Architectural Characterisation

CHAPTER 2
Architectural Characterisation
Before economical and cost-effective seismic retrofit solutions can be identified and de-
veloped for New Zealand’s earthquake risk unreinforced masonry (URM) buildings, it
is first necessary to accurately assess their structural seismic performance. Before the
structural characteristics of such buildings can be ascertained and in order to develop a
comprehensive understanding of the nature of the URM building stock, the architectural
characteristics of these buildings must be defined. This chapter outlines general struc-
tural configurations (typologies) which apply specifically to New Zealand URM buildings.
Other characteristics are outlined, in terms of materials and wall geometries, and how
these relate to the overall architectural typologies. The outcome of this characterisation
research is used for finite element modelling of New Zealand URM structures as well de-
signing and evaluating appropriate retrofit interventions.
2.1 Introduction
Within the architectural characterisation of URM buildings, the broadest and most impor-
tant classification is that of the overall building configuration. The seismic performance of
-5- Alistair P. Russell

Chapter 2. Architectural Characterisation
a URM structure depends on its general size and shape. A small, low-rise square building,
when subjected to induced seismic forces, will behave differently than a long, row-type,
multi-storey building (Robinson and Bowman, 2000). In addition to this, retrofit inter-
ventions which may be appropriate for one type of building may not be appropriate for
another, different type of building. From a heritage perspective, The New Zealand His-
toric Places Trust recognises the importance that the configuration of a building has on
its behaviour during an earthquake. It states that buildings up to three storeys high with
many walls can generally withstand a reasonable amount of earthquake shaking, whereas
buildings higher than three storeys with open-plan frame-type geometries are much more
vulnerable (Robinson and Bowman, 2000). And in a more general sense, Magenes (2006)
notes the importance of building configuration in the overall seismic performance of ma-
sonry buildings. He states that “the main factor that produces the variation of the OSR
(overstrength ratio) is the geometric configuration (plan and number of storeys).” The
OSR arises from the analysis of the inelastic cyclic deformation and energy dissipation
capacity of a structural unreinforced masonry system.
The purpose of classifying URM buildings in New Zealand into typologies has two facets.
The first of these is to provide a framework for the Seismic Retrofit Solutions research
project, and in particular, for the aspect of the project on URM. In order to conduct
research into both the seismic assessment and seismic retrofit of URM buildings it is nec-
essary to first know the nature of what is being assessed or retrofitted. Research into
seismically improving the performance of structures needs to focus not only on generic
retrofit science, but on providing tailored retrofit solutions for New Zealand-specific struc-
tures. The second facet of classifying URM buildings is how the typologies are applied.
When considering an assessment of a specific building’s expected seismic performance, the
typology which that structure is classified into can be made use of in the early stages of the
analysis. The performance of a “prototype” or generic structure in a given typology can
be applied to the specific structure under consideration and gives a basic understanding
of the expected seismic performance. In this context, most typological assessments will
be for the purpose of initial assessment, as a first step, and will provide some parameters
from which to conduct further detailed analysis. Generalising buildings into categories
also provides a basis to analyse structures through finite element modelling (FEM). This
Alistair P. Russell -6-

2.1. Introduction
enables an analysis of structures through modelling without needing building-specific pa-

rameters. At the same time, while the models are generic, they are realistic because of
the detail and comprehensiveness of the typological characterisation.
A systematic, consistent and convenient method is needed to accurately capture the state
of the building stock on a national, regional and local level. In New Zealand the whole
country is under the jurisdiction of the Department of Building and Housing, and legisla-
tion for structures and construction is governed by the Building Act 2004 (DBH, 2004).
The country is divided into districts over which Territorial Authorities (TAs) have author-
ity and responsibility for new and existing structures. The Building Act 2004 required
TAs to adopt a policy on earthquake-prone buildings1 by the 31st of May 2006 (DBH,
2004). The first step in the policy developed by each TA requires a preliminary assessment
of their building stock to determine the number and types of buildings with the potential
to be earthquake-prone. This involves both reviewing earthquake hazards within the TA
district in accordance with NZS 1170.5:2004 (Standards New Zealand, 2004) and consid-
ering the structural performance of the building stock in an earthquake taking account
of age, construction, location and use in relation to the earthquake hazard in that dis-
trict (DBH, 2005). The option is given for TAs to take an active or passive approach
in the implementation of their policies, and the appropriateness of each depends on the
size of the district under the jurisdiction of that particular Territorial Authority and the
number of potentially earthquake-prone buildings in it. An active approach involves the
TA carrying out an initial evaluation of buildings in its district to identify those likely
to be at high risk, and then establishing priorities for further, more detailed evaluations
and guidelines of required performance levels for upgrading. A passive approach is more
ad hoc, where evaluations of buildings are conducted on a case-by-case basis, and usually
triggered by an application to the TA under the Building Act for building alteration,
change of use, extension of life or subdivision. For larger TAs, an active approach would
1
“A building is earthquake prone for the purposes of The Building Act 2004 if, having regard to its
condition and to the ground on which it is built, and because of its construction, the building — (a)
will have its ultimate capacity exceeded in a moderate earthquake (as defined in the regulations); and
(b) would be likely to collapse causing — (i) injury or death to persons in the building or to persons
on any other property; or (ii) damage to any other property,” (quoted from The Building Act 2004
(DBH, 2004)). Furthermore, “a moderate earthquake is legally defined as: in relation to a building, an
earthquake that would generate shaking at the site of the building that is of the same duration as, but
that is one-third as strong as, the earthquake shaking (determined by normal measures of acceleration,
velocity and displacement) that would be used to design a new building at the site.”(DBH, 2007)

be more appropriate as it enables the adoption of the best possible risk-reduction pro-
gramme and sets and controls the level of any work required to mitigate risk. From this
point of view, classifying buildings into typologies is a useful tool for territorial authorities
in New Zealand to assess their building stock. While it is not envisioned that a “one-size-
fits-all” solution is viable for all URM buildings, for initial assessments and vulnerability
analyses, classifications of buildings into groups is a useful and necessary exercise. Also it
is an achievable first step, before having to undertake the daunting task of analysing every
building under a TA’s jurisdiction. It is envisioned that the use of typologies developed
here will assist TAs throughout New Zealand to classify their building stock for seismic
vulnerability assessments and the economic implications of their at-risk building stock.
2.2 Background of Building Typologies
The word typology is used as a classification according to a general type, and in the
sphere of architectural characterisation different groupings of buildings can be classified
correspondingly. Categorising buildings into typologies is a useful way of grouping them
according to common features or elements. The study of typologies is widely used in the
field of architecture to define buildings according to spatial or formal qualities, context or
function.
Building typologies are generally used as a method of categorising the building stock on
a large scale, although there are different levels for distinguishing typologies. Some re-
searchers have utilised the concept of typologies on a regional level, some for urban/rural
distinctions, some applied to specific towns, and others have looked at particular parts
of towns. While the use of typologies in some locations is specific to that location and
may not have the same applications elsewhere, there are common principles that can be
applied generally.
Quatremère de Quincy, the editor of the Dictionnaire d’Architecture between 1788 and
1825, first explained the idea of typology in his seminal text Dictionnaire historique
Alistair P. Russell -8-

2.2. Background of Building Typologies
d’Architecture (1832), and drew attention to the contrast between model, which is an ex-
act and replicable plan for a building, and typology which is a category or rule for grouping
specific models (Lavin, 1992). To explain typology, Pontius (2007) draws a parallel with
the concepts in biology of phenotype and genotype, where genotype is the genetic consti-
tution of an individual organism, and phenotype is the set of observable characteristics of
an individual resulting from the interaction of its genotype with the environment. Argan
(1996) and Rossi (1982), in particular, investigated a path of architectural theory where
instead of looking at a building as an isolated element on the landscape they shifted their
focus towards a study of cities and context. For them, the building’s function can adapt
to the dynamic life of the surrounding city, but the basic formal relationships within
the structure and to adjacent structures remain the same. Pontius (2007) recognises the
strength of categorising buildings according to typology without depending on the build-
ing’s use, because the inherent structural properties are much less likely to change over
time than the function of the building.
Much previous research involving typologies in structural engineering has been on case
studies — villages, cities (or parts of a city) or regions. It is a way of characterising
the building stock, and has been done frequently for this purpose. In Germany building
typologies have been used for the purpose of identifying energy consumption and demand
(Neidhart and Sester, 2004). This research was performed with residential housing as the
focus, and buildings were classified as one-family house, row house, small more-family
house or large more-family house, and then classified according to building age. This al-
lowed specific heat demand to be determined on a systematic basis. In Portugal, Lourenço
et al. (2006) investigated moisture and weathering effects on historic building stock, and a
useful tool was classifying such buildings into typologies, rather then examining individual
buildings. This was a case-study, and again was used with mainly residential dwellings as
a focus, and clear distinctions could be made to distinguish building typologies. Common
details and features within typologies allowed common defects to be detected.
The most relevant research (to seismic assessment and retrofit) which has been published
in the field of structural engineering involves the use of typologies in seismically active
areas. The classification of historic buildings into typologies has been used for the pur-

pose of seismic vulnerability analysis, and in this case, many such historic buildings are
constructed of unreinforced masonry. The purpose of typological analysis is to find as
much information on buildings as possible. This information used in the context of the
seismicity of the region, for example, can show the seismic vulnerability of those struc-
tures. The concept of classifying buildings into typologies has been used by researchers
in Italy, Portugal, Greece, Slovenia, Canada, The United States and Iran. The object of
the vulnerability assessment can vary in domain from villages, to towns, to regions and
even to whole countries.
2.2.1 European URM Typology Studies
Italy
As part of developing a procedure to quantify the reduction in seismic risk for historic
structures, D’Ayala and Speranza (2003) analysed four case-study towns in the Marche
region in Italy and characterised their building stock into typologies. It was found that
although the typologies differed between towns in terms of several factors such as overall
dimensions, foundations, and wall-floor connections, similarities could be seen in terms
of the horizontal structures and masonry characteristics. Typologies were analysed and
assigned vulnerability classes (low, medium, high, extreme), according to the level of seis-
mic risk associated with the overall configuration and structural details.
Similarly Binda has characterised buildings into typologies in Italy (on various levels), for
the purpose of assessing the seismic vulnerability of historic towns in earthquakes (Binda,
2005, 2006a; Binda et al., 2005). The approach was taken to analyse whole historic towns,
and not single buildings, and the purpose of this was to develop a database with rescue
plans and design interventions for the preservation of cultural heritage. One facet of the
research focused on four sample areas in the province of Perugia (Binda, 2005; Binda
et al., 2005), while another facet enabled more general conclusions to be drawn on the
overall state of URM buildings in Italy as a whole (Binda, 2006a). It was noted that
representative typologies of a town can be easily recognisable through similar features
Alistair P. Russell - 10 -
(number of storeys, exposure, type of façade, material and structural elements). On a

regional scale in the case of Perugia, three typologies were identified and the structures
were classified as either simple isolated buildings, row buildings, or complex buildings
(Binda, 2005; Binda et al., 2005). On a more general scale, which could be applied to
the whole of the historic Italian building stock, seven typologies were identified (Binda,
2006a) (see also Figure 2.1):
• Typology A: isolated houses and/or dwellings;
• Typology B: row houses;
• Typology C: palaces;
• Typology D: bell towers;
• Typology E: arenas;
• Typology F: churches and cathedrals; and within this typology,
– Typology F1: churches, plan based on Latin cross scheme;
– Typology F2: churches, with a central plan.
Although Argan (1996), Pontius (2007) and Rossi (1982) remark that buildings generally
retain consistent structural configurations while the function of the building may change
over time, Binda et al. (2005) points out that the evolution of the building may still have
a significant effect on its seismic vulnerability: “born as an isolated building, it could
have become a row building or a complex one. . . The more complex the building is, the
more difficult the detection of its vulnerability is; therefore, its structural evolution should
be known as much as possible.” It was also noted that a geometric survey may not be
sufficient to ascertain a comprehensive understanding of the site or building, and effort
should be made to research possible modifications over time. Nevertheless, a “minimalist
approach” can still yield useful and significant information by sampling from buildings
representative of the whole, particularly with regard to materials used on an urban scale.
This may also be useful to detect local and compatible materials and techniques for rep-
resentative prevention and repair measures.
Valluzzi et al. (2005a) performed a simplified yet comprehensive assessment of two towns
- 11 - Alistair P. Russell
in the Treviso province in Italy. Through the use of kinematic models describing the loss
of equilibrium of structural macro-elements and classification of buildings into typologies,
vulnerability assessments for the whole of each town and proposed intervention techniques
were presented, without needing to analyse each individual structure. The criteria used
to define the typologies was based on dimensions, presence of contiguous constructions
(isolated or row buildings) and the presence of colonnades (a row of columns supporting
a roof) at ground level.
The limitations of general classifications have also been reported. As part of a similar
study into seismic vulnerability in Perugia, Valluzzi et al. (2005b) indicates that com-
plex buildings in seismic areas have specific vulnerability aspects which can make it very
difficult to apply or adopt common structural assessment procedures usually applied for
regular buildings. Still, if there are a sufficient number of complex buildings, common
features and details can be made use of as Binda (2006a) notes in the case of churches.
Portugal
Vicente et al. (2006) analysed traditional building stock in the old city centre of Coimbra
in Portugal. A vulnerability assessment was conducted in conjunction with a GIS map-
ping tool, which allowed a global view of the site in question. The authors determined that
without this, inadequate decisions in terms of rehabilitation and refurbishment policies are
often made. In the area under consideration, the historic buildings were classified mainly
according to dimensions and construction materials. A relational database of building
characteristics that govern the structural behaviour of historic buildings, with particular
focus on their seismic vulnerability, was connected with a GIS database. This allowed
the mapping of different damage scenarios and risk reduction actions associated with
strengthening interventions. Also, the GIS tools were used to provide a global overview
of building quality, current conservation status, risk scenarios and potential damage map-
ping for different seismic intensities. The authors intended that the GIS and database
system could be applicable for other regions and old city centres, and if necessary easily
adapted and modified for specific building features.
Figure 2.1: Typologies for the Italian URM building stock (Reproduced from Binda
(2006a))
Greece
Karantoni and Bouckovalas (1997) performed a vulnerability study considering almost

all unreinforced masonry buildings in the town of Pyrgos in Greece after four significant
earthquakes in or near the town. Buildings were classified according to the masonry ma-
terial (adobe, stone or clay brick) and number of storeys (one or two). Another parameter
which was recorded for each building was the time period of construction. This was an
important factor as it not only determined the masonry material but also the structural
form of the building. Four basic time periods were distinguished: 1800–1850, 1850–1900,
1900–1940 and 1940–present. It was determined that old adobe masonry buildings ex-
hibited poor seismic performance, while more recent stone and brick masonry buildings
exhibited a much better seismic response. Perhaps the most useful conclusion was that
Prygos was thought to be a typical Greek town, and that buildings of other Greek towns
would behave almost with the same manner during the actions of near-field earthquakes.
Thus it was thought that the results of this study could be applied more widely through-
out the country and would allow more reliable planning of seismic scenarios for the future.
Slovenia
Tomaževič and Lutman (2007) researched typologies in Slovenia and recognised that ty-
pologies vary from region to region and from urban to rural areas. In the context of
Slovenia, construction materials are usually those which are locally available – limestone
and slate, and sometimes brick elsewhere in the country. It was found that in cities,
stone masonry houses are mostly three or four storeys (Figure 2.2), but are limited to
two storeys in rural areas. The distribution of walls is usually uniform in orthogonal
directions and because of the thick walls, the wall/floor area ratio is large, sometimes up
to 10 %. Floor structures are usually wooden and few ties are used to connect them with
the walls. Roof structures are also wooden and are usually covered with ceramic tiles.
The authors note that although the typology of heritage buildings varies from region to
region and country to country, similar observations can be made regarding the response
to earthquakes. Also, damage occurring in such buildings can be classified in a uniform
way. It was stated that the two main causes of damage to URM buildings is inadequate
structural integrity (particularly connections between walls at floor levels), and inade-
quate structural resistance, and can result in diagonal shear cracking and disintegration
of walls and partial or total collapse of buildings.
Figure 2.2: Typical masonry buildings in Ljubljana, Slovenia (Reproduced from

(Tomaževič and Lutman, 2007))
2.2.2 North American URM Typology Studies
Canada
Nollet et al. (2005) analysed historical buildings in Old Montreal. Old Montreal has been
declared an historic district in Quebec, and contains buildings of significant architectural
and cultural heritage, as well as being in a seismically active zone. Montreal is somewhat
older in its original European settlement than most of New Zealand. Nevertheless, there
are important similarities in terms of the principles of determining typologies and their ap-
plications to the locality in focus. Much of the early construction in Western Canada has
greater parallel to European (particularly French) construction of the time than to other
American construction, particularly in the settlement period between 1680 and 1860. In
the earliest periods of European settlement in New Zealand, construction bore greater
similarity to British construction practices, before its own practices developed (Ingham,
2008).
The historical buildings in Old Montreal were considered to be those built before 1929,
and an inventory of such buildings was classified into typologies according to construction
type (masonry, wood, steel, concrete), number of storeys, and the period of construction.
The date of construction is important, although Binda et al. (2005) and Blaikie and Spurr
(1992) take this further and state that the age per se 2 and evolution of the structure are
important factors for evaluating the condition of the structure.
It was noted by Nollet et al. (2005) that typological classification is used in many ap-
proaches to assess the seismic vulnerability of a group of buildings, and although it is
generally defined for the population of buildings under study it is sometimes used at a
larger scale. The implication of this is that sometimes larger scale classifications are not
necessarily applicable for a smaller population. For example, the Canadian typological
classification is based on the descriptions given in the report ATC-21 of the Applied
Technology Council of California (ATC, 1996). It was recognised that while this may be
reasonable for parts of North America, including Western Canada, direct application of
this classification to the population of buildings in Old Montreal was questionable. It was
concluded after the analysis of the inventory that there were notable differences between
structural characteristics of those buildings and the North American typologies identified
in ATC-21. This confirmed the necessity to develop an adapted method for evaluation of
the seismic vulnerability for historic areas of Quebec. It was acknowledged that a more
general classification system was a useful starting point, from which tailored typological
analyses could then be developed.
2
Blaikie and Spurr (1992) state that, “a common observation after recent earthquakes has been that
newer buildings designed to modern day standards performed much better than older types of construc-
tion. This type of statement is often directed at “old” buildings constructed in the 1930’s and earlier.
However almost identical statements were made shortly after the 1931 Hawke’s Bay earthquake. Some
of the new buildings referred to are now considered old and beyond redemption. The implication of this
is that age per se, not just date of construction, is an important factor even for purportedly permanent
building materials.”
The United States
Erbay and Abrams (2004, 2007) recognised a need to identify the relative impacts that
rehabilitation may have on performance of large populations of buildings, instead of indi-
vidual buildings, so that regional losses can be estimated and mitigated. This information
is useful for public officials, emergency response managers and other public stakeholders.
In their study, the authors state that “as a gross extrapolation, structural characteris-
tics are generalised for a wide range of possible masonry buildings inherent in a region.”
Nevertheless, it was thought that the estimated response and loss of the entire popula-
tion of buildings in the sample region was not sensitive to probable deviations of some
buildings from such generalised characteristics. Within this simplified model a number of
building specific parameters were varied as input to the damage model. These parame-
ters included number of storeys, floor area, storey height, floor aspect ratio, wall area to
floor area ratio, average pier height, uniform load over storey, elastic modulus of masonry,
wall thickness and masonry density. The sensitivity analysis identified that hazard-loss
relationships which were unacceptably scattered for individual building loss calculations,
could be utilised to estimate risk/loss at a regional level (Erbay, 2003).
2.2.3 Iranian URM Typology Studies
Fragility curves have been reported as a useful method of estimating the probability that
buildings will exceed a specific state of damage for a given seismic intensity. Karimi and
Bakhshi (2006) investigated the development of fragility curves for URM buildings before
and after seismic upgrade in Iran. Although the typologies which were used in the mod-
elling consisted only of single-storey and three-storey structures, useful conclusions on the
expected damage in URM structures could be drawn, which could then be extrapolated
and applied to structures on a general basis.
A study was conducted in Iran to classify existing rural houses of Iran (Bakhshi et al.,
2005; Ghannad et al., 2006; Mousavi Eshkiki et al., 2006). From field inspections, nine
typologies of rural houses were recognised as representative of all rural houses in Iran.
The study classed these typologies according to building material (brick/cement block,
cement block, stone, adobe, wooden) and geometry of roof (flat, arched). For each ty-
E-W
pology, the constituent percentage of the overall building stock and their general location
N-S
were determined, as well as typical deficient details. Finite element modelling was con-
ducted on structures representative of each typology (see Figure 2.3) and the dominant
s, earthquake in modes
failure N-S were ascertained.
Figure 15: Normal stress, earthquake in E-W
on direction
Figure 2.3: Typology in Iranian rural residential building stock; comparison of model and
d roofs Figure 17:
prototype (Reproduced fromComparison
(Ghannad et al.,of2006))
prototype and model
4. CONCLUSION
2.2.4 Previous Research into URM Typologies in New
ant types of rural Zealand
houses in Iran was presented. To have a rough estimation of
baked and unbaked adobes, tests were performed on samples taken from some
to field investigations,
When analysing adobe walls withofarched
the performance roof (model
URM buildings 1) and
in previous brick walls
earthquakes, withand
Blaikie
were known as dominant vulnerable dwellings damaged in Zarand earthquake. To
Spurr (1992) made observations on the influence of overall configuration on building col-
mechanisms of typical rural houses of Iran, elastic numerical models of these two
veloped andlapse.
the Itresults
was noted
werethatcompared
square shaped buildings
with have aexamples
the actual lower susceptibility
observed to earthquake
by the
damage than either irregular or rectangular buildings. The same authors also reported
partially on the nature of New Zealand’s URM building stock, by surveying buildings in
Wellington City, Nelson City and the Petone suburb of Lower Hutt City. These three
locations were thought to be representative of a main city, a provincial city and a subur-
ban shopping district respectively. The data used was last updated in May 1988. It was
concluded that the bulk of the sample population’s earthquake risk URM building stock
2.3. New Zealand URM Building Typologies
had one or two storeys, and that three or more storeys were found almost exclusively in
Wellington City. This was extrapolated to also conclude that research into New Zealand’s
existing URM buildings should be directed towards buildings having three or fewer storeys.
2.3 New Zealand URM Building Typologies
It has been identified that the New Zealand building stock warrants seven typologies,
which are outlined in Table 2.1. Buildings are separated according to storey height, and
whether they are an isolated, stand-alone building or a row building made up of multiple
residences joined together in the same overall structure. A suggestion of the expected
importance level of the structure is also given, according to AS/NZS 1170.0:2002 (Stan-
dards New Zealand, 2002). The prevalence of URM structures according to the number
of buildings in that typology which make up the overall building stock is shown in Table
3.3. Typology A are single storey isolated structures, typology B are single storey row
structures, typology C are two storey isolated structures, typology D are two storey row
structures, typology E are three and higher storey isolated structures, typology F are
three and higher storey row structures and typology G structures are religious, institu-
tional and industrial structures. For all row-type structures, (B, D and E) the dimensions
of each individual substructure can vary, that is, they are often heterogeneous.
Typologies are distinguished according to two factors – storey height and building foot-
print. The storey height significantly influences the way the ground floor walls behave
under lateral loads. This is because different axial loads can cause different failure modes
of those walls. Walls of multi-storey buildings are usually thicker on the ground floor than
on higher storeys, and also much thicker than single storey walls. The second distinction
between typologies is whether they are row or isolated buildings. This has significant
implications for torsional effects from lateral loads. A long row building can be thought
to be more torsionally stable than an isolated structure, which tends to have a ground
footprint with aspect ratio of between 1:1 and 1:2. A row building could have a ground
footprint aspect ratio of up to 1:10, or even higher. Also, isolated structures tend to have
an open front with corresponding low stiffness in that direction. This means the centre
of stiffness is not in the centre of the building. Pounding effects can be significant in
row buildings, particularly where floors are made from concrete or are at different levels.
Pounding is usually not a problem in isolated structures but can occur where there is
insufficient clearance between neighbouring structures.
A prototype (“mean”) structure for each typology has been established and FEM mod-
elling can be conducted on each structure. This enables a “standard” modelling procedure
to be applied to any structure based on its typology. The information for these typolo-
gies has been obtained from building surveys throughout New Zealand. The locations of
buildings and origins of photographs is shown in Figure 2.4.
Auckland
Paeroa/Thames
Taihape
Bulls
Palmerston North
Nelson
Wellington
Christchurch
Ashburton
Temuka
Timaru
Oamaru
Dunedin
Figure 2.4: Locations of buildings surveyed throughout New Zealand
Table 2.1: URM Typologies
Typology Description Prevalence Importance Description

(rank) level
(AS/NZS
1170.0)
A One storey, iso- 4 2 One storey URM buildings, of-
lated buildings ten with an open front. Exam-
ples include convenience stores
in suburban areas and small of-
fices in a rural town.
B One storey, row 3 2 One storey URM buildings with
buildings multiple occupancies joined with
common walls in a row. Typical
in main commercial districts, es-
pecially along the main street in
a small town.
C Two storey, iso- 2 2/3 Two storey URM buildings, of-
lated buildings ten with an open front. Ex-
amples include small cinemas, a
professional office in a rural town
and post offices.
D Two storey, 1 2 Two storey URM buildings with
row buildings multiple occupancies joined with
common walls in a row. Typical
in commercial districts.
E Three or more 7 3/4 Three + storey URM buildings,
storey, isolated for example office buildings in
buildings older parts of Auckland and
Wellington.
F Three or more 6 3/4 Three + storey URM build-
storey, row ings with multiple occupancies,
buildings joined with common walls in a
row. Typical in industrial dis-
tricts, especially close to a port
(or historic port).
G Institutional, 5 3/4/5 Churches (with steeples, bell
industrial and towers etc), water towers, chim-
religious neys, warehouses. Prevalent
throughout New Zealand.
Within the above typologies, further distinctions have been made. For example, Typology
A buildings can be divided into those which have a dividing wall in the centre (Typol-
ogy A1), and those which do not (Typology A2). Typology G buildings are generally
monumental
AS/NZS structures
1170.0:2002 and those which do12not fit easily into the other categories, and
A2 usually
3.4 for such structures
ANNUAL unique OF
PROBABILITY problems are presented, and unique analyses are neces-
EXCEEDANCE
sary. 3.4.1
Nevertheless
Ultimate there are useful sub-classifications which can also be made within this
limit states
For ultimate
grouping. limit statesTypology
For example, for structures
G1ofbuildings
importanceare
levels 1 to 4, the
religious annual probability
buildings of
in New Zealand,
exceedance (P) for wind, snow and earthquake loads shall be as given in Table 3.3.
Typology G2 are
The design warehouses
working and factories
life of structures with
that are veryfor
erected large tall sides
a number and periods
of short large open
of usespaces
insideand
anddismantled between each, is equal to the total of the periods of use.
typology G3 are institutional buildings such as ferry terminals and train sta-
3.4.2 Serviceability limit states
tions. Further detail on each typology is given in subsequent sections.
Serviceability limit states shall include—
A4 (a) SLS1—the structure and the non-structural components do not require repair after the
SLS1 earthquake, snow or wind event; and
AS/NZS
(b) 1170.0
SLS2 —Table 3.1 is maintains
the structure reproduced here ascontinuity
operational Figure 2.5
afterto
thegive
SLS2an indication of the
earthquake.
relative
For importance
serviceabilitylevels
limit of URM
states for typologies (further
structures of details
importance can2 be
levels found
to 4, the in Chapter
annual
probability of exceedance (P) for wind, snow and earthquake loads shall be determined as
3 of AS/NZS 1170.0
given in Table 3.3. (Standards New Zealand, 2002)). All structures identified fall into
NOTE: Guidelines for limits associated with serviceability events are given in Appendix C.
importance level 2 or higher, with medium to high consequences for loss of human life.
TABLE 3.1
CONSEQUENCES OF FAILURE FOR IMPORTANCE LEVELS
Consequences Importance
Description Comment
of failure level
Low consequence for loss of human life, or
Minor structures (failure not likely to
Low small or moderate economic, social or 1
endanger human life)
environmental consequences
Medium consequence for loss of human life, or
Normal structures and structures not
Ordinary considerable economic, social or environmental 2
falling into other levels
consequences
High consequence for loss of human life, or 3 Major structures (affecting crowds)
High very great economic, social or environmental Post-disaster structures (post disaster
consequences 4
functions or dangerous activities)
Circumstances where reliability must be set on a
Exceptional 5 Exceptional structures
case by case basis
Figure 2.5: Building importance levels from AS/NZS 1170.0 – Table 3.1
2.3.1 Parameters for Differentiating Typologies
Storey Height
Typologies are separated according to whether the buildings are one storey, two storey
or three or more storeys. Buildings taller than three storeys are not common in New
--`,``,,`,`,,,```,,``,,``,,```,-`-`,,`,,`,`,,`---
COPYRIGHT
Copyright Standards New Zealand
Provided by IHS under license with SNZ Licensee=UNI OF AUCKLAND/5936938001
No reproduction or networking permitted without license from IHS Not for Resale, 06/05/2006 18:13:17 MDT
Zealand and exist mostly in the central business districts (CBD) of the largest city, Auck-
land, and the capital, Wellington, as well as some other South Island cities, such as Timaru
and Dunedin. While one and two storey buildings are approximately evenly distributed
throughout the country, three and higher storey buildings are few in number and a sin-
gle typology to classify all URM buildings that are three storeys or higher is sufficient.
Moreover, the difference in expected seismic behaviour between a three and four storey
building is significantly less than the difference between a one and two storey building,
for example. This is particularly because three and higher storey buildings tend to be of
masonry frame construction (on at least one face of the building, usually the front and
back faces), in contrast to solid wall construction, and have much thicker walls in the lower
storeys. As a broad generalisation, in masonry frames rocking of piers between windows
and openings is the expected in-plane behaviour when subjected to lateral seismic forces
(Abrams, 2000), and diagonal shear failure is less likely. For walls without openings (or
with small openings), and depending on the magnitude of the axial load, the expected
in-plane failure mode in an earthquake is likely to be either sliding shear failure, diagonal
tension (shear) failure, or rocking of the wall itself.
Another reason why buildings are distinguished according to storey height is because of
the axial loads acting in the walls. One storey buildings with low axial loads are less likely
to exhibit diagonal shear failure and are more likely to rock or slide. The bottom storey
walls in a taller building are more likely to fail in shear because of the higher axial loads
that they are subjected to. But the thickness of the ground floor walls is usually greater
than the thickness of the walls higher up the building in typology E and F structures,
meaning that large lateral forces are necessary in order to generate sufficient shear stresses
in these walls to cause cracking or failure.
Building Footprint
The second primary characteristic for separating buildings into typologies is the building
footprint. That is, whether the structure is considered a stand-alone, isolated, (almost)
square building, or a row building made up of multiple residences joined together with
common walls. This accounts for Typologies A – F, whereas those buildings with a non-
uniform ground footprint (for example, many URM churches) will fit into the Typology
G classification.
In row structures containing walls common to neighbouring residences, pounding can in-
crease the lateral forces on an adjoining structure during earthquake loading. This is
especially likely when floor or ceiling diaphragms in neighbouring residences are at dif-
ferent levels. Different heights for the force transfer into the common wall can result in
punching shear failure of the wall or diaphragm detachment and collapse. The effects
of pounding are greater when concrete floor diaphragms are in the structure, compared
with timber diaphragms. Conversely when diaphragms connected to shared walls are at
consistent heights within the structure, the seismic resistance is greatly enhanced due to
the increased stiffness in one direction. Essentially square or round buildings with well
distributed walls generally have a greater torsional resistance than buildings with less
evenly distributed lateral force resisting walls (Robinson and Bowman, 2000). Long row
structures have different torsional properties than isolated buildings.
A significant difference between isolated buildings and row buildings becomes evident at
the time of upgrading the structure. An isolated structure usually contains few residences,
perhaps two shops for example, or occasionally more. Row structures may contain many
residents, even ten or more. An isolated structure is generally considered just that — a
single structure. A row structure, despite behaving in an earthquake as a single inter-
connected structure, may be perceived as different buildings. It may be more difficult to
conduct earthquake strengthening on an entire row structure at one time compared with
seismic retrofit of an isolated structure. If seismic retrofit interventions are implemented
on only a part of a structure, such an intervention may be ineffective.
2.3.2 Details of New Zealand URM Typologies
Typology A – One Storey Isolated Buildings
Typology A buildings are one storey isolated structures, and are the fourth most prevalent
in New Zealand, out of the seven typologies identified. Equally common in both small
towns and large cities, these structures are today often used as convenience stores, or
small commercial premises. Typically they have an approximately square ground foot-
print, but may have longer side walls making the building rectangular. Sometimes there
is a dividing wall through the centre of the building, usually with no piercings, but some-
times there is a doorway width opening. That dividing wall can be one or two leaves
thick, but single leaf dividing walls are non-load bearing. Typology A1 refers to buildings
with a dividing wall, and Typology A2 refers to those buildings without a dividing wall.
Generally the external side walls have no openings. The front may be largely open, with
up to 90 % of the width of the front open, with glass windows, ranch-sliders or doorways,
and the rear wall usually has openings in the form of doorways. It is thought that many
front walls originally had closed-in fronts, but that legislation changed which allowed the
whole front wall to be removed. This was perhaps not part of the original design (Smith,
2007). The external side walls usually have no openings. A parapet is usually on top of
the front wall, and often the side walls slope from the top of this parapet to the top of
the rear wall, meaning that the side walls are of trapezoidal shape. Also the central wall,
if present, follows the same shape as the side walls and slopes from the top of the parapet
to the back wall. Sometimes there is no parapet on the front. The side walls are usually
either two or three leaves thick, as are the front and back walls. The ceiling diaphragm
is usually timber with a suspended ceiling underneath, and light cladding over the top,
usually either tiles or corrugated iron. Typology A structures usually contain either one
of two residences, and this usually corresponds to either A1 or A2 structures. In the long
direction (L1 in Figure 2.6) the building is stiff, but because of the large openings on the
front face, the building is less stiff in the L2 direction.
Figure 2.6 and Table 2.2 show typical dimensions of a Typology A1 structure, with an
internal wall in the centre of the building, and Figure 2.7 and Table 2.3 show typical
dimensions of a Typology A2 structure, with no centre wall. Upper and lower bound
dimensions are shown, as well as those of a typical Typology A building. The mean
dimensions of such a building can be used for finite element modelling, and will enable
expected seismic behaviour patterns to be established for this Typology. Photographic
examples of Typology A buildings are shown in Figure 2.8.
Figure 2.6: Overall dimensions of a Typology A1 Building
Table 2.2: Typical Dimensions of a Typology A1 Building
Dimension Lower bound “Mean” Upper bound

structure structure structure
(m) (m) (m)
L1 Overall length 6 8 16
L2 Width of residence 4 6 8
P1 Height of parapet at front 1 1.2 2
P2 Height of parapet at rear 0 0 1
H1 Height from ground to 3 4 6
bottom of parapet
H2 Height of front opening 2 2.5 4
F1 Width of front opening 3 4 6
G1 Depth of rear wall below 0 0.5 2
front street level
Figure 2.7: Overall dimensions of a Typology A2 Building
Table 2.3: Typical Dimensions of a Typology A2 Building

(m) (m) (m)
bottom of parapet
front street level
(a) Parnell (Auckland) (b) Ashburton
(c) Temuka (Canterbury) (d) Seatoun (Wellington)
(e) Epsom (Auckland) (f) Thames
Figure 2.8: Typology A buildings – single storey isolated
Further photographic examples of typology A buildings are shown in Appendix A.
Typology B – One Storey Row Buildings
Typology B buildings are one storey row structures, differentiated from typology A struc-
tures in that they are essentially made up of multiple typology A buildings joined together.
They are slightly more prevalent throughout New Zealand (the third most prevalent out
of the seven typologies identified) than typology A structures, and often occur in small
provincial towns, especially along the main street which is the commercial centre of the
town. Single storey row structures tend to be smaller in the overall length of the block
than corresponding two storey row structures (Typology D). Generally there is a uniform
front to the building which faces the street, but the location of the rear wall in each resi-
dence may vary depending on what extra rooms are appurtenant to the main structure –
a toilet and bathroom block, for example. The “amenities” block could be separated by
a structural wall, or it might only be behind a non-structural party wall. Each residence
can vary in width and height, and especially the height of each parapet. That is L2, H1
and P1 in Figure 2.9 can vary for each individual tenant. There can be anywhere from
three to ten or more tenancies joined together in the one overall structure. Usually the
overall width of the building is much longer than its length from the front to the back
(L5 and L6 respectively in Figure 2.9). In the original construction the individual ten-
ants were separated by non-pierced structural walls, see Figure 2.9. Similar to the side
walls, these internal structural walls may slope from the top of the parapet at the front
of the building to the back, or the top of the wall may be horizontal (refer to P1 and P2
in Figure 2.9). There are often individual parapets on the front of the building, which
usually are unique for each residence, although the original parapet may not reflect the
current tenants. For example, there might be a single parapet over two tenants which are
separated by non-structural party walls, which were not part of the original construction.
Inside each individual occupancy there may be various non-structural walls, for example,
in a clothing retail shop, there may be a fitting room separated from the rest of the shop
floor. Similar to typology A structures, the external side walls are usually two or three
leaves thick and not pierced, the frontages may be up to 90 % open and there are sus-
pended ceilings underneath roofs of either tiles or cast iron cladding. Typically typology
B structures are occupied by commercial tenants – retailers, hairdressers, cafes etc. Row
structures like typology B and D structures were often built on sloping ground. The
ground surface slopes away from the road, so that the rear of the building is lower than
the front. Sometimes there is a basement level below ground floor at the back, even if
the building is actually single storey at the front. See Figure A.2(d) in Appendix A. Fig-
ure 2.9 and Table 2.4 show typical dimensions of a Typology B structure. Photographic
examples of Typology B structures are shown in Figure 2.10.
Figure 2.9: Overall dimensions of a Typology B Building
Table 2.4: Typical Dimensions of a Typology B Building

(m) (m) (m)
L4 Length of “amenities block” 2 4 5
bottom of parapet
front street level
(a) Ashburton (b) Paeroa
(c) Thames (d) Timaru
(e) Royal Oak (Auckland) (f) Epsom (Auckland)
Figure 2.10: Typology B buildings – single storey row
Further photographic examples of typology B buildings are shown in Appendix A.
Typology C – Two Storey Isolated Buildings
Typology C buildings are two storey isolated structures, and buildings of this height are
the most prevalent form of URM building throughout the country — both isolated (ty-
pology C) and row (typology D). Examples of typology C buildings include small town
cinemas, a small professional office in a rural town, small town pubs and small hotels.
They are prevalent throughout New Zealand, in both smaller rural centres and in city
“village” areas (mainly older early settled parts of town). These structures are usually
square, but on occasions they may have a trapezoidal or triangular ground footprint,
particularly when the building is situated at an intersection of two streets on an acute
angle. Such corner-type structures are classed Typology C3 structures, see Figure 2.14.
Typology C1 structures are those with a dividing wall in the centre of the ground floor
and C2 are those with an open-plan ground floor. Typical dimensions of each classifi-
cation of typology C structures are shown in Figures 2.15 – 2.17 and Tables 2.5 – 2.7.
The most important feature of typology C buildings is the floor and ceiling diaphragms
and their connections to the walls. It is common for the bottom storey walls to be three
leaves thick, and the upper walls to be two leaves thick. Timber joists supporting the
floor are often only simply supported on the resulting ledge, with no positive anchorage.
In other instances the joists may be embedded in the walls, but the depth of embedment
is usually very small, providing little anchorage, see Figure 2.11. Often buildings exhibit
spikes on their exterior (Figure 2.12(a)), which appear to be the fixing for internal timber
diaphragms, and may have been subsequent additions and not part of the original con-
struction. In other instances these spikes appear to have also been subsequent additions,
but for the purpose of tying multi-leaf walls together, and possibly cavity walls (Figure
2.12(b)). Typology C buildings occasionally have concrete floor diaphragms instead of
timber, that are cast on top of the ground storey wall. The second storey masonry wall
is built directly on top of the concrete. Sometimes the floor diaphragms are timber, but
are embedded into a concrete ring beam cast on top of the ground storey wall (Figure
2.13). Other features of these structures are that the side walls are generally not pierced,
there are window openings on the top storey at the front and back, and the front of the
building on the ground floor may have large openings. The back wall is usually similar to
the front, but the openings may be smaller, and sometimes in the upper storey there are
no openings or just small frame windows. There are usually parapets on the top of the
front wall, which are sometimes very ornately decorated. As with one-storey structures,
the side walls can have a horizontal top or can slope from the top of the parapet at the
front to the top of the rear wall, and roofs are generally a truss structure, supporting cast
iron cladding. Typology C3 structures are those which usually occur at a corner between
two roads, and the structure results in an overall wedge shape, as shown in Figure 2.14(a).
Sometimes C3 buildings occur between obtuse intersections, or at right-angle intersections
also, see Figures 2.14(b) and 2.14(c). Typology C3 (corner) structures tend to be on flat
ground, and the parapets also tend to be of consistent height around the building. There
are usually no piercings on the back face of typology C3 buildings. Figures 2.15 – 2.17
and Tables 2.5 – 2.7 show typical dimensions of Typology C structures. Photographic
examples of typology C structures are shown in Figure 2.18, and Figure 2.19 shows ex-
amples of two storey corner structures.
(a) Joists embedded in wall (b) Joists seated on wall
Figure 2.11: Timber joist details
(a) (b)
Figure 2.12: Spikes on exterior wall for fixing internal floor diaphragm
(a) (b)
Figure 2.13: Concrete ring beam between floors
(a) Acute intersection
(b) Obtuse intersection
(c) Right angle intersection
Figure 2.14: Location of Typology C3 building at junction of two roads
Figure 2.15: Overall dimensions of a Typology C1 Building
Table 2.5: Typical Dimensions of a Typology C1 Building

(m) (m) (m)
P1 Height of parapet at front 1 2 2.5
bottom of second storey
H2 Height of front opening 3 5 5
H3 Height of second storey 3 5 6
G1 Depth of rear wall below 0 0 2
front street level
W1 Height of windows 0 0 2
above top of ground storey
W2 Height of windows 1 2 2.5
W3 Width of windows 0.8 1 1.5

(m) (m) (m)
P1 Height of parapet at front 1 2 2.5
front street level
above top of lower storey

(m) (m) (m)
L2 Width of front of building 2 5 8
L3 Width of rear of building 8 10 16
P2 Height of parapet at rear 1 1.2 2
(a) Kingsland (Auckland) (b) Bulls
(c) Lyttelton (Christchurch) (d) Taihape
(e) Onehunga (Auckland) (f) Northland (Wellington)
Figure 2.18: Typology C buildings – two storey isolated
(a) Newmarket (Auckland) (b) Bulls
(c) Kingsland (Auckland)
Figure 2.19: Typology C3 buildings – two storey corner
Further photographic examples of typology C buildings are shown in Appendix A.
Typology D – Two Storey Row Buildings
Two storey row structures are the most prevalent of the seven typologies identified. Like
typology B structures, they are common in commercial districts and on the main streets of
small towns, but they are especially common in larger centres such as Auckland, Welling-
ton and Christchurch. Many characteristics described for typology B structures are also
applicable to typology D buildings. Some typology D structures are mostly uniform, with
consistent floor, roof and parapet levels resulting in a structure which is essentially ho-
mogeneous but divided into sections. Other structures are very much heterogeneous and
have the appearance of individual and distinctive buildings joined together. There can
be many variations between adjoining tenancies, both in appearance and dimensions. For
example, the height of the first floor may be different between adjoining tenancies (which
may significantly contribute to the effects of pounding) and parapet levels and heights can
also vary. Floor diaphragms may be concrete or timber (timber is more common) and the
diaphragm seating issues identified for typology C structures are also true for typology D
structures. Openings are similar to those in Typology C structures, and the number of
windows and their spacing can vary between the adjoining parts of the overall structure.
In these structures, each tenancy is long compared to its width, and (in Figure 2.21) L1
is usually twice as long as L2. Photographic examples are shown in Figure 2.22.
Further photographic examples of typology D buildings are shown in Appendix A.
Whilst structures which do not fit the criteria of typology A – F classifications are classed
as typology G buildings, there can be further non-conformity to the above framework,
particularly where buildings of two typologies are joined. Figure 2.20 shows a one storey
building adjoined to a two storey building.
Figure 2.20: One storey building adjoining a two storey building
Figure 2.21: Overall dimensions of a Typology D Building
Table 2.8: Typical Dimensions of a Typology D Building

(m) (m) (m)
L1 Length of front ground floor 8 12 16
building
L4 Length of ground floor 4 4 6
rear amenities block
P2 Height of parapet at rear 0 0.5 1.5
of each residence
of each residence
front street level
(a) Central Christchurch (b) Timaru
(c) Devonport (Auckland) (d) Grey Lynn (Auckland)
(e) Ponsonby (Auckland) (f) Central Auckland
Figure 2.22: Typology D building – two storey row
Typology E – Three + Storey Isolated Buildings
Typology E structures are isolated buildings of three storeys or more. It was felt there
were too few four, five and six or more storey buildings in the country to warrant a sepa-
rate class of building for each. These structures tend to be located in the larger centres,
particularly Auckland, Wellington and Dunedin. They are situated mainly around older
commercial areas and also close to the ports. The most important feature of these struc-
tures is that their front and back faces have multiple and consistently spaced window
openings. This means that these faces tend to behave as masonry frames and the piers
between windows can rock when subjected to lateral displacements. Large openings on
the upper floors are uncommon. The side walls tend not to have any openings. The
ground storey walls may be up to nine leaves thick, with each subsequent wall above usu-
ally decreasing in thickness. As described for typology C structures, the floor and ceiling
diaphragms often rest on the ledge resulting from the decrease in wall thickness. Again,
the floor or ceiling diaphragms may be of timber or concrete construction, with the same
seating issues as identified above. For Typology E and F structures, the upper floors
are often open inside, possibly with intermediate columns supporting the roof structure
above. See Figure 2.23 for example. Photographic examples of typology E structures are
shown in Figures 2.25 and 2.26.
Further photographic examples of typology E buildings are shown in Appendix A.
Figure 2.23: Open plan interior of upper storey in typology E structure
Figure 2.24: Overall dimensions of a Typology E Building
Table 2.9: Typical Dimensions of a Typology E Building

(m) (m) (m)
building
L2 Width of building 12 16 20
of each residence
of each residence
B1 Width of rear opening 3 4 5
of each residence
front street level
W2 Height of lower floor 1 2 2.5
windows
W4 Height of upper storey 1 2 2.5
windows
W5 Height of top storey 0.8 1 1.5
windows
W6 Height of rear windows 0 0 2
W7 Height of rear windows 1 2 2.5
rear windows
W9 Width of rear windows 0.8 1 1.5
(a) Central Christchurch (b) Timaru
(c) Timaru (d) Dunedin
Figure 2.25: Typology E buildings – three + storey isolated
(a) Newtown (Wellington) (b) Central Wellington
(c) Central Wellington
Figure 2.26: Typology E buildings – three + storey isolated
Typology F – Three + Storey Row Buildings
These structures are three and higher storey buildings which, similar to Typology B and
D, are row structures. These structures were often built as warehouses, granaries or stor-
age facilities and are common both in industrial and commercial districts, especially close
to a port (or historic port), for example parts of Timaru and in Auckland’s waterfront.
There can be a blurred distinction between Typology E and F, but generally typology
F structures are very long and may form a whole block of a street. Most characteristics
of these buildings are similar to those of typology E buildings, in particular, typology
F structures can be considered as “frame” type structures. The main difference is that
neighbouring (joined) residences may have different floor and overall heights resulting in
pounding or lateral displacement incompatibility issues. This type of building might be
owned by a single owner, but there are usually many different tenancies, both between
floors and within floors. As noted in section 2.3.1 this needs to be taken into considera-
tion when designing a retrofit for the overall structure. Typical dimensions of typology F
structures are shown in Figure 2.27 and Table 2.10 and photographic examples are shown
in Figure 2.28. Figure 2.27 shows only the back of a single equivalent component typology
E structure.
Further photographic examples of typology F buildings are shown in Appendix A.
Figure 2.27: Overall dimensions of a Typology F Building
Table 2.10: Typical Dimensions of a Typology F Building

(m) (m) (m)
building
L2 Width of building 12 16 20
of each residence
of each residence
B1 Width of rear opening 3 4 5
of each residence
front street level
W2 Height of lower floor 1 2 2.5
windows
windows
W5 Height of top storey 0.8 1 1.5
windows
W6 Height of rear windows 0 0 2
W7 Height of rear windows 1 2 2.5
rear windows
W9 Width of rear windows 0.8 1 1.5
(a) Central Christchurch (b) Central Christchurch
(c) Central Dunedin (d) Central Wellington
(e) Central Auckland (f) Central Auckland
Figure 2.28: Typology F buildings – three + storey row
Typology G – Religious, Institutional and Industrial Buildings
Typology G structures are characterised by structures for which a generic form is unreal-
istic, and those URM buildings which do not have a simple or uniform ground footprint.
These are distributed all around the country. The list of typologies is comprehensive and
any monumental or unique structure which does not fit into Typologies A – F will be a
Typology G structure. It is not the purpose of these classifications to define buildings by
their use, but typology G buildings tend to be churches, warehouses and factories, or civil
buildings such as a town hall, ferry building or post office. Because of this, Typology G has
been further divided into G1 for religious buildings (Figure 2.29(e)), G2 for warehouses
and factories with large tall sides and large open spaces inside (Figure 2.29(d)), and G3
for institutional buildings, such as ferry terminals and train stations (Figure A.17(c)).
For typology G structures unique analyses are necessary on a case-by-case basis. Photo-
graphic examples are shown in Figure 2.29.
Further photographic examples of typology G buildings are shown in Appendix A.
(a) Lincoln (Christchurch) (b) Dunedin
(c) Oamaru (d) Seatoun (Wellington)
(e) Palmerston North
Figure 2.29: Typology G building examples – Religious, institutional and industrial
2.4. International Comparisons with New Zealand Typologies
2.4 International Comparisons with New Zealand

Typologies
URM construction in New Zealand shares many similarities with practices elsewhere in
the world. This is true particularly for British Colonies settled contemporaneously with
New Zealand, and also for the United States. European and Asian construction practices
extend back much further, and other types of masonry construction, such as stone and
rubble, are much more prevalent there than in New Zealand.
Australia and New Zealand are considered to have had very similar building practices up
until the 1931 Hawke’s Bay earthquake (Russell et al., 2006). URM construction typical
of this era in Australia consisted of clay brick walls, supporting timber floor systems and
timber roof trusses with metal or clay brick roofing. The brick walls were typically cavity
walls, with no rubble fill, and were supposed to be connected with metal wall ties. The
bricks themselves were usually pressed bricks of a regular size. The timber floor/roof
systems typically were supported with bearing supports on the inner leaf of the cavity
wall without any significant positive connection to the walls.
Comparison of New Zealand URM structures with those in California is important be-
cause of the similarities in terms of construction types, histories and seismic context. In
both places, most URM buildings are pre-1935. Outside the central business districts of
the main cities, most are one or two storeys high and typically have timber floors and roof
framing and a two or three leaf wall construction (Blaikie and Spurr, 1992). It has been
found that the use of unreinforced masonry walls with timber diaphragms for floors or
roofs was commonplace in pre-1945 construction in America (Bruneau, 1994; Peralta et al.,
2004). Timber diaphragms typically consist of three components; sheathing, framing and
chords. It was found that in California buildings constructed before 1945, sheathing was
made of straight members, nailed onto the framing, rather than plywood, as would be
expected in current practice. Occasionally iron wall anchors were used to connect the
diaphragm to the wall, but only on about every 4th joist. Usually joist-to-wall anchors
were designed without taking seismic forces into consideration.
Figure 2.30 shows examples of typologies in Australia and South Africa, similar to those
identified for New Zealand.
(a) Typology C structure Cape Town (South (b) Typology D structure Cape Town (South
Africa), near historic British naval base Africa), near historic British naval base
(c) Typology D structure in Newcastle (Australia), (d) Typology B structure in Newcastle (Australia),
(British penal colony) (British penal colony)
Figure 2.30: International Comparisons with New Zealand Typologies
2.5 Supplementary Characteristics of New Zealand

URM
New Zealand’s masonry construction heritage is comparatively young, spanning from 1833
until the present time – a period of less than 200 years (Ingham, 2008). Consequently,
a study of New Zealands masonry building stock has a narrow scope in comparison with
2.5. Supplementary Characteristics of New Zealand URM
international norms (see for instance Binda et al., 2005; Lourenço et al., 2006; Magenes,
2006). This comparatively narrow time period has the advantage of facilitating the doc-
umentation and reporting of New Zealand masonry construction practice with a greater
degree of accuracy than is often possible in countries with an older and more diverse
history of masonry construction (Binda, 2006b).
2.5.1 Bond Pattern
Bond pattern determines how bricks in a masonry wall are connected. This has a signifi-
cant effect on the strength of the wall and how the wall acts together as a unit. Stretcher
units are bricks laid in the plane of the wall, and header units are laid across the wall,
joining leaves together. In cross-section, a wall that is three bricks thick is a three leaf
wall. Each leaf needs to be adequately connected with header bricks at appropriate in-
tervals. A course is row of bricks along the length of the wall. Thus, a wall may be three
leaves thick, and 30 courses high.
New Zealand URM buildings are mainly constructed with either Common bond or English
bond. English bond has alternating header and stretcher courses, as shown in Figure
2.32. Common bond (sometimes referred to as American bond or English Garden Wall
bond) has layers of stretchers with headers every 3 – 6 courses, and is shown in Figure
2.33. Running bond (only stretcher courses) often suggests a cavity wall and is shown
in Figure 2.34. Flemish bond (alternating headers and stretchers in every course) is
the least common and is often found in New Zealand between openings on an upper
storey, for example on piers between windows. Flemish bond is shown in Figure 2.35.
Common bond is the most frequently occurring bond pattern in New Zealand, but English
bond also appears in New Zealand buildings often, especially in the bottom (ground)
storey. Common bond can vary with headers at different levels in different buildings, and
sometimes there is variation within a building. For example, at the bottom of the wall
the headers may be every second course, but at the top of the third storey wall they may
be every 3rd course. See Figures 2.36(a) and 2.36(b): both photos are from the same
building wall. Header courses may be irregular and made to fit in at ends of walls and
around drainpipes with half widths and other cut bricks, see Figure 2.31.
Figure 2.31: Uneven bricks made to fit
(a) (b)
Figure 2.32: English bond pattern
(a) (b)
Figure 2.33: Common (or American or English Garden Wall) bond pattern
(a) (b)
Figure 2.34: Running bond pattern
(a) (b)
Figure 2.35: Flemish bond pattern
(a) Ground storey, header bricks every second course (b) Second storey, header bricks every fourth course
Figure 2.36: Bond pattern varying over height of a single wall
2.5.2 Wall Height and Thickness
Generally most bearing walls in New Zealand are at least three leaves thick (330 mm),
and parapets are two leaves thick (220 mm) with the walls having increasing thickness
down the height of the building (often reducing by one leaf every 2 storeys) (Oliver, 2007).
Typical storey heights generally range from approximately 3.2 m to 6.0 m (see details on
individual typologies above). “The Architect’s and Builder’s Pocket-Book” (Kidder, 1905)
from California gives guidance on wall thickness for mercantile buildings in large cities.
As noted in Section 2.4, New Zealand URM construction is thought to have been similar
to other British colonies at the time, and also similar to URM construction practice in
The United States, especially California. In 1905 Brick buildings in San Francisco were
permitted to be up to six storeys high, and the top storey was prescribed to have walls
13 in. thick, the next three storeys down were to be 17 in. thick, and the next two were
to be 21 in. thick. If a structure was only three storeys high, the bottom two storeys
would have 17 in. thick walls and the top storey walls would be 13 in. thick (Kidder,
1905). This prescription corresponds approximately to five or six leaf thick walls at the
bottom, and three leaf walls at the top of a typology E or F structure, which has also
been observed in New Zealand.
Taller buildings (typically those of typology C, D, E & F) have tall slender window open-
ings on the front side usually facing the street. Above the windows, lintels are generally
formed with perpendicular bricks extending from the face of the wall. More sophisticated
façades may have other decorative features around the windows and lintels. Slender wall
sections between windows can act as piers which rock during lateral loading. This is seen
to be a more favourable failure mode than shear failure or sliding, as residual displace-
ments of the walls are small. However this can lead to significant non-structual component
damage from the large displacements.
2.5.3 Age
The age of existing New Zealand URM buildings is dealt with more fully in Chapter 3.
URM was prohibited as a structural material in most parts of the country in 1965 with the
introduction of NZS 1900 (New Zealand Standards Institute, 1965). Most URM buildings
were built between about 1880 and 1930. As settlement grew in the late 19th Century,
many new buildings were constructed, and the Hawke’s Bay earthquake in 1931 (Dowrick,
1998) was a turning point after which few URM buildings were built. Conveniently, many
URM buildings display their age on the front façade, as shown in Figure 2.37.
(a) (b)
(c) (d)
Figure 2.37: Date of construction on façade
2.5.4 Materials
URM buildings were constructed using bricks generally of a uniform size, which is differ-
ent from the dimensions of both modern concrete blocks and modern clay bricks. Bricks
manufactured in the early 20th Century were solid (uncored) and sometimes had frogging
on one face (an imprint often with the manufacturers stamp). Typical dimensions of
early New Zealand bricks are shown in Figure 2.38, but as can be seen in Figure 2.40,
not all bricks were laid to a uniform quality. Bricks can be of varying qualities now due
to deterioration, depending especially on proximity to sea air. Even some inner-city walls
display bricks which are substantially eroded, as shown in Figure 2.39. The condition of
bricks does not seem to correspond with the colour of the bricks.
Many buildings which look like they could be constructed of URM based on character-
istics such as geometry, configuration, occupancies, age and location, but appear to be
made of concrete because of the nature of their external finish are often found to be brick
on closer inspection, because of a concrete plaster over the bricks (see Figure 2.19(c)).
This plaster can be up to 20 mm thick, although it often varies in thickness because the
bricks were not a smooth surface. In some places the plaster is eroded.
Figure 2.38: Typical brick dimensions
Figure 2.39: Eroded bricks in centre of wall
Figure 2.40: Broken Bricks
In binding masonry walls to the foundations, different mortar was used from that in the
walls themselves. A specification from a URM residential building in central Auckland
constructed in 1919 states that “ground mortar is to be two parts clean red scoria ash,
one part sand, and one part hydraulic lime, finely ground together in a mortar mill”
(O’Connor, 1919). For mortars used between bricks, it is thought the constituent ma-
terials that were used varied considerably. Some mortars were lime based, and others
were cement based. Mortars varied from 1:3 lime:sand to 1:2:9 cement:lime:sand, to 1:2
cement:sand. Near the sea shore, poorly graded beach sand was usually used (sometimes
with entire shells used in the mix, see Figure 2.41), but inland finer river sand was gener-
ally used. Lime was used as a binder in the mortar, and also to improve workability when
using cement as the main binder. Cement based mortar tends to be in better condition
now than lime based mortars, due to the leaching of lime particles from the mixture. This
often leaves a powdery sand with no bond strength and generally occurs where water in-
filtration has been a problem.
Figure 2.41: Sea shells in mortar
The performance of mortar in existing buildings is extremely variable. Some masonry

exhibits considerable strength in terms of bond and cohesion between bricks and mor-
tar, while others exhibit very little tensile strength at the brick-mortar interface. Mortar
condition also varies according to its location in the building. Internal walls not exposed
to weathering are frequently in better condition than walls exposed to external elements.
Mortar in buildings close to the sea is often eroded due to salt carried from sea breezes,
and also from salt in the sand from the original mortar construction.
Workmanship and mortar quality varied widely in early NZ URM construction. Public
works buildings generally used good quality mortar, often cement based, while private
buildings tended to use cheaper and more easily available materials for mortar (Stil,
2007). Today, the quality of workmanship in URM buildings also shows large variabil-
ity. When comparing non-structural visible façades with non-visible internal structural
2.6. Conclusions
walls, it is often evident that the structural components were built with lower quality
control than parts which were seen. Internal walls are sometimes painted or plastered,
but cracks and movement in the walls shows the often low quality of workmanship. Some
URM buildings were built with great care taken, and are still in good condition today. A
residential building specification in 1919 shows that there were some quality controls in
place, “Bricks throughout the structure are to be sound, square, well burnt, and to have
of an even shape and colour, and be obtained form an approved yard,” (O’Connor, 1919).
These specifications were left to the bricklayer to meet, but it is not clear how strongly
quality control was met or enforced.
2.6 Conclusions
Characterisation of the URM building stock into typologies is one of the key parameters
for defining and understanding the nature of URM buildings in New Zealand. The overall
configuration of a building influences its performance in an earthquake. Seven typologies
have been identified to categorise configurations in the New Zealand URM building stock.
Separations between typologies are made on the basis of building height and the geometry
of the building’s ground footprint. When URM was commonly used in New Zealand in
the late 19th Century and early 20th Century, construction practices were similar to other
countries of British origin, such as Australia and the United States. New Zealand also
shares common architectural characteristics and and building typologies with some Euro-
pean countries, particularly Italy. Assessment and analysis of the structural performance
of buildings within the seven identified typologies will enable targeted and cost-effective
retrofit solutions to be implemented for the retention of New Zealands heritage URM
buildings.
2.7 References
Abrams, D. P. (2000). “Seismic Response Patterns for URM Buildings”. TMS Journal,
18(1):71–78.
Argan, G. C. (1996). On the Typology of Architecture. In Nesbitt, K., editor, Theorizing a

New Agenda for Architecture: An Anthology of Architectural Theory 1965 – 1995,
pages 240–246. New York: Princeton Architectural Press.
ATC (1996). ATC-21: Rapid Visual Screening of Buildings for Potential Seismic Hazards
Training Manual. Applied Technology Council, Redwood City, California.
Bakhshi, A., Bozorgnia, Y., Ghannad, M. A., Khosravifar, A., Mousavi Eshkiki, S. E.,
Rahimzadeh Rofooei, F., and Taheri Behbahani, A. (2005). “Seismic Vulnerability
of Traditional Houses in Iran”. 1st International Conference on Seismic Adobe
Structures, Lima, Peru.
Binda, L. (2005). “Methodologies for the vulnerability analysis of historic centres in

Italy”. Advances in Architecture Series, 20:279–290.
Binda, L. (2006a). “Investigation on the masonry quality and mechanical behaviour

(Clinic 1)”. The First International Conference on Restoration of Heritage Ma-
sonry Structures, Cairo, Egypt, April 24 – 27, 2006.
Binda, L. (2006b). “The Difficult Choice of Materials Used for the Repair of Brick and
Stone Masonry Walls”. The First International Conference on Restoration of Her-
itage Masonry Structures: Clinic 1, Cairo, Egypt, April 24 - 27, 2006.
Binda, L., Cardani, G., and Saisi, A. (2005). “Application of a multidisciplinary investiga-
tion to study the vulnerability of Castelluccio (Umbria)”. Advances in Architecture
Series, 20:311–322.
Blaikie, E. L. and Spurr, D. D. (1992). “Earthquake Vulnerability of Existing Unreinforced
2.7. References
Masonry Buildings”. EQC Report, December 1992, Works Consultancy Services,

Wellington, New Zealand.
Bruneau, M. (1994). “Seismic evaluation of unreinforced masonry buildings - a state-of-

the-art report”. Canadian Journal of Civil Engineering, 21(3):512–539.
D’Ayala, D. and Speranza, E. (2003). “Definition of Collapse Mechanisms and Seismic

Vulnerability of Historic Masonry Buildings”. Earthquake Spectra, 19:479–509.
DBH (2004). Building Act 2004. Department of Building and Housing – Te Tari Kaupapa
Whare, Ministry of Economic Development, New Zealand Government, Date of
assent: 24 August 2004.
DBH (2005). Earthquake-Prone Building Provisions of the Building Act 2004: Policy
Guidance for Territorial Authorities. Department of Building and Housing – Te
Tari Kaupapa Whare, Wellington, New Zealand.
DBH (2007). Earthquake-prone buildings. Department of Building and Housing – Te Tari

Kaupapa Whare, http://www.dbh.govt.nz/bomd-earthquake-prone-buildings.
Dowrick, D. (1998). “Damage and Intensities in the Magnitude 7.8 1931 Hawke’s Bay,
New Zealand, Earthquake”. Bulletin of the New Zealand Society for Earthquake
Engineering, 30(2):133–158.
Erbay, O. O. (2003). A Methodology to Assess Seismic Risk for Populations of Unrein-

forced Masonry Buildings. PhD thesis, University of Illinois at Urbana-Champaign.
Erbay, O. O. and Abrams, D. P. (2004). “A Methodology to Assess Seismic Risk for Popu-
lations of Unreinforced Masonry Buildings”. 13th World Conference on Earthquake
Engineering, Vancouver, British Columbia, Canada, August 1 – 6, 2004.
Erbay, O. O. and Abrams, D. P. (2007). “Modeling seismic risk for populations of un-
reinforced masonry buildings”. Tenth North American Masonry Conference, St.
Louis, Missouri, USA, June 3 – 6, 2006.
Ghannad, M. A., Bakhshi, A., Eshkiki, S. E. M., Khosravifar, A., Bozorgnia, Y., and
Behbahnani, A. A. T. (2006). “A study on seismic vulnerability of rural houses in
Iran”. 1st ECEES, Geneva, Switzerland, September 3 – 8, 2006.
Ingham, J. M. (2008). “The Influence of Earthquakes on New Zealand Masonry Construc-

tion Practice”. 14th International Brick and Block Masonry Conference (14IB-
MAC), Sydney, Australia, February 17 – 20, 2008.
Karantoni, F. V. and Bouckovalas, G. (1997). “Description and analysis of building

damage due to Pyrgos, Greece earthquake”. Soil Dynamics and Earthquake Engi-
neering, 16:141–150.
Karimi, K. and Bakhshi, A. (2006). “Development of fragility curves for unreinforced ma-
sonry buildings before and after upgrading using analytical method”. 1st ECEES,
Geneva, Switzerland, September 3 – 8, 2006.
Kidder, F. E. (1905). The Architect’s and Builder’s Pocket-Book: A Handbook for Archi-
tects, Structural Engineers, Builders, and Draughtsmen. New York: John Wiley &
Sons, 14th edition.
Lavin, S. (1992). Quatremère de Quincy and the Invention of a Modern Language of

Architecture. The MIT Press, Cambridge, Massachusetts.
Lourenço, P. B., Luso, E., and Almeida, M. G. (2006). “Defects and moisture problems
in buildings from historical city centres: a case study in Portugal”. Building and
Environment, 41(2):223–234.
Magenes, G. (2006). “Masonry Building Design in Seismic Areas: Recent Experiences

and Prospects From a European Standpoint”. 1st ECEES, Geneva, Switzerland,
September 3 – 8, 2006.
Mousavi Eshkiki, S. E., Khosravifar, A., Ghannad, M. A., Bakhshi, A., Taheri Behbahani,
A., and Bozorgnia, Y. (2006). “Structural Typology of Traditional Houses in Iran
Based on their Seismic Behaviour”. 8th US National Conference on Earthquake
2.7. References
Engineering, San Francisco, California, USA.
Neidhart, H. and Sester, M. (2004). “Identifying building types and building clusters
using 3D-laser scanning and GIS data”. Geo-Imagery Bridging Continents, XXth
ISPRS Congress, Istanbul, Turkey, July 12 – 23, 2004.
New Zealand Standards Institute (1965). NZS 1900:1965, Model Building By-Law. New
Zealand Standrds Institute, Wellington, New Zealand.
Nollet, M. J., Chaallal, O., and Lefebvre, K. (2005). “Seismic vulnerability study of
historical buildings in Old Montreal: Overview and perspectives”. Advances in
Architecture Series, 20:227–236.
O’Connor, J. (1919). Courtville Apartments Architectural Specifications. Auckland.
Oliver, S. (2007). Typical storey heights for URM walls. Personal communication.
Peralta, D. F., Bracci, J. M., and Hueste, M. B. D. (2004). “Seismic Behavior of Wood
Diaphragms in Pre-1950s Unreinforced Masonry Buildings”. Journal of Structural
Engineering, 130(12):2040–2050.
Pontius, S. E. (2007). “genotype | phenotype: investigations in typology”. Master’s thesis,

The School of Architecture and Interior Design, The University of Cincinnati.
Quatremère de Quincy, A.-C. (1832). Dictionnaire historique d’Architecture. Paris. Pub-

lished by Charles-Joseph Panckoucke.
Robinson, L. and Bowman, I. (2000). Guidelines for Earthquake Strengthening. New

Zealand Historic Places Trust - Pouhere Taonga, Wellington, New Zealand, 13
pages.
Rossi, A. (1982). The Architecture of the City. Cambridge: The MIT Press.
Russell, A., Ingham, J., and Griffith, M. (2006). “Comparing New Zealand’s Unreinforced
Masonry Details with those Encountered in Other Seismically Active Countries”.

7th International Masonry Conference (7IMC), London, 30 Oct – 1 Nov 2006.
Smith, P. (2007). Personal Correspondence. Spencer Holmes, Wellington, New Zealand.
Standards New Zealand (2002). AS/NZS 1170.0:2002, Structural Design Actions Part 0:
General Principles. Standards New Zealand, Wellington, New Zealand.
Standards New Zealand (2004). NZS 1170.5:2004, Structural Design Actions Part
5:Earthquake actions – New Zealand. Standards New Zealand, Wellington, New
Zealand.
Stil, J. (2007). Nikau Contractors. Personal Communication.
Tomaževič, M. and Lutman, M. (2007). “Heritage Masonry Buildings in Urban Settle-

ments and the Requirements of Eurocodes: Experience of Slovenia”. International
Journal of Architectural Heritage, 1(1):108–130.
Valluzzi, M. R., Bernardini, A., and Modena, C. (2005a). “Seismic vulnerability assess-
ment and structural improvement proposals for the building typologies of the his-
toric centers of Vittorio Veneto (Italy)”. Advances in Architecture Series, 20:323–
332.
Valluzzi, M. R., Cardani, G., Saisi, A., Binda, L., and Modena, C. (2005b). “Study of
the seismic vulnerability of complex masonry buildings”. Advances in Architecture
Series, 20:301–310.
Vicente, R., Varum, H., and Mendes da Silva, J. A. R. (2006). “Vulnerability assessment of
traditional buildings in Coimbra, Portugal, supported by a GIS tool”. 1st ECEES,
Geneva, Switzerland, September 3 – 8, 2006.

Architectural Characterisation

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Architectural Characterisation

Uploaded by

Copyright:

Available Formats

CHAPTER 2

-5- Alistair P. Russell

Alistair P. Russell -6-

enables an analysis of structures through modelling without needing building-specific pa-

-7- Alistair P. Russell

2.2 Background of Building Typologies

Alistair P. Russell -8-

-9- Alistair P. Russell

2.2.1 European URM Typology Studies

(number of storeys, exposure, type of façade, material and structural elements). On a

• Typology A: isolated houses and/or dwellings;

• Typology B: row houses;

• Typology D: bell towers;

• Typology F: churches and cathedrals; and within this typology,

– Typology F1: churches, plan based on Latin cross scheme;

– Typology F2: churches, with a central plan.

Karantoni and Bouckovalas (1997) performed a vulnerability study considering almost

Figure 2.2: Typical masonry buildings in Ljubljana, Slovenia (Reproduced from

2.2.2 North American URM Typology Studies

The United States

2.2.3 Iranian URM Typology Studies

2.3 New Zealand URM Building Typologies

Figure 2.4: Locations of buildings surveyed throughout New Zealand

Table 2.1: URM Typologies

Typology Description Prevalence Importance Description

2.3.1 Parameters for Differentiating Typologies

2.3.2 Details of New Zealand URM Typologies

Typology A – One Storey Isolated Buildings

Figure 2.6: Overall dimensions of a Typology A1 Building

Table 2.2: Typical Dimensions of a Typology A1 Building

Dimension Lower bound “Mean” Upper bound

Figure 2.7: Overall dimensions of a Typology A2 Building

Table 2.3: Typical Dimensions of a Typology A2 Building

Dimension Lower bound “Mean” Upper bound

(a) Parnell (Auckland) (b) Ashburton

(c) Temuka (Canterbury) (d) Seatoun (Wellington)

(e) Epsom (Auckland) (f) Thames

Figure 2.8: Typology A buildings – single storey isolated

Further photographic examples of typology A buildings are shown in Appendix A.

Typology B – One Storey Row Buildings

Figure 2.9: Overall dimensions of a Typology B Building

Table 2.4: Typical Dimensions of a Typology B Building

Dimension Lower bound “Mean” Upper bound

(a) Ashburton (b) Paeroa

(c) Thames (d) Timaru

(e) Royal Oak (Auckland) (f) Epsom (Auckland)

Figure 2.10: Typology B buildings – single storey row

Further photographic examples of typology B buildings are shown in Appendix A.

Typology C – Two Storey Isolated Buildings

(a) Joists embedded in wall (b) Joists seated on wall

Figure 2.11: Timber joist details

Figure 2.13: Concrete ring beam between floors

(a) Acute intersection

(b) Obtuse intersection

(c) Right angle intersection

Figure 2.14: Location of Typology C3 building at junction of two roads

Figure 2.15: Overall dimensions of a Typology C1 Building

Table 2.5: Typical Dimensions of a Typology C1 Building

Dimension Lower bound “Mean” Upper bound

Figure 2.16: Overall dimensions of a Typology C2 Building

Table 2.6: Typical Dimensions of a Typology C2 Building

Dimension Lower bound “Mean” Upper bound

Figure 2.17: Overall dimensions of a Typology C3 Building