See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/320411000

Seismic retrofit of masonry walls using timber strong-backs

Article · January 2017

CITATIONS READS

4 1,426

4 authors:

Dmytro Dizhur Marta Giaretton

University of Auckland University of Auckland
152 PUBLICATIONS   877 CITATIONS    47 PUBLICATIONS   115 CITATIONS   

SEE PROFILE SEE PROFILE

Ivan Giongo Jason Ingham

Università degli Studi di Trento University of Auckland
42 PUBLICATIONS   143 CITATIONS    464 PUBLICATIONS   3,281 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Earthquake response of wineries View project

Understanding How Building Owners Make Resilience Decisions View project

All content following this page was uploaded by Marta Giaretton on 30 November 2017.

The user has requested enhancement of the downloaded file.

 SEPTEMBER 2017
JOURNAL
 SESOC Journal

Structural Engineering Society New Zealand (Inc.)

JOURNAL September 2017

A Technical Group of the Institute of Professional Engineers, New Zealand

Vol. 30, No. 2 Contents

Page
SESOC Information........................................................................................................................................................ 3
SESOC President’s Report............................................................................................................................................. 5
Note from the Editor....................................................................................................................................................... 7
Obituary.......................................................................................................................................................................... 7
SESOC Committee ....................................................................................................................................................... 9
Letters to the Editor........................................................................................................................................................ 10
Seismic tests on moment resisting stainless steel connections (HERA)......................................................................... 12
SESOC Conference 2017.............................................................................................................................................. 14
The Seismic Restraint of Building Services or the Lack Thereof
An article by John Scarry, Structural Engineer, Auckland, and SESOC Education and Training Coordinator....... 16
Parts DESIGN to 1170.5 2004 Section 8: Face Loaded Precast Panels in Industrial Buildings
Article prepared by David Southwick of Batchelar McDougall Consulting, for SESOC......................................... 21
NZS 1170.5 P-delta Example
Article prepared by Stuart Oliver of Holmes Consulting Group for SESOC........................................................... 25
Seismic Retrofit of Masonry Walls using Timber Strong-backs
Dmytro Y. Dizhur, Marta Giaretton, Ivan Giongo, Jason M. Ingham...................................................................... 30-44
A Direct Displacement-Based Seismic Design Procedure for Moment Frames with Non-Linear Viscous
Dampers – Part 2: Validation of the Design Procedure
Dion Marriott.......................................................................................................................................................... 45-56
Estimation of the Static Vertical Subgrade Reaction Modulus ks from CPT for Shallow Foundations on
Cohesionless Soils
Nick Barounis, Johnny Philpot.............................................................................................................................. 58-63
Soil Classification in the Palmerston North Area using Strong Ground Motion Record
Jonathan Dymoch................................................................................................................................................. 64-68
Improvements in Finite Element Analysis for Complex Structural Design
Ben Horsfall, Max Waters, Matt Bishop................................................................................................................ 71-78
Using Intumescent Coatings to Achieve Building Code Compliance on Steel Structures
Allan Jowsey, AkzoNobel....................................................................................................................................... 79-84
ASEC Conference 2018................................................................................................................................................. 86-87
SCOSS report ................................................................................................................................................................ 88-90
MBIE and Standards New Zealand Update.................................................................................................................. 91-92
News from the Regional Structural Groups.................................................................................................................... 93-94
The Institute of Structural Engineers Report................................................................................................................... 95
SESOC Membership...................................................................................................................................................... 96
SESOC Treasurer’s Report............................................................................................................................................. 97
SESOC Membership Application Form.......................................................................................................................... 98

Front Cover:
Photo with compliments Dmytro Dizhur: Timber retrofit to masonry wall, refer p.30 et.seq.

Volume 30 No.2 September 2017 1

SESOC Journal

SEISMIC RETROFIT OF MASONRY WALLS USING TIMBER

STRONG-BACKS
Dmytro Y. Dizhur1, Marta Giaretton2, Ivan Giongo3 and Jason M. Ingham4

ABSTRACT:
A high proportion of New Zealand’s clay brick unreinforced masonry (URM) structures have not been retrofitted to
resist earthquake forces, and in particular to prevent out-of-plane failures which are the most critical deficiencies
of URM buildings. Despite a number of seismic improvement techniques having been applied previously there is a
significant lack of experimentally validated simple and cost-effective solutions that also consider the impact on the
building tenants, aesthetics and heritage building fabric. The main objectives of the research presented herein were to
develop and validate seismic securing techniques for URM solid- and cavity-walls that satisfied the above conditions.
Full-scale shake-table testing of two cavity and three double-leaf solid clay brick URM walls was undertaken. The
vertical timber framing that is typically considered to be a non-structural support for the inner wall lining was used
as part of the retrofit solution and was fixed to the wall using mechanical screw-ties in order to form a strong-back.
The intended outcomes of the research reported herein included (i) measuring via laboratory testing the improvement
of seismic capacity in terms of peak ground acceleration (PGA) achieved and maximum out-of-plane displacement
experienced by the URM solid- or cavity-walls due to strong-back retrofit installation; (ii) comparing the performance
of different strong-back configurations; (iii) providing construction details, and providing analytical formulations for
response prediction.
Keywords: unreinforced masonry, seismic retrofit, seismic rehabilitation, strong backs, masonry anchors, cavity
walls, brick walls, out-of-plane, earthquake resistance

1. INTRODUCTION inadequate, the wall effectively becomes a tall unrestrained

cantilever and results in collapse when inertial forces
The earthquake vulnerability of buildings constructed
result in the wall being pushed beyond its point of static
using unreinforced clay brick masonry (URM) prior to the
instability. If the anchorage to the diaphragm is adequate,
introduction of modern earthquake loading standards is
the wall will fail in a ‘beam type failure’ and cracking occurs
well-known in New Zealand (Dizhur et al. 2011, 2013;
above mid-height leading to rocking of the two parts as
Ingham and Griffith 2011). Furthermore, URM walls with
two separate rigid bodies. The inertial forces on the wall
cavities (i.e., continuous air gaps separating leaves of
are distributed to the diaphragm, causing the wall to act
brick from one another) are fairly prominent in the New
in bending and being reliant on the limited tensile strength
Zealand URM building stock and are likely to be even more
of the mortar (Meisl et al. 2007). Out-of-plane beam-type
vulnerable to out-of-plane (OOP) collapse than are solid
failure can either occur in one-way bending or two-way
URM walls (Giaretton et al. 2016c; Walsh et al. 2015), see
bending. One-way bending tends to occur in longer walls
Figure 1. A URM wall subjected to out-of-plane loading
or walls without side supports whereas two-way bending
can be idealised as a one-way vertically spanning strip with
requires the support of at least one vertical edge (Dizhur
a horizontal crack developing at some height within the
et al. 2014). Factors affecting the response of the wall
span (Giaretton et al. 2016c; Graziotti et al. 2016; Penner
include (Penner and Elwood 2016): the height to thickness
and Elwood 2016). There are two basic failure modes
ratio of the wall, overburden weight carried by the wall,
describing how out-of-plane failure occurs: cantilever
quality of mortar and construction, strength and stiffness of
failure and beam type failure. The determining factor
diaphragm, and displacement response of the structure.
between which failure mode occurs is the adequacy of the
anchorage at the floor and roof connections (Russell and The method studied herein for the out-of-plane seismic
Ingham 2010). When connections to the diaphragm are retrofit of URM solid- and cavity-walls consists of

PAPER CLASS & TYPE: RESEARCH REFEREED

(1) Lecturer, Department of Civil & Environmental Engineering, University of Auckland, Private Bag 92019, Auckland 1010, New Zealand, ddiz001@aucklanduni.ac.nz - corresponding author
(2) Research Fellow, Department of Civil & Environmental Engineering, University of Auckland, Private Bag 92019, Auckland 1010, New Zealand, mgia506@aucklanduni.ac.nz
(3) Assistant Professor, Department of Civil, Environmental and Mechanical Engineering, University of Trento, via Mesiano 77, 38123 Trento, Italy, ivan.giongo@unitn.it
(4) Professor, Department of Civil & Environmental Engineering, University of Auckland, Private Bag 92019, Auckland 1010, New Zealand, j.ingham@auckland.ac.nz

30 Journal of the Structural Engineering Society of New Zealand Inc

 SESOC Journal

connecting a series of vertical members (termed herein as
strong-backs) to the interior surface of the wall (see Figure
2) at sufficient spacing to ensure that the width of wall
between supports is capable of resisting the out-of-plane
forces. Strong-backs act in flexure in order to transfer wall
loads to the adjacent floor diaphragms, breaking up a
large planar wall into a number of buttressed segments. A
similar approach was proposed by (King et al. 2009) using
steel strong-backs as a retrofit strategy to protect masonry
cladding structures from blast loading. Strong-back
members are connected to the URM material via adhesive
anchors or through-plate anchors, which allows a high
level of reversibility should the need to remove the retrofit
system arise. Design considerations when using strong-
backs include height to thickness of the URM wall and
spacing of strong-backs. The demand and capacity of the
wall will dictate the spacing of the anchorages between the
URM wall and strong-backs, as well as the considerations
of connection detailing to transfer loads into the diaphragm
(FEMA 547 2006).
Two shake-table experimental campaigns are reported
herein, that investigated the performance of timber
strong-backs applied to clay brick URM cavity- and solid-
walls as a cost-effective seismic securing solution. The
choice of using timber as a retrofit material comes from
investigations of existing URM buildings in New Zealand
that showed that a large proportion of these buildings
had timber framing lined with plasterboard as the interior
finish. Hence, validating a securing solution that connected
the masonry to the timber framing as a load path into the
diaphragm would provide a practical and low-cost seismic
securing method. The concept also aligns with the timber
framing that was used as an earthquake-resistant system
for masonry buildings during the Minoan era (Tsakanika-
Theohari 2008), which was then later extended to the
entire Mediterranean area (Ruggieri et al. 2015). The
authors also experimentally validated the combined use
of timber strong-backs and bracing for retrofitting URM
parapets, with further details available in (Dizhur et al.
2017b; Giaretton et al. 2016a).

(a) Solid-wall (b) Cavity-wall

Figure 1. Examples of out-of-plane collapse of URM walls showing presence of internal non-structural timber framing.

(a) External vertical strong-backs applied following the 2010/2011 (b) Internal vertical strong-backs
Canterbury earthquakes
Figure 2. Examples of steel members used as strong-backs for out-of-plane retrofit of URM walls.

Volume 30 No.2 September 2017 31

SESOC Journal
2 TIMBER STRONG-BACK RETROFIT
SOLUTION
The vertical strong-backs used in the reported study
consisted of 90 x 45 mm standard timber studs located
at 550 mm spacings and secured to the URM wall using
Ø12/230L mm mechanical screws, see Figure 3. This
retrofit solution was investigated herein for application in
both clay brick URM cavity, see Figure 4 (a) and solid-walls,
see Figure 4 (b) with the aim of being a multi-purpose
seismic mitigation system and as support to the inner
wall lining as well as allowing space for the installation of
electrical and plumbing systems.
Based on previous airbag testing of cavity-walls (Walsh et
al. 2015), Ø12/230L mm mechanical screw were identified
as the most effective retrofit solution in terms of increased
out-of-plane cavity-wall capacity when compared with
stainless steel helical rods and chemical ties. The screws
had a hexagonal washer type head and a total threaded
length of 160 mm and were installed with a spanner in a
pre-drilled Ø12 mm hole. The masonry was drilled using
a low-impact drill, making sure to limit vibrations in the
walls, see Figure 3 (a). High torque was required during
installation of the Ø12 mm screws and from pull-out
testing it was observed that in the case of weak bricks
in lime-based mortar the screws had a tendency to split
the bricks, resulting in a lower pull-out capacity of 18
kN (6 samples, COV 4%). In the cavity-walls the Ø12
mm mechanical screw also serves as a tie to secure the
(a) Hole drilling (b) Cavity view of the final installation
(d) Finish for solid-wall with parapet (e) 5 mm steel brackets to fix
base of strong-backs
Figure 3. Installation process for timber strong-back retrofit
32
two wall leaves together, see Figure 3 (b) and hence the
vertical timber studs were fixed using steel angle brackets
mounted onto the wall at a spacing of 400 mm through
the screws, see Figure 3 (c) and Figure 4 (a). In this way
the screws reached at least half the depth of the bricks of
the opposing masonry leaf Figure 4 (a), thereby ensuring
an adequate embedment length without modifying the
external appearance of the building. In the solid-walls the
screws were located at the centre of the timber studs with
a vertical spacing of approximately 500 mm, see Figure 4 (b).
When necessary to provide a smooth surface for the wall
lining, a washer countersunk 10 mm into the strong-back
can be adopted as shown in Figure 3 (e-f). The base of
the strong-backs was fixed to the floor-diaphragm (shake-
table) using a 5 mm thick steel bracket and two Ø12 mm
standard timber screws to allow the transfer of shear
induced in the strong-back, see Figure 3 (e) and Figure
5 (d). For testing set-up constrains, see Figure 3 (e) the
5 mm thick steel brackets were installed differently from
what would be recommended in a real case application,
see Figure 5 (d). The top of the strong-back was fixed to
the roof-diaphragm using steel brackets and 30 mm long
Ø5.5 mm standard timber screws, see Figure 5 (c-d).
Standard GIB plasterboard was fixed to the timber strong-
backs to demonstrate the aesthetic finish achievable using
the securing technique, as shown in Figure 3 (f). A 3D
schematic of the complete strong-backs installation on
both cavity- and solid-cavity walls is presented in Figure 5.
(c) Finish for cavity-wall
(f) Plasterboard finish
Journal of the Structural Engineering Society of New Zealand Inc
 SESOC Journal

(a) Cavity-wall (b) Solid-wall

Figure 4. Cross-sectional schematics of the timber strong-back retrofit technique

(c)

(d)

(a) Installation on a cavity-wall (b) Installation on a solid-wall

(c) Detail of the top connection, (d) Detail of the bottom

Figure 5. Three dimensional schematics of the strong-backs installation

Volume 30 No.2 September 2017 33

 SESOC Journal
3 EXPERIMENTAL VALIDATON
3.1 Test programme and set-up
The strong-back retrofit technique was validated during
two experimental shake-table campaigns: one focusing
on cavity-walls (Giaretton et al. 2016b) and the second
focused on the performance of solid-walls (Dizhur et al.
2017a). Two tests were undertaken for cavity-walls in
order to consider both as-built (W1) and retrofitted (W5)
conditions. The retrofit was undertaken using vertical 90
x 45 mm standard timber studs @ approximately 600
mm horizontal spacings and secured with Ø12/230L
mm mechanical screws @ 400 mm vertical spacing. Five
tests were performed on solid-walls: (i) as-built condition,
URM-p, with the parapet being secured in order to identify
the response of the wall and avoid premature failure of
the URM parapet, (ii) installing 90 × 45 mm timber strong-
backs with mechanical screws @ 500 mm vertical spacing
and securing the parapet, 45SB-p, (iii) using 90 × 90 mm
timber strong-backs and mechanical screws applied in
three different configurations. The three configurations
were: (i) 90SB, wall strong-backs and as-built parapet, (ii)
90SB-p1, wall strong-backs and parapet secured only on
the top and inner side (version 1), (iii) 90SB-p2, eccentric
wall strong-back and parapet secured on the top and both
inner and outer sides (version 2). Table 1 summarises the
characteristics of the tested walls including schematics and
photos of each retrofit configuration.
The test set-ups were designed to replicate common
in-situ boundary conditions for a single-storey wall portion
located at the top floor of a perimeter load-bearing URM
Table 1. Test matrix
Wall Securing Type
W1 URM (with original wire ties)
CAVITY-WALLS
W5 2 x (90 x 45 mm) timber strong-back
URM-p URM + parapet securing
45SB-p 2 x (90 x 45 mm) timber strong-back + parapet securing
SOLID-WALLS
90SB 2 x (90 x 90 mm) timber strong-back (URM parapet)
90SB-p1 2 x (90 x 90 mm)timber strong-back + parapet securing
90SB-p2 1 x (90 x 90 mm) timber strong-back + parapet securing
34
wall, fixed at the base for continuity with the lower walls or
foundation and allowing rotation and vertical displacement
at the top based on typical seating arrangements observed
at the roof level. During cavity-wall testing, the top wall
restraint consisted of horizontal beams applied to both
sides of the wall and into the air-cavity, see Figure 6 (a).
Conversely, solid-walls were restrained at the top to
simulate a timber diaphragm anchored to the wall using
Ø12/230L mm mechanical screws and 50 mm square
washers and composed of four 1500 mm lengths of 190 x
45 mm timber joists, see Figure 6 (c). For both test set-ups
the top restraint was fixed to the purpose built protection
frame and the base of the wall panel was secured with
strong mortar between two stiff steel angles to prevent
lateral movement of the wall base.
Accelerometers were installed at the bottom, middle,
and top of the wall and on the shake-table (denoted as
ACC B, ACC M, ACC T, and ACC ST and three string
potentiometers were attached at the middle and top of
the wall and on the shake-table to measure differential
displacement of the wall (denoted as DW M, DW T, and
DW ST in Figure 7).Two additional accelerometers were
mounted on the solid-wall at three quarter-height and onto
the parapet. A single-axis acceleration-controlled sinusoidal
test transitioning from 0.5 Hz to 50 Hz was applied with
increasing acceleration of approximately 0.05 g every 15
seconds and constant amplitude at 50 mm. All walls were
tested until displaying signs of instability and within the
range of the maximum possible load generated by the
shake-table.
Spacing
Schematic Photo example
(mm)
-
600
-
600
600
600
1200
Journal of the Structural Engineering Society of New Zealand Inc
 SESOC Journal

(a) Cavity-wall test set-up (b) Damaged building (c) Solid-wall test set-up
Figure 6. Laboratory simulation of in-field condition

(a) Cavity-wall (c) Solid-wall

Figure 7. Schematic of instrumentation locations

Volume 30 No.2 September 2017 35

 SESOC Journal
3.2 Wall characteristics
TThe most commonly encountered boundary condition,
geometric characteristics, and material properties for
URM cavity-wall (Giaretton et al. 2016c) and solid-wall
(Russell and Ingham 2010) arrangements were selected,
and four walls that closely mimicked in-situ conditions
were constructed using recycled clay bricks obtained
from demolished vintage URM buildings. Brick dimensions
were of size (230L × 110W × 75H mm) and compressive
strength of 26.4 MPa (10 samples, COV 22%) for cavity-
walls and 30.5 MPa (6 samples, COV 27%) for solid-walls,
estimated using the half brick compression test (ASTM C67
2016). The compressive strength of masonry prisms was
7.4 MPa (3 samples, COV 16%) and 8.2 MPa (3 samples,
COV 20%) respectively, in accordance with (ASTM C1314
2016). The mortar mix was made from sand and lime in
the ratio of 3:1 by volume respectively. 50 ×50 ×50 mm
mortar test cubes were prepared during wall construction
and were tested in compression after 28 days to obtain
an average compressive strength of 0.54 MPa (6 samples,
COV 18%, (ASTM C109 2013).
Test cavity-walls were constructed with two single URM
leaves in a running bond pattern with a mortar joint
thickness of approximately 10 to 15 mm. Masonry leaves
were interconnected using 4 mm diameter horse-toe
metal wire cavity-ties to replicate as-built field conditions
(Giaretton et al. 2016b; c). Notches were cut into the wire
near the outer-leaf to simulate the rusted and deteriorated
condition of typical wire cavity-ties. Wire cavity-ties were
laid based on the most common tie arrangement observed
during preliminary surveys (Giaretton et al. 2016c), being
two ties per row (600 mm) and one row of ties for every
six masonry courses (450 mm). Cavity-wall samples
were 3000 mm high, 1190 mm wide, and 270 mm thick
including a 50 mm air-cavity.
Solid-wall test samples were two-leaf-thick common brick
pattern walls with a mortar joint thickness of approximately
10–15 mm. The panels were 3300 mm high including a
300 mm high parapet above, and were 1200 mm wide and
230 mm thick.
(a) W1 (b) W5
36
3.3 Dynamic response
For all tests, cracking was mainly concentrated in the top
quarter of the wall height, see Figure 8. As-built cavity-
wall W1 exhibited differential movement of the two wall
leaves, resulting in bending of the original cavity-ties and
significantly reduced air-cavity width, see Figure 8 (a). The
rocking capacity was influenced by mortar strength, with
ultra-weak mortar tending to crumble, thereby reducing the
contact area at the point of rotation and expediting rocking
and, consequently, wall failure which occurred at 0.45g, see
Table 2. The maximum displacement registered was 26 mm
at mid-height and 17 mm at the top, see Figure 9 (a).
The use of 90 x 45 mm strong-backs in W5 significantly
reduced the lateral mid-height displacement during testing,
preventing the occurrence of both failure from rocking
due to flexural tensile crack formation and out-of-plane
bed-joint shear failure. The increased capacity resulted in
the initiation of stiff global flexural behaviour at high levels
of table acceleration (1.3 g, see Figure 8 (b) and Table 2)
without reaching instability or collapse within the range
of the maximum possible load generated by the shaking
table. W5 exhibited a displacement profile comparable
to W1, with 31 mm at mid-height and 13 mm at top, see
Figure 9 (a), but at an acceleration that was three times
higher than for W1.
As-built solid-wall URM-p displayed a typical one-way
out-of-plane bending failure with major cracking at three
quarter-height and minor cracking at mid-height, see
Figure 8 (c). The formation of the three quarter-height crack
caused a hinge effect, with the wall parts above and below
beginning to rock as two separate almost-rigid bodies,
inducing a large increase in acceleration at this level. As the
three quarter-height displacement increased the flexural
capacity of the wall was exceeded, causing the wall to
collapse at 0.46 g. The maximum displacement recorded
near-collapse was 55 mm at top and 186 mm at mid-
height, see Figure 9 (b).
45SB-p was retrofitted using 90 x 45 mm timber strong-
backs from wall base to parapet top. Cracking and
(c) URM-p
Journal of the Structural Engineering Society of New Zealand Inc
 SESOC Journal

(d) 45SB-p (e) 90SB (f) 90SB-p1 (g) 90SB-p2

Figure 8. Screenshots showing crack-pattern survey and failure progression

consequent falling of bricks occurred at the parapet
edges, external to the strong-backs, as shown in Figure
8 (d). With increasing motion intensity, a flexural behaviour
was observed, leading to crack formation at three
quarter-height followed by bricks being expelled from the
surrounding area which involved only the outer leaf. The
acceleration reached 1.33 g, which was three times higher
than that for the as-built condition. The displacement
profile was linear, with 34 mm being recorded at top and
23 mm at mid-height, see Figure 9 (b), corresponding to a
reduction with respect to the as-built condition of 39% and
87% respectively.
Walls 90SB and 90SB-p1 were both retrofitted using
90 x 90 mm timber strong-backs, with the parapet
being un-retrofitted in 90SB and being retrofitted in
90SB-p1. The linear displacement profile in Figure 9 (b)
clearly shows that the 90 x 90 mm timber strong-backs
significantly increased the wall monolithic behaviour
and prevented any cracks from forming. In wall 90SB
the un-retrofitted parapet exhibited rigid-body rocking
behaviour after cracking formed at the roof diaphragm
level (parapet base), see Figure 8 (e). In wall 90SB-p1 the
parapet was retrofitted with strong-backs and a single-
side horizontal top restraint, preventing rocking failure but
allowing the parapet to slide outwards on the existing
cracking plane as motion intensity increased, see Figure
8 (f). 90SB and 90SB-p1 behaved similarly in terms of
acceleration and displacement along the wall height, see
Table 2. Summary of results
profiles in Figure 9 (b). Instability due to parapet rocking
or sliding was reached at approximately 0.96 g (average
value), corresponding to a maximum displacement of
approximately 15 mm at mid-height and 23 mm at top.
The recorded PGA was twice the value reached in the
as-built condition and the reduction in displacement
was 85% at mid-height and 77% at top. In wall 90SB-
p2 the eccentricity caused by the strong-back position
increased the stiffness of one end of the wall configuration
in comparison to the other end, resulting in the initiation of
torsion. A crack formed at the wall base, starting from the
side without strong-backs and eventually propagated all
the way through the base as the shake-table accelerations
increased, see Figure 8 (g). The crack at the base allowed
rocking to develop in the whole wall, which led to an
increase in the displacement at the roof diaphragm level.
The ultra-weak mortar did not provide enough friction
against the increasing displacement, enabling brick pull-
out where the mechanical screws were tied and resulting
in the formation of a 15 mm gap between the wall and
the roof diaphragm. Consequently, the displacements
registered were approximately twice those experienced
by 90SB and 90SB-p1, even though the PGA was lower
(0.82 g), see Table 2. The single strong-back provided
a sufficient increase in stiffness to prevent any cracks
developing at the three quarter-height and mid-height,
hence providing securing from out-of-plane failure. The
parapet had a double-sided horizontal top restraint and
hence did not present further damage.

Wall ID PGA Max mid-height Max top Failure mode

displacement displacement
W1 0.45g (-) 26 mm (-) 17 mm (-) One-way bending
W5 1.31 g (291%) 31 mm (119%) 13 mm (76%) Flexural behaviour
URM-p 0.46 g (-) 186 mm (-) 55 mm (-) One-way bending
45SB-p 1.33 g (289%) 23 mm (13%) 34 mm (61%) Flexural behaviour
90SB 0.95 g (205%) 14 mm (7%) 20 mm (36%) Rigid body behaviour and parapet rocking
90SB-p1 0.97 g (209%) 15 mm (8%) 25 mm (46%) Rigid body behaviour and parapet sliding
90SB-p2 0.82 g (177%) 28 mm (15%) 38 mm (69%) Rigid body and torsional behaviour

(%) comparison to the as-built value

Volume 30 No.2 September 2017 37

SESOC Journal

(a) Cavity -walls (b) Solid-walls

Figure 9. Displacement profiles

4 ANALYTICAL PREDICTIONS a uniform stress distribution in compression. Under such

assumptions, wall bending capacity 𝑀𝑅 at the generic (𝑥)
An analytical formulation to predict the response of solid
cross-section can be obtained as recommended by (DM
URM walls retrofitted with timber strong-backs was
14/01/2008 2008) where a stress-block distribution of the
developed based on the theory of ‘composite beams
compression stresses is adopted:
with incomplete interaction’ that was originally developed
by Newmark et al. (1951). Because of the presence of (1)
mechanical screws, composite action between the URM
wall and the timber strong-backs is expected as a result of
the shear force transfer at the masonry-to-timber interface. where:
The formulation assumptions are the same as per 𝜎0(𝑥) is the average compression stress at the (𝑥) cross-
(Newmark et al. 1951): i) materials (i.e. timber, masonry section location (MPa);
and connections) are linear elastic; ii) displacements/
𝑓𝑐 is the masonry compression strength (MPa);
deformations are small in comparison with wall/strong-
back relevant dimensions (small displacement/deformation 𝑏 is the wall length (mm);
theory); iii) beam components have equal curvature (no t is the wall thickness (mm).
separation or penetration allowed); iv) Planar sections
(classical beam-theory); v) beam cross-section is constant
along the longitudinal axis; vi) fasteners are uniformly
distributed (constant stiffness k along the longitudinal axis).

4.1 URM wall

URM wall behaviour when loaded out-of-plane by
earthquake action was schematised by the trilinear model
suggested by (Doherty et al. 2002) and reported in Figure
10. Δ1 is the wall maximum horizontal displacement at
the point of plastic hinge formation in the most stressed
cross-section (section bending capacity is engaged),
while Δ2 represents the wall displacement capacity which
corresponds to the beginning of the softening branch
(strength decrease due to local crushing and/or out-of- Figure 10. Schematisation of OOP wall behaviour showing
plane bed-joint shear failure). Due to the masonry tensile theoretical and experimental representation
capacity being significantly smaller than the masonry
compression capacity, the initiation of cracking is expected
well before the cross-section capacity is reached (Figure
10), despite the vertical arching action provided by the
presence of the overburden from the parapet. Figure 11
shows the wall cracking profile calculated for URM-p at Δ1
condition by assuming zero masonry tensile strength and

38 Journal of the Structural Engineering Society of New Zealand Inc

 SESOC Journal

𝑀𝑚𝑎𝑥, which occurs at 0.375ℎ (from the top of the wall), is

equal to 𝑓ℎ2 and consequently l𝑀𝑠l > l𝑀𝑚𝑎𝑥 l
(Figure 12).

However, if the vertical stresses due to overburden are

small (as for the tested specimens), then the wall bending
capacity is reached at a critical section 𝑥 located above the
wall mid-height and the fixity condition at the base can still
be considered valid. Therefore, it was possible to adopt the
distance ℎ (Figure 12) between the wall top support and the
location where the bending moment is null, as the reference
height for the simply supported composite system required
to satisfy Möhler’s assumption (Möhler, 1956).

Figure 11. Example of wall (URM-p) cracking at Δ1 condition

(calculation assumptions: uniform stress distribution in
compression; zero tensile strength)

For assumption i to be valid, an equivalent elastic wall

characterised by a mean effective thickness has to be
determined. Mean effective thickness tcr was determined
via finite element modelling by imposing displacement
compatibility at the Δ1 condition between the equivalent
elastic wall (thickness = tcr) and the real wall where crack
formation over the wall height is simulated. Numerical
modelling was undertaken using TNO DIANA software
(DIANA FEA BV 2014). In order to be consistent with
the masonry assumptions previously described, elastic Figure 12. Bending moment vs. wall bending capacity
perfectly-plastic material behaviour in compression (material properties as per specimen URM-p)
was selected. No-tension line-interface elements were A reference wall-width b2 corresponding to the strong-back
introduced at regular intervals (30 mm) to simulate masonry spacing was assumed during the calculation process.
zero tensile strength. Obtained tcr values ranged between The maximum admissible acceleration (capacity-based
approximately t/4 and t/3 depending on the wall conditions procedure) was determined according to the lower bound
in terms of overburden and masonry compression strength. case scenario reported in Figure 13, where the masonry
wall is loaded on the same side as the location of the
4.2 Timber-to-Masonry Composite strong-backs.
(TMC) wall
The solutions to the problem of composite beams with
partial interaction, as proposed by Newmark et al. (1951),
are characterised by long and complicated expressions that
make them unsuitable for engineering design practice. The
problem solution was simplified by Möhler (1956) assuming
simply supported boundary conditions and sinusoidal load
distribution.
Because of the limited overburden, boundary conditions
of the tested specimen (and more generally boundary
conditions of masonry panels located at the upper-most
level of a building) can be schematized as pinned at the top
and fixed at the base.
If a uniform lateral-force distribution f is adopted, then the
Figure 13. Timber-to-masonry composite wall section
minimum bending moment 𝑀𝑠 at the wall base is equal to
𝑓ℎ2 (ℎ is the wall height) while the maximum moment
8

Volume 30 No.2 September 2017 39

SESOC Journal

In order to be consistent with Möhler’s simplification of the 𝛾1 is a coefficient that reduces coupling action
problem (Möhler, 1956), the inertial load on the Timber-to- because of the connection shear deformability at
Masonry Composite (TMC) wall is reproduced by using a the wall to strong-back interface.
sinusoidal distribution f(x) (N/mm): 𝛾1 was determined as the following.

Eq.(1)
Eq.(8)

where:
where:
𝑊 is the wall weight (N);
𝑘 is the distributed stiffness of the connection and is
ℎ Is the wall reference height (mm); calculated as:
𝛼 is the ratio of the wall acceleration to the
acceleration of gravity.
The bending moment generated by the inertia force Eq.(9)
becomes the following:
The Δ1 condition in the TMC wall is reached when
Eq.(2) 𝛼 = 𝛼𝑝𝑒𝑎𝑘 and the bending moment 𝑀2(𝑥) at the critical
section 𝑥 equals the bending capacity 𝑀𝑅(𝑥). When
Hence, the shear force is: determining 𝑀𝑅(𝑥), the wall capacity reduction due to the
axial tension force induced by the composite action was
Eq.(3) disregarded. Hence, the acceleration 𝛼𝑝𝑒𝑎𝑘 that engages the
TMC full capacity can be determined by using the following
Because of the curvature compatibility assumption, the expression:
following expression can be written:
Eq.(10)
Eq.(4)

where: where 𝑥 identifies the critical section.

𝑀2𝑥 is the bending moment on the masonry wall at The acceleration 𝛼𝑚𝑎𝑥 that determines the wall ultimate
section x (Nmm); condition (Δ2) is calculated as:
𝐸2 is the wall modulus of elasticity (MPa); 𝛼𝑚𝑎𝑥=𝛼𝑝𝑒𝑎𝑘∙𝑞 Eq.(11)
𝐽2 is the wall moment of inertia (out-of-plane) (mm4); where q is the behavior factor of the wall. The relation
𝐸𝐽𝑒𝑓𝑓 is the effective flexural stiffness of the timber-to- proposed by (Newmark and Hall 1982) (Equation (12))
masonry composite structure. that links the q factor to the ductility ratio is
adopted.
𝐸𝐽𝑒𝑓𝑓 can be determined as follows:
𝐸𝐽𝑒𝑓𝑓=𝐸𝐽0+𝐸𝐴0∗a2 Eq.(5)
Eq.(12)
where: Doherty et al. (2002) provided experimental values for both
Δ1 and Δ2. From the average values of Δ1 and Δ2, μ = 3.6 is
𝐸𝐽0=𝐸1𝐽1+𝐸2𝐽2 Eq.(6)
obtained and consequently 𝑞≈2.5 is derived.
Eq.(7) At the Δ2 condition, the maximum out-of-plane
displacement is measured at 0.375ℎ from the top of the
wall and corresponds to:
where:
𝐸𝑖𝐽𝑖 is the flexural stiffness of the i-th component;
𝐸𝑖𝐴𝑖 is the axial stiffness of the i-th component; Eq.(13)
𝑎 is the distance between the barycenter of the
strong-back element and the barycenter of the
wall;

40 Journal of the Structural Engineering Society of New Zealand Inc

 SESOC Journal

4.3 Connection verification 4.5 Equation summary & procedure

The shear force at the masonry wall to timber strong-back
interface is related to the shear load 𝑉( ):
Eq.(14)
The fastener shear capacity RfS must be greater than the
maximum shear force SfS, which can be estimated as:
application example
Table 3 summarises the analytical formulations required
to predict the response of solid URM walls retrofitted
using timber strong-backs and mechanical connections.
Table 4 shows the results obtained when applying the
analytical procedure to specimen 45SB-p. The predicted
capacity and displacement values appeared to be in good
agreement with the test results reported in Table 2.

Eq.(15)

where sp is the fastener spacing.

The fastener withdrawal capacity RfT must be greater than
the maximum tension force TfT, which can be calculated by
using Equation (16):

Eq.(16)

where 𝛽 is: Table 3. Equation summary ( = /2)

Eq.(17)

4.4 Strong-back verification

The maximum bending moment 𝑀1,𝑚𝑎𝑥 and axial
compression/tension force 𝑁1,𝑚𝑎𝑥 in the timber element
are located at 𝑥=0.5ℎ (from the top of the wall) and are
determined as follows: Table 4. Procedure application example for TMC wall
specimen 45SB-p

4.6 Analytical
Eq.(18) prediction sensitivity to
parameter variations
The sensitivity of the predicted TMC response to variations
Eq.(19) in the values of the key parameters are reported in the
following graphs. Figure 14 shows how the acceleration
Shear is maximum (𝑉1,𝑚𝑎𝑥) at the wall ends (with reference αmax changes in relation to the strong-back spacing b2,
to ): Figure 14 (a) and the fastener spacing sp, see Figure 14 (b).
Figure 15 shows how the efficiency η of the timber-to-
masonry composite system is influenced by the variation
Eq.(20)
of the same parameters (b2 and sp). Similarly to what
is typically done for timber-concrete or timber-timber
composites η is determined as :

Eq.(21)

where:
𝐸𝐽∞ is the flexural stiffness of the ideal composite system
where no slip is allowed at the masonry-to-timber interface:
𝐸𝐽∞=𝐸𝐽0+𝐸𝐴0∙𝑎2; 𝐸𝐴0=1/(1/𝐸1𝐴1+1/𝐸1𝐴1) Eq.(22)

Volume 30 No.2 September 2017 41

SESOC Journal

Figure 14. Specimen 45SB-p: Effect of variations in strong-back spacing (a) and fastener spacing (b)

Figure 15. Specimen 45SB-p: Effect of variations in strong-back spacing (a) and fastener spacing (b) on η (efficiency of
the timber-to-masonry composite system)

Figure 16. Specimen 45SB-p: Effect of variations in

strong-back cross-section geometry (t1 and b1) on
acceleration αmax

42 Journal of the Structural Engineering Society of New Zealand Inc


4 CONCLUSIONS
Shake-table tests were undertaken in order to
experimentally validate the use of timber strong-backs
as simple and cost-effective seismic retrofit solutions for
clay brick URM cavity and solid-walls and the following
conclusions were drawn:
• The critical failure mode for URM walls in the as-built
condition was one-way bending in the out-of-plane
direction with a crack formation at three quarter-height
enabling the wall to act as two separate rocking bodies.
In the cavity-walls, bending of the original cavity-ties and
subsequent differential movements between masonry
leaves was observed.
• All of the tested retrofitted walls sustained increased
PGA values with reduced lateral displacements
experienced up the height of the wall. The most effective
mitigation system was the use of 90 x 45 mm timber
strong-backs from wall base to parapet top, which
allowed flexural behaviour with a significant reduction in
displacement and an increased PGA of three times the
as-built condition for both cavity- and solid-walls.
• The use of 90 x 90 mm timber strong-backs further
decreased the lateral displacement experienced,
resulting in rigid-body behaviour. The parapet failure
induced earlier instability with respect to the dynamic
loading sustained by 45SB-p.
• Timber strong-backs were the most cost-effective
and simple-to-install securing technique implemented.
Standard 90 x 45 mm timber framing can be used
as strong-backs, and do not require a specialist
construction contractor to install.
• The roof diaphragm interaction with the wall provided a
weak plane for cracking to form and the parapet to fail.
• Mechanical screw ties provided adequate wall-to-roof
diaphragm connection during dynamic loading. Brick
pull-out was observed prior to screw pull-out from
bricks.
Analytical formulations required to predict the response of
solid URM walls retrofitted using timber strong-backs and
mechanical connections were also provided.
Volume 30 No.2 September 2017
SESOC Journal
ACKNOWLEDGEMENTS
The authors would like to express their sincere gratitude for
the funding and support provided by the Building Research
Association of New Zealand (BRANZ), the New Zealand
Natural Hazards Research Platform, the QuakeCoRE
Centre for Earthquake Resilience, and the 2017 ReLUIS
Project framework (funded by the Italian Emergency
Management Agency). Students who participated in the
laboratory testing efforts include Chengliang (Liang) Qian,
Xi (Kevin) Jiang, Kevin Crowe, Timothy Cleaver, and Sean
Bissett. The authors are also thankful to ITW New Zealand
for proving components for the test frame and to Steel
Stuff Limited for proving fabrication assistance.
REFERENCES
ASTM C109. (2013). Standard test method for
compressive strength of hydraulic cement mortars.
American Society for Testing and Materials, USA.
ASTM C1314. (2016). Standard test method for
compressive strength of masonry prisms. ASTM
International, American Society for Testing and Materials,
USA, 10.
ASTM C67. (2016). Standard test methods for sampling
and testing brick and structural clay tile. American Society
for Testing and Materials, USA.
Dizhur, D., Giaretton, M., Crowe, K., Cleaver, T., and
Ingham, J. (2017a). “Securing solutions for face-
loaded clay brick URM walls.” 13th Canadian Masonry
Symposium - CMS, (Halifax, June 4th-7th) Canada.
Dizhur, D., Giaretton, M., and Ingham, J. (2017b).
“Seismic restraints for clay brick URM parapets validated
using shake-table testing.” 16th World Conference on
Earthquake Engineering - WCEE, (Santiago, January 9th-
13th) Chile.
Dizhur, D., Griffith, M., and Ingham, J. (2014). “Out-of-
plane strengthening of unreinforced masonry walls using
near surface mounted fibre reinforced polymer strips.”
Engineering Structures, Elsevier Ltd, 59, 330–343.
Dizhur, D., Ingham, J., Moon, L., Griffith, M., Schultz, A.,
Senaldi, I., Magenes, G., Dickie, J., Lissel, S., Centeno,
J., Ventura, C., Leite, J., and Lourenco, P. (2011).
“Performance of masonry buildings and churches in the 22
February 2011 Christchurch Earthquake.” Bulletin of the
New Zealand Society for Earthquake Engineering, 44(4),
279–296.
Dizhur, D., Moon, L., and Ingham, J. (2013). “Observed
performance of residential masonry veneer construction
in the 2010/2011 Canterbury earthquake sequence.”
Earthquake Spectra, 29(4), 1255–1274.
43
 SESOC Journal

FEMA 547. (2006). Techniques for the seismic rehabilitation
of existing buildings. Prepared by R&C Consulting Engineers
and NIST. Federal Emergency Management Agency (USA).
Giaretton, M., Dizhur, D., and Ingham, J. (2016a). “Shake
table testing of seismically restrained clay brick masonry
parapets.” Earthquake Spectra, (under review).
Giaretton, M., Dizhur, D., and Ingham, J. M. (2016b).
“Shaking table testing of as-built and retrofitted clay brick
URM cavity-walls.” Engineering Structures, 125, 70–79.
Giaretton, M., Dizhur, D., da Porto, F., and Ingham, J. M.
(2016c). “Construction details and observed earthquake
performance of unreinforced clay brick masonry cavity-
walls.” Structures, Elsevier, 6, 159–169.
Graziotti, F., Tomassetti, U., Penna, A., and Magenes, G.
(2016). “Out-of-plane shaking table tests on URM cavity
walls.” Engineering Structures, Elsevier Ltd, 125, 455–470.
Ingham, J. M., and Griffith, M. C. (2011). Report to the Royal
Commission of Inquiry: The Performance of unreinforced
masonry buildings in the 2010/2011 Canterbury Earthquake
Swarm.
King, K. W., Wawclawczyk, J. H., and Ozbey, C. (2009).
“Retrofit strategies to protect structures from blast loading.”
Canadian Journal of Civil Engineering - Special Issue on
Blast Engineering, NRC Research Press, 36(8), 1345–1355.
Meisl, C. S., Elwood, K. J., and Ventura, C. E. (2007).
“Shake table tests on the out-of-plane response of
unreinforced masonry walls.” Canadian Journal of Civil
Engineering, 34(11), 1381–1392.
Penner, O., and Elwood, K. J. (2016). “Out-of-plane dynamic
stability of unreinforced masonry walls in one-way bending:
Shake table testing.” Earthquake Spectra, 32(3), 1675–
1697.
Ruggieri, N., Tampone, G., and Zinno, R. (2015). Historical
earthquake-resistant timber frames in the Mediterranean
area. Springer.
Russell, A. P., and Ingham, J. M. (2010). “Prevalence of New
Zealand’s unreinforced masonry buildings.” Bulletin of the
New Zealand Society For Earthquake Engineering, 43(3),
182–202.
Tsakanika-Theohari, E. (2008). “The constructional analysis
of timber load bearing systems as a tool for interpreting
Aegean Bronze Age architecture.” Proceedings of the
Symposium “Bronze Age Architectural Traditions in the
Eastern Mediterranean. Diffusion and Diversity,” (Munich,
May 7th-8th) Germany, 127–142.
Walsh, K. Q., Dizhur, D. Y., Shafaei, J., Derakhshan, H.,
and Ingham, J. M. (2015). “In situ out-of-plane testing of
unreinforced masonry cavity walls in as-built and improved
conditions.” Structures, Elsevier B.V., 3, 187–199.

TIP: Concerned About Building Clause B2

Durability and Producer Statements?
If you are submitting designs and the relevant Building Consent Authority has asked you to show explicit compliance for
B2 Durability, you may have been wondering how to do that. You may have been asked to show compliance for B2 by
completing a Producer Statement (PS1).
IPENZ and ACENZ have been working with Auckland Council, Christchurch City Council and MBIE to explore whether
existing regulations allow engineers to show compliance for durability for all materials used, and whether Producer
Statements are suitable for showing compliance to B2.
We recommend that compliance for durability be provided only on a material-by-material basis. To do this, submit a letter
to the Building Consent Authority tailored to the materials that you have specified in the design. Download sample letter
from IPENZ website https://www.ipenz.nz/home/news-and-publications/news-article/concerned-about-building-clause-
b2-durability-and-producer-statements which details the durability standards relevant to timber treatment, concrete cover
and steel protection. Current regulation provides acceptable solutions only for a limited number of materials.
If you are being asked to show compliance for B2 during the construction phase, supply a letter detailing only those
materials you can verify have been correctly used in the construction, according to the consented design. Poured
concrete is one material that an engineer could verify. IPENZ and ACENZ are developing an example letter for
construction monitoring and durability, and we will let you know when this is available.
In our view, Producer Statements are not suitable for showing compliance to B2.
We are keen to work with Building Consent Authorities on your behalf, so please contact us if you would like help working
through any B2 compliance matters.
IPENZ
(Auckland Council have recently agreed to the letter format for B2 compliance – Editor)

44 Journal of the Structural Engineering Society of New Zealand Inc

View publication stats