wal l , a thin layer of adobe
came o f f with the sampl e . The inj ected adobe samples had
shrinkage patterns . Mixes that , indicated by excessive initial
shrinkage , improper mixing procedures and water content ,
displayed shrinkage patterns that were j udged to be excessive .
The mixes that had low initial shrinkage in the cyl indrical
sample had a pattern of cracks on the wa l l o f about . 0 1 cm
( . 0 0 4 - inch ) width between uncracked pieces about 2 . 5 cm ( one
inch ) in s i z e ; shrinkage indicated by thi s type of crack pattern
was j udged to be acceptable .

Permeabi l ity of the modi fied mud was compared with the
adj acent compacted soil by placing the sample cross section in
sha l l ow water . Saturation times for soil and cured mod i f ied mud
were virtual ly the same .

The hardened grout has a smooth texture and is a very l ight


grayish tan , in color . The l ightness of color is largely due to
l ime in the mix . It i s not comparable in appearance to adobes or
mud mortars we have encountered .

Conclus ions

The mud inj ection procedure appears to be a good alternative to


convent ional crack repa ir methods such as rebuild ing damaged
wal l s , removing and resetting adobe bl ocks , or f i l l ing cracks by
hand . Such methods are labor intens ive and require ski l l s that
are no l onger common , especially in urban areas l ike Southern
Cal i fornia .

The use of l ime and fly ash as mud mod i f iers contribute
bene ficially to both the fluid mud and to the hardened f i l l i n
the cracks . I n the fluid state , l ime gives the mud good water
retention qua l ities that w i l l help mainta in fluidity as it f l ows
across adobe surfaces . Fly ash , cons isting principa l ly of minute
ba l l s of glass , contributes lubrication propert ies to the fluid
mud . Fly ash , a pozz olonic material , reacts chemically with l ime
to form a cement icious component that improves the strength and
hardness of the crack f i l l ing material and reduces shrinkage .
The resul ting material , being basically a soil material , has
propert ies compatible with the adobe it i s repa iring .
S ince the mud was observed to flow into approximately 0 . 16 cm
( one-s ixteenth- inch-wide ) voids behind the plastic panel , it
should be expected that the procedure may be used to repa i r
Seismic Mitigation 33 9

cracks as sma l l as approximately 0 . 3 2 cm ( one-e ighth inch) i n


thickness . Shrinkage w a s observed to b e insign i f icant in the 5 . 7
cm ( two-one-quarter- inch ) diameter samples , indicating that
repa i r of cracks of up to 5 cm ( two inches ) in width may be
appropriate . The crack patterns noted in the plastic panel
samples indicate that water content and mixing procedures need to
be strictly control led to minim i z e shrinkage of the inj ected
material , especially when wide cracks are being repa ired .

Improved abrasion resi stance may be attainable i f coarser


sand or natural sand i s used . However , use o f coarse sand w i l l
greatly reduce the l i fe of a Moino pump .

Proposed Appl i cation

It is now proposed to use the inj ection of mod i f ied mud to


stab i l i z e the wa l l s of the pio Pico Mans ion . The stab i l i z at ion
proj ect w i l l be conducted us ing a s i lty sand soi l . The selected
s o i l w i l l be tested for suitab il ity for the proj ect by making
mod i f i ed mud samples to be evaluated for shrinkage and hardness .
The test program for the selected soil w i l l be conducted by
making samples us ing packed soil molds as was done for the
demonstrat ion proj ect . Proposed proport ions by weight for the
samples are as follows :




6 0 parts soil
(20
20 parts s i l ica sand
18 parts fly ash ( type F)
2 parts l ime ( type S )
grit )

water a s required for proper consistency .

Proportions by weight instead of volume are required because o f


the tendency of s o i l t o bul k .

The wa l l cracks have been mapped and inj ection locat ions have
been des ignated on proj ect drawings .

To augment the resistance of the wal l s to shear-related


cracks adj acent to wal l intersect ions , 1 . 9 cm ( three-quarter
inch ) diameter threaded fiber-glass rods , rod w i l l be embedded to
cross the intersect ions . The rods w i l l be instal led 5 cm ( two
inch) diameter holes dril led that w i l l be pumped ful l of the
modi fied mud .

Proj ect spec i f ication are as fol lows :

1. The speci fications for the test procedure are :

a) Prepare packed soil molds on the ground ins ide a 15


cm ( s ix- inch-diameter) by 3 0 cm ( twelve inch ) high
metal cyl inder of the type used for concrete
compress ive strength testing . The metal cyl inder sha l l
have top and bottom removed . Form the mold with a 5 cm
( two-inch) diameter standard pipe at the center of the
cyl inde r . Place moistened s o i l in approximately 2 . 5 cm
( one inch ) l i fts around the pipe . Tamp each l i ft of
the soil firmly with a rod of about 2 . 5 cm ( one inch)
diameter . Bui l d mold to a depth of 30 cm ( s ix inches ) .
Finish the top of the mold to a level surface . Rotate
and l i ft the pipe often to keep it free from the
compacted soil surface .

b) Mix grout in proport ions and flow diameter to be


tested .

c) Pour grout ioto the molds in three l i fts , puddl ing each
l i ft with a wooden stick . Top the sample with a 1 2 . 5
cm ( five- inch ) diameter puddle about 2 cm ( three­
quarter inch ) over the top of the compacted soil mol d .

d) The fol lowing day , note whether settl ement o f grout has
occurred . If the top surface of the puddle drops below
the top surface of the mol d , there .is excess ive initial
shrinkage requiring correction of mix or mixing
procedure before proceeding .

e) After twenty-eight days , saw the mold in sections


without disturbing the sample and observe the packed
soil -grout interface . A crack or separation between
the sample and the packed soil is an indication o f
excess ive shrinkage after set , which sha l l b e corrected
before proceed ing . ( Be sure that an observed crack is
not caused by the sawing operation . )
340

f) store sawn samples i n a dry place a t room temperature .

g) Fourteen days a fter sawing compare the hardness of the


sample with the hardness of the adobe at the building
by scraping with a metal tool . Acceptable hardness i s
equal t o o r s l ightly harder than the adobe . Adj ustment
of the proportion of cement it ious material ( fly ash and
l ime ) to soil and sand may be required to achieve
acceptable hardness . The proportion of l ime to fly ash
sha l l be kept constant .
h) An acceptable grout mix sha l l have the fol l owing
characteristics as determined above :

acceptable initial shrinkage


no shrinkage after set
acceptable hardness
materials and procedures for making grout for the
proj ect sha l l be identical to those that produced
the acceptable samples

2 . Spec i f i cations for the inj ection work are :

a) Mixing
Combine the dry ingredients in a s l owly rotating drum
mix . Add fly ash and l ime to the s o i l -sand mixture .
Mix unt il uni formly blended . Add water gradua l ly .
Test flow diameter when mixture reaches a uni form
consi stency .

b) Mixing time
Grout not used within one hour after adding water sha l l
be d i scarded .

c) Pumping
Pump grout from a mortar pump with agitat ing hopper
that is capable of developing a pressure of 3 0 p s i .
Pump through a flexible hose to a 1 . 3 cm ( one-ha l f­
inch) diameter thin-wa l l metal tube noz z l e .

d) Consi stency
Cracks less than 1 . 3 cm ( one-ha l f inch ) in width :
14 . 6 cm ( f ive-an-three-quarter inch ) f l ow d iameter .
Cracks 3 . 8 cm ( one-ha l f to one inch ) width : 1 3 . 3 cm
( five-an-one-quarter-inch) flow diameter .
Cracks 2 . 5 cm ( one inch) and greater in width :
repointing consi stency ; shal l ma intain its shape and
not flow when deposited by the noz z l e .

Flow diameter
The measure of fluid grout consistency sha l l be
measured as fol lows :

F i l l a 5 cm ( two- i nch ) diameter by 10 cm ( four- inch )


high cyl inder with grout . Pour the grout from the
cyl inder from a he ight of 15 cm ( s ix inches ) onto a
smooth plastic sheet on a flat surface to form an
approximately c ircular sample . Measure the width o f
the sample o n two perpendicular axes , one o f which
sha l l be the maximum width of the sample . The f l ow
diameter should be taken as the average of the two
widths .

e) Preparation of cracks

1. Expose the ful l length of the crack to be


inj ected � Remove plaster and other obstructing
material s , being careful not to disturb framing
members .

2. Remove accessible loose pieces o f adobe and mortar


from the crack .

3. Blow dust and small particles from the cracks with


compressed a i r .
Seismic Mitigation 341

4. Caulk cracks on both s ides of wall with wet


newspapers . Caulk in lengths of about 15 cm ( s ix
inches ) between 1 . 3 cm ( one-hal f - inch ) diameter
openings . Press the wet newspaper into the cracks
at a depth about equal to the cracks width .
Alternatively , caulk with sti f f grout to which two
parts Portland cement have been added . When caulk
i s hard , drill 1 . 3 cm ( one-hal f- inch ) diameter
holes into the crack at 15 cm ( s ix inches ) on
center . Use non- impact dri l l ing equipment .

f) Inj ection
Inj ect each crack from one s ide of the wal l onl y .
start from the bottom and inj ect into 1 . 3 c m ( one­
hal f- inch ) diameter openings in sequence working up .
Plug inj ected holes with wet newspapers . Flow of grout
sha l l be observed from 1 . 3 cm ( one-hal f- inch ) diameter
opening on the oppos ite s ide of the wal l . I f grout
does not appear on the opposite s ide of the wal l , the
reason w i l l be determined and corrected before
proceeding . As grout flows from a hole on the opposite
s ide of the wal l , the hole sha l l be plugged with wet
newspape r . Proceed with inj ection unti l the crack i s
f i l led with grout .

g) Newspaper caulking shal l be left i n placed unti l grout


has set ( approximately two or three days ) . Remove
newspaper caulking before the grout hardens
( approximately seven to ten days . )

h) Clean-up
Remove a l l hardened sp i l l s and unused grout from the
s ite and d i spose of it legally . Leave floors broom
clean .

3. Spe c i fication for the threaded fiberglass rods are :

a) Fiberglass ,rods sha l l by 3/4 " diameter threaded rods o f


glass f iber i n vinyl ester res in ; Fibrebolt as
manufactured by Morrison Molded Fiber Glass Company .
Nuts sha l l be fiberglass . Washers sha l l be galvani zed
standard cut steel washers .

b) Dri l l holes for 2 " diameter fiberglass anchors with


non- impact tool s .

c) Holes for fiberglass anchors sha l l b e d r i l l ed a fter


cracks in the vicinity have been inj ected and the grout
has set .

d) Inj ect ion holes for fiberglass anchors sha l l b e 1/ 2 "


d iameter and sha l l intercept the bottom o f the hole
into which the anchor i s inserted .

e) Inj ect ion grout into the inj ection hole , f i l l ing the
anchor hole and embedding the fiberglass rod in one
operat ion . Plug the inj ection hole with wet newspaper .
Remove the plug a fter grout has set and before it
hardens .

Further Research

Wal l s repaired by the proposed inj ection process have not


been subj ected to in-s itu testing , nor to actual earthquake
shaking . A test program to evaluate how wel l the method
actual ly restores the integrity of repaired wal l s would be
an important next step . Future testing should focus on the
e f fect the grouting process has on resistance to formation
of shear-related cracks adj acent to wal l intersections as
well as the e f fectiveness on restoring the bond of loose
assembl ies into the wa ll .
34 2 Adobe 90

ABSTRACT " CR ITERIOS Y TECNICAS DE RESTAURAC ION APL I CADAS EN LOS MONUMENTOS
DE ADOBE EN EL PERU "

The author outl i n e s


geograph i c features o f the Robe rto Sama n e z Argumedo
Peruv i a n landscape M . S . Arqui tecto
previous l y occupied by Apart ado Postal 2 8 9 Cusco-Peru
ancient civi l i z at i o n s that
left behind a substantial
legacy o f structure s made I ntroduccion :
of e arth . The 3 0 0 years o f
Spanish coloni z ation El Peru e s t a s i t uado en la parte central y occ ide ntal de Ame r i c a
added to thi s legacy . de l Sur , e n un territorio e n el q u e l a Cord i l l e ra And i n a presenta
d i s t i ntos planos a l t i tudinale s , con marcadas di feren c i a s de c l ima ,
Thi s paper presents an topog r a f i a y vegetacion que determinan la e x i s t e n c i a de un l i toral
anal y s i s of s o c i a l and desertico , de valles i nterandino s , a l t iplanos y de se lva trop i cal .
e conom i c factors bearing Grac i a s a su cercania a la l i nea ecuator i a l , l a s partes a l t a s de
on the survival o f these los Andes tienen cl ima s benignos que han perm i t i do su ocupacion
structures over the last desde epo c a s muy antigua s .
1 5 0 years . F i ve spe c i f i c
i ntervent i o n s a r e descr ibed Los grupos humanos que ocuparon e s e territo r i o desde hace d i e c io
and documented , with cho m i l anos se adaptaron a sus con d i c i o n e s geograf i cas , obtenien=
d i s cu s s ion o f the do el mej or provecho de sus variados e c o s i stema s y a l c a n z aron un
theore ti cal ori entation and alto grado de orga n i z a c ion y desarro l l o tecnolog i co .
resul t s . F i n a l l y , the El emp l e o de bloque s de t i e rra s e cado s al sol 0 t i erra conf i nada
exper ience of restoring y api sonada , con s t i tuyo el procedimiento construct ivo mas ut i l i za­
archeo logical and col o n i a l do e n el Peru a n t i guo y 1 0 encontramos e n l a costa arida y sin llu
monuments made w i th adobe vi a s , donde se combi naba la t i erra con canas y fibras 0 en l a s zo­
is summa r i z e d , w i th the n a s l luvi o s a s de l a cordi l l era , donde exi ste abundante p iedra que
re commendation that more se emp l e aba en la c imentacion de l a s e d i f i c a c ione s .
research is needed on the
preservation of the Andean Desde 2 0 0 0 anos ante s de C r i s to se l evantaron compl e j as edi f i ca­
mural paintings that ciones de t ierra creando una arqui tectura monumental expr e s ada e n
adorn the wal l s o f adobe p i r amides truncada s , temp l o s y conj untos urbanos d e g r a n cal idad .
hou s e s and churches bui l t Al producirse la conqu i s t a e spanola en el s i g l o XVI se introduj e -
duri ng the colon i a l pe riod . ron camb i o s fundame n t a l e s en l o s a spectos tecnolog i cos y cultura­
l e s , pero se continuo construyendo con adobes , por ser este un ma­
terial al que tamb i e n e s taban habi tuados los europeo s , que 1 0 se­
KEYWORDS gui r i a n ut i l i z ando durante los c a s i t r e s c i entos anos que duro el
Virreynato del Peru .
En el periodo que s i guio a la i ndepende n c i a del dominic e spanol
Earthquake damage , ocurrida e n l a primera mi tad del s i glo pa sado y casi hasta la de­
recon struction , cada de los anos cuarenta del p r e s ente s i glo , e l adobe cont i nuo
restorat i on , improved s i endo el mater i a l b a s i co para l a s edi f i caciones y en la actual i ­
adobe blocks , mural dad todav i a representa l a uni c a a l ternativa d e baj o costo para que
paintings . hi story . gran parte de la pob l a c i o n pueda construir sus vivi enda s .

Las pr imeras i nterve n c i o n e s de res taurac i o n :

Como en pocos lugares del mundo en el territo r i o d e s c r i to f l o ­


reci eron nume r o s a s c i v i l i z a c i o n e s y culturas q u e a trave s de los
siglos dej aron un cuantioso l egado monumental , que deber i a haber­
se conservado celosamente para convert i r s e en un e j emp l o de l a
capac idad tecn i c a y art i st i c a d e l a s generaciones p a s adas y con s ­
t i tu i r un hito d e identidad p a r a los peruano s . S i n embargo , por
mucho t i empo , esa heren c i a del pa sado fue olvidada e i ncomprendi­
da sin mere cer n inguna accion de protec c i on .

En la decada de 1 9 3 0 a 1 9 4 0 se e f ectuan l o s pr imeros trabaj o s


para conservar los monumentos colon i a l e s mediante i nterve n c i o n e s
Forta l e z a de Paramonga con s­
que denotan emp i r i smo y falta de c r i terios adecuado s . Por 1 0 g e ­
truida con t i erra , en la c o s
neral s e r e speto p o c o l a o r i g i nal idad de los mate r i a l e s y s e
ta peruana , antes de la l l e =
pre f i r i o l a recon struccion antes que l a r e staurac ion . A r a i z del
gada de l o s e spano l e s .
terremoto que a fe cto a la capital del P e ru en 1 9 4 0 , caus ando
cuan t i o s o s danos en los monumentos r e l i g i o s o s y en la arquite ctu­
ra c ivi l , se res tauro un numero importante de i g l e s i a s colon i al e s
construida s con adobe· y e s tructuras l i geras d e c a n a y barro , que
en su mayor parte hab i an s ide mod i f i cadas y remodeladas en el
s i g l o X I X . En e s o s trabaj os la preocupacion se encamino a la re­
cuperacion de l a f i sonomia o r i g i n al de las f achadas e inter iore s ,
s a c r i f i cando l a autent i c i dad de los test imon i o s e n aras de
reconstrucc ione s hi sto r i c i stas .

A partir de la dec ada de los ano s sesenta se l l evaron a cabo l a s


prime r a s labore s en l o s monumentos arqueologicos d e l a c o s t a p e ­
ruana e ncaminadas a su reconstruc c i o n , c o n el obj e t i vo de conser­
varlos y sobre todo de permi t i r que los vi s itantes perciban como
Construcciones de adobe de fueron e s o s t e s t imon i o s ante s de su abandono y ruina . Los resulta
epoca I nca en el s i t i o de dos fueron muy d i s cutibles y obj etados por los arqueologos porque
Tambo Col orado , con restos se e f ectuaron comp l etando los mat e r i a l e s o r i g i n a l e s de los edi f i ­
de p i ntura de vivos colores . cio s , s i n contar c o n l a s evide n c i a s que j us t i f iquen la reproduc­
c ion de formas hipotet i ca s .
Seismic Mitigation 343

S i tuacion actual y tareas futuras de conserva c i on :

Por 10 general los p l anes y accione s estata l e s para conservar y


recuperar el patrimon i o monumental peruano , han s urgido a r a i z de
catastrofes como los terremotos que periodi camente dej an sent i r
sus e f ecto s . A s i como e l s i smo q u e a f e cto L ima en 1 9 4 0 determino
las primer a s obras de restauracion en las i g l e s i a s colon i al e s , f ue
a r a i z del terremoto que se produ j o en Cus co en 1 9 5 0 que se conto
con l a p r imera m i s ion que e f e ctuaba l a UNESCO fuera de su s ede ,
para ori entar la recuperacion del conj unto monumental .

En 1 9 7 0 la c i udad de Truj i l l o ub i c ada en el norte del pai s , fue


dafiada por un nuevo terremoto y para contribu i r a l a recuperacion
de sus monumentos se orga n i z � otra m i s ion de l a UNESCO que prepa­
ro un conj unto importante de proyectos , re cogi endo los criterios
contemporaneos de restaura c i on e introduci endo l a modal idad del
traba j o i nterdi s c i p l i nario . Con l a creacion del I n s t i t uto Nacional
de Cultura en 1 9 7 3 se organ i zaron l a s areas encargadas de l a con­
servacion del patrimon i o h i storico y s e encaminaron adecuadamente
los criterios para la res tauracion .

En e s e periodo se recibio una importante part i c ipacion de l a


UNESCO para desarrollar u n plan en l a zona de Cus co p o r un periodo
I g l e s i a Jesuita de los De­ de s iete afios . Mas ade l ante en el marco de un p l an de desarrollo
samparados edi f i cada ·en tur i st ico que conto con fondos de un prestamo del Banco I nterame­
1 6 7 0 , demol ida en 1 9 3 7 . r i c ano de Desarrollo se empe z aron los traba j o s directos en l o s
monumentos arqueologicos y colon i a l e s i n c l uyendo una amp l i a l abor
de conservacion de bienes mueble s .

L a mayor parte de los monumentos i ntervenidos en e s a epoca e ran


de adobe y planteaban problemas nuevos que motivaron traba j o s de
inve s t i ga c ion para encontrar l a tecnologia apropi ada , desarro l l an­
do c r i t e r i o s para interpretar las l e s i ones y e s tablecer los proce­
dimi entos metodologicos para trabaj ar correctamente en e s e t ipo
de e s t ructura s . Los resul tados fueron muy al entadores y determi­
naron la creacion de cursos semestrales de capacitacion que , a su
ve z , perm i t ieron l a pre sencia en Cusco de destacado s e s pe c i a l i s tas
en l a mate r i a .

E s e proceso tuvo el mer i to de permi t i r una evo l uc i on de los co­


nocimientos y lograr que un con s i de rable numero de tecnicos y pro­
f e s i on a l e s compartan conceptos comunes con respecto a las teor ias
y criterios de r e s taura c i o n , dej ando de f i n i t ivamente l a s contro­
vertidas ide a s aceptadas hasta la decada de los afios se senta .

En 10 relativo a la res taura c i on de los monumentos h i s toricos


cons truidos con adobe , l a expe r i e n c i a que se obtuvo en e s o s afio s ,
con los trabaj os real i z ados e n Cus co , permite r e s um i r los s iguien­
tes principios bas i cos :

I g l e s i a de La Merced en L i ­ _ En los monumentos h i st o r i cos edi f i cados con adobe s , estos e l e ­


ma , mostrando su aspecto neo mentos cons truct ivos tamb i e n s o n test imon ios d e l a tecnologia
c l a s i co antes de l a r econs: original y son parte de l a documentacion h i s to r i c a que debe
trucc ion de 1 9 3 9 . conservarse en 10 posible , por 10 que la s u s t i tucion del ma­
terial en forma i n d i s c r iminada 0 l a reconstruccion son prac­

t _
t i c a s que no se pueden aceptar .

Al reemp l a z ar los adobes deteriorado s como parte del proceso de


res taura cion , n e c e s a r i amente se debe mantener la homogene idad ,
10 cual e x c l uye el uso de adobes e s t ab i l i zado s .

_ Es recomendab l e que para la res tauracion de edi f i ca c iones de


adobe que aun es tan en uso , como los inmuebl e s colon i a l e s y
repub l i cano s , que demand an mayores requerimientos de e s tabil idad
e s t ructural , se u t i l i cen adobes nuevos de dimens i ones igual e s
a l a s d e los o r i g i n a l e s pero fabri cados con tecnologia mej orada
en sus caracte r i s t i cas mecani ca s y f i s i ca s .

_ La preparacion de adobes mej orados para ser uti l i zados en reem­


p l a z o de aque l lo s cuya sustitucion es impre s c i ndible , se hara
med i ante la sel eccion cuidado sa de los componentes y un adecua­
do p ro c e s o de control de l a e l aboracion y secado .

_ Antes de emprender un trabaj o de restauracion es impre s c i ndible


e stablecer un di agno s t i co para conocer los s i ntomas , l a natura­
l e z a y las caracte r i s t i c a s de las e structuras a f ectadas . E s por
eso que adema s de l a i nvestigacion de l a s l e s i ones y s ignos ex­
Aspe ctos de la mi sma I g l e s i a ternos de deteri oro se debe intuir cual es l a relacion entre
reconstruida tomando como r e causa y e f ecto , v i s l umbrando los remedios mas e f i caces para
ferenc ia grabado s y fotogra: anul ar l a cau s a .
f i as del siglo X I X .
_ E s recomendable que l a s so luciones de consol idacion y refuerzo
e s t ructural se bus quen i ncorporando al ternativas locales y pro-
344 Adobe 90

p i a s , pre f i r i endo la rna no de obra y l o s materi a l e s trad i c ionales


exi stentes e n el lugar . Excepcionalmente y cuando l a s cond i c i o -
n e s 1 0 j us t i f iquen se podra recurrir a t e c n i c a s y materi a l e s i n ­
dustri a l i zado s .

Aun conoc i endo l a s causas que origi nan los danos en los monumen­
tos de adobe , e x i s te n c a s o s en que el di agno s t i co es comp l e j o y
no hasta el anal i s i s matematico de l o s p r i n c i p i o s e structur a l e s ,
s i no la i ntui c ion del e s pe c i a l i sta experimentado . Como e s o s e s ­
pec i a l i stas s o n e s c a sos y s e deben formar en 1 a prac t i ca , e s
conve n i e nte e l trabaj o d e equipo c o n l a part i c ipacion d e l arqui ­
Huaca del Drag6n en la cos­
tecto res taurador .
ta norte , reconstru ida en
1 9 6 3 comp l e tando muros y re
_ Por ult imo e s fundamental tener en cuenta que no se puede part i ­
l i eves .
c ipar e n u n proyecto y menos e n una obra de restaura c i o n , s i no
se conoce a fonda que son y por que t i enen valor l a s edi f i cacio­
ne s a n t i gua s . La forma c i o n y capa c i t a c i o n espec i a l i z ada e n 10
conceptual y pract ico , e s indi spensab l e .

Palacio pre- I nca de Puruchu­


co cerca de L ima reconstru i ­
d o compl etando arbitrariamen
te muros y techo s .
<D �...ZO Y
RE n.l
T
P\. RU
POR T RAMO
l.J(JIt
ll:) O£ LA CIMENTACION
II Al..TD'lNAOO$

@ MII'uatZOS EN LOS DlNT!!:,,!:S O£ LOS VANOS


Por 1 0 general e n los monumentos que se i nterv i n i eron en la zo­
na de Cusco y otros lugares del p a i s de sde 1 9 7 5 , se adoptaron l a s
recome ndaciones a n t e s mencionada s , cons iguiendo resul tados s a t i s ­
factori0 3 , e j ecutando los traba j os a part i r d e l refuerzo de l a c i
Palacio de epoca I n c a e x i s ­
menta c i o n , l a consolidacion d e l o s muros s u s t i tuyendo puntualmen=
tente en Yuca y cerca d e Cus
te los e l ementos danados , l a co1 0cacion de " l l ave s " de madera pa­
co , restaurado con c r i t e ­
ra re for z a r e l e ncuentro de muros y l a adicion de d i n t e l e s mej ora
r i o s adecuado s en 1 9 7 6 .
dos en los vano s .

S i n embargo , l a r e s t auracion de edi f i cac ione s a n t i guas construi­


das con adobe s puede presentar c a s o s de mayor comple j i dad , que no
s i empre pueden ser resuel tos a base de. los e l ementos c i tado s , que
son parte de la tecnologia tradiciona1 mej orada . En muchos c a s o s
l o s requer imientos para l a nueva func ion de l o s inmuebles anti -
guo s , la seguri dad de los usuarios 0 de l a s coleccione s museogra­
f i ca s , obligan a tener e n cuenta solucion e s que sean mas seguras
ante el r i e sgo s i smico .

Deben aceptar se , por 10 tanto , a l ternativas que i ncorporen e s ­


tructur a s adi c i o n a l e s d e refuer zo , d e diseno especial para que
sean compatibles con las normas y p r i n c i p i o s de la res tauracion
y que permitan garant i z ar las con strucciones con adobe e n areas
I nterior del mi smo p a l a c i o s i smi ca s .
mo strando l a di ferencia en
tre 1 0 original y 1 0 re- - Se han exper imentado con exito e s tructuras de concreto armado
pue sto e squemati came nte , o madera i n corporadas a l o s muros de adobe mediante disenos e s ­
en base a l a s evide n c i a s p e c i a l e s para evitar e l dana d e l mater i a l d e refuerzo sobre los
arqueolog i c a s ( a la dere­ adobe s que en caso de s i smo , t i en e n comportamientos di f e r e nt e s .
cha ) .
En el caso de l o s monumentos pre-hispanicos construidos con
adobe s , que por su cond i c ion de t e s t imon i o s arqueo logicos no se
r e s t i tuyen a una funcion actual , los reque rimientos de conserva­
c ion han e stado e ncaminados a proteger los edi f i c i o s e n forma
dec i dida , aunque esto imp l i c a poner cubiertas temporal e s que
nada tienen que ver con las o r i g i n a le s .

Para me j orar la r e s i stencia de los adobe s , se han i ntroducido


importantes ava n c e s tecni cos , apl i c ando emul s i on e s a c r i l icas con
10 cual se logro endure cer e impermeabi i i z ar e l mater i a l antiguo .
Seismic Mitigation 345

Un e s pe c i a l i s ta de la Unive r s i dad de Tur i n enviado al Peru por


l a UNESCO probo la apl i cac ion de un producto reminera l i z ante , el
eter de a c i do s i l i c i co para reforzar l o s s i l i catos que se encuen­
tran en l a compo s i cion del barro , hac i endolo impermeable . E l s i l i ­
cato d e e t i l o se di suelve e n al cohol y se caracte r i z a por tener
gran capac idad reticular y por 1 0 tanto se obti ene buena penetra­
cion en f r i s o s y muros .

E s e producto se ut i l i z� con exito para pre servar l o s adobes y


morteros de barre de reve s t i miento de monumentos arqueolog i c o s en
la zona de Cusco y en var i o s s i t i o s de l a costa peruana .

Otro a specto que empezo a cobrar importa n c i a en la decada de


F r i so con f igur a s en r e l i e ­ los anos s etent a , fue el de la con serva c i on de la p intura mural
v e en l a c iudad pre - hi spani a s o c i ada a l a s e s tructuras de adobe , tanto del per iodo pre- h i s ­
ca de Chanchan , tratado s - panico como d e la epoc a colonial . Grac i a s al cl ima s e c o y arido
con s i l i cato de e t i l o . de la c o s t a peruana se han con servado importantes t e s t imon i o s de
pintura mural , en los s i t ios arqueologicos de Panamarca , Garagay ,
Paramonga y S e ch i n . Lamentablemente muchas de e l I a s se deter i ora­
ron por haber l a s dej ado al de s cubi erto 0 porque fueron restaura­
das emp i r i camente .

CORTE TRANSVERSAL o!""=�"""""�


Rel evamiento de la i g l e s i a de Oropesa , cerca de Cus
c o , construida a l f inal de� siglo XVI . Conserva im=
portantes pi nturas mural e s de varios periodo s .

Ma s abundante es la pintura mural exi stente en l a s i g l e s i a s co­


lon i ales , en e s p e c i a l en l a s numerosas parroq u i a s y capi l l a s ru­
ELEVACION FRONTAL L rales s i tuadas entre Cusco y el l ago T i t i caca . Encontramos a l I i
pi nturas renace n t i s t a s e j e cutadas entre 1 5 7 0 a 1 6 3 0 , con intere­
s ante s tema s mitolog i co s , tamb i e n p i nturas barrocas del S i g l o
XVI I y mur a l e s de fuerte i n f l ue n c i a indigena p i ntados desde el
S iglo XVI I I hasta el inicio del s i glo pas ado .

La profusion de pi nturas mura l e s se expl i c a por su ut i l i za c ion


como i n s trumento de cateque s i s y por s u valor didactico para i n ­
c u I car l a rel igion c r i s t iana a los indigenas . La e x i s t e n c i a de
e s a s p inturas tambien en c a s a s colon i a l e s demue stra que se trata
de una expre s i on art i s t i c a de gran aceptacion y d i f u s i on .

Los e s tudios sobre la pi ntura mural son relativamente recien­


te s , en p a r t i cul a r , l o s referidos a las p inturas al templ e , po­
l i cromada s , que se apl i can directamente sobre una base de prepa­
r a c i on enc ima de los enlucidos de barre que revisten los adobe s .
Las re s t aura c i ones que se emprendieron en Cus co a mediado s de los
anos setent a , introdu j eron tecni c a s de exploracion que permi t i e ­
ron de scubr i r gran c antidad d e p inturas cubiertas p o r capas p o s ­
Baut i sterio de la i g l e s i a teriores y recuper a r l a s � r a s p a c i ente trabaj o .
d e Orope sa durante el pro
ceso de recuper acion de - S i bien se lograron buenos avances en e l conocimiento de l a s
la pi ntura mural del s i ­ t e c n i c a s y componentes d e l a p intura mural , hab i endo se expe r i ­
g l o XVI I , cubi erta c o n e s mentado y probado d i v e r s o s mate r i a l e s y productos quimi cos p a r a
tuco . su con s 0 1 i da c ion , sera nece s a r i o h a c e r un g r a n es fue r z o p a r a
emprender a c c iones d i r igidas a conservar l a s p i nturas mural e s ,
que d e otro modo desapareceran e n pocos anos .

La e s t ructura s o c i a l que perm i t i a a l a s i g l e s i a s rura l e s contar


con t i e r r a s de cul t i vo para sufragar los gastos de su cui dado y
mantenimiento des apare c i o con la re forma agrar i a , quedando e s t a s
en el m a s grave abandono , q u e e s t a produci endo su r u i n a a c e l e r a ­
da .
Adobe 90

Lamentabl emente , por ahora , s o l amente se e s t a procediendo a in­


ventariar las igl e s i a s , cap i l l a s y casas con pi ntura mura l , pero
no e x i s ten recursos para su conserva c i o n .

Dramatico e s tado de conser


vacion de l a mayor parte
de l a s igle s i a s decorad a s
c o n pi ntura mura l .

Adecuada re stauracion y r e i n tegra­


cion de la pi ntura mur a l en Andahuay
l i l l a s y con sol idac ion prevent iva en
e l Mol ino de l o s I ncas , de Acomayo .

Por otro l ado , l o s traba j os que en e scaso numero se vienen l l e ­


vando a cabo p o r organi smos e s tatal e s , para conservar a l gunas de
e s a s i g l e s i a s , e s tan a cargo de tecnicos y p e r s on a l que no t i ene
una idea cabal del a l t i s imo valor de estos tes t imon i os . Obreros
de re stauracion que conocen e l trabaj o de con s o l idacion en l o s
muros de adobe , a f a l t a d e adecuada or ientac ion , interv i enen en
muros p i ntados de l a mi sma forma como 1 0 hacen en paramentos s i n
p intura , o c a s ionando danos i rreparab l e s .

Const ituye pues , un reto para e l futuro i nmedi ato e studiar me­
didas de proteccion para l a s p i nturas mura l e s y metodos de i n ­
tervencion que se adapten a l a s c i rcun stancias y no c a u s e n danos
a ese t ipo de expre s iones pi ctoricas .

Las tecnicas que pueden


ser correc t a s para l a con­
sol idacion de los muros de
adobe , pueden afectar l a
p intura mura l .

Lamentable i ntervenc i o n reciente para


reparar f i suras e n e l muro de adobe
s i n r e spetar la pi ntura . A la derecha
se ven l o s danos en la pi ntura al re­
Es impre s c i ndible pre ser­ parar la cubierta .
var l a edi f i cacion a n t i ­
g u a para salvar l a p i ntu­
ra mural que es parte de
el l a .
CONCLUS IONES :

Hemos senal ado brevemente , al i n i c i ar l a ponen c i a , l a s e s pe c i a­


l e s c i r cuns t a n c i a s geograf i c a s que h i c ieron po s ib l e que en el P e ­
ru antiguo se den grupo s humanos d e " al t a cultur a " convi r t i endo se
en uno de l o s l ugares del mundo donde se desarro l l aron grandes c i
vi l i z a c i one s . E n e l d i f i c i l territorio de l Peru e x i s t e n a l rededor
de 5 0 , 0 0 0 s i t i o s arqueologicos y un numero menor , pero no menos
importante , de test imon i o s h i storicos de epoca colonial y repub l i
cana . Gran parte d e e s e patrimon i o e s t a e n pe l igro de destrucc i on
a falta de una pol i t i ca adecuada de di fusion y conocimiento de
l o s valores prop i o s y a l a f a l t a de acciones de defensa y conser­
vac i on .

A 1 0 l argo de l o s u l t imo s 1 5 0 anos de vida repub l i cana , parte


de ese importante patr imon io cultura l , por 1 0 general construido
con adobe , ha s i de depredado y destruido .
Seismic Mitigation 347

S i n embargo , la expe r i e n c i a r e c i ente en el campo de la restaura­


cion de monume ntos de adobe e n el P e ru e s import ante y puede ser
muy ut i l no solo para futuras i ntervenciones locale s , s i no para
quienes busquen i nformacion para emprender tareas s imi l are s , pero
a s u vez los restauradores de monumentos y obra s de arte de e s t e
p a i s requieren todo e l apoyo y colaboracion de l a comuni dad i nter­
nacional , para pres ervar su patrimonio hi stori co .

Uno de l o s rubro s mas n e c e s i tados de a c c i on i nmediata , por s u


caracter s i ngular y su g r a n v a l o r hi stor i co , e s l a p intura mural
andina del per i odo coloni a l que ame r i t a el apoyo y el e s fuerzo
de i n s tituciones y tecnicos empefiados e n l abores s im i l are s , para
evitar que prosiga su deter ioro y para s e ntar l a s bases c i e nt i f i ­
Salon con f r i so y arteso­ c a s d e su correcta conservacion .
nado p intado s , e n la casa
del s i g l o XVI I de Don Fer
nando de Ver a , e n Cusco .
B I B LIOGRAFIA :

1 .- Jose Garcia Bryce , " La arqui tectura en el Peru desde 1 8 3 9 " ,


e n s up l emento especial del diario " E l Comercio " , Lima 4 de
mayo de 1 9 8 9 .

2.- Harold E . Wethey , " Co l o n i a l Archi t ecture and S culpture i n


Peru " , Cambridge , 1949 .

3.- Rogger Ravines , " Chanchan metrop o l i Chimu " , I n s t i tuto de


E studios Peruanos , Lima , 1 9 8 0 .

4.- A l f re do B a rbacc i , " 1 1 r e s t auro dei monumen t i i n I t a l i a " ,


I n s t i tuto P o l i g ra f i co del l a Stato , Roma 1 9 5 6 .

Deta l l e de la p i ntura mu­ 5.- Enrique del Moral , " De f en s a y conservacion de l a s c iudade s
ral re staurada cuidado sa­ y conj untos urbanos monumentale s " , Academia de Arte s , Mex i co ,
mente di ferenc iando l a s 1977 .
parte s re i ntegradas de
las o r i g i nale s . La r e s ­ 6.- Roberto Saman e z , The restoration of mud-bri ck structures i n
taura c i o n , c o n a s i s t e n c i a hi stor i cal monuments o f t h e Andean region of Peru ,
d e la UNESCO , concluyo "Appropriate technologi e s in the conservation of cul tural
en 1 9 7 8 . property " , the UNESCO P re s s , Par i s 1 9 8 1 .

7 .- Dani e l S chave l z on , " Cambio y transforma c i o n : la restauracion


arqueolog i c a e n Ame r i ca Latina entre 1 9 7 0 y 1 9 8 0 " , en
Ana l e s No . 2 5 , Unive r s idad de Buenos Aire s , Arge n t i n a 1 9 8 7 .

8.- Roberto Samane z , " Mural P a i n t i ng on adobe wal l s dur ing Pe ­


ruvi an colonial t ime s - i t s restoration and conservation " .
C a s e studies i n the conservation of stone and w a l l p a i n t i ngs ,
I I C , Bologna Congre s s , 1 9 8 6 .
Adobe 90

ABS TRACT

Th i s paper provide s an over­


SEISMIC
AN OVERVIEWSTRENGTHENING OF HISTORIC ADOBE BUILpINGS IN CALIFORNIA:
v i ew of modi f i c at i ons by the
e a r l i e st adobe bu i l ders and G i l Sanche z , FAIA, and D aryl Al len
de scribes pres ent - day , more Gi lbert Arn o l d S anchez Architect Inc .
sophi s t i cated, s e i smi c 3 0 2 2 Glen Canyon Road
strengthening s y s t ems . Af­ Sant a Cruz , C a l i fornia 95060
ter the 1 9 6 0 s s e i sm i c
st rengt hening s y s t ems were Running paral l e l with the devel opment o f adobe bui l ding i n
emp l oyed u s ing a comb i n at i on C a l i fornia i s a h i s t ory o f periodic de struct i on b y e a rthquake s . The
of grade beams , c o l umn s , and use of adobe as a bu i l ding mat e r i a l was introduced into Alta (upp e r )
bond beams . But when the C a l i forn i a by the Span i sh Crown . T h e chain o f Franc i s can m i s s i o n
State H i s t o r i ca l Bu i l ding sett lement s that extend northward from B a j a Cal i forn i a are located
Code was adopt e d i n 1 9 7 6 , i t in an area which coinc ides with what is t oday des ignated on the
al l owed s y s t ems that are far sei smic ri s k map as Zone 4-the area of great est damage and area
less dest ructive to h i s t o r i c c l o s e st to ma j o r earthquake fau l t s ( see f i g . 1 ) [ 1 ) . Dest ruct i on by
fab r i c . Three e x amp l e s o f seismic forces was a t remendous threat t o the e arly bui lders o f
structural s t rengthening adobe m i s s i on and ranch bui lding s , a s the resu l t s o f years o f
syst ems in h i s t o r i c adobes painstaking l abor could be de st royed i n s e conds .
s i nce 1 9 7 5 are descr ibed .
The f i rst examp l e document s A great e a rthquake which struck in 1 8 1 2 in southern C a l i forn i a
a high- impact s y s t em ( n ow a f fected m i s s i on s from a s f a r south as San D i e g o t o a s f a r north a s
out-moded and p o s s i b l y un­ Mi s s i on La Puris ima in cent ral C a l i forn i a . Fr . Lui s G i l and Fr .
necce s s a ry ) . The other two Marcos Ame stoy o f Mi s s ion Santa Barbara wrote in the i r annual re­
examp l e s ut i l i z e s y s t�ms port , "In the terrible earthquakes of December 2 1 st and the days
that r e l y prima r i l y on the foll owing, the m i s s i on was considerably damaged, nece s s it at i ng a
concrete bond b e am . The care ful in spect i on and s omewhat ext ens ive repai r s . " [ 2 ) D i stressed
S ant a Cruz M i s s i o n Adobe repo rt s s i m i l a r t o the above are repeated over and over in the an­
resto rat i o n ( a bond beam nual repor t s of the Franci s can father s .
s y s t em) was put to the t e s t
during t h e 1 9 8 9 e arthquake As a result o f the 1 8 1 2 earthquake a bel fry and a roo f o f the
and performed we l l . The stone church at Mi s s i on San Gabriel ( known as S an Gab r i e l de los
evidence in t h i s pro j e ct Temb l o re s ) c o l l apsed [ 3 ) and the Great St one Church with vau lted
support s the hypothe s i s that ce i l i ngs at San Juan C ap i s t r ano was in ru ins a fter only six years o f
a structural s y s t em con s i st ­ use [ 4 ) . The adobe church at Mi s s i on San Buenaventura was damaged
ing o f a bond b e am i s s u f f i ­ by earthquakes in 1 8 0 0 , 1 8 0 8 , and a new church under construct i on
ci ent t o w i t h s t and moderate was damaged by the 1 8 1 2 temb l o r requ i ring repa i r work and modi fica­
t o severe e arthquak e s . t i ons t o the bui l ding de s i gn [ 5 ) .

KEYWORDS In central C a l i fornia in 1 8 2 5 , an earthquake with many a fter­


shocks batt e red the m i s s ion church at Mi s s ion S anta Cruz . Due t o a
Sei smic st rengthen i n g , ret ­ lack o f worker s , it was never adequate l y repaired and when a ma j or
rofit , adobe p r e s e rvat i o n , eart hquake struck in 1 8 5 7 , the church, by then de scribe d as " in ru­
h i s t o r i c architecture , in s , " was de stroyed [ 6 ) . To the north at Mi s s ion San Jo s e , a new
earthquake s , C a l i f o rn i a m i s ­ adobe church was under construct i on in 1 8 0 8 when an earthquake
s i ons struck the S an F ranc i s co Bay Area . Because of this quake and i t s
a fte rshoc k s , i t w a s decided t o const ruct a l ow be l l t ower i n s tead o f
the t a l ler o n e o r iginally p l anne d . Thi s m i s s i on church had adobe
s i de wa l l s 9 m ( 3 0 ft . ) high and 1 . 5 to 2 . 7 m (5 to 9 ft . ) in thick­
ness with ma s s ive butt resses . An enthu s i a s t i c French priest
" imp roved" the s imple church bu i l ding by removing the but t r e s s e s and
cut t ing t a l l , Gothi c-in fluence windows . This modi ficat i on no doubt
contributed to the inab i l ity of the church wa l l s to res i st l ateral
mot i on when a ma j o r quake rolled through the area in 1 8 6 8 . The
church was de st royed [ 7 ) . In the Los Ange les area , the church at
Mi s s i on San Gab r i e l i s present l y c l osed due to damage from an
earthquake in 1 9 8 7 .

Adaptat oped to
ions Devel With stand Eart hquake rces
Fo

The e a r l i e s t adobe bu i l ders on the Cali fornia coast , educated


through t r i a l and erro r , unde r st ood the ir material we l l enough t o
make the f o l l owing adapt at i on s t o improve re s i s t ance t o sei smic
force s :
A rat i o o f wall thi ckne s s to wal l hei ght o f 1 to 5 was t yp i c a l ly
LEGEND
maint ained . F o r examp l e , a 9 m high wall would be bu i l t 1 . 5 m in
thickne s s [ 8 ) .
Mission Sites Openings in wa l l s such as doors and windows were kept sma l l and

c=:J� ZONE

ZONE
12 Minor Damage

Moderate Damage
to a min imum .
Be l l t ower s and b e l l wal l s were o ften bu i lt l ow , not extending
above the roo f .
Large st one o r adobe but t re s s e s were bui l t t o brace l ong wa l l s o r

r:;,.t<8-ffiM ZONE

ZONE
34 Major Damage

Major damage and


corners , such a s at Mi s s i on L a Puri s ima and a t Mi s s i on S an Jose
church .

with proximity to
certain major fault
systems

Rgure 1 . Spanish mission locations


superimposed on seismic risk map (Seis­
mic information from BOCA National
Building Code, 1987).
Seismic Mitigation 349

Aft e r the demi se o f the mi s s i on system in the 1 8 3 0 s , many o f


t h e mi s s i on st ructures fe l l i n t o di suse o r were r e legated for ut i l i ­
t a r i an purposes such as storage rooms or stab l e s . L i t t l e interest
was shown in t h e s e aging r e l i c s unt i l a ft er 1 9 0 0 when act ivity which
was primar i l y reconst ruct i on occurred at 5 o r 6 mis s i on s it e s i n the
state ( 9 ) .

On t he crest o f renewed interest in the rest orat i on o f h i s t o ri c


adob e s aft e r t h e 1 9 6 0 s , architects working w i t h st ructural engineers
devised s e i sm i c st rengthening methods t o s at i s fy bui l ding o ff i c i a l s
and owners o f the bu i l dings . A s e r i ous drawback t o t h e e a r l y types
of s t ructural s y st ems which emp l oyed grade beams , vertical co lumn s ,
and bond beams w a s the i r great impact upon h i s t o r i c fabr i c .
Howeve r , at the t ime , both pub l i c o ff i c i a l s and de s ign pro fe s s i o n a l s
were unde r s t andably caut ious about t h e s a fety o f t h e pub l i c in
bu i l dings constructed o f archa i c mat e r i a l s . C a s e studi e s fo l l ow
later in the art i c l e which des cribe early s e i sm i c st rengthening s y s ­
tems de s i gned b y t h e author s ' f i rm .

The Development of the State Historical Building Code in California


Work on formu l at ing a code to e stab l i sh acceptable s a fety s t andard s ,
whi l e recogn i z in g t h e unique qua l i t i e s o f h i s t o r i c a l st ructure s , had
begun in 1 9 6 3 with rec ommendat i ons from the Ca l i fornia Landmarks
Advi s o ry Committ e e . The Committee expre s sed a need to protect pub­
Rgure 2. Mission San Juan Bautista, l i c health and s a fety with a new bu i l ding code that a l s o o f fered
Convento Wing. View of damage after the "enough f l e x ib i l it y t o a l l ow restoration of a h i s t o r i c feature whi l e
1906 earthquake. (Photograph from San st i l l ret a in ing i t s h i s t o r i c integrit y . " The Department o f Parks
Juan Bautista Historical Society) and Re creat i on work ing together with the O f fice o f the State
Arch itect deve loped st atewide input into a draft bill which passed
the l e g i s l ature in 1 9 7 5 . The bill ( SB 9 2 7 , Mi l l s ) be came e f fective
January 1 , 1 9 7 6 [ 1 0 ] .

When C l arence Cu l l imore , Jr . o f the Office o f the S t at e


Arch it e ct was c a l l e d upon t o prepare the Adobe sect i on o f t he State
Historical Bui lding Code , he drew upon the e xp e r ience o f h i s fat h e r ,
Clarence Cul l imore , S r . T h e senior Cu l l imor e , a n architect pract ic­
ing in Bak e r s f i e l d , Ca l i fo rn i a , in the 1 9 2 0 s and 1 9 3 0 s , had de s i gned
and bui lt some two hundred new adobe bu i l dings which emp l oyed the
use of the concrete bond beam (or tie o r c o l l ar beam) t o e n c i r c l e
t h e wa l l s at the t op . A concrete bond beam r e i n forced with steel
reinforcing bars acts a s a band t o stab i l i z e wa l l s as i t re ceives
the weight o f the roof ( and s e i smic force s when they occur) and d i s ­
t r ibut e s t h e load even l y , p a s s ing i t down through t h e wa l l s . The
roof st ructure is attached to the bond beam so that it w i l l not be
disp l aced during wa l l movement from s e i smic forces [ 1 1 ] .

In addi t i o n , Cu l l imore , Jr . had f i r st -hand know l edge o f an


early bond beam ret rofit to an h i st or i ca l adobe in Long Beach,
Cali forn i a-Rancho Los Cerritos-whi ch had undergone a s t rong
earthquak e . L . T . Eva n s , st ructural engineer , had de s i gned a bond
beam ret ro fit for both the two-story r e s i dence and the one - story,
1 4 0 feet l ong wings in a remodel o f the 1 8 4 4 ranch house i n 1 9 3 1
[12] . In 1 9 3 3 , two years l at e r , an e a rthquake me asuring 6 . 3 on the
Richter Scale spread dest ruct ion t hroughout the Long Beach area .
The wa l l s o f many unre in forced brick buildings fe l l into the
street s . When C l arence Cull imore , S r . inspected the adobe s t ructure
after the quake , he was pleased t o find t he adobe ranch house prac­
t i c a l ly unscathe d . A few cracks were obs e rved where the one-st ory
wings had racked aga i n st the t wo - s t o ry section, and t h e re was a
crack over the front do o r . But there we re virtual l y no cracks at
a l l in the one - st o ry wings and the two- story ma in sect i on ( 1 3 ) .
This " fi e l d test ing" veri fied that wal l s st rengthened with bond
beams cou l d inde ed withstand moderate eart hquake s .
3 50 Adobe 90

I
I
I
I
I
I
I
I
I
I
I
1+---- concrete column

grade beams
I I
?. '7�,� ""�£-� /u,J:J.
Rgure3.
1975. At this tim e, both vertical and horizontal concrete supports were required.
FAIA, architect.
Gil
The Peralta Adobe, ca. 1800. Restoration and seismic strengthening completed in
Sanchez,

Case Studies of Seismic Strengthening TechniQ;ues


Thi s art i c le w i l l not att empt to do cument every st rengthening method
emp l oyed in the stat e , but w i l l out l in e those t e chniques which are
most fami l i ar-those de s i gned and constructed by the authors o f thi s
art i c l e .

I n 1 9 7 3 , one o f the pre sent authors was retained by the C i t y o f


San Jose t o rest ore the P e r a l t a Adobe , a s imple one-story r e s idence
bu i l t b e fo r e 1 8 0 0 which i s the only extant s t ructure from the Pueblo
de S an Jose [ 1 4 ] . Because t h i s was prior to the adopt i on o f the
Historical Bu i l ding Code and s i n ce the Un i form Bui l ding Code of 1 9 7 0
gave l i t t l e o r n o ment i on o f construction u s ing unfired masonry
unit s , st ructural engineers at the t ime could a s s i gn no value at a l l
t o t h e l oad-be aring ab i l ity o f unstab i l i z e d adobe wal l s . There fore ,
it was cons idered nece s s ary to de s i gn a structural s y stem which by
its e l f cou l d support the roo f , with adobe wa l l s being treat ed s imp l y
as i n fi l l .

For the P e ralta Adobe p r o j ect , Ken Yue n , structural engineer ,


de s i gned a system o f s t e e l r e i n forced concrete grade beams , co lumn s ,
and a bond beam a t the roof l ine ( s e e fig . 3 ) . Adobe bricks were
removed at a l l four corners of the bu i l ding t o a l low for pouring the
vert ical concrete column s . At the base of interior and ext e r i o r
wal l s , cont inuous stee l - re i n forced concrete footings ( grade beams )
were p l a ce d . At the t op o f the wal l s , a rece s s was e xcavated into
the adobe mat e r i a l for pouring the continuous bond beam around the
perimeter of the bu i lding . Vert ical co lumns we re t ie d into the
grade beam and bond beam so that , in the o ry , the concrete s t ructural
e l ement s would cont inue t o support the roof and provide for human
s a fety even i f a l l the adobe wa l l s were t o c o l l apse during an
earthquake . Al l new s t ructural work was concea l e d behind adobe and
mud p l a s t e r , s o that i t could not be det e cted visua l l y .

The disadvantage o f t h i s type o f st ructural system, part i cu­


larly for such a sma l l bu i l ding ( 1 2 . 8 m x 6 m) , is the disturbance
t o the h i s t o r i c adobe wa l l s . Th i s stab i l i z ation i s what Randolph
Langenbach wou l d de scr ibe a s a "Vietnam approach , " i n which the
bu i l ding i s pract i ca l l y de st royed in the name o f " s aving" it [ 1 5 ] .
Seismic Mitigation

�---;77"J11�;::--- new roof truss system

t------t-+- reconstructed side aisles --t----t----�


Rgure 4. Mission San Juan Bautista Church, 1803- 12. Restoration (supervised by Harry Downie

Gil
1975-78) included side aisle reconstruction by Michael Taylor, general contractor, and main roof
reconstruction by Sanchez, FAIA, architect.

Baut staStren�thenin� and Roof Reconstruction at Mission San Juan


seismici

Thi s pro j e ct was begun under the direct i on o f Harry Downie and the
Cathol i c D i ocese of Monterey and cons isted of re construct ing exte­
rior church wal l s which had been lost due t o damage i n the 1 9 0 6
earthquake . I n the o r iginal const ruct i on the wal l s between the nave
and s i de a i s l e s were p i e rced with an arcade of arche s , but they were
soon f i l l ed in with adobe t o become s o l i d wal l s [ 1 6 ] .

The first port ion o f the work was carried out in 1 9 7 5 (by
Michael Tay l o r , gener,al cont ractor) which included i n s t a l l ing a re­
info rced bond beam at the t op of the newly cons t ructed e xt e r i o r
wal l s . A f t e r t h e H i s t o r i c a l Bu i lding C o d e came int o u s e , unstab i ­
l i z e d adobe wa l l s could n o w be con s i dered a s l o ad bearing wal l s and
l e s s structural work was deemed neces sary . In addit i on to removing
thick concrete but t r e s s e s (bu i l t after the l o s s of the exterior
wa l l s ) and early adobe s placed t o fill the arched opening s , epoxy
was i n j ected into e a ch mortar j o int of the arches for extra fort i f i ­
cat i on [ 1 7 ] .

The pre s ent author ( G . S . ) was called in to s o lve the problem o f


a s a gging c e i l ing framing sy stem i n the church ( s e e f i g . 4 ) . The
original s ystem de s i gned with beams and a king post was exhibit ing
exce s s ive de f l e ct i on . The h i s t o r i c system was documented and re­
moved and a new sy stem was designed which was connected t o the new
bond beam . A new roo f was installed over the church nave and s i de
a i s l e s and the h i s t o r i c roof t i l e s were reinst a l led . The m i s s i on
bui l dings sustained no damage during the October 1 7 , 1 9 8 9
earthquake , but for some reason which i s a s yet unexp l ained, the
earthquake sho cks a l s o bypassed the l i t t l e t own of San Juan
Baut i sta .

Seis mic Strengt hening and Restoration ssio Santa Cruz Adobe
of Mi n

Thi s adobe bui l d ing owned by the State o f C a l i fornia i s the only
structure remaining from the original Santa Cru z m i s s i on comp l e x .
It was bui l t in 1 8 2 2 - 2 4 to house native Ameri can fami l ie s . Afte r
seve ral y e a r s o f h i s t o r i ca l rese arch and archae o l og i c a l invest iga­
t i on t o document h i s t o r i c fabr i c and determine const ruct i on det a i l s ,
the pre sent author was retained by the State t o prepare construct i on
document s for a restorat i on back to i t s m i s s i on period appe arance
[18] .

For s e i sm i c st rengthening, it was decided t o insta l l a concrete


bond beam at the t op of all perimeter wa l l s and all cross wa l l s at
the p l at e l ine . The bond beam was de l iberate ly not keyed in t o the
wa l l so that the adobe wal l s could flex i f needed beneath the r i g i d
beam . C r o s s wa l l s that had been removed through the years were re­
constructe d i n the i r h i s t o r i c l ocat i ons and keyed in at the corne r s .
The adobe cross wa l l s extended upwards to the r i dge o f the roo f and
the ,ridge and ro o f beams rested on the s e adobe wa l l s . To better
dist ribute the w e i ght of the roof beams , rein forced concrete seats
were inst a l led at the l o cat i on where the beams rest on the adobe
cross wa l l s . The s e were not connected to the bond beam beneath them
but were l e ft " fl o ating , " supported only by adobe ( s e e f i g . 5 ) .
3 52 Adobe 90

Figure 5. Mission Adobe at Santa Cruz Mission State Historic Park, 1822-24. Restoration and

had been removed. Gil


seismic strengthening with bond beam. Project included reconstructing four crosswalls which
Sanchez, FAIA, architect.

The st ruct ural work at Mi s s ion Santa Cruz repre sent s a system
in which the adobe wal l s are a l l owed to t ake the fu l l s t r e s s e s o f
the roof l o ad; i t a l s o recogn i z e s the abi l it y o f the thick adobe
wal l s to absorb energy from s e i smic shock . Randolph Langenbach sug­
ge st s , in h i s de s cript ions of vernacular const ruct i on i n Ka shmi r ,
Indi a , that a l a ck o f rigidity may b e a posit ive fact or t o withst and
earthquake force s . Speaking o f h i s t o r i c bui ldings in Kashmir that
have survived numerous earthquakes he state s , "Becau s e of the primi­
t ive mat e r i a l s and means of const ruct i on in Kashmir ( masonry and
t imber runners combined) , st rength was not p o s s i b l e , so f l e x ib i l it y
was nece s s ary" [ 1 9 ] .

The s e i sm i c st rengthening work at Mi s s ion Sant a Cruz was t e sted


on October 1 7 , 1 9 8 9 when an earthquake me asuring 7 . 1 roc k ed the
area . Although the re storat i on had not been fu l l y comp l eted at that
t ime , a l l st ruct ural reinforcing was in p l ace inc luding st ructural
plywood at the roo f . Thi s adobe-the olde s t bui l ding i n Sant a Cruz­
survived the event extremely we l l , in contrast t o many unreinforced
brick bu i l dings in the downtown area which were damaged beyond re­
pair . As could be expected, the short er cross wal l s , l e s s flexible
than the l ong s i de wal l s , exhibited ten s i on cracks in the "X" pat ­
tern [ 2 0 ] and cross wa l l s which were pie rced with doo r openings were
cracked more seve rely than s o l id ones . The long s i de wal l s of the
mi s s i on were vi rtua l l y uncracked . One end wa l l of the m i s s i on
bui lding was s l i ght l y displ aced, moving a few inches outward from
under the bond beam at the p l ate l in e , but it was determined that it
could be j acked back into p l umb p o s i t i o n . Subs equent repa i r work
con s i sted o f in j ecting cracks with a mix of mud mort a r and fly ash,
in addi t i on t o j acking one end wa l l . F ibergl a s rods were i n serted
at an angl e at damaged building corners and wa l l inters ect i on s .
Compared with the heavy l o s s e s to h i s t o r i c bu i l dings in Sant a Cruz
from the 1 9 8 9 quake , this repair work can be cons idered minimal .

Conclusions
S e i smic act ivity has threatened adobe bui ldings s ince the int roduc­
t i on of the buil ding t ype into Cal i forn i a . I n the 1 9 7 0 s highly
rigid s e i smic st rengthening systems with co lumns support ing the r o o f
were thought t o b e nece s s ary f o r pub l i c s a fety . After the
Historical Bui lding Code was adopt ed, sei smic st rengthening systems
ut i l i z ing only a bond beam were encouraged . In terms o f preserva­
t i on , this sy stem is much more favo rable becau s e i t destroys l e s s
h i s t o r i c fabri c . The S anta Cruz Mi s s i o n Adobe , a one - s t ory st ruc­
ture st rengthened only by bond beams at perimeter and cro s s wal l s ,
was " fi e l d t e st e d" i n tb..e October 1 7 , 1 9 8 9 earthquake and performed
exceedingly we l l . The evidence in this examp l e supports the hypoth­
e s i s that a st ructural system con s i st ing primar i l y of a bond beam
(whi ch a l l ows for some wa l l movement ) is suffic ient to withstand
moderate to severe earthquake shocks .
Seismic Mitigation 353

NOTES

1. Uniform Building Code , 1 9 8 5 edit ion (Whit t i e r , C a l i forn i a :


I nt e rnat i on a l Conference o f Bui lding O f fi c i a l s , 1 9 8 5 ) , 1 3 5 .

2.
California Jou Mines Geology
rnal of
Fabricas,
E l i s abeth L . Egenho f f , edit o r ,
and
Supplement t o the
for April 1 9 5 2 ( St at e o f
Cal i fo rn i a , Department o f Natural Resourc e s , 1 95 2 ) , 1 6 5 .

3.
( Lincoln and London :
dian Life the Old Missions
Edith Buckland Webb , In at
Unive r s i t y o f Nebraska P re s s , 1 95 2 ) , 1 3 5 .

4. I b id . , 131 .

5.
Cal i forn i a :
The Cal Missio
Dorothy Kre l l , edi t o r , i fornia
Lane Publ i shing Company , 1 9 7 9 ) , 55 .
ns (Menlo P ark

6.

7.
Indian Life the OldTheMissiHistory
Webb , ons, of Mission San Jose·
at

Franc i s F l o rence McCarth y ,


133 .

California 1797-1835
1 95 8 ) , 2 1 8 , 2 4 2 .
( Fresno , Cali forn i a : Academy Library Gu i l d ,

8. G i l Sanche z , FAIA, personal commun i cat i o n , 1 990 .

9. Krel l , The Cali fornia Missions, 66.

10 . State Hi storical Bui lding Code , T i t l e 2 4 , Bui lding St andards ,


( O f f i ce of Administrat ive Hearings , Department o f General S e rvi ces ,
State o f C a l i forn i a , 1 9 8 5 ) , 8 - 3 .

1 1 . Clarence Cull imore , O ffice o f the State Architect , State o f


Cali forn i a , personal communicat ion, 1 9 9 0 .

1 2 . Kenneth Win g , Architect , Const ruct ion Do cument s for Renovat ion
o f Rancho Los Cerritos , (Rancho Los Cerr i t o s Archive s , Long Beach ,
Cali forn i a , 1 9 3 0 ) .

Structure Repo
1 3 . G i l Sanche z , FAIA, and Daryl Al len ,

1 98 7 ) , 53 .
Ran
cho Los Cerri Hi
rt (Unpub l i shed ms . , Rancho Los Cerr i t o s archive s ,
tos sto ric

California First City,


1 4 . Edwin A . Bei lharz and Donald O . DeMe r s ,
's
P re s s , 1 9 8 0 ) , 3 6 .
( T u l s a , Oklahoma :
Jr . , San Jose ,
Continental Herit age

1 5 . Randolph Langenbach, "Bricks ,


Bulletin XXI , no . 3 8 4 ( 1 9 8 9 ) : 3 2
Mortar, and Earthquake s , " AE:r.
1 6 . Ground F l o o r P l a n , H i s t o r i c Ame r i can Bui lding Surve y , Survey no .
3 8 - 4 , ( Un i t ed S t at e s Department o f the Interior, O f f i ce o f Nat i onal
Parks Bu i l dings and Re servat i on s ) .

1 7 . Michael Taylor, general contract o r , S ant a Cru z , personal commu­


n i cat i on , 1 9 9 0 .

1 8 . David L . F e lt on , Sa
Arch itectu ral and Ar chaeontaogi vest gatState sto
l
Cruz Mission
cal I n i
Hi ric Park
ions 1984-85, Unpub l i shed
ms . , ( Cultural Herit age Sect i on , Resource Protect i on Divi s i o n ,
Department o f Parks and Recreat ion , State o f C a l i forn i a , S a c rament o ,
1 9 85 ) , 7 .

1 9 . Langenbach, "Bricks , Mortar and Earthquake s , " 3 6 .

Adobe med
( Tucson :
Earth Bui ldi ngs
2 0 . Ge rald W . May , " S t ructural Engineering for Earth Bu i ldings" in
and Ram by Paul Graham McHen ry , Jr .
Unive r s i t y of Ari z ona P re s s , 1 9 8 4 ) , 1 7 8 .
Current Field Research
Current Field Research 3 57

ABSTRACT

This study consisted o f


ADOBE : CONSTRUCTIVE METHOD AND THERMIC C

Maria Auxil iadora Afonso Alvarenga


HARA CTERISTICS

identifying the 1 1 1 , Pra9a Governador Valadares


constructive methods Araxa , Minas Gerais , 3 8 . 1 8 0
using adobe and a hand Bra z i l
operated b l ock press .
The thermic
characteristics o f adobe Introduction
were analyzed and show
that using adobe wal l s The technol ogies that use earth as bui lding material for domestic
with high thermal inertia architecture have been used in Bra z i l from the t ime o f their
it is possible to provide d iscovery unti l the present . The Portuguese and Africans brought
comfortabl e internal their constructive methods , cal l ed adobe and p ise . By integrating
temperature . them with those used by natives of Braz i l , they developed another
method cal l ed "taipa" or "pau-a-pique" (watt l e ) .

KEYWORDS Whol e cities in Bra z i l --Ouro Preto , Diamantina , Sao Joao del
Rei and others--were constructed using earth techno l ogies . The
Adobe , earthen preserved buildings of these cities indicate that with sound
architecture , Bra z i l . construction methods , careful selection of earth , together with
proper protection of the wal l s from erosion by storms , this type
of construction is capable of l asting for centuries .

These technologies were used intensively in Bra z i l unt i l the


coming of D . Joao VI , and the royal family to Rio de Janeiro . At
that time , with the opening of the ports to " friendly nations , "
Eng l ish materials such as Portland cement and glass came to be
used indiscriminately .

with the advent of modern architecture and the international


style , materials such as glass , steel , and concrete became
symbo l s of " status" and modernity .

The o i l and energy crisis of 1 9 7 3 brought about a


reevaluation of these modern material s , which are produced at a
high cost of energy and resources . Engineers and architects who
were interested in bui l ding comfortab l e houses at l ow cost had to
look in other d i rections .

As a result , research began into the use of those o l der


technol ogies that adapted so wel l to l ocal cl imate and culture .

stil l , the biggest problem in a country l ike Bra z i l is that


peop l e prefer imports from foreign countries because they have
more " status . " Earthen architectural systems are considered to
be the l ot of the poor who have no other choice . Peop l e do not
see that building with earth is a good alternative in a country
l ike Bra z i l because it is wel l adapted to the l ocal c l imate and
social conditions . Because of these misconceptions , it becomes
more and more d i fficult to find builders who know how to select
the soil and are knowledgable about the constructive method .

Adobe is one of the technologies that continues to be used


because o f its simplicity . The f irst part o f this study provides
an overview of the whol e process , based on data obtained from
many cities in northern Minas Gerais where these technol ogies are
sti l l used . Through personal contacts , photographs , and by
participating in construction , it was possible to identify the
principal e l ements in the adobe constructive method .

Adobe or mud brick

Adobe comes from the Arabic word "Adob , " which means mud brick
dried in the sun . It is a constructive method that uses a mud
brick molded in a wooden or steel form .

It is probably th� most popu l ar earth building method used


in Bra z i l at this present t ime . Its relative s imp l i c ity , which
a l l ows for " autoconstruct ion , " and its l ow cost exp l a in why it is
ADOBE so widely used .
Mud brick dried in the
sun . Another factor is that adobe is an ecological building
material . It uses natural elements and does not interfere with
Adobe 90

the environment , which is a great concern to people today . with


wal l s of the right weakness , adobe can provide internal comfort
with stab i l i zed temperatures independent of external conditions .

Descript ion

The process involves us ing a proper s o i l , with some kind of


stabil i ze r , puddled with water . This mud can be mixed us ing bare
feet or a mechanical process . It is necessary to test the s o i l
QUANTITY OF WATER
in order t o identi fy what kind o f stab i l i zer w i l l be used . The
The best mud cons istency
mud mixture is l e ft undisturbed for 2 4 hours in order to attain
for moulding contains 1 5 %
the proper cons istency . In northern Minas Gera i s , the bui lders
and 2 0% o f wate r .
bel ieve it is necessary to make adobes in "becrease moon , " in
order to have some cracks . It is necessary to pay attention to
the quantity of water used ; for adobe , the ideal consistency i s
1 5 % to 2 0 % water b y weight i n i t s dry state . When i t has
attained its proper consistency , the mud is put into a wooden
form , which may be dusted with a fine sand in order to faci l itate
removal of the finished adobes . After each adobe is finished , it
is necessary to wash the form to make sure that no mud remains in
the corners .

A man us ing a s ingle mould can make between 2 0 0 and 5 0 0


adobes in a day . For a sma l l house o f about 6 0 square meters ,
about 3 , 0 0 0 adobes are required . When they come out of the
moulds , the adobes must rema in undisturbed for three days ; then
they must be l e ft face upwards for another day in order to dry
out . The adobes need to be completely dry before being used for
construction . Thirty days seems to be sufficient in a normal
c l imate without storms . It is important to cure them in the
shade , not in the sun , without protection , which can cause
shrinkage and consequently cracks .

Moulds

It i s possible to use a two-brick , four-bric k , or s ingl e mould .


The s impl est and most practical method is the s ingl e , bottomless
MOULDS wooden mould . It can be l ined with sheet metal to a l l ow the
The adobes can be mould to s l ip easi ly away from the adobe when l i fted . The adobe
manufactured us ing a s i z e depends on the mould ; the most common has dimensions of 10 x
c i nva-ram press or a 2 0 x 3 5 cm . The s ingle mould has been used in a quick way that
wooden mou l d . produces the same quantity of mud bricks as the multiple moulds .
spec ial closer brick w i l l be required for bonding , for this kind
o f adobes it is necessary to fix suitably shaped p ieces o f wood
inside the mou l d .

Soil Selection

The proper selection o f the soil is essential for a good earth


bui lding . It is necessary to investigate the kinds of soil that
are available in the region . If the soil is unsuitable for the
kind of construction desired , it can be mixed with others in
certain proport ions and used to make adobes . s imple
identi f ication of the texture and spec ific tests can be used to
make a proper analys is and select the suitable s oi l . Observation
o f the s o i l texture helps to determine the s o i l compos ition .
When one rubs dry s o i l between the fingers , the sand particles
are gritty to the touch , the silt and fine particles adhere
closely to the skin and have a s i l ky feel when the sand particles
are d i scarded . Lateritic soi l , normal ly red and reddish-brown
indicating the presence of iron ( hematite) or yel l ow and
yel l owish-brown indicating the presence of l imonite , often do not
shrink or swe l l much upon wett ing . They have high stab i l i z ing
qual ities which are apparently connected in s ome way with their
iron compounds and colloids . This kind o f soil is des irable for
earth construction . In order to manufacture adobes , it is
necessary to select a soil without organic sUbstances . S o i l s are
usua l ly graded into divisions according to the s i z e of the soil
particles . According to the International System o f s o i l -texture
c l a s s i fication :

I-- C L A v -- I -- S I L T -- I - F I N E SAND --
O . 0 0 2 0 ( mm ) O . 02 0 ( m m )
f--
O . 2 0 ( mm )
- C O A R S E SAND --

Genera l l y , s o i l s containing l ess than 2 0 % clay are classed as


I
2 . 0 ( mm )

sand and grave l , l oamy sands , sandy loams , and l oams , depending
on the c l ay content . S o i l s containing from 2 0% to 3 0 % c l ay are
c a l l ed clay loams and those over 3 0 % clay are classed as clays .
The s i l t l oam and clay loam are cons idered suitable for adobe .
But it is necessary to have the right proportion of clay to sand .
Adobes made of s o i l with too great a proportion of c l ay are
vulnerable to shrinkage and consequent cracking . The presence o f
sand in the selected soil is important because o f i t s compression
resistance and because sand is an inertial material in the
presence of wate r .
Current Field Research 35 9

Analys is

A simple test can identify the percentage o f c l ay and sand in a


given s o i l samp l e . A sma l l quantity of the selected earth is
p l aced in a glass f i l l ing it to one third . Then measure the
height that corresponds to the earth sample . Put water into the
TESTS other two-thirds of the glass . Turn it over with a tampion and
Before selecting the wash it repeatedly until the water runs c lear because the clay
soil , it is necessary to and silt have f l oated off . Now measure again . Suppose the first
know the percentage o f measure was 5 cm , and the last was 3 cm . The fol l owing formula
c l ay and sand . expresses the relationship :

53 100%x
-
x -
3005- 60%
The test shows that the percentage of sand is approximately 6 0 % .
In this way the correct percentage of sand and clay can be
checked . The samples selected are triturated , put in a dish ,
then heated . The dried soil is weighed . Using a no . 2 0 0 U . S .
sieve , the sample is washed , e l iminating the c l ay and s i l t . The
residue is dried again and then weighed . By comparing the
measures the exact percentage of sand and clay can be determined .
Genera l ly the d i f ference is 5 % to 1 0 % between the tests .

S o i l preparation

After analysis and correct selection o f the soil , it is time to


prepare the mud . The soil should be reasonably free from humus
or any quantity of organic material . A steel s ieve can be used
to make the soil homogenous and el iminate the l arger stones . A
sha l l ow pit can be prepared in which the material is mixed .
Based on the results of the tests described above , it may be
necessary to add sand or to mix the selected soil with another
kind in order to obtain the e ffective mo isture . other admixtures
can also be used in order the stab i l i z e the material .

Earth stabil i zation

The clay contained in the soil is susceptible to present


f d i f ferences and volume changes when mixed with wate r .
Alternating cycles o f wetting and drying causes gradual
dis integration at the surface , especially at the point of contact
with the water . The sand is inert and the particles do not
absorb moisture , and it l imits the volume changes of adobes
because of the shrinkage of clay . It is possible to use
admixture to stab i l i z e the soil . Attaining the correct
proport ion of clay and sand is the first step . I f the soil has
too much sand , it is necessary to mix clay in as a binder or
cementing agent , and i f there is too much clay , it is necessary
to add sand in order to l imit the volume changes which cause
shrinkage and cracks .

There are other kinds of stabil i z ation :

1. cementing : using an agent such as Portland cement to reduce


the volume changes of the clay .

2. The addition of fibers : straw , hai r , sisal , and other


s imilar materials can be used in order to reinforce the mud and
reduce the volume changes .

3. water-resistant stab i l i z ation : use of certain materials such


as asphal t , mixing o i l , to give permanent water-resistant
stab i l i z ation .

4. Compaction : This increases weather resistance in machine­


made b locks .

5. Lignin and tropical p l ants ; Euphorbia l actea , with a hard


elastic white f ilm and the f i lm of banana can be used mixed with
the water that puddles the soil in order to give protection from
tropical storms . It can be used in wal l paints , when the plants
are cut and chopped up in containers . This produces a sticky
l i quid which can be mixed with l ime before using .

To select the proper stab i l i z ation and the quantity that is


necessary it is better to make tests with different substances
and d i f ferent quantities before starting construction .

The qua l ity of stab i l ized adobe bricks can be estab l i shed
recogni z ing testing methods :

1. General conditions : Cured mud bricks sha l l be reasonably


true to s i z e with para l l e l sides and free from excessive cracks
and other defects .
2. Moisture : Content of mud brick when dried and ready for use
STABILI ZATION sha l l not be more than 4 % .
The adobe bricks can be
stab i l i zed using a press . 3. Shrinkage cracks : Shal l not be more than 1/8 inch i n width
and 3 inches in l ength .

4. compression strength : Sha l l average 3 0 0 pounds or more


under recogn i z ed test methods , per square inch with tolerance to
2 5 0 pounds for one brick in a test series of five .

5. Absorption : Sha l l average less than 2 1/2 percent of dry


weight in 7 days .

6. Erosion : Bricks shal l not be appreciably pited or eroded in


two hours under a fine spray of water under 20 pounds of
pressure .

7. Modulus of rupture : Shal l average not l ess than from 4 0 to


5 0 pounds per square inch with tolerance to 3 0 pounds for one
brick in a test series of five .

Mortar

Mud mortar is most commonly used for adobe wal l . It is more


appropriate and gives a more homogenous final result . Earth
found in ant and termite houses can be used to guarantee its
qual ity . using some kind of stab i l i zation , the same mix should
be used as in the mud bricks . In the state of Minas Gerais , the
most common mortar is 4 parts sand , one of ant or termite earth ,
and half of cow dung ( 4 : 1 : 1/ 2 ) .

Bui lding the Wal l s

S ince the adobes a r e completely cured , they can be used in


building wal l s . S ide preparations , foundations , and provisions
BUILDING THE WALLS against dampness are requirements that must be comp l eted before
When the adobes are beginning the adobe wal l s . It is necessary to provide a special
comp l etely cured , they protection against moisture between the foundation and the adobe
can be used in building wal l . Stone , brick , or wood can be used to prevent infiltration
wal l s . of water on the wal l s . The height of the foundation can b e about
15 cm above the ground to protect the wal l against ra in and
consequent eros ion . When the construction is f inished ,
protective wal l coverings can be appl ied , mud p laster or mud
mixed with a sma l l quantity of cement and l ime . I f cement and
l ime is used , it is essential to test the quantity because a
mortar that is too strong can diminish adhesion between the adobe
wal l s and the covering .

Considerations about Earthen Homes

certain considerations must be kept in mind when building a home


of earth : views , protection against inclement weather , a suitable
slope to assist drainage and provide for sewage l ines . I f the
site is s l oping , it w i l l be an advantage , if the fal l is in a
direction suitable for drainage . In selecting the site it is
necessary to consider the kind of soil that is ava i lable and the
constructive method that best suits the particular s ite and
building l ay-out . Other local materials such as wood , stone ,
kiln- fired brick can be incorporated into the house design i f
they are ava i labl e . The orientation of the house must consider
the best view , the solar and wind directions as passive means of
obtaining comfortable internal temperatures . In the case of a
tropical country l ike Bra z i l , the orientation must , in some
cases , be chosen with a preference for the dominant winds . In
Araxa , with a tropical cl imate but high altitude ( 9 0 0 m) , there
is a thermic ampl itude , which makes the solar orientation most
important .

To protect the wal l s from weathe r , verandas can be used to


keep out the water during rainstorms and to protect from hot
sunshine during summer . The roof should be extended considerably
from the wal l s giving complete protection from ra in . Impervious
masonry is desirab l e to a height above the f inish grade which
w i l l prevent erosion from the splash of rain water . Wood or
stone can be used before starting the adobe wal l s to reduce
cap i l l ary action . A SUbstantial foundation and footing , designed
for the unit compressive soil bearing capacity is essential .
Other considerations depend on the special requirements of the
c l imate or the builder ' s desires .
Current Field Research

Thermic Characteristics

S ince the 1 9 7 3 oil cris is , the home as energy consumer has come
under consideration . Bardou ( 19 7 9 ) states that approximately 1/ 3
of the total energy consumed is used by houses . Using solar
energy and b iocl imatic architecture , it is possible to heat or
cool houses using pass ive methods without any further costs and
providing comfortable internal temperatures . using cl imate data ,
the designer can understand better the local meteorological
parameters and orient the house adequatel y . Knowing the
s imultaneous action of temperature , sun , and wind , it is possible
to have a b iocl imatic conception of the proj ect . There are two
fundamental aspects to the thermal action of a building . First ,
it can be assumed that the external conditions are permanent and
the internal temperature is constant ; the thermal conductivity of
the material , its resistance and conductance are considered .
But , in fact , the temperature regimen is variab l e , and it is
necessary to consider radiation and convection . In this case ,
the thermal characteristics that better exp l a in the real ity are
heat capacity ( C ) and thermal inertia ( ) . The heat capacity can
be obtained by the specific heat of the bui lding material c -
( J/Kg C ) .

Calculat ion

For adobe the characteristic time constants found for the


d i f ferent wal l thicknesses was : 3 . 4 7 , 6 . 17 , 1 8 . 9 , and 3 8 . 59 days
respectively ; using brick , the characteristic time found was 2 . 2 ,
3 . 9 1 , 1 1 . 9 8 , and 2 4 . 4 6 days respectively .

INTERNAL TEMPERATURES Conclusion


using adobe you can
provide comfortable using adobe as a building material , it is important to know the
internal temperatures . constructive method , its special requirements in order to inform
the design cons iderations , selection of the better kind of soil ,
simple tests to certi fy it , and each step of the construction .
Knowing its thermal capacity , us ing local material s , and
considering the cl imate , an adobe house can be constructed at low
cost and comfortable internal temperatures using passive and
natural means .

"Man has always had a creative instinct , which reveals


itse l f in one way or another , " and a house can express the
bui lder himsel f , his personal ity , his dreams and determination to
resolve the problem of building a shelter for hims e l f and his
family . Earth , with its universal avai l ab i l ity , its thermal
properties , low cost and ecological considerations , seems to be
one of the best materials for home construction . Nonethel ess ,
people are neglecting this method because of misconceptions .
M i l l ions of people today do not have homes . us ing earth as a
building material could help to solve this housing problem whi l e
revita l i z ing these building techniques .

Autoconstruction experience

Practical experience in using the above principal s was offered in


the construction of an adobe house outside of Araxa , an area of a
sma l l c ity in Braz i l . In every aspect , from the design to
construction , the architect worked with the community to take

techniques .
obtained .
us ing CINVA RAM,
advantage of its knowledge of amb ient comfort and adobe building
a hand-made brick press was
The community used the press to manufacture the
adobes , but they preferred to use the single wooden form because
of its s impl icity .

The f irst d i f ficulty in the construction was the builders '


dissatis faction with the adobe . They considered it synonymous
with poverty and the lack of any other choice . After many
discussions , they were convinced that they could construct a
comfortabl e house using adobe . The steps of the construction
were detailed in photographs .
REFERENCES

1. American Institute of Architects . La Casa Passiva - C l ima Y


Ahorro Energetico . Trad . J . Corral do original Regional
Guidel ines for Building - Passive Energy conserving Homes , 1 9 8 0
Hermann Blume , Madrid , 1 9 8 4 .

2. Arzoumanian , V . and Bardou , P . Sol Y Arquitectura . Trad . de


M . T . Trias de Bes do original Archi de Soleil , 1 9 7 8 , Editorial
Gustavo G il l i , Barcelona , 1 9 8 0 .

3. Cornoldi , A . and Los , S . Habitat Y Energia . Editorial


Gustavo Gil l i , Barcelona , 1 9 8 2 .

4. Dreyfus , J . Le Con fort dan l ' habitat Tropical . Eyrol les ,


Pari s , 1 9 6 0 .

5. Fonseca , M . R . et a l l i i . Desenho Solar . Proj eto Editores


Associados , Sao Paulo , 1 9 8 3 .

6. Goldemberg , J . Energia no Bras i l . Livros Tecnicos e


Cienti f icos , Rio de Janeiro , 1 9 7 9 .

7. Gri f fiths , J . F . C l imatology - An Introduction . Oxford


University Pres s , second edition , oxford , 1 9 7 6 .

8. Grupo de Estudos de Arquitectura Alternativa ( GAA) .


Arquitectura e Energia - Uma Alternative de Proj etos . Editora da
UFMG Bel o Horizonte , 1 9 8 1 .

9. Grupode Estudos de Arquitetura Biocl imatica ( GEAB ) .


Relatorio de Pesquisa : s istemas Construtivos Vernaculares de
Minas Gera is e sua AdequaQao ao Fator Cl ima . Belo Horizonte ,
1983 .

10. Guyot , A . y I zard , L. Arquitectura Biocl imatica . Trad . M . T .


Trias de Bes do original Archi Bio, 1 9 7 9 , Editorial Gustavo
Gil l i , Barcel ona , 1 9 8 0 .

11 . Lynch , K . Plan i ficacion del Sitio. Trad . J . F . Calaya do


original s ite Planning , 1 9 8 0 , Editorial Gustavo Gil l i , 1 9 8 0 .

12 . Mascaro , L . and Vianna , M . I luminaQao Natural nos Edificios .


PROPAR/UFRS , Porto Alegre , 1 9 8 0 .

13 . Mascaro J . Consumo de Energia e ConstruQao de Edi f iviod .


SECOVI , Sao Pau l o , 1 9 8 0 .

14 . Mascaro , L . Energia na EdificaQao - Estrategias para


Minimizar seu Consumo . Proj eto Editores Associados , Sao Paulo ,
1985 .

15 . Mota , S . Planej amento Urbano e PreservaQao Ambiental .


PROEDI , Forta l e z a , 1 9 8 1 .

16. o l iver , P . Cob i j o Y Sociedad . Blume ediciones , Madri , 1 9 7 8 .

17 . Page , J . K . The Fundamental Problems of Building C l imatology


considered from the Point of View of Decis ion-Making Architect
and Urban Designer , in : Building Cl imatology .

18 . Rivero , R . Arquitetura e Cl ima - Acondicionamento Termico


Natura l . Trad . J . M . Aroztegui do original em castelhano , 1 9 8 5
D . C . Luz z atto Editores , 2 a . ediQao , 1 9 8 6 .

19 . Val e , B . and Val e , R . La Casa Autonoma - Diseno Y


Plani f icacion para la Autosuf iciencia . Trad . A . R . Llobera do
original The Autonomous House - Design and Planning for S e l f­
Sufficiency , 1 9 7 5 , Editorial Gustavo Gi l l i , Barcelona , 1 9 7 7 .

20. varios , Cobij o . Trad . J . Corral do original Shelter . 1 9 7 3 ,


H . Blume Ediciones , segunda reimpres ion , Madri d , 1 9 8 1 .
Current Field Research

ABSTRACT INFLUENCE DE L ' HUMID I TE SUR LES PROPRI ETES THERMIQUES DU MATERIAU
TERRE : PROBLEMAT IQUE , METROLOG I E , RESULTATS EXPERlMENTAUX .

Con s idering the examp l e o f


some earths used to rea l i z e Jean-Paul Laurent
"Pise" ( r ammed e a rt h ) , we I n s t itut de Mecanique de Grenoble
p r e sent h e r e e x p e r iment a l BP 53 X, 3 8 0 4 1 , Grenoble-Cedex, France
result s concerning t he i n­
f luence of t e mp e r at u re and
wat e r - c ont e n t on
propert ies o f earth building
mat eria l . After an int roduc­
t he rm a l Introduction
Les t e r r e s ut i l i s a b l e s en t ant que mat e r i a u de con s t ruct ion sont
t i on on t he st ructure cha­ de s m i l i e u x poreux granu l a i r e s dont l a cohe s i on e s t a s s u r e e
rac t e r i z a t ion me t h o d s of n a t u re l lement p a r une c e r t a in e p r op ort ion d e pha s e a r g i l eu s e .
t h i s part icu lar granu lar po­ Que l l e que soit sa c'omposit ion minera logiqu e , ce l l e - c i poss ede une
rous mate r ia l , t he ma in me­ s t ructure mic roporeu s e qui con fere au mat e r iau terre un caract e re
c a n i sms of ads orpt ion and hyg r o s c o p i qu e . P rat i quement , c e l a s i gn i f i e qu e , meme apre s la
migrat i on of wat e r i n earth p e r iode de s e chage qu i suit l a mi s e en reuvre et ,n l ' absence de
wa l l s are reviewed . The ex­ t o u t e a r r i ve e d ' eau " p a r a s i t e " , une p a r o i en t e rre cont ient
perimen t a l device , a " t he r­ tou j o u r s une c e r t a ine quant i t e d ' eau adsorbee p a r l' argile en
mal shocks probe " , we use t o p r e s ence de l ' humidite de l ' a i r . Sous l ' e f fet de s o l l i c it at i ons
determine t h e t he rma l p a ra­ c l imat ique s , not amme nt thermique s , cette eau e s t s u s ce� ible de
me t e r s , h e a t - c o n du c t i v i t y changer d' e t at ( vapo r i sat ion/ conde n s a t i o n ) et / ou de migrer dans
and heat - c a p a c i t y i s a l s o l ' espace des pore s , ce qui mod i f i e evidemme nt l e b i lan t hermique
de s c r ibe d . The ma in conc l u­
s ion we have drawn from t h i s
study i s t hat , i f th� wat e r
c o n t e n t i s n o t too impor­
final . 11 est indispensable d ' etud i e r , de mode l i s e r
qu ant i f i e r ce couplage ent re t ra n s fe rt d' eau et t r a n s fe rt de
chaleur dans une paroi en t e rre car il est c l a i r qu ' i l cond it i onne
e s sent i e l lement le confort t hermique propre a ce t ype d ' habitat . Le
et de

t ant , the influence o f tem­ t rava i l que nous pres entons i c i s ' inscrit dans cette per spect ive .
perature on heat-conduct ivi­
t y can be neglected and the Ap res un rapp e l sur la caract e r i sat ion de la st ructure poreuse
i n f luence o f wat e r - c on t e n t du mat e r i au t e rre et s e s prop r i e t e s thermiques a l ' etat s e c , nous
o n therma l parameters mode­ p a s s on s e n revue les dive r s phe nome n e s phys i q u e s qu i s ont a
l i z e d by l i n e a r r e l at i on­ l ' orig ine de la f ixat ion et du t rans fert de l ' eau dans ce materiau .
ships . F igures and nume r i c a l Nous int rodu i s on s e n s u i t e la not ion de " p a rame t r e s t he rmiques
va lues a r e given to est imate apparen t s " d ' un mi l i e u poreux humide e t nous de c r i vons la
pract i ca l ly the therma l pa­ method o l o g i e de me s u r e de l a condu c t i v i t e t h e rmique et de l a
rameters of an earth buiding c ap a c i t e c a l o r i f ique p a r " s ondes a c h o c s t h e rmiqu e s " que nous
mat e r i a l in n o rma l condi­ u t i l i sons . Enfin , nou s pres entons et nous d i s cuton s l e s prin c i paux
t ion s . re s u l t at s e x p e r iment a u x que nous avon s obt e n u s concernant
l ' i n f luence de l a t empe rature e t de l a teneur e n eau s u r l e s
prop r i e t e s thermiques du materiau terre .
KEYWORDS
Pour cet expo s e , nous avons cho i s i d ' i l lu s t rer notre propos en
con s ide rant le cas de t e rres "a p i s e " d ' une reg ion de France : le
EARTH BUILD I NG MATER IAL , PO­ Dauphine ( Vo i r Fig . 1 et Tab . I) , que nous avons plus
ROUS ME D I U M , HEAT & MAS S p a r t i c u l i e rement etudie [1, 2] . T ou t e f o i s , il est clair que
TRAN S F E R , E XP E R I MENTAL ME­ l ' approche methodologique que nous avons adoptee est t ransposable a
THOD S , HEAT COND UCT I V I T Y , d ' aut re s c l a s s e s de mat e r iaux . Nous avons d ' a i l l eu r s , nous -meme s ,
STRUCTURE , CLAY, HUMID ITY . egalement t rava i l le s u r de s "torchis " ou s u r de s t e rres a f r i ca ines
ut i l isees en " fa conn age direct " [ 1 , 2 ] .

MOTS CLES
Structure
thermiQ)1es aetl'Hatcaracterisation
sec du materiau terre. propriet6s
MATE R I AU TERRE , M I L I E U PO­
Une t e rre crue peut e t re emp loyee comme mat e r iau de st ructure s i ,
REUX, TRANSFERTS DE MAS SE ET
par u n procede que l conque , o n peut atte indre u n n iveau de de ns ite
DE CHALEUR, METHODES EXPERI­
qui lu i confere une res i s t ance mecanique s u f f i sant e . Par a i l leur s ,
MENTALE S , CONDUCTIVITE THER­
u n cert a i n nomb re d ' iden t i f icat ions pre l imina i r e s ( g ranu lome t r i e ,
M I QUE , S TRUCTURE , ARG I LE ,
teneur en argi l e , e s s a i " P roct o r " , e t c ... ) doivent e t re e f fectuees
HUMI D ITE .
pour savoir si une te rre donnee peut e f fect ivement etre ut i l isee
pour con s t ruire [ 3 , 4 ] . Que l le que soit la t e chnique const ruct ive
f ina lement ut i l i s e e , la s t ructure phys i que du mat e riau t e r re , en
s ituat ion norma le de fonct ionnement , s ' apparente plus a ce l l e d ' une
roche pore u s e qu ' a c e l l e d ' un s o l b i e n que ce l u i - c i s o i t , en
l ' o c c u r en c e , l a mat i e r e premiere con s t i t u t ive . Global ement , on
p o u r r a donc le c a r a ct e r i s e r au n i veau ma c r o s c o p i qu e par de s
parame t r e s comme la den s i t e s e che ou la poro s i t e ( Vo i r Tab . I I ) .
L ' e xp e r ience montre que l a conn a i s sance de ces s e u l s parametres
suffit souvent pour prevo i r corre ctement certaines caract e r i s t i ques
fon c t i onn e l l e s . Par exemp l e , conce rnant les prop r i e t e s t hermiques
e n p a rt i c u l i e r , nou s avo n s pu mont r e r que l a de n s i t e s e che
determine e s sent i e l lement la conduct ivite thermique " a s e c " ( Vo i r
F i g . 2 ) ce dont on p e u t rendre compte phys i quement [ 1 , 2 ] . Par
cont re , l o r s que l ' on s ' i n t e r e s s e aux propr i e t e s hydr i que s , ou a
l ' i n f luence de s c ondit ions hygrothermiques s u r le comportement
me can i que ou thermique d ' une paroi en terre , il devient abso lument
necessa ire de con s ide rer d ' aut res facteurs st ructuraux .

En premier l i e u , il importe de caract e r i s e r , qua l itat ivement et


q u ant i t a t iveme n t , la phase argileuse qu i
t ie l lement l e s propr i e t e s hygro s cop iques (Voir §
c o n d i t ionne e s s en­
su ivant ) . O r , l a
min e ralogie d e ce type de roches sediment a i r e s e s t s ingu l i e rement
complexe [ 5 ] . Chimi quement , l e s argiles sont des s i l i co-a luminat es
Adobe 90

hydrat e s dont la st ructure c r i s t a l l in e p o s s ede la p a rt i c u l a r it e


d ' e t re o rga n i s e e e n f e u i l l e t s . Ces f e u i l l e t s s o n t e u x- meme s
c o n s t i t u e s de couches d ' atomes dont l ' emp i l ement fo rme deux t ypes
de mot i f s octaedriques autour de s a t om e s d ' a l um i n i u m ou
t e t ra e d r i qu e s autour des atome s de s i l i c ium . 1 1 est p o s s ib l e de
d i f f e r e n c i e r l e s d i f fe rent s groupes d ' a r g i l e s su ivant la s e quence
d ' emp i l ement de s couches d ' a t ome s dans l e f eu i l let a l a qu e l l e
c o r r e spond u n e d i s t a n c e i nt e r f o l i a i re caracte r i s t i que ( Vo i r Tab .
III) . L ' ident i f i ca t i on de ces mineraux s ' e f fectue par
c r i s t a l l ograph ie aux rayons X ma i s , dans l e s terres nature l l e s , de
nomb r e u x types de mineraux a r g i l eux peuvent e t re pre sents
s imu l t anement ( Vo i r exemp l e Tab . I ) et l e u r d i f fe renc iat ion peut
s ' ave rer d i f f i c i l e d ' autant plus qu ' i l s sont sus cept ibles de former
de s c o mp o s e s interst rat ifies ( i l l i t e / m o n t mo r i l l o n i t e ou
montmo r i l l on i t e / ch l o r i t e p a r e xemp l e ) et q u e l e u r s d i s t an c e s
inte rfoliaires ne sont pas t o u j ou r s tres stables ( ca s des
smect i t e s ) . D e p l u s , i l e s t s ouvent impo s s i b l e d e de t e rmine r
p r a t i quement le pourcentage en ma s s e que r e p r e s e n t e la phase
a r g i l eu s e . On peut alors se cont enter d ' un critere granulometr ique
en a s s im i l ant phase argi leuse et fract ion ma s s ique de diametre de
gra in inferieur a 2 �m .

P a r a l l e lement a l ' ident i f i cation des a r g i le s , i l e s t ega lement


imp o r t a n t de caract e r i s e r la s t ru c t u r e du reseau de p o r e s
intergranula ire s . E n e f fet , c ' est dans cet e space que s ' operent les
phenomenes de t rans fert hydr ique et sa morpho logie i n f lue donc sur
l e s va leurs de s coeffi c ients de t ransfert corre spondant s . P l u s ieurs
te chniqu e s complement aires peuvent e t re u t i l i s e e s pour caract e r i s e r
c e res eau poreux : vo i r Tableau I I .

eau dans le materiau terre. notion de parametres thermiQues


L'
apparents
Comme nous l ' avons dit en int roduct ion, l e s mater iaux auxque l s nous
nous int e r e s s ons i c i contiennent t ou j ours de l ' e au . C e l l e - c i pourra
etre
avec
carac t e r i s e e par (0)
sa proport ion en ma s s e (w) ou en volume

W= . o = Masse d'eau
Masse totale
ou
Volume d'eau
Volume total
(1 )
0= w.ds (ds : densite sache)

En fonct i onnement norma l , l ' eau adsorbee par le mat e r i a u t e r re


p r ovient de la vape u r d' eau contenue dans l ' a i r et e l l e e s t
loca l i s e e e s s e nt i e l lement dans l a pha s e argileuse qu i p o s s ede une
t re s grande surface spe c i f ique et l e s pores les p l u s p e t it s . Pour
caract e r i s e r cette tendance a f i xe r l ' eau ( - hygro s c op i c it e H ) , on
r e a l i s e en labo rat o i re de s - i sotherme s d' adso rpt i o n H relat ion
t eneur en eau a l ' e qu i l ibre en fonct ion de l ' humidite r e l at ive de
l ' a i r ( H . R . ) . La F i gure 3 en p r e s e nte des exemp l e s qu i mont rent

l'
b ien qu ' e n l ' o ccurence ce n ' est pas la de ns i t e s e che l e parametre
important ma i s bien le type de terre . A e qu i l ibre a une teneur en
eau donne e , on aura - conden sat ion c ap i l l a i r e H ( appa r i t ion d' eau
l i qu ide ) dan s les pores de rayons i n f e r i e u r s a un s e u i l r ( q) que
l ' on peut e s t ime r , conna i s sant l ' i s ot he rme d ' adso rpt i o n , par la
relat ion de Laplace-Kelvin :

_ 20"(T) PI RT
r(O)
=
M
Log H.R.(O) (1)

Avec :

0"Pl
T
coefficie nt de t en s ion supe r f i c i e l l e a i r / eau
la t empe rature du mil ieu ( oK ) .
ma s se volumique de l ' eau ( 9 9 8 , 3 kg /m3 a 2 0°C ) .
( 0 , 0 7 3 N/m a 2 0 °C)

R con stante des gaz parfaits ( 8 , 3 1 4 3 J/mole . oK )


M ma s s e molaire de l ' eau ( 1 8 , 0 1 5 3 4 1 0 - 3 kg/mo l e ) .

Sous l ' e f fet d' un gradient thermiqu e , not amment , le pote n t i e l


d e l ' eau d a n s le mat e r iau t e r re p e u t e t r e mod i f i e d e man i e re
diffe rent i e l l e et divers me cani smes de t ransfert s et /ou changement
d ' e t at app a r a i s s e nt alors . P h i l ip et De Vries ont , les p remiers ,
fo rmu l e l e s equa t i on s gene rales t radu i sant c e s t rans fert s coup l e s
de ma s s e et d e cha leur dans u n mil ieu poreux humide [ 6 ] . Depu i s , de
t re s nomb reux t ravaux ont porte sur ce t heme ( Vo i r par exemple :
[ 7 ] , [ 8 ] ou [ 9 ] ) . Nous nous contenterons , i c i , de rappeler quelques
r e s u l t a t s en r e l a t i on avec la not ion de pa ramet r e s t h e rmiques
apparents .
Current Field Research

S i l ' on supp o s e ve r i f i e e s un certain nomb r e d ' hypot hes e s dont


l e s p l u s i mp o r t an t e s s o n t mi l i e u homoge n e , mat r i c e s o l i de
i n de fo rmab l e et phase liqu ide i n comp r e s s ib l e , fluides parfaits
( vapeur d ' eau e t a i r ) , pas d ' hystere s i s , pas d ' e f fet d e gravit e ,
p a s d ' e changes radiat i f s , e qu i l ibre the rmique i n s t antane , e t c... , Ie
p remi e r p r i n c ip e de la t he rmodynamique ( conse rvation de l ' energ i e )
pe rmet d ' e c r i r e l ' e quat ion de l a cha l e u r , pour un m i l ieu poreux
humide , sous la forme :

Var i at ion d ' enthalpie (a) (b) (c)

Ou ( a ) rep r e s ente Ie t ra n s f e rt d e chaleur purement condu ct i f , (b)


I e t e rme d e t ransport p a r chaleur latente e t ( c ) c e l u i p a r cha leur
s en s ib l e avec :

t Ie temps ( s ) .
C capacite calorif ique ( J/m30K ) .
L enthalpie de vaporisat i on ( 2 4 5 4 , 3 J/kg a 2 0°C ) .
W " chaleur de mou i l lage " ( J/ kg ) .
A conduct ivite thermique ( W/moK ) .
8v ,81
qv
'
: t eneurs en eau volumiques l iquide et vapeur
q : den s i t e s de f lux de vapeur et de flux tot a l ( Kg / s . m2 ) .
(m3/m3 ) .

Cm : capa c i t e calorif ique du melange a i r + vapeur d ' eau ( J/m30 K )

S ' i l n ' e x i s t e dans I e mil ieu que des forces mot r i c e s d ' origine
thermique , une analyse d ' ordres de grande u r p e rmet de n e g l i g e r Ie
t e rme ( c ) devant les deux aut re s . En devel oppant I e t e rme (b)
l ' e quat ion ( 2 ) prend alors l a forme suivante :

d ' un
Ou Dtv
gradient
e s t Ie coe f f i c ient de diffus ion de la vapeur sous l ' e f fet
t h e rmique et Dav Ie coe f f i c i ent de di f fu s ion de la
vapeur sous l ' e f fet d ' un gradient de t eneur e n eau . S i , de p lu s , on
r e t i e n t l ' hypothe s e d ' un e humidite repart ie de man i e r e u n i forme ,
( 3 ) peut se rame n e r a une exp r e s s i on " c l a s s i qu e " de la l o i de
Fourier :

A* et C* s e ront co n s i d e r e s comme e t ant les p a r ame t r e s


t h e rmiques appa rent s d u mi lieu poreux auxqu e l s on a a c c e s p a r la
me sure .

Dispositif de mesure des parametres thermiQues


Pour me s u r e r la conduct ivite thermique et la capacite ca l o r i f i que ,
nous ut i l i s ons des methodes de me s u re en reg ime tran s itoire impose
par " sonde s a chocs thermique s " .

Rapp e lon s que Ie p r i n c ip e de ce t yp e de methode con s i s t e a


creer un choc thermique dan s Ie mat e r i au , dont on veut me surer l e s
prop r i e t e s t h e rmique s , par l ' intermed i a i r e d ' un e l ement chauf fant
a l imente e l ec t r iquement . Une repon s e a cette excitation e st me suree
e n t r a n s i t o i re par l ' i n t e r me d i a i r e d' un capteur p l a c e da n s Ie
mat e r i a u ou dan s l ' e lement chau f fant . La conn a i s s ance de la forme
ana lyt ique de la " f onct ion de transfert" ent ree / sortie
c o r r e s p o n dant a une geome t r i e donnee p e rmet d ' i d e n t i f i e r l e s
parame t r e s thermiques a part ir des courbes expe rimentales [ 1 0 ] .

Ce s met hode s e x i s t ent depu i s p l u s d ' un demi- s i e c l e et s o nt


l a rgement ut i l i s e e s dans d i f f e rent s doma i n e s car e l l e s p r e sentent
de nombreux avantages : s imp l i c ite de l ' app a r e i l lage , rap idite de
l a me s u r e , aut omat i s a t i o n p o s s i b l e et , s u r t out , po s s ib i l it e
d ' op e r e r i n - s itu dan s d e s condit ions hygrot hermiqu e s que l conqu e s
e n r e s p e ctant l ' etat phy s i que du m i l ieu . Pour notre part , n o u s
avo n s c h e r c h e a opt imi s e r des s ondes e n f o n c t i o n de l ' u s age que
nous s ouha it ions e n f a i r e . En part i cu l i e r , nous avon s coneu une
s onde "monot i ge " , pou r l a me sure de la conduct ivite t he rmique , et
une s onde " b it ige " p e rmett ant ega lement l a det e rminat ion de la
cap a c i t e c a l o r i f ique [ I , 1 0 , 1 1 ] . C e s sondes ( Vo i r F i g . 4 ) sont
mises e n p l ace par per eage prealab l e . L e s per formances atteint e s ,
g r a c e a un c o r p s d e chau f f e b r e v e t e d e t r e s f a i b l e i n e rt i e
Adobe 9 0

thermique , nous ont p e rmis d ' att e i ndre une p r e c i s ion de l ' ordre de
±5% que l l e que s o it la r e s i s t an c e t h e rmique de contact
sonde/mat e r iau .

Mesures
1.
la teneurn desen parametres
en fonctionde l'deeyolutio
Methodo l o g i e
temperaturedu materiau terre
eall et de lathermiQues
P o u r me n e r a b i e n c e t t e e t ude , n o u s avons
integre nos s ondes au s e in d ' un dispo s i t i f aut omat i s e a s s oc i ant les
fonct i o n s de c ommande , d ' a c qu i s i t i on e t de t ra it ement , p ou r la
me sure des parame t r e s thermique s , au p i l otage en t emp e rature d ' une
etuve .

Pour de s ra i s on s d ' o rdre p ra t i qu e , nous avons cho i s i d ' etudi e r


l ' evo lution des parame t re s thermiques en fonct ion de la t empe rature
a de s teneurs en eau donnee s . Nous avon s donc realise
s u c c e s s i vement de s teneurs en e au qu e nous avons cho i s i e s
i n t e r me d i a i r e s entre l ' etat sec et la teneur en eau a la
fabr icat ion : (w= O , 1 , 2 , 4 e t 8 % ) c e qui couvre p l u s que l a gamme
des t en e u r s en eau qui peuvent etre at t e in t e s "nat u r e l leme n t " par
l e mat e r iau t e r r e e n f o n ct i onnement no rma l c omme l e mon t r e la
F i gu re 3 . La gamme de t emperature exploree allait de 0 ° a 6 0 °C . Ces
choix ont e t e mot ive s par le souci de s e placer dans de s condit ions
proches de c e l l e s du mat e r iau reel en reuvre .

P o u r obt e n i r c e s t e ne u r s en eau avec une repart it i on au s s i


homogene que p o s s i b l e , nous avons ut i l i s e la methode s u ivante
ap re s in j e ct i o n dans les e chant i l l o n s ( cube s de 10 cm d' a re t e
enve l oppe s d a n s de s hou s s e s e n mat i e r e p l a s t i que ) , de l a ma s s e
d ' eau l i qu ide correspondant a la teneur e n eau souhaitee , o n opere
une redi s t r ibut ion par vapo r i s at i on s ou s l ' e f fet d ' u n chau f fage
m i c r o - o n de s . L ' e f f i c a c i t e de c e t t e me t hode a ete c on t r o l e e en
e f fectuant un certain n omb re de me s u re s au banc gammame t r ique de la
repart it ion de l ' e au dan s l e mat e riau . Le s t eneurs e n eau f i n a l e s
etaient det e rmine e s par p e s e e , par re ference a u po i d s s e c , avec une
pre c i s ion de l ' ordre de 0 , 1 % .

2. Re s u lt a t s D i x e chant i l l o n s de t e r r e crue n o n - s t a b i l i s e e
re a l i s e s a part i r de s ept t e rre s d i f ferentes ont e t e s e l e c t i onne s .
Pour ne p a s a l o u r d i r notre p r e s entat ion , nous ne reportons pa s ,
ici , l ' integralite des resu ltat s obtenus que l ' on pourra ret rouver
dans [ 1 ) . Nou s pre s enton s , a t it re d ' exemple s , les di f fe rents types
de comp o r t eme n t s ob s e rve s e n ce qu i concerne l ' e vo l u t i o n de la
conduct ivite thermique (Vo i r F i g . 5) ou de la capac ite c a l o r i f ique
( Vo i r F i g . 6 ) . La F i gure 7 e f fectue une synthe s e , pour tous l e s
e chant i l l o n s d e t e r r e s d u Dauph ine , de la va r i a t i on r e lat ive de
conduct ivite t hermique par rapport a l ' etat s e c en fonct i on de la
t eneur en eau ponde ra l e .
Te rre c (J/kgOK)
M o re st e l 891 3. I n t e rp r e t a t i on d e s r e s u l t a t s o b t e n u s Tout e s l e s me s u r e s
r e a l i s e e s e n f o n c t i o n d e l a t e neur e n eau o n t mon t re q u e 1 ' on
Isle d'Abeau 81 5
pouvait admettre une variation l i neaire de la capa c i t e c a l o r i f ique
C D300 N°2 81 3 en fonction de la teneur en eau avec une pente egale a la capacite
C D300 N ° 1 81 8 ca l o r i f i que de l ' e au (Vo i r exemple Fig . 6 ) . On peut donc est imer la
capa c i t e c a l o r i f i que d ' un mat e r i au terre humide a p a rt i r de la
M archand 807
va leur obtenue a sec (Vo i r Tab . IV) par :
L a Verpi l l ie re 820
Chumide = Cs ee S . Ceau = 1 OOO.ds.cse e 4, 1 8 1 06 ds W (5)
+ +
Tab . IV :
Cha leurs spe c i f ique s a s e c (J/m3°K) (J/kgOK)
me surees a u calorimetre
Si l ' on c o n s idere ma int enant l ' evolu t i on de l a conduct ivite
the rmique e n fonct ion de la teneur e n eau ponde rale (Vo i r exemple
F i g . 5 ) , on constate egalement une va riat ion prat iquement l ineaire
aux f a ib l e s t e ne u r s en e au . De . p l u s , ce comp o r t e m e n t est
independant d e l a den s i te s e che du mat e r iau cons idere . O n pourra
donc e s t imer s imp l ement l a condu ct ivite thermique d' un mat e r i au
terre humi de , que l l e que s oit sa den s i t e , par :

A(W) ..
= As ee (1 K w.w ) pour 0% � W � 4% (6)
+
Ou K ..
w est une constante dependant un iquement du type de t e rre .
Nou s avon s r e leve , pour l e s t e r r e s du Dauph i n e , des va leurs de K ..w
comp r i s e s e n t re 0 , 0 8 5 et 0 , 1 4 independant e s de la den s i t e s e che .
Compte t e nu de la dive r s it e de s comport eme nt s hyd r i que s que l ' on
p e u t att r i b u e r a d e s t e n e u r s v a r i ab l e s en a rg i l e s de nature
di ffe rent e s , un coe f f i c ient de + 2 0 % par rapport a la va leur a s e c ,
a de faut d ' autres rense ignement s , donnera u n e e s t ima tion plausible
de la conduct ivite thermique du mat e r iau t e rre e n reuvre dan s une
atmosphere d ' humidite relat ive moyenne .

A s e c et pour l e s t e n e u r s en eau f a i b l e s , la condu ct ivite


the rmique e s t t ou j ou r s lege rement de c r o i s s ante en f o n c t i o n de la
Current Field Research

t empe r a t u r e . C e c i e s t dO. a la va r i at ion de l a condu c t i v i t e


the rmique d e la pha s e sol ide , en part icu l ie r d u quart z . Aux teneurs
en eau p l u s e l evee s , par cont re , la condu ct ivit e thermique peut
c r o i t r e faib lement en fonct i on de la temperatu re . I I faut vo i r la
l ' inf luence de la phase flu ide contenue dans l ' e s pa ce poral ( a i r +
vapeur d ' eau et , e v e n t u e l l e me n t , eau c o n de n s e e ) , dont la
condu c t ivit e the rmique augmente e n fon ct ion de l a t empe rature .
Tout e fo i s , ce s va r i at ions restent faible s , ce qu i montre que l ' eau
est fort ement liee dans la phase argileuse d ' un mat e r iau t e r r e , et ,
p rat i quement , on pourra negl iger l ' in fluence de la temperature sur
les prop r i e t e s thermiques ut i le s .

Conclusion
Le t r ava i l que nous venons de presenter peut e t re con s i de re comme
Ie p r e m i e r vo l e t d' un p r ogr amme p l u s amb i t i e u x v i s ant a l a
mode l i s at ion comp l e t e d u comp o r t e ment hygrothe rmique r e e l du
mat e r iau t e r r e . P ou r pouvo i r r e s ou dre l e s e qu a t i o n s cou p l e e s de
t r a n s f e r t de c ha l e u r et de ma s s e dans ce mat e r i a u , i l faudra
e g a l eme nt d i s p o s e r de va l e u r s d e s coe f f i c i e nt s de t ra n s f e r t s
hydriques et d e mode l e s d e l e u r evolut ion e n fonct ion du t yp-e de
terre et des condit ions de temp e rature et de t eneur en eau . C e c i
con s t itue actue l lement un d e n o s axe s d e recherche .

Bibliographie
1. J . P . Laurent , " Cont ribut ion a la caract e r i s at ion the rmique des
mil ieu poreux granu laires : opt imi sation d ' out i l s de mesure in-s itu
des p a rame t r e s t h e rmique s , app l i ca t ion a I ' e t ude des p rop r i e t e s
t h e rmiqu e s d u mat e r i a u t e r r e " , ( Th e s e d e l ' I n s t i t u t N a t i on a l
P o l yt e chn ique d e Grenob l e , 1 9 8 6 ) .

2. J.P. Laurent , " P rop r i e t e s t h e r m i que s du mat e r i a u t e r re " ,


Cahiers du CSTB. Etudes et Recherches,
( 1987 ) .
l ivra i s on 2 7 9 , cahier 2 1 5 6 ,

deuxieme edition", "Earth construction


3. P. Doat et al . ) , "Construi re en terre.
( G roupe CRAt e r re
( P a r i s : Edit ions a lternative s , 1 9 8 3 ) .

4 . H . Houben , H . Guil laud,


( B ru s s e l s " I nt e rn a t i o n a l
primer. volume 8",
c o l l oqu ium on E a r t h con s t r u c t ion
te chnolog ies appropriate t o deve loping count r i e s " , 1 9 8 4 ) .

5. S. Cai l le re ,
deuxieme edition",
6.
S. Mineralogie des argiles.
Henin , M . Rautu reau ,
( P a r i s , Ma s s on , 1 9 8 2 ) .

J . R . P h i l ip , D . A . De Vri e s , " Mo i s t ure


"

movement in porous
mat e r i a l s under t emperature gradi ent s " ,
3 8 , p . 2 2 2- 2 3 2 , ( 1 9 57 ) .

7. P . C r au s s e , " E t ude fondame n t a l e de s t r a n s f e rt s coup l e s


..IT. ,

de
c h a l e u r et d ' hum i d i t e en m i l ieu poreux non s a t u r e " , ( Th e s e de
Docteur e s - S c ience s , INP Toulou s e , 1 9 8 3 )

8. J . F . Da 'i.an , " P r o ce s s u s de conden sat ion et de t ra n s fert d' eau


dans u n mat e r iau me s o et macroporeux . E t ude exper iment a l e du
mor t i e r de c iment " , ( These de Docteur e s - S cie nce s Phys ique s , INP
Grenoble , 1 9 8 6 )

9. C . Moyne , "Trans fert s coup l e s chaleu r-ma s se l o r s du s e chage


p r i s e en comp t e du mouvement de la phase g a z e u s e " , ( The s e de
Docteur e s - Science s , I NP L , Nancy, 1 9 8 7 )

10 . J . P .
poreux :
Laurent , " Eva luat ion de s parametres thermique s d ' un m i l ieu
opt imi s a t i on d ' ou t i l s de me sure in- s it u " ,
Heat & Mass Transfer, P 32 , 7 , 1247-1259, ( 1989) .
Int. Journal of
1 1 . D . Quen a r d , H . S a l l e e , " D e t e rminat ion rap i de de s p a rame t r e s
the rmiqu e s de s mat e r iaux par sonde a c h o c et thermofl uxme t r e s " ,
Cahiers du CSTB. Etudes et Recherches,
(1988) .
l ivra i s on 2 9 4 , cahier 2 2 9 5 ,
3 68 Adobe 90

Terres Ita pise" du Dauphine


0a
(f)W-.J
aen
aCD
a"
r--- �l"-.
��
-� 5 �
:::l:::lL
U�
a<D
If)aa
""
fuseau inf(
:-::.1,,',-::"'--:::..�3 I'-4 \!\.
- fuseau sup.

t<laa
Na
a 1 00 10
\ 1-".1""":: .01 -r--
Fig . 1
DIAMETRE (mm)
Courbes granulome t r i ques de 7 t e r res "a p i s e " du Dauphine .

P HASE SOLi D E GRANULA I R E PHASE ARGILEUSE


TERRE 1::N '"<II �'iQ.ii :8m<II 'i'"<IIii E '"<IIc �i cciE .2�Qi .2'"<II CI)
.:;,gE11
.
VJ

N' 0'"::J J.!()'" :2U.<II D.�. '<I::t:EI � �O�v ��50 §<II �<II �0 � �()
Designation
.
.B
.�
N
. . Yen'"<IIE
. .
>Q;
1 Morestel 88 2 4 6 0 0 0 0 0 0 50 0 0
Marquis 1 00 0 0 0 0 0 6 tr 20 30 15 0 15 0 15
2
3
64Isle d'Abeau
CD300 n02
61 26
19
50
13
5
0
5
0
0
0
7
9
9
25
0 1 00
0
0
25
0
0
0
0
0
0
tr
0
0
4
5
6
CD300 n01
Marchand
88
80
2
0
5
4040
0
5
20
0
0
tr
0
15
13
0
10
57
20
0 0
0
0
20
0
10
43
0
0
tr
7 Verpilliare 49 0 51 0 0 0 20 20 0 0 0 0 40 0

Tab . ! Carac t e r i s t iques de s t e rres du Dauphine etudiee s .

CARACTERISTIOUE DEFINrrlON OBTEN TION REMARQUE

(/)w...J
«m
Porosite tot ale n- VouVou......... . tllal
des pores
1 --ds
d
g
pycnometrie
porosimetrie mereure
eau, Mlium ...
limitee au petit+
pore accessible

0 = Masse sec he pycnometrie


Cl...J Densite sache
ds
1000. Volume tltal pesee etat sec de reference
Masse seche
Densite de grain d pycnometrie glmeralement : 2,65
g 1000.Volume solide
Distribution de Fraction massique de tamisage grains supposes
w0:J diamatres de grain grains de diamatre d<do
en fonction de do
+
sedimentometrie
spMriques

g�a:!;( Composition
mineralogique
% relatijs et nature
des mineraux analyse X diagram me de poudre
microscopie electronique

::< Composition
chimique
%relatijs et nature
chimique des mineraux analyse X generalement exprimee
en oxydes elementaires

Fraction volumique de porosimetrie mercure pores cylindriques

x:>w
Distribution de
diamatres de pores
pores de diamatre d<do
en fonction de do adsorption moleculaire
analyse d'image
par calcul (BJH)
-
a:
�ffl Surface developpee porosimetrie mereure parcalcul
:>«w
a:
Surface specijique
11
des pores ramenee
I'unite de masse
adsorption moleculaire
analyse d'image
direct (BET)
-
Geometrie du reseau Morphologie du reseau performances liees
poreux contig u"ile entre phases analyse d'image aux algorithmes
coordinence des grains d'analyse utilises

Tab . I I P r i ncipa les methodes de caract e r i s at ion d ' un mil ieu poreux
granula i re .
Current Field Research

TERRES A PISE DU DAUPHIN E (FRANCE)


1 ,6

1 ,4
L mbda 0,083 • 1 0'(0 57d)
"\. D .IV
()..•
1 ,2

o �� +
0
� /f +D

La Verpilliere
CD300 N"2
Isle d'Abeau

1 ,0 ... tt x CD300 N°l

.,)V.
+ o Marquis

D+ A Marchand
0,8

0,6
V I9tl -
1 ,6 1,7 1 ,8 1 ,9 2,0 2,1 2,2 2,3
Denslte sache

Fig . 2 : Inf luence de la den s it e s eche sur la conduct ivite


the rmique "a sec" [ 1 , 2 ] .

Distance interfoliaire FAMILLE FORMULE CHIMIQUE

7A Kaolinite Si 2 05 AI �OH)4

Montmorillonite ( 1
Si4 01 o AI 2-x Ridi ca l ! IOH) 2 Catio nx nH 20
10 A Illite (Micas) ISi 4-x AI J 01 oIAl� (OH) 2 Kx
Vermiculite ( � �i (OH)2CEx_y
( Si4_x Al x ) 010 R �y R

'" 1 4 A Chlorite ( Si4_x Al x ) 010 ( R�+ ) (OH) 2 ( R:+ R��J (0I-1) 6

Tab I I I . Les d i f fe rent e s fami l I e s d ' argiles rencontrees dans Ie


mat e r iau terre [ 5 ] .

3
"C0300 N°2"

d s = 1 ,71
3
"La Verpil iere"
ds = 1 ,89

+0 ds =
0 0

2
ds = 1 ,92
2, 1 0 2
0+ ds ds = 1 ,99
= 1 ,83
�� l
1
� 1

o 20
( % )40
H.R.
60 80
0
0 20
( %
40
H.R. ) 60 80

3
o
"Marchand"
ds = 1 ,99
�===0===�----1 3

0
"Isle d'Abeau"
ds = 2,02
ds = 1 ,79
2 2 + ds = 1 ,91

l

o 20 40
H.R. (%) 60 80
0
o 20 40
H.R.
��-
(%) 60 80

Fig . 3 I s othermes d ' adsorption pour 4 te rre s "a p i s e " du Dauphine .


370 Adobe 90

Alimentation eleclrique
de I'element chauffant
Sorties

Support deuxieme
thermocouple
(Inox 0 3, e�O,1 mm) Fig . 4 :
Sondes u t i l i se e s pou r la
� placement thermocouple E
element chauffant -. III
Emplacement
deuxieme
thermocouple E
mesure des parametres
thermiques

III
III
III�II
20mm

"C0300 N°2", densite 1 ,79 "La Verpilliere", densite 2,01


2,0

1 ,8
:r--� --u.. -... � � ..... ..e-!l W 8%
2,5e+6

1 ,6

A. 1 ,4 r--- - - - ...-! 4%
C
2,Oe+6

1 ,2 l ,5e+6

2%
1 ,0
����� 1%
....... � ... sec
0,8
o 10 20 30
T (OC)
40 50 60
l ,Oe+6
0 2 4
w 6
(%)
8 10

Fig . 5 A. = f ( T , w) Fig . 6 : C = f (w) a 2 0°C

'
Fig . 5 & 6 Exemp l e s d ' evolut ions me surees des parametres
:
thermiques en fonction de la temperature et de la teneur en eau

"C0300 N °2" " La Ve r p i l l iere"


� A.
1 ,2 1 ,2
A.sec
x 0 d=2,01
1 ,0

0,8
0 1 ,0

0,8
x =
d l 8,
0,6 0,6

0,4 0,4
0 d = 1 ,79
0,2 x 0,2
d=2

0,0 0,0
0 2 4 6 8 10 0 2 4 6 8 10

"Isle d 'Abeau" " M archand "


1 ,2 1 ,2

1 ,0 1 ,0 0 d=l ,88

0,8 0,8

0,6 0,6

0,4 0,4

0,2 0 d=2 0,2

0,0 0,0
0 2 4 6 8 10 0 2 4 6 8 10

Fig . 7 variat ion relat ive de la conductivite the rmique en


:
fonction de la teneur en eau . Synthese de s resultat s obtenus pour 4
w (%)
t yp e s de terres "a p i s e " du Dauphine
Current Field Research 37I

ABSTRACT CONSERVACION DE UN SITIO ARQUEO LOGICO CONSTRUIDO EN T IERRA

S i t e Tulor 1 has become one * Edua rdo Munoz G . Mon i c a Bahamonde z P .


of the most out standing Universidad d e Anto fagast a Centro N a c i on a l de Conservacion y
archaeo l o g i c a l discoveries C a s i l l a 1 7 0 Ant o f a gast a Rest aurac ion . Av . Kennedy 9 3 5 0
in r e c ent ye a r s in t h e north Chile Sant i ago , C h i l e
of' Chi l e b e i n g t h e o l dest
examp l e o f e a rt h techn o l o gy
used for b u i l ding a v i l l age I . Sit i o Arqueologico Tulor
in t h i s c ountry .
Efforts to preserve this E l S i t i o Arqueo l ogico T u l o r 1 , denominado t amb ien A l d e a de Tulor ,
s it e '5 ruins h a v e generated esta ub i cado en el S a l a r de At ac ama , a unos 1 0 kms . en direcc ion
j o int rese arch work whi c h S . O . d e l Pue b l o San Pedro de At a c ama , Region de Ant o fagast a , Norte
wi l l c omb ine t h e f o l l owing de C h i l e .
studie s :
Asse ssment of i t s present S e gun l a s investigaciones a r queo l o g i c a s que se han e fectuado ,
condition . l a cron o l o g i a d e l S i t i o se ext i ende desde l o s 4 0 0 anos A . C . h a s t a
Ident i f i c a t i on of degrading l o s 1 0 0 a n o s D . C . , y corresponderia a l p e r i odo intermedio t emprano
agen t s t h reatening its ex­ en la secuencia cultural de las regiones at a c amena s . Se t rat� r i a
t inction . de uno d e l o s a sentamientos permanent e s de l o s i n i c i o s d e l periodo
The construct i on of an agropecua r i o de e s t a c omp l e j a cultura l oc a l c on a c t i v i dades de ca­
exper iment a l po lygon . za , recolecc ion e inc i p i ent e agricultura ( A . L l agost e r a et a l . ,
Charact e r i z a t i on of earth 1 9 8 4 ; A . M . Baron , 1 9 8 6 )
b a s e d mat e r i a l s .
Exper iments wi l l be o a r r i e d L a A l d e a de Tulor c o n s i s t e en 2 3 e s t ructura s de p l ant a c i rcu­
o u t using v a r i ous t e chniques l a r , de las que se generan p a s a d i z o s y construc c i ones anexas de
a imed at st ab i l i z i n g the formas mixt a s hasta un t ot a l de 1 0 6 est ructuras dispuest as en c om­
e a rt hen v i l l age ' s wa l l . p l e j a t r ama de crec imiento c e lu l a r . Se accede a l o s recintos p o r
At present , research work v a n o s de puert a de dos t ipos : l o s o r iginales c on dint e l y l o s h o r a
being conduct e d on these d a d o s en l a s paredes de t i erra en d i s t i n t a s epocas , e v i denc i ando
e arthen h e a d wa l l s involves una gran dinamica en su ocupac i on .
among other techniques the
use of c hemi c a l sur f a c e E l conjunto esta defendido por un muro de mayor vo lumen que
t r eatment s ( cons o l i dant s ) , el de l a s estructuras , di spuesto en semic irculo en di reccion de
ethyl s i l i c ate b a s ed . sur a noroest e .

L a const ruccion fue r e a l i zada con b loques de t ierra mode l a do s


KEYWORDS i n situ , di ferenci andose h i l a das superpuest a s . De l o s restos , que
aun se conservan , s e e l evan muros h a s t a una a ltura de 1 , 8 5 m . L a
superf i c i e t o t a l d e l S i t i o que s e c onoce en l a actua l i dad a l c an z a
Conservat ion , C onso l i dat ion , a 2 . 8 0 0 m2
Exper iment a l Po lygon
Aunque este S it io ya e r a c onoc i do desde la dec ada de 1 9 5 0 (G.
L e P a i ge , 1 9 5 7 - 5 8 ) , l a c omun i dad c i ent i f i c a c h i l ena puso e n r e l i e ­
ve s u gran import anc i a s o l o de spues que fue r a redescub i erto en e l
a n o 1 9 8 1 . L o s t rabaj os arqueo logicos se i n i c i aron e n e l a n o 1 9 8 2
( Ve r graf . 1 , fot o 1 )

I I . Prob l emas de Conservacion

Este yac imi ento se ub i c a en un medio amb i ente det e r iorado por e l
av ance d e l a desert i f icac ion , l a que , a s u v e z , h a s i do c ausada
por los camb io s fluv i a l e s ocur r i do s en los ult imos s i glos en e s a
zona . La r e gion del S a l a r de Atacama corresponde a u n desi ert o d e
altura ( 2 5 0 0 m . sobre e l n i v e l del ma r ) . E l c l ima e s t a i n f l u i do
por l a s condiciones que r i gen en Los Andes Mer i dional e s , con un
comportamiento p luvioso c i c l i c o en l o s me ses de d i c iemb r e , enero
y feb rero . L o s meses de i n v i e rno s e c a racteri zan p o r l a ausenc i a
d e l luvi a s y b a j a s t emperaturas . En pr imavera , hay t empor a l e s de
viento cuya fue r z a puede a lcanzar h a s t a 1 0 0 km . por hora .

Graf . N ° 1 (1) 1
Sitio Tu10r P 1 anta de Conjunto

Pol igono de Ensayos


Escal a : o"--=10=' m
� E structuras en cuya forma se bas6 1a reconstrucci6n
372 Adobe 9 0

E l m e d i o amb i ente e s e l mayor agente de degradacion d e l S i t i o


d e Tulor . En sus comienz o s l a desert i f i c a c ion sepulto l a A ldea con
arena a r r a st rada por e l v i ent o . L a f o rmac i o n y avance de dunas h a n
h e c h o e v o l u c i o n a r e l det e r i o r o de l o s restos de l a s c onstruc c i ones
de acuerdo a sus propios camb i o s . El mat e r i a l que desp l a z a el v ien
t o e s arena c on gran c ant i da d de s a l e s s o lub l e s , las que s e deposi
t an p r i n c i pa lmente en l a s c a b e z a s de muro . L a s precipitac i ones di=
suelven las s a l e s . Luego los muros absorben la solucion que cont a ­
mina e l mat e r i a l d e t i e rr a provoc ando e f e c t o s degene r a t i v o s i r r e ­
v e r s i b l e s , t a le s como c r i st a l i z a c i on , sub f lore scenc i a y exf o l i a ­

suma una enorme gradiente de t emperatura entre el d i a y la noche ,


10
c i on . E s t e estado f avorece amp l i amente e l efecto a b r a s i v o d e l v i en
to y arena sobre l a s c a b e z a s de muro a fectadas . A

y una humedad r e l at iv a b aj a , f a c t o res que propician l a f o rmacion


ant er i o r se

de gr i e t a s y fractur a s , asi como l a met e o r i z a c ion d e l mat e r i a l ,


l o s cua l e s , a su v e z , actuan como agent e s mul t i p l icadores d e l dete
r i oro .

L a s excav a c i ones a rqueo l ogi c a s , pract icadas en t emporadas su­


ces ivas entre los anos 1 9 8 2 y 1 9 8 5 ( 1 0 % de la supe r f i c i e de scub i e r

s i glos estuv i p ron cub i ertas por la arena , 10


t a d e l S it i o ), expusieron a l i nt empe r i smo est ructuras q u e durante
que , indudab l ement e ,
h a provocado un incremento de su degradacion en l o s u l t imos ano s .

Otro factor import ante de dest rucc i on 10


const ituye l a a c t i v i
dad t u r i s t i c a que , d e s d e h a c e a l gunos ano s , se h a inten s i fi c a do
en la zona .

I I I . C r i t e r i o s de Int ervencion

L a s invest i gaciones han determinado que debido a l o s e lementos c l i


mat i co s advers o s , e n l a actua l i da d result a improb ab l e c r e a r l a s
condi c i on e s d e conservacion d e l S i t i o excavado p a r a s e r exhibido .
No obstant e , l a preocupacion va mas l e j os a l tener que enfrent a r
e l h e c h o de que l a s estructuras que no h an s i de excavadas y que
se encuent ran enterradas en f o rma natur a l t amb ien sufren e l e f e c t o
d e l desga st e progre sivo de l a s c a b e z a s de muro , ya q u e l a a rena

vos cumpl a n su efecto devast a dor . P r u e b a de 10


que las cub re no e s suf ic i ente barrera para que los v ieritos a b r a s i
ant e r i o r e s l a d i fe
renc i a de a ltura de muros entre e st ructuras l oc a l i zadas en e l sec=
t o r nort e y sur , 0 , 2 0 y 1 , 8 5 m . r e spect ivament e . Consi derando 10
ant e r i o r se preve proxima l a ext i n c i o n d e l S i t i o e n su e s t a do a c ­
tua l .

La import anc i a de e s t e yacimiento para l a invest i gac ion a r ­


queologica y de o t r a s d i s c i p l inas c ient i fi c as exige agot a r l o s e s ­
fue r z o s p a r a s u c onservac i on . D e e s t e modo , l o s c r i t e r i o s d e inter
vencion s e han hecho sobre l a a lt ern a t i v a de c onservar e l S i t io
b a j o t i e rra , conscient e s ademas de otras expe r i enc i a s que recomien
dan e s t a a l t ernativa como l a mas opt ima en s i t i o s con prob l ema�
insolub l es para la t ecnologia actual ( A . A l v a et ---- al . , - 1 9 8 4 ; N .
Stanley , 1 9 8 4 )

IV . C onservacion del S i t i o

IV . 1 . P o l i gono de Ensayo : Reconst ruccion i d e a l de e st ructura s t i ­


picas

La nec e s i da d de comenzar l a s inve s t i ga c i ones para l a conservacion


del S i t i o mot ivo l a const rucc ion de una estac ion exper iment a l de
campo que , a su vez , se pudi e s e ut i l i za r c omo P o l i gono de Ensayo s .
Adema s , de acuerdo con l a s invest i gaciones a rqueologicas , se p o ­
d r i a n reconst ruir hipoteti c amente l a s c o n d i c i on e s i de a l e s de fun ­
c io n a l i d a d de l a s estructur as c i rculares agregando e l e l ement o t e ­
chumb r e ( inexi stent e ) y e l acceso a n i v e l d e l sue l o a d i ferenc i a
d e l a p o s i c ion semienterradas en que s e encuent ran e n l a actua l i ­
da d .

S e escogio l a forma de l a s est ructuras ( E ) c i rcu l ares t ip i c a s


Graf. N ° 2 Plano de las estructu­
ras N ° s 2
das
y 3 utili za­
como modelo en la
N ° 2 y 3 ( Ver graf . 1 y 2 ) que son l a s mej or conservadas d e l c o n ­
j unt o , p a r a h a c e r l a s rep l i c a s . La reconstruccion inclui r i a , a d e ­
mas , e l e s p a c i o que se int erpone entre amb o s c i rculos u n i d o s p o r
muros de e n l ac e semi c i rcul a re s .
reconstrucci6n ideal
usada como pol igono
de ensayos . A poco mas de 3 0 m . desde l a s ult imas construc c i ones de l a
Aldea , en di reccion oest e , s e a l l ano u n e s p a c i o de 3 0 0 m2 para
Escal a : o-==<2 m construir las r e p l i c a s y s e rep lantearon las d o s c i rcunferen c i a s .
La b a s e se e s t ab i l i z o p i c ando l a capa de t i erra que luego se hume­
decio y a p i sono l ogrando se una base de fundacion c ompa ct a .

El S i t i o de Tulor , e s t a construido sobre un depo s i t o a luvi a l


d e t i erra a r c i l l os a , e l mat e r i a l emp l e a do e n l a construc c i on de
la A l de a procede del mi smo lugar . De i gu a l modo para el P o l i gono
Current Field Research 373

(1) Mor a l e s G. , Ricardo de Ensayo , se escogio el mat e r i a l de un sector inmediato al t e rre­


( 1 9 8 3 ) define l a t 'cnica " c a no a constru i r . Deb i do a l a b a j a granulomet r i a y exc e s i v a p l a s t i c i
pping" p a r a l a e s t ab i l i za-=­ dad d e l mat e r i a l de t ierra f u e n e c e s a r i o h a c e r una me z c l a agregan-=­
c ion y recomp o s i ci on de l a s do un 3 0 % de arena .
cabezas d e muro expues t a s
a l int empe r i smo . P a r a aproximarse a l a t 'c n i c a de b l o ques modelados in situ
de l o s mur o s o r i ginal e s , s e opto por l a t'cnica de t a p i a l er a con
( 2 ) C h i ar i , Giacomo ( 1 9 8 3 ) ; un mo l d e sob redimen s i on ado de 0 , 4 0 x 0 , 4 5 x 0 , 6 0 , dan do lugar al
Schwartzb aum M . , Paul et a l . desbaste p o s t e r i or a fin de dar l a aparienc i a de l o s muros o r i gi n a
( 1 9 8 3 ) y Mo r a l es G . , Ricardo l e s . Fue necesario superar a l gunas d i f i cultades p a r a e s t a t 'cn ica
( 1983)
s ant e s
comun i ca ron int e r e ­
expe ri enc i a s
e l uso de c onso l i dante s en
sobre

base a s i l i c a t o de e t i l o .
10
de t a p i a l con mate r i a l de b a j a granulomet r i a . Esto o b l i go a h i d r a ­
t a r l a me z c l a s o b r e l o s n i v e l e s r e que r i do s , que s e man ifesto
en e l agrietamiento de l o s b l o ques durant e e l proceso de secado .
Este inconveniente se supero a j ust ando l a cant idad de a gua en l a
mez c l a y c o n l a c o l oc a c ion d e cobert i zo s de hule para imp e d i r l a
rapida evaporac ion e n e s t e amb iente d e s e quedad ext rema . A l c o n ­
c l u i r l a a lb a n i l e r i a a l a altura de 1 , 7 5 en l a E 2 y 1 , 6 0 en l a E 3 ,
s e desbastaron l o s muros dej ando una leve i n c l inacion h a c i a e l i n ­
t e r i or . Ademas , se c orto u n v a n o de ventana que se c omuni c a con
e l e s p a c i o intermedi o . En l a E 3 se corto en el muro un vane de
puert a , imitando este r a s go o r i gi n a l de l a dinamica de ocupac ion .

La reconstruccion h i pot't i c a de l a t echumb re se h i z o en b a s e


a e v i denc i a s a r queologicas que mostraron indicadores d'b i l e s d e
u n a armadura de t echumbre e n p o s i c ion r a d i a l . E s t o fue r a p i damente
acept ado , ya que resulta la s o lucion mas logica a la f orma c i rcu­
lar de las e st ructuras y p e rmite su c onstruc c i on de f orma c o n i c a
aprovechando l a Menor l ongitud de l a s v i ga s . E l uso de p i e s d e r e ­
chos en e l interior se pudo advert i r por hue l l a s dej a das en l a E 1 6
que e s l a de mayor d i amet ro d e l conjunto ( 7 m . )

Se h i c i e ron l a s armaduras de t echumb r e radiadas y en f o rma


conica con maderos de chanar ( gour l i a e a c o r t i c a n ) . En la E 2 se c o ­
locaron c inco p i e s derechos a media l u z de l a s v i ga s r a d i a l e s p a r a
aminorar l a f lexion . En l a E 3 no fue n e c e s a r i o e l u s o de p i e s dere
chos por su Menor t amano . Las v i gas se unie ron con v a r i l l a s c o s t a-=­
neras de 2 , 5 a 5 c m . de di ametro formando c i rcu l o s conc'nt r i c o s .
Todas l a s p i e zas de madera se ama rr aron con cuero de c am' l i do , s e ­
gun l a t 'c n i c a vernacu l ar andina .

L a s cub iertas se h i c i eron c on una capa de 0 , 1 0 m . de ramas


de la espe c i e arbus� i v a brea ( at r i p l e x ) , c o l ectadas en los a l r e d e ­
d o r e s d e l S i t i o . E l r e l l eno s e h i z o con ramas v e r d e s y s e c a s que
fueron c o l ocadas en paquete s en p o s i c ion radi a l . Sobre l a s ramas
se c ol o c o una capa de barre de 0 , 0 5 m. de espesor , cuya dos i s fue
de un 7 0 % de arci l l a , un 30% de arena y un t er c i o de vo lumen con
p a j a de t r i go . Fina lmente s e c o l oco una cub ierta de t e rminacion
que cons i s t i o en una me z c l a de un 2 5% de arc i l l a mas un 7 5% de are
na . E s t a u l t ima capa se l l ante de gr i e t a s y f i suras debera ser man-=­
tenida c onst antemente ( E . Muno z . , 1 9 8 7 , describe en det a l l e l a
construcc ion d e l a estac ion d e c ampo y p o l i gono d e ensayo )

� En uso c omo Estac ion de c ampo , l a s c a s a s reconstrui d a s o f r e -


cen una exc e l ente adapt a c i on a l c l ima d e l desiert o , tanto por l a
inerc i a t ' rmica d e sus mat er i a le s , e l espesor d e sus muros , e l t i ­
p o d e t e c humb re y qui za , t amb i 'n , p o r s u f orma c i rcula r . E n e l am­
b iente de Tulor de grandes o s c i l a c i ones de t emperatura , las est ruc
turas se mant i enen frescas de d i a y c a l i das de noche .

E l usa como P o l i gono de Ensayo se ha divers i f i ca do entre medi


c iones t 'rmi ca s , medic i on e s del desgaste de los muros expuestos
a l o s v i ent o s abrasivos y al intempe r i smo en general . T amb i'n s e
han ensayado s oluci ones " c apping" ( 1 ) previ ament e a su a p l i c a c i on
en l o s mur o s o r i gi na l e s . De mayor i nt er's han s i do l o s t ra t amien­

cato d e et i l o ( 2 ) , gra c i a s a 10
t o s supe r f i c i a le s y pruebas c on sust anc i a s quimicas a base de s i l i
cua l podrian ser a p l i c adas sin
r ie s go en los muros o r i gina l e s ( Ver gra f . 3 , fot o 4 )

IV . 2 . Inves t i gac ion : c ar act e r i z a c ion de mat e r i a l e s y ensayos e spe ­


cificos

L a s invest i gaciones prel iminares c o n s i s t i eron e n l a caract e r i z a ­


c ion f i s i c o - quimica d e l mat e r i a l de t i e rra usado en l a construc ­
cion de l a s e st ructura s . E s t e mat e r i a l de procedenc i a a l uv i a l se
Graf. N ° 3 Plano del polfgono de
compone de un mayor porcent a j e de a rc i l l a , li mo y arenas f i n a s ,
ensayos y estaci6n de
por 1 0 que su granulomet r i a es b a j a y su p l as t i c i d a d exagera da .
campo .
La man i f e st ac ion f i s i c a es un a l t o indice de retraccion y c o n s i ­
gui ente a g r i e t amiento durante e l proceso de fraguado . L o s anal i s i s
Escal a : Ldm det erminaron a l t a s concent r aciones d e s a l e s solub l e s , que v a r i a n
en l o s dist intos lugares d e l yac imiento y t amb i 'n en l o s muros d e
l a s estructur as . Ot r a s invest i ga c i ones se han o r i ent ado a l a e s t a ­
b i l i za c i on d e l a s cabezas de muro , l a s que , c omo s e h a menc i onado ,
sufren un c onst ante proceso det e r i o rant e .
374 Adobe 90
IV . 2 . 1 . E s t r at i gra f i a de muros

Los t r ab a j o s de invest igacion y sondeo permi t ie r on comprobar l a


existenc i a de una estrat i g r a f i a c l arament e definida , c on d i st int a s
propiedades mecan i c a s dentro de u n mi smo muro , s i tuac i on que s e
encuentra d i rect amente r e l a c i onada con l o s mecani smos de det e r i o ro
que e s t e sufre .

P a r a e l e studio de dicha estrat i gr af i a se h i z o un pequeno c o r


t e en e l muro norte d e l a e s t ructura N ° 5 7 y se ext r a j eron l a s s i ­
gui entes muest ras :

Tab l a I . E s t ra t i gr a f i a d e l muro :

Muest r a Ub i c a c i on Espesor C aract e r i s t i c a s


1 supe r f i c i a l 2-4 cm. costra dura
2 2 ° estrato 2-4 cm. p o l v o muy fino , s i n c o h e s i on
3 3° estrato 6 - 1 0 cm . muy duro y c ompacto
4 muro no alte no determ . mate r i a l homogeneo
r a do

Tab l a I I . Ana l i s i s de mue s t r a s por espect romet r i a de rayos Este


ana l i s i s permite i dent i f i c a r cuant i t a t i v amente l o s componente s mas
import ant e s de la muest r a .
X:
Muestra E l ement o s a n a l i zados % en peso
F e 2 03 T i 0 2 C a O K 2 0 S i OZ A 1203 MgO Na 2 0 S03 Cl NaN03
1 3 ,8 0,5 3 , 3 1 , 9 39 9,1 1,5 15 0 , 4 1 0 ,6 0, 14
2 3,8 0,4 4,6 2,0 41 9,6 2,0 1 3,5 0 , 2 10,0 0, 19
3 3,3 0,4 3 , 2 1 , 6 34, 5 7,5 1,4 17,6 0 , 5 1 3,6 0,12
4 4, 5 0,6 4, 1 2 , 8 59,4 1 4, 8 2,4 4,4 0 , 5 2,0 0 , 005

Tab l a I I I . Dete rminacion granulomet r i c a ( % en peso que pasa por


las ma l la s , Norma ASTM ) pH de muest r a s .

Tami z N ° Muest r a Muestra 2 Muestra 3 Muest ra 4


4 100 1 00 100 1 00
10 98 99 1 00 100
20 96 98 99 98
40 94 96 97 92
60 92 94 94 86
200 67 64 69 51
pH 6,5 6,7 6,6 6,3

IV . 2 . 2 . C a r a c t e r i z a c ion de l mat e r i a l de arrastre

Se r e a l i z o una s e r i e de ana l i s i s que p e rmi t i eran , mediante una


aproxima c i on vo lumet r i c a , conocer la fue r z a del impacto del mat e ­
r i a l d e supe r f i c i e c ausado p o r e l v i ento sobre l o s muros d e l a A l ­
dea .

Tabla IV . Determinacion de granulome t r i a de l mat e r i a l superf i c i a l


d e arrast r e .

Tami z 4 10 20 40 60 200
% peso que
Foto N ° 1 : Deta1 1 e de estructuras
pasa 100 68 35 32 28 8
excavadas en pleno pr�
ceso de deterioro .
T a b l a V . Constantes f i s i c a s del mat e r i a l superf i c i a l de a r r a st r e .

Dens idad Mat e r i a l ret enido Mat e r i a l b a j o


( Kg/m 3 )
- Real saturada
- Real sec a
- Neta
en ma l l a #
2 . 50 9 , 0
2 . 406 , 0
2 . 608 , 0
10 ma l l a # 10
2 . 6 06 , 0
2 . 59 3 , 0
2 . 7 37 , 0
Absorcion % 2 . 51 2 . 04
- C o e f i c iente volumet r i c o med i o C = 0 , 30
( pa r t i c u l a s retenidas en ma l l a 10)#
IV . 2 . 3 . Soluc i on e s capping exper iment a l e s

Se exper iment aron soluciones c apping con mat er i a l e s t r a d i c i on a l e s


como t i e r r a - a rena -cement o y c a l . Los mej o res resultados s e c o n s i ­
guieron con u n a me z c l a de l a s i guient e dosi s : t i erra 2 0 , 0% a rena +
7 5 , 0% + cement a 2 , 5% c a l 2 , 5% = 1 0 0 , 0 % ma s un t e r c i o d e l volu­
+
Foto N° 2 : E structuras sin exca­ men e n f i b r a veget a l .
var ·casi extinguidas
por los agentes degra­ E st e c apping se degrado c a s i t o t a lmente 2 6 me s e s despues de
dantes . su co l o c a c ion deb i do a l desgaste d e l v i ento abrasivo . S i n emb a rgo ,
t i e�e l a vent a j a de un desgaste parej o y no se produj o h o r a da c ion
eo l ic a b a j o la c a p a , efecto c o l at e r a l muy f recuente cuando s e a p l i
c a e s t a t ecnica deb ido a l a diferencia de dure z a c o n l o s mat e r i a�
l e s de la c a b e z a de muro y la me z c l a capping . Esta mez c l a se probo
Current Field Research 375

Mat e r i a le s en altura de 1 , 7 5 m. ( en e l P o l i gono de Ensayo ) con resultados op­


t imos c omprobados 4 afios despues de su co l o c a c ion .
1 . Cement o Port l and
2 . Cal h idraul i c a IV . 2 . 4 . Ensayo de con s o l i da c i on e h idrorrep e l en c i a en muros del
3 . - Agente cons o l i dant e O R p o l i gono de ensayo s
( Wacker Strengthener
OR ) De acuerdo a l o s o b j e t i v o s p a r a cua l e s f u e c re a do e l p o l i gono d e
- Ridrorrepel ente 090L ensayo , se h a n r e a l i z ado s o b r e sus muros nume rosas prueb a s de t r a ­
y 0 9 0S ( S i l ic ona Wacker t amiento superfi c i a l c on product o s a b a s e d e s i l i cato d e et i l o fun
0 9 0 L y 0 9 0S ) dament a lment e . Dichas prueb a s , cuya f i n a l i da d ha s i do la de v e r i fi
Wacker Chemi e : GMBR Werk car e l a decuado c omport amiento de l o s product os en el c l ima deser�
Burghausen 8 2 6 3 . Burg­ t i c o d e l norte de C h i l e , ha permi t i do a l a ve z , opt imi z a r e l met o ­
nausen/OBB West Ge rma ­ d o d e a p l i cac ion y determi na r , previo e studio d e l a s condi c i on e s
ny . T e l efono ( 08677) amb ient a l e s , l a s horas del d i a adecuadas para r e a l i z a r l o .
8 3 2 2 2 2 , Telex 56944 .
C omo un pr imer acercamiento al problema , el uso d e l p o l i gono
de ensayo h a s i de alt amente provechoso , pudiendo comprobarse en
prime r a instancia que la protecc ion de las zonas t r a t a d a s en l o s
muros h a s i do efectiva c ontra l o s agent e s det e r iorant e s .

IV . 2 . 5 . Ensayos de con s o l i da c i on e h i dr o rrepelencia de s o luciones


capping con tecnica de a p isonado

En v i s t a de los result a do s obte n i do s en e l p o l i gono de ensayo , s e


profun d i z o en l a invest i gacion , a p l i c ando e l t rat ami ent o sup e r f i ­
c i a l a mue s t r a s r e a l i z a da s con t ec n i c a de api sonado ( so lu c i on c a ­
pping en p r o c e s o de i nvest i gacion ) Para t a l e f e c t o s e f a b r i c a ron
30 cubos de t ierra de Tulor de 64 cm3 ( 4 x 4 x 4 )
1 0 cubo s muestras t e st i go
1 0 cubo s c o n s o l i dante ( Wacker St rengthener OR )
1 0 cubos cons o l i dante + h idrorrepelent e ( Wacker S . OR S i l i cona
+
Wacker 0 9 0 L )

E st o s ensayo s , actualment e en desarro l l o , t ienen c omo o b j e t i ­


vo c ompa rar e l comport amiento d e l mate r i a l a l ser some t i do a proce
sos de e r o s ion y comprension , asi como impregnacion con niebla s a�
l ina .

De l o s result ados que se obtengan de e s t o s ensayos dependera


l a a p l ic a c ion de los product o s mencionados a l a s o lucion capping
de f i ni t i v a .

V . C onclusiones

E l S it i o Tul o r 1 proyecta su import anc i a a diversas areas de inte­


res c i ent i f i c o . Adema s de contribuir con import ant e s indicadores
para l a a r queo logia como a sentamiento s e dent ar i o t emprano ; para
l a h i s t o r i a de las t ecnologias represent a e l Sitio conoc i do mas
ant i guo construido de t i erra h a s t a ahora en Chi l e .

L a s invest i gac i ones rea li z a das sobre l a s c ondiciones y probl�


mas de c onservacion que afectan a l yac imient o , conl levan a p r e su­
mi r que e s p o c o probab l e c r e a r las condi c i on e s nece s a r i a s p a r a que
permane z c a como s i t i o excavado para ser exh i b i do . No obstant e , l a
degradac ion de l a s estructur a s , aun e st ando enterradas , h a c e nece­
saria l a invest i gacion de t ecnicas especificas para e s t ab i l i z a r
l a s c a b e z a s de muro en c o n s t a n t e det e r i o r o p o r l o s a gent e s men c i o ­
nados . E s t a s s o luciones se deberan compl ement ar c on l a co l oc a c i on
de b a rr e r a s de c ont r o l e o l i c o y l a e l ab o r a c ion de un p l a n p e rmane�
te de preservacion d e l S i t i o post intervenc ion .

L o s a v ances que se han l ogrado h a s t a ahora y que aun se inves


t igan se r e l a c i onan con l a t ec n i c a p a r a l a estab i l i z a c i on de l as
c a b e z a s de muro mediante e l r e emp la z o d e l mate r i a l de t i erra , c o n ­
t aminado y degradado i r r e v e r s i b l emente , por u n a me z c l a de a rc i l l a
y arena d e granulomet r i a cont r o l a da . E s t a e s a p l i ca d a e n l o s mur o s
mediante e l a p i sonado , l ograndose u n a e fect i v a integrac ion ent re
e l mat e r i a l d e l muro y e l de reemp la zo . L a s cabezas de muro estabi
l i z adas s e r i an somet idas a t rat amiento a b a s e de s i l i c ato de et i�
1 0 , en c a s o de obtener buenos resu l t a do s con este product o .

Por t ratarse de un proyecto actua lmente en desarro l l o , l o s


resultados s e encuent ran e n proceso de e l aboracion , p o r 1 0 cua l ,
seran dados a conocer posteriorment e .

Notas

1. A . L l a gostera M . , A . M . Baron P . , .y L . Bravo V . , " Invest i gac i o ­


n e s a r queo logicas e n Tulor 1 , " E studios Atacamefios 7 , ( 1 9 8 4 ) : 1 33-
151 .

2. A . M . B a ron P . , " Tu l o r : p o s ib i l i dades y l imit a c i on e s de un e c o ­


s i stema , " Rev i st a Chunga ra 1 6 - 1 7 , ( 1 9 8 6 ) : 1 49 - 1 58 .
Adobe 90

3. G . L e P a i ge , " Ant i guas Culturas At a c amenas en la Cordi l le r a


Chi l ena , " Anales d e l a Universidad C at o l i c a de Va l p a r a i s o 4 - 5 ,
( 1 9 57 - 58 ) : 1 5- 1 4 3 .

4. A . A l va B . , y G . C h i a r i , " Prot eccion y conservacion de e s t ruc ­


turas excavadas de a dobe , " La Conservacion en Excavac i ones Arqueo­
logicas , ed . N . P . Stanley ( Roma : ICCROM , 1 984) 162 .

5. S . Stanley P . , " Excavacion y conservac i on , " La Conserv a c i on


en Excavaci ones Arque o l ogi c a s , ed . N . P . Stanley (Roma : ICCROM ,

6. E . Muno z G . , " Ruinas de Tul o r , conservac ion y restaura c i on , "


Hombre y Desi ert o 1 , ( 1 9 8 7 ) : 37-52 .

7. R . Mor a l e s G . , " La conservacion en estructur as y decora c i ones


de adobe en Chan Chan , " EI Adobe : Simpo s i o Internac i on a l y Curso­
T a l l e r Sobre Conservac i on del Adob e , e d . PNUD!UNESCO ( L ima : PNUD ,
ICCROM , 1 985) 153.

8. G . C h i a r i , " C aracter i z a c ion d e l adobe como mat e r i a l de cons­


t rucc ion . T � c n i c a s de preservacion , " E I Adobe : S impos i o Internac i o ­
nal Cur s o - Ta l l e r Sobre Conserv a c i 6n d e l Adob e , ed . PNUD!UNESCO
L ima : PNUD , ICCROM , 19 5 153.

9. P . M . Schwart zb aum , y S . Z . Lewin , " Invest i ga c i on sobre e l e f e c ­


t o a l argo p l a z o d e l uso de u n c o n s o l i dante a b a s e d e s i l i cato d e
e t i l o para e l a dobe , " E I Adobe : Simpo s i o Internacional y Cur s o - T a ­
l l e r S o b r e Conservaci6n de l Adobe , ed . PNUD!UNESCO ( L ima : PNUD ,

I�C�C�R
�O�M
� ,� 1�
9�
8�5�)r-�1�
5�
3-.--------------�---

10 . Mor a l e s , " L a conservacion de e s t ructura s , " 113.

Foto N ° 3: M.
Vista aerea del S i t i o Tulor 1 , en donde se aprec i an las estructuras
excavadas entre los aiios 1982-85 . Foto : Ana Baron .

Fotograf!a
Ottorino Ovalley fotomecanica
o. Foto N ° 4
3.
Pol fgono de Ensayos construido a semejanza de las es tructuras orig�
nales N ° s 2 y
Current Field Research 377

ABSTRACT AN I NVE S T I GAT I ON I N TO THE PATTERN OF DETER I ORAT I ON OF DA GA ( EARTH )


STRUCTURES AT Z I MBABWE TYPE MONUMENTS
The m a j o r i t y o f p r e h i s t o ri c
s t ruc tures in southern Africa We b b e r N d o r o
were constructed o f daga , a N a t i o n a l Mu s e ums a n d Monume n t s
l ocal c l ayey soi l which is s t i l l G r e a t Z i mb a bwe R u i n s
being used for bui lding houses P . Bag 9 1 5 8
in the rural areas . Al though Ma s v i n g o
the ma terial does not normal l y Z i mb a bwe
survive f o r l ong periods , at
same archaeol og i cal s i tes the I N TRODU C T I ON
s truc tures remain as evidence
of the I ron Age c omm un i t i es . If The Z i mb a bwe - t yp e monumen t s s c a t t e r e d a l l o v e r S o u t he r n A f r i c a
these ruined s t ruc tures are t o a r e we l l - kn own by b o t h a c a d em i c r e s e a r c h e r s and t o u r i s t s f o r
be presented t o the publ i c as t he i r p r e h i s t o r i c d r y s t o n e w a l l s t r u c t u r e s . The r u i n e d monumen t s ,
s i t e museums , there is a need who s e s e t t l emen t p a t t e r n s r e f l e c t t h e s o c i o e c o n o m i c a r r a n g emen t
to preserve them . In southern and c u l t u r a l e t h o s o f t h e A f r i c an I r o n A g e c ommun i t i e s , a r e d a t e d
Africa , conserva t i on of archae­ AD 1 1 0 0 - 1 5 0 0 . M o r e t h an 1 5 0 o f t h e s e d r y s t o n e o r Z i mb a bwe ­
o l o g i cal s i tes has pl aced empha­ t yp e monumen t s have b e e n r e c o r d e d i n s o u t he r n A f r i c a . The mo s t
sis on dry s t one wal l s and f am o u s and l a r g e s t o f t he s e I r o n A g e monumen t s a r e t h e G r e a t
neglected the daga s t ru c t ures . Z i mb a bwe r u i n s , wh kh oompr i s e t he d r y s t o n e w a l l s and nume r o u s
The goal of the current research daga ( e ar t h ) s t r uc t u r e s o f vary i n g s i z e s , c o n s t ru c t ed over a n
progr amme is to redress this area o f approxima t e l y 7 2 0 hec t a r e s . Th i s a n c i e n t s e t t l eme n t ,
s i tuat i on . The project seeks t o wh i c h p r o v i d e d mo s t o f t he d a t a i n t h i s p a p e r , i s l o c a t e d 2 9 km
underst and the ma terial and the s o u t h - e a s t o f Ma s v� n g o , a m o d e r n t own i n s o u t h - c e n t r a l Z i mb a bwe .
causes of de teriora t i on of the
prehi s t ori c s t ructures by doc­ The d r y s t o n e w a l l s t o wh i c h t he a c a d em i c r e s e a r c h e r and
ument ing the symp toms o f dis­ t he v i s i t o r s have b e e n a t t r a c t e d can o n l y b e d e s c r i b e d a s the
t ress and nature o f decay as so­ s k e l e t o n o f t he p r e h i s t o r i c monumen t s . The f l e s h i s t h e , d we l l i n g
c,i a t i oti wi th them. Thi s approach s t r u c t u r e s , b u i l t w i t h A f r i c a ' s mo s t c ommon i nd i g e n o u s b u i l d i n g
is a l ready yielding resu l t s ma t e r i a l , da g a , a pudd l e d c l ayey s o i l , b i n d i n g t o g e t h e r n a t u r al l y
which are he lping t o unders t and we a t he r e d g r a n i t e g r ave l a g g r e g a t e . When d r y , t h e m i x t u r e
the behaviour o f the ma terial f o rms a d u r a b l e ma t e r i a l , wh i c h i s de s c r i b e d a s " d a g a c emen t "
and s t ruc tures bui l t o f daga . o r " g r an i t e c e me n t " b e c a u s e o f i t s c h a r a c t e r i s t i c s . In pre­
h i s t o r i c t i m e s , t he b u i l d e r s u t i l i z e d t he p l a s t i c p r o p e r t i e s o f
t he ma t e r i a l wh i c h w a s we t , t o c o n s t r u c t s u b s t a n t i a l r o und
ho u s e s and mo u l d ed f i t t i n g s o n the w a l l s and f l o o r s . The f i t t ings
w e r e m a i n l y b e n c he s , k e r b s and b a s i n s . At t i m e s , d e c o r a t i ve
Conserva t ion , prehis toric mo t i f s w e r e p l a c e d on t h e w a l l s or f l o o r s [ 1 ] . S ome o f t h e d a g a
s t ructures , Excava ted s t ructures, f e a t u r e s a r e l e s s e a s i l y i de n t i f i a b l e t o d a y . D i f f e r en t d a g a
Pa t terns of decay , Wa t t le and s u r f a c e t ex t u r e s a n d c o l o u r chan g e s w e r e a c h i e v e d by e x p l o i t i n g
daub , Southern Afri can vernac­ t he v a r i e d c l a y m i n e r a l d e p o s i t s wh i c h w e r e d e r i ve d f r om t he
u l ar archi tecture , Phys i cal and l o c a l paren t g e o l ogy . The m o s t c a r e f u l l y s e l e c t e d and w o r ke d
chemical compos i t ion o f daga . d a g a p r o d u c e d a h a r d , s m o o t h s u r f a c e wh i c h w a s a b l e t o s u r v i ve
c o n s i d e r a b l e w e a r and exp o s u r e . The ma t e r i a l w a s a l s o u s e d f o r
c o n s t r u c t i n g sme l t i n g f u r n a c e s and g r a i n b i n s . S ome o f t he d a g a
s t r u c t u r e s h a v e p a r t i a l l y s u r v i ved i n v a r i o u s f o rms f o r 5 0 0 - 6 0 0
years . I n p r e h i s t o r i c t i m e s , t he d ome s t i c d a g a s t r u c t u r e s w e r e
e n c l o s e d by t h e d r y s t o ne w a l l s , i n o r d e r t o d i v i d e s p a c e i n t o
a r e a s and f o rm c o u r t y a r d s and e n c l o s u r e s . I n s ome s e c t i o n s o f
t he s e t t l emen t , t he s t o n e w a l l s w e r e a l s o p l a s t e r e d w i t h d a g a
s o t h a t t he e n c l o s u r e p r e s e n t e d a homo g e n o u s appe a r a n c e w i t h
t h e dwe l l i n g s .

BACKGROUND TO I NV E S T I GAT I ON S

A l t h o u g h s ome a t t emp t s have b e e n m a d e o v e r t h e p a s t d e c a d e t o


c o n s e r v e t he s t r u c t u r e s a t Z i mb a bwe - t ype monumen t s , mo s t c o n s e r ­
v a t i o n e f f o r t s have f o c u s e d o n t h e d r y s t o ne w a l l s t o t h e e x ­
c l u s i o n o f t he d a g a f e a t u r e s [ 2 , 3 ] . Ho weve r , i t s h o u l d b e
r e c o g n i s e d t h a t t he d a g a s t r u c t u r e s a r e p ar t o f t he mo n umen t s ,
and i t i s d i f f i c u l t t o v i s u a l i s e t he s i t e s i n t h e i r p r o p e r
h i s t o r i c a l p e r s p e c t i v e w i t h o u t t hem . Ove r emph a s i s o n t h e d r y
s t o n e w a l l s g i ve s v i s i t o r s a n d r e s e a r c h e r s a s l i g h t l y d i s t o r t e d
i n t e r p r e t a t i o n o f t h e Z i mb a bwe - t yp e monumen t s .

U n l i k e t h e d r y s t o n e w a l l s , v e r y f ew d a g a s t r u c t u r e s r ema i n
above g r ound . M o s t o f t h e s t r u c t u r e s a r e c o n c e a l e d b y v e g e t a t i o n,
s o i l and r u b b l e d e p o s i t s . O f t h o s e a b o v e g r o und , o n l y p a r t i a l
s t r u c t u r e s s u r v i ve a s ev i de n c e o f t h e pr o t o -h i s t o r i c f e a t ur e s .
The e v i d e n c e o f t h e unde r g r o u nd s t r u c t u r e s a r e t he nume r o u s
mounds s c a t t e red i n s i d e a n d ou t s i d e t h e d r y s t one enc l o s u r e s .
The m o u n d s a r e t he r e s u l t s o f t h e de t e r i o r a t i o n o f t h e o n c e
c omp l e t e d a g a s t r u c t u r e s . Ar c h a e o l o g i c a l exc av a t i o n o f t h e s e
r e ve a l t he r em a i n s o f t h e p r e h i s t o r i c d a g a f e a t u r e s . I n s ome
c a s e s , t h e s e w i l l b e f l o o r f i t t i n g s , d i v i d i n g w a l l s and , a t
t i m e s , a r t i s t i c a l l y mo u l d e d and d e c o r a t i ve f e a t u r e s o f h o u s e s
exp r e s s i n g s ome t h i n g o f t h e s ymbo l i c a n d f i g u r a t i v e e t h o s o f t h e
p r o t o - h i s t o r i c c ommun i t i e s . I f t h e s e s t r u c t u r e s a r e t o be m a d e
t he s u b j e c t o f p r e s e n t a t i o n o r r e s e a r c h w o r k , t h e r e i s a n e e d t o
r e t a r d t h e p r o c e s s o f d i s i n t e g r a t i o n and de t e r i o r a t i o n . Th i s i s
even m o r e c r i t i c a l f o r t he r u i n e d s t r u c t u r es a b o v e g r o u n d . If
l e f t t o t he r a v a g e s o f n a t u r e , t h e y w i l l b e c ome m e an i n g l e s s
mounds o f s o i l d o t t ed i n and a r o u n d t he d r y s t o ne m o n um e n t s .
378 Adobe 90

At t he G r e a t Z i mb a bwe r u i n s s ome a t t emp t s w e r e made two d e c a d e s


a g o t o p r e s e rve t h e d a g a s t r u c t u r e s u s i n g chemi c a l w a t e r - r e p e l ­
l en t s , Th e s e p r e l i m i n a r y a t t emp t s f a i l e d , p a r t l y d u e t o i n a d ­
e q u a t e und e r s t an d i n g o f t h e de t e r i o r a t i o n m e c h an i sm s , a n d t o t h e
h i gh c o s t s o f t he t r e a tmen t . The p r i m a r y a i m o f t h e p r o j e c t
r e p o r t e d i n t h i s p a p e r i s t o d o c um e n t and q u a n t i f y t h e n a t u r e
o f d e c a y a s s o c i a t e d w i t h d a g a s t r uc t ur e s . I t i s hoped tha t
t h i s w i l l g i v e an i n s i g h t i n t o t h e f a i l u r e m e c h a n i sm and p r o v i d e
p r a c t i c a l g u i d e l i n e s f o r t he c o n s e r va t i o n o f p r e h i s t o r i c r e ma i n s .

F I ELD STUDY

G e n e r a l l y , phys i c a l we a t he r i n g , a t mo s phe r i c c o n d i t i o n s , moveme n t


o f s o l u a b l e s a l t s and b i o de t e r i o r a t i o n a c c o u n t f o r mo s t o f t he
decay ass o c i a ted w i t h daga s t r u c t ures . Howe v e r , t he r a t e o f
a ,
d e t e r i o r a t i o n i s a f u n c t i o n o f t he d a g a c omp o s i t i o n , t ex t u r e ,
c o n s t r u c t i o n me t h o d s and s u b s e q u e n t u s e o f t h e s t r u c t u r e s .

\'\
" \ P r e h i s t o r i c d a g a s t r u c t u r e s r e p r e s e n t an e n d - p r o d u c t o f t h e
s e q u e n c e o f e v e n t s r a n g i n g f r om c o n s t r u c t i o n a n d o c c u p a t i o n t o
a b a n d onme n t and t r a n s f o rma t i o n t o a p a r t i a l l y e r o d i n g a r c h a e o ­
l og i c a l f e a t ure . I n t he c a s e o f e x c a v a t e d s t r u c t u r e s , a new
p r o c e s s o f t r a n s f o rma t i o n and d e c a y w i l l b e g i n a g a i n a s s o o n a s
t he y a r e expo s e d t o t h e " n e w env i r onmen t " a b o v e g r o und . In
order t o under s t and t h e pa t t e rn o f d e c a y i t was n e c e s s ary t o
b e g i n i nv e s t i g a t i o n s by a n a l ys i n g t n e de s i g n s and me t h o d s o f
c o n s t r u c t i o n emp l o y e d by c ommun i t i e s s t i l l u s i n g t he ma t e r i a l .

I n v e s t iga t i o n s on c o n t e mpo r a ry d aga c o n s t r u c t io n s

I n s o u t h e r n A f r i c a d a g a h o u s e s a r e s t i l l b e i n g c on s t r u c t e d ,
a l b e i t w i t h s ome m o d i f i c a t i o n o f t he p r o t o -h i s t o r i c me t h o d s .
Howe v e r , s t u dy o f t h e c o n t empo r a r y c o n s t r u c t i o n me t h o d s c an
g i v e i n s i g h t i n t o t h e f a i l u r e me chan i s ms a s s o c i a t e d w i t h s u c h
s t r u c t u r e s a n d a n i n d i c a t i o n o f t h e na t u r e o f s t r u c t u r a l pr o b ­
l ems e n c oun t e r e d when u s i n g d a g a a s t he p r i m a r y b u i l d i n g
ma t e r i a l . S u c h s t u dy p r o v i d