Scientific tools for assessment of the degradation of historic
brick masonry.
K. Van Balen
R.Lemaire Center for Conservation, K.U.Leuven, Leuven (Heverlee), Belgium.
Abstract
The aim of the research on brick masonry degradation supported by the D.G. XII is
presented.
A systematized questionnaire has been conducted that allows the collection of
expertise from different sources in relation to damage on historic brick masonry structures.
The main initial sources are literature, data from in situ investigation carried out on historic
brick masonry buildings and laboratory simulation tests. This information is collected in an
“Atlas of damage to historic brick structures” s containingu uniform terminoloev and a
uniform description of damage types and damage origins. This information is schematized
into relations between damage types, causes and the physical phenomena related to a set of
deterioration processes. Particular attention is paid to the interaction between the different
materials of which ancient brick masonry is composed and the effect of environmental
factors.
The deterioration causes can only be understood with a knowledge of the different
historic brick masonry construction types and the historic, climatical and geographic context
in which those masonries have been built; used and eventually restored.
A Masonry Damage Diagnostic System is developed that - through systematization
of this knowledge and relations - will guide the person using this expertise toward the
assessment of damage type and damage cause. It also provides information on the type of
investigation that has to be carried out on the masonry to be diagnosed, in order to improve
the quality of the information about the damage type and damage origin. Therefore, appropriate (monument-friendly non-destructive) testing techniques are evaluated on their
likelihood of increasing the precision of the diagnosis. For the creation of the expert system
an expert’s system shell is used based on decision tables.
The scientific methodology used for the development of both tools is given.
Key words: Masonry damage diagnostic system, expert system, historic brick structures,
masonry, damage atlas.
UJ
1
Introduction.
Conservation and restoration of historic structures depend on various factors and include
aspects related to different disciplines. For architectural historians involved in decision
taking processes for the conservation of historic buildings the decay problem is often
evaluated in a different way as by engineers or material scientists. From the other hand
historical information (e.g., severe damage which occurred in the past by an accident) is a
very valuable source of information if existing damage is to be evaluated.
Multi-disciplinary approach and possibility of exchange of information between
different partners are hence crucial when choices between options have to be made.
Diagnostic tools to evaluate the condition of brick monuments will have to take this into
account.
A research group coordinated by the author developed instruments like the Masonry
Damage Diagnostic System (MDDS) and a Atlas of Damages of historic brick structures
which aimed at improving the diagnosis before intervention but which also can be used to
evaluate previous interventions
2
Masonry Damage Diagnostic System and Atlas of damage to historic brick
structures.
21
Objective
The objective of the project is to improve our knowledge of the effects of environmental
factors on damage to Europe’s cultural heritage and to guarantee better treatment and
protection of our heritage, by providing the professionals who work on the analysis of ancient
buildings with an expert system including a damage atlas.
22
Scientific approach
Tie aim of the project was to create an instrument bases on scientific information which
could increase the number of persons to execute general monitoring of historic buildings. By
increasing their number there should be a shift toward maintenance type interventions on
historic buildings instead of more “heavy” restoration interventions [I]. Therefore groups
of technicians, architects, engineers should be helped in executing correct analysis of the
major part of (more simple) damage cases, leaving the more difficult and special cases to the
smaller group of leading professionals. The latter will be thus only involved in those cases
where high specialism is required. It has also been noticed that even for specialists those
instruments could be a helping guide when they have to deal with fields in which they do not
have very specialized knowledge. To define the group of less specialized professional
(nevertheless having a broad background) the profile of the “Monumentenwachters”
(Monument Watchers) of the Netherlands or in Flanders has been taken as a reference. The
specificity of the scientific approach can be summarized by the following headings.
2.2-l Multidisciplinary approach:
The approach takes into account different dimensions of the problem from material sciences,
environmental sciences, biological sciences, historic sciences, architecture and urban
development;
2.2.2 Complexity:
The durability of the masonry does not only depend on the behaviour of one material but on
a complex interaction of different materials and different boundary conditions (e.g., traffic,
pollution, precipitation, orientation, etc.);
2.2.3 Scope:
The evaluation and interpretation of damage do not so far include a proposal for treatments,
but is limited to the diagnosis. It gives precise descriptions of damage types and the process
responsible for this damage in the given circumstances.
2.2.4 Damage approach:
The relation between damage definition (type) and damage cause (mechanism) is based on
a thermodynamical approach: resistance against stress defines the damage (function). In that
respect approaching from visible damage, the process of decay is the right way to make tools
that can be used for monitoring.
2.2.5 Usefulness:
The definition and terminology are based on maximum use of visual analyzing techniques
and maximum collection of relevant data at the site. Additional guidance is given through
testing techniques from non-destructive to destructive techniques if necessary.
23
Methodology
The methodology used is based on the scientific principles of research. In the problem stated,
damage to historic brick structures, and the development of the Masonry Damage Diagnostic
System defined the scientific description and deductions of the physical mechanisms causing
the damage. Inherent choices have been made in relation to the order of input taken into
account for the deduction: it started with the most easy way of identification which is visual
analysis (this also explains the usefulness of the damage atlas) and then included in order the
in situ and laboratory tests. A thermodynamical model of decay of materials as developed
in the project [2] defines damage as a result of stresses and resistance allowed damage types
and damage causes to be linked. This was then developed in terms of processes within the
Masonry Damage Diagnostic System. The development of the Masonry Damage Diagnostic
System allowed us in many cases to limit the problem only to those parameters which are
described really necessary (goal oriented), thus omitting irrelevant elements while the link
with the practice remains guaranteed.
This approach is unique in this field and the experience of the project demonstrated
the scientific interest of this engineering approach for the evaluation of deterioration of
ancient brick structures. It is an applied scientific approach producing practical results for
the conservation of historic brick monuments.
2. 4
Research results.
2.4.1 Terminology.
In the first stage of the work common definitions had to be set up which should be useful
within the different instruments of the project. A first set of damage types has therefore been
defined which is useful for the questionnaire but could be developed within the Masonry
Damage Diagnostic System. The hierarchic concept allows the user to narrow his definitions
gradually. This approach was based on the assumption in the project that defining goes hand
in hand with increasing knowledge. Vague knowledge needs broader terms while added
information has to narrow the set of terms. In the questionnaire the group of damage types
is given with the subset of more precise terms. The logic in the definition was based on the
visual discrimination. It has been defined so that in the order of the analysis the visual
appreciation was the frrst “instrumentls used. As the questionnaire and the related damage
atlas are related to this first analysis, it was logical that the definition of damage types should
use visual criteria. Within the Masonry Damage Diagnostic System the same terminology
has been used while the “instrument” itself allows the user to check his interpretation of the
terminology by answering questions defined by the conditions contained in the definition of
the term. The structure of the terms of the damage types and definitions given in terms of
conditions will be illustrated later.
2.4.2 Questionnaire.
The original questionnaire with which the project was started and which stemmed from the
collaboration of the experts of the NATO-CCMS pilot study on Conservation of Historic
Brick Structures has been modified considerably as described in one of the project reports [3].
The main reasons for this development are found in the interaction with the setup of the terminology which was developed in relation to the Masonry Damage Diagnostic System.
Another argument was the subdivision of the questionnaire according to the order of
investigation of the damage type into its possible cause. The first type of analysis is based
on visual analysis and in site measurements while the second part has been made for the integration of laboratory results leading in the third part to a synthesis and an interpretation.
It was experienced that even experts contacted by purpose are hesitant to formalize
the collection of information in a questionnaire if they see no direct interest or reply leading
them to the diagnosis. As an instrument to monitor the questionnaire seems to have the
disadvantage to be to lengthy, not precise and synthetic enough. This seemed very
discouraging but it was realized that in fact this event proved the advantages of the Masonry
Damage Diagnostic System. It has therefore been concluded that the questionnaire should
be developed electronically. So that the user, after giving some basic information on the
building and according to the previous answers he has given, is requested to answer only
those questions which are relevant to the problem, which is what the Masonry Damage
Diagnostic System does. The latest version of the questionnaire itself became simpler and
has been developed with a view to the rapid collection of the necessary data related to
pictures for the damage atlas.
It was also seen that in practice very few professionals and owners are willing to
deepen the diagnosis with laboratory tests and even with tests in situ. It is the experience of
most of the partners in the project however that in many cases only those types of analyses
can provide the necessary information for a full diagnosis. There are many evidences as the
example described in on the case of a neo-gothic church in Pamel, near Brussels [4] that the
analyses of the pore structure of the mortars and bricks, and the chemical composition of the
different mortars used, are necessary to define whether the damage to the brick was due to
the mortar or to the composition of the bricks. In this matter the project has the task of
proving the interest of different types of analysis for the diagnosis of damage and to stimulate
proper diagnostic methods in that respect.
Within the project the questionnaire is also used as a structured way of collecting
documentation on different cases from all the partners. Examples from Belgium, Germany,
Italy and the Netherlands have also been documented in this way. The full-scale masonry
models of the Politecnico di Milan0 set up for monitoring damage processes are also included
to check the possibility of evaluating the effectiveness of the questionnaire from a dynamic
viewpoint, i.e., can the questionnaire inform us about the speed of the degradation process
PI7.
2.4.3 Damage atlas.
A classification of damage patterns found in brick masonry has been set up [6]. The terminology used follows the same structure and setup as the questionnaire and the Masonry
Damage Diagnostic System. The definitions are more extensive than the definitions
presented in decision table form in the Masonry Damage Diagnostic System and possible
damage causes are also given. A complete set of illustrations of the different damage types
is provided in the atlas with an explanation about the possible causes of the damage.
Consistency with the Masonry Damage Diagnostic System was guaranteed as they were
developed to be used together. An example of a page of the damage atlas as it will be
published is given in Fig. I.
2.4.4 Masonry damage diagnostic system.
The Masonry Damage diagnostic System (MDDS) [7] is a Knowledge-based System (KBS).
It is a database with information and additional knowledge to create relations according to
given answers and questions. It is the translation of expertise into a system computers can
handle. The knowledge is structured and can be consulted using a Decision Table System
Shell (DTSS) developed by TNO-Bouw. It is actually a programme running on Macintosh
computers but a Windows version is on its way. The Masonry Damage Diagnostic System
is a prototype and is not for sale. It is the aim of the partners of the project to create links
with organisations that would like to develop the existing system for their use and co-finance
in this way the updating and extension of the system.
‘&b: &:f+~ ::o.b:C: (.iLi:.j:& .q ~(4 .cV:o&p*):
.---‘----.-.;-....-.---‘----.....”.”^’-.---~,-ml!% 1;: ;T;mr’.r 1ygi:.
C~lxcx~~-,tL;~~s u:ii~riq hir:txk.d ::::G:~I +&.
i?PaZ!S!P),Ill!>'
TC9.j tara *itl:~!ir:~S~ :fiJfX idis
W;yeT:tir.!; 9f 55,:; !v;!i.i 2!'S;r.:C.',!.i."?. c-,:.&*
y&-t*-:
--&w\t,- *-2---I.- w-m. .y’
. cc.
----‘fi.c cl* .i.-r:
.&. I--t; r::e: r-ttr ,;t’ if,.; !;v&tt;r~
'ck?C !*Jkl.:,
I
.a
: !,-a ?h+. I;r:Jc ]:,:.+rLu’ :z pcz:cc <,:a:,,
:c, :Q<.:*l
;.i:.;: :L ?i:;c c3?XF!d iYili:>r is &y,: (,):,y, :Q c,:i:,l,v ::: ;i:- iqjr).>.
I.. pfli!,:;+ r)G Gw:tf,~c occ:013j
F<i
c,i C:-c
r,-.i -.
:..I _A.
-*‘iom!!: v&xl: i 2 2. >a&;? iq !‘+>b:G3:& :<. a:(( f-y :*!.:
-- c,-. c I..
. .
d31!VC:i:, X3X LG:O;c$ c;Q;tr,rg&~.
lo tte *:dkr)L a;’’ ::+-h?;*rl: .>>~t*$)t;:y
'.~ I>.,X?
COG. I.3:. il:e @);9&->;rC r:r'r,y;:.l&:$
tb;c+l $2;. ;w.-: i’.r, l .h.,J& t;.’8.3- t& z _ zd \fi:: i8q it*:: :q jc:.* :*);Q;pr+d.
‘- - - -L-m-.L-.--C-CCCI_..
. . . . -----------...-------- ---__--...---.
_._-____
___________ ..-.-..----------_.....,
I
I.-------_.----..____---.......----..
. -.-...-m-.-.-.5.__ _ _. . . . . . _.
----___._.--..-_.---------.-....-,
Fig. 1: : Example of a page of the damage atlas.
The knowledge is translated into a large set of decision tables themselves containing
sets of conditions and actions. The set of condition tables is structured in a hierarchic way
and can be presented as a decision table tree [S].
Fig.2 presents the starting branches of the tree with the sub-tree of processes checked
by the system. Fig.3 shows the tree of the damage types.
~~~~~
Fig. 2: : Tree structure of the main branches of the Masonry Damage Diagnostic System:
determination of the type of damage (detail in next figure) and checking of damaging
process responsible for the damage.
The computer programme aims to make the consultation of the Masonry Damage
Diagnostic System user-friendly. In contrast to the questionnaire the Masonry Damage
Diagnostic System will adapt its questioning according to previous answers given.
Comparative data as pictures and comparative results of test trends are also given so that if
there is uncertainty or when the answer to be given is not clear, additional information can
be consulted. A typical screen picture is given in Figure 4.
discoloration
\e pi$i jq
Fig. 3: : Structure of the decision tables containing the knowledge about the determination
of damage types; the hierarchic approach corresponds with the structure of the damage
atlas.
.I’
:-:
.:a
.*I
11
‘7’
.;:
:;:
i
fj
)+
r:::
fg
.*<
&
3”
.:,
82
‘p
:,:’
.>I
:.-.
I.
g:;‘:
.
I
,.
.
.
.
...
:
;,
..
.
:
,
..
.
..
.
.
:
..
.
.
:
:
.
i
CH
Fig. 4: : Example of a consultation screen using the Masonry Damage Diagnostic System.
The output of the Masonry Damage Diagnostic System is given immediately on the
screen, but a printed report of the consultation is also generated. This document can be used
as a report by the user. The results of a consultation can be stored in a file which allows
progressive and step-by-step consultation in different stages. During the progress of the
project it has been shown that it is very useful to create, using the Masonry Damage
Diagnostic System, an interesting collection of examples of damage. The structure of the
KBS within the Masonry Damage Diagnostic System obliged the users to work in a very
systematic way while developing the system and using it afterwards.
3
Conclusions
The description of the damage types on historic brick structures has been improved. The link
between terminology, questionnaire, damage atlas and Masonry Damage Diagnostic System
is guaranteed by accurate definitions and the hierarchic concept of the damage types
descriptions. In the same way the description of possible damage causes and the processes
leading toward the above-mentioned damage types have been improved. The increase in
systematization which was sought by the project is a necessary and interesting by-product of
the creation of Knowledge Base systems (KBS).
The Decision Table System Shell (DTSS) has proved to be a very interesting tool for the
development of the KBS.
The collection of different types of information using the questionnaire and other
related summary sheets allowed the research team to make an interesting collection of
degradations of different types. Those examples are useful for the elaboration of the damage
atlas.
Reactions on the complexity of the questionnaire and uncertainty on the part of the
user about which information should be mentioned in this document prove the advantage of
the Masonry Damage Diagnostic System. It is certain that in the dialogue between the expert
system and the user the latter will feel more comfortable as he will think that the questions
which are asked are relevant to the problem, though on the other hand the expert system will
limit his questioning only to that information it can consider in its reasoning. Further
development of the Masonry Damage Diagnostic System could be guaranteed by the
integration of research results from other projects and by the inclusion of data available on
the degree and effect of pollution. Let us not forget however that no expert system will ever
be as competent in each of the domains of expertise as real experts. Nevertheless the
Masonry Damage Diagnostic System will be of great interest for many routine operations
where leading professionals are now wasting their time and could be replaced by low-profile
professionals if they are helped with a system such as the Masonry Damage Diagnostic
System, provided that one issue of the system remains: “for this (difficult and complex) case
I have no answer. Please call an expert”.
Acknowledgement:
To D.G. XII of the E.C. which supported the Environment R&D project EV5V-CT92-0108
and to the partners in the project: from the Politecnico di Milan0 (Italy): Prof. G. Baronio,
Prof. L. Binda; from the Building research Institute (The Netherlands): ir. R. Van Hees, dr.ir.
L. Lucardie, drs S. Naldini, ing. L. Van der Klugt; from the Technische Universitat
Hamburg-Harburg (Germany): Prof. L. Franke, drs. geol. min. I. Schumann and at our own
institute ir. J. Mateus.
References:
PI
PI
PI
PI
PI
PI
Van Balen K., (1997) Monitoring of degradration in view of the selection of
treatment strategies, in: BAER N. S., SNETHLAGE R. (Eds) Saving Our
Architectural Heritage: the conservation of historic stone structures, J. Wiley and
Sons Ltd., Chichester, pp* B 67-180;
Van Balen, K., Mateus, J., (1993) Broad categories for causes/agents of damages,
unpublishedproject document, K.U.Leuven, 12 p
Van Balen, K., Mateus, J., (1994) What damage can we see and how?, unpublished
Project report on task P.2.4., 78~.
Van Gemert, D., and K. Van Balen, (1988) The influence of pointing mortar on the
alteration of brick and masonry, Proceedings of the 2Vh international Brick, Block
Masonry Conference, p. 175 1-1758, Elsevier Applied Science, London and New
York.
Binda, L., Baronio, G., (1993) Presentation of the full scale masonry models and the
monitoring of damage process, Nato-CCMS Pilot study “Conservation of Historic
Brick Structures” meeting, Venice 22=23/I l/93.
Francke, L., Schumann, I., (1993) Classification of damage patterns found in brick
masonry, unpublished project report, TUHHarburg, September 1993, 43~.
Van Hees, R., (1995) Entwicklung eines Systems fur die Diagnose von Schaden an
historischem Ziegel-Mauerwerk auf der Grundlage eines Knowledge Based System,
in E.G. Niel (Ed.), Instanhetzen von Mauerwerk, WTA-Schriftenreihe, Heft 6,
Aedification Verlag, pp. 12- 18
Van Hees, R., and Silvia Naldini, 1995b. The masonry damage diagnostic system,
International Journal for restoration of buildings and monuments, Vol. 1, No.6, pp.
46 l-474.