Contents lists available at ScienceDirect

International Journal of Disaster Risk Reduction

International Journal of Disaster Risk Reduction 7 (2014) 51–67

2212-42http://d

n CorrE-m

sn12@ii1 M2 M

journal homepage: www.elsevier.com/locate/ijdrr

Review Article

A critical review of retrofitting methods for unreinforcedmasonry structures

Subhamoy Bhattacharya a,n, Sanket Nayak 1,b, Sekhar Chandra Dutta 2,b

a Department of Civil and Environmental Engineering, University of Surrey, Guildford,Surrey GU27XH, United Kingdomb School of Infrastructure, Indian Institute of Technology Bhubaneswar, Bhubaneswar 751013, Odisha, India

a r t i c l e i n f o

Article history:Received 26 March 2013Received in revised form6 December 2013Accepted 9 December 2013Available online 19 December 2013

Keywords:Unreinforced masonryEarthquakeLateral loadsBrittleRetrofit

09/$ - see front matter Crown Copyright & 2x.doi.org/10.1016/j.ijdrr.2013.12.004

esponding author. Tel./fax: þ44 148368953ail addresses: s.bhattacharya@surrey.ac.uk, stbbs.ac.in (S. Nayak), scdind2000@gmail.comobile: þ91 94373 10275; fax: þ91 674 2306obile: þ91 78944 07830; fax: þ91 674 230

a b s t r a c t

Unreinforced masonry (URM) buildings are common throughout Latin America, theHimalayan region, Eastern Europe, Indian subcontinent and other parts of Asia. It hasbeen observed that these buildings cannot withstand the lateral loads imposed by anearthquake and often fails, in a brittle manner. Methods for retrofitting URM buildings toincrease the time required for collapse and also to improve the overall strength widelyvary. This review has collated information on various types of retrofitting methods eitherunder research or early implementation. Furthermore, these methods are categorized andcritically analyzed to help further understand which methods are most suitable for futureresearch or application in developing countries. The comparison of the different methodsis based on economy, sustainability and buildability and provides a useful insight. Thestudy may provide useful guidance to policy makers, planners, designers, architects andengineers in choosing a suitable retrofitting methodology.

Crown Copyright & 2013 Published by Elsevier Ltd. All rights reserved.

Contents

1. Background to the problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521.1. Importance of studying unreinforced masonry (URM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

2. Types of URM buildings vulnerable to collapse during earthquakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 532.1. Adobe buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542.2. Brick masonry buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542.3. Stone masonry buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

3. Common failure mechanisms of URM buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 553.1. Adobe buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 553.2. Brick masonry buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563.3. Stone masonry buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

4. Existing URM retrofitting technologies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 574.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 574.2. Surface treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

013 Published by Elsevier Ltd. All rights reserved.

4.ubhamoy.bhattacharya@gmail.com (S. Bhattacharya), sanketiitbbsr@gmail.com,, scdutta@iitbbs.ac.in (S.C. Dutta).203.6203.

Table 1Earthquakes ca100 years.

Year L

1908 M1920 G1923 K1927 Q1932 G1948 A1970 P1976 T2001 G2003 B2004 S2005 K2008 S2010 H2011 S

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–6752

4.2.1. General surface treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584.2.2. Application of shotcrete. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

4.3. Stitching and grout/epoxy injection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584.4. Re-pointing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594.5. External reinforcement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

4.5.1. Bamboo reinforcement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594.5.2. Seismic wallpaper or glass fibre-reinforced polymer (GFRP) reinforcement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

4.6. Strengthening of junction of URM walls by using L-shaped reinforcement and poly-propylene (PP) band . . . . . . . . . . . 594.7. Post-tensioning using rubber tyres. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604.8. Confinement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614.9. Mesh reinforcement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

4.9.1. Polymer mesh reinforcement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614.9.2. PP packaging strip mesh reinforcement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614.9.3. Steel reinforcement in Peru. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624.9.4. Comparing various methods of mesh reinforcement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

5. Summary and conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63Acknowledgments: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

1. Background to the problem

Earthquakes are results of the deformations of tectonicplates on two sides of a fault resulting from the tendencyof relative displacement between the two tectonic plates.These deformations take place over the years, strainenergy keeps on accumulating in the tectonic plates.Finally, a slip occurs at the fault when the plates slipsback to their original undeformed shape suddenly releas-ing a tremendous amount of energy. The above pro-cess generates earthquakes as per the classical elasticrebound theory. Such earthquakes are called inter-plateearthquakes. However, if there is a weak zone in the plateitself then during the process of accumulation of deforma-tion, a crack may suddenly occur resulting in instanta-neous energy release. This also causes earthquake knownas intra-plate earthquake. Further, volcanic eruption andremoval of mineral ores without adequate protectivemeasures may also be the reasons for generation ofearthquake. However, irrespective of the cause of theearthquake, the consequences can be devastating tohuman lives, see for example Table 1, which lists thecasualties from a few past major earthquakes. For some

using the greatest number of casualties in the last

ocation Casualties

essina (Italy) 70,000 to 100,000ansu (China) 200,000anto (Japan) 143,000inghai (China) 200,000ansu (China) 70,000shgabat (Turkmenistan) 110,000eru 66,000angshan (China) 255,000ujarat (India) 20,000am (Iran) 30,000umatra (Indonesia) 220,000ashmir (Pakistan) 73,000ichuan (China) 69,197aiti 230,000ikkim (India–Nepal border) 150

developing countries, there exists a vicious cycle wherebythey do not possess the wealth to develop their infra-structure sufficiently to withstand the damages caused byearthquakes and conversely, earthquake damage affectsthem from developing their economy.

1.1. Importance of studying unreinforced masonry (URM)

Earthquakes are one of the most deadly forms ofnatural disaster, yet human fatality does not occur directlybecause of ground motions; people die as a result of fallingstructure. A vast amount of research has been carried outover the last few decades to prevent the collapse of tallbuildings, resulting in new building codes and guidelinesbeing written. Such structures are found in wealthycountries and are now reasonably designed for seismicloading that they do not cause many casualties. However,in rural areas of developing countries where people aregenerally poor and equipped with little knowledge ofengineering or construction, very little work has beendone to help protect housing against the dangers ofearthquakes. In these areas masonry becomes the majorform of habitat. These remote areas are difficult to reachfor the emergency services meaning that most of thefatalities will occur at the time of the earthquake as rescueis very unlikely. However, the sustainability of masonrystructures has been questioned during past earthquakes.These structures performwell under gravity loading due tosatisfactory compression carrying capacity of masonry.However, it is a challenging task for the engineeringcommunity to improve the shear and tension carryingcapacity of masonry structures for achieving better sus-tainability of such structures during earthquakes. It istherefore vitally important that the engineering commu-nity is made aware of this problem, because solving it willsave hundreds of thousands of lives.

A history of exposure to the effects of serious earth-quakes has allowed the engineering community to pro-gressively increase its knowledge of how buildingsrespond to seismic activity. Fig. 1 shows how the worldhas been subjected to large number of earthquakes whoseepicenters are distributed all over the world. Table 2 shows

Fig. 1. Earthquake epicenters in world from 1963 to 1998 [1].

Table 2Historic development of earthquake engineering practice [2].

Earthquake Remarks Engineering developments

Messina (Italy), 1908 Following 100,000 fatalities, committee of engineersand professors appointed by Italian government tostudy failures and set design guidelines

Base shear equation evolved (i.e. lateral forceexerted on structure is typically 5–15% ofbuilding0s dead weight)

Kanto (Japan), 1923 Bridges and buildings destroyed. Foundations settled,tilted and moved

Seismic coefficient method was incorporated indesign of Japanese highway bridges

Long Beach (USA), 1933 Destruction of building, particularly schools First earthquake for which acceleration recordswere obtained from recently developedaccelerograph

Niigata (Japan), 1964 Soil can also be major contributor to earthquakedamage

Soil liquefaction studies started

San Fernando (USA), 1971 Failure of bridges and dams and soil effects observed Liquefaction studies intensified and bridge retrofitstudies started

Mexico City (Mexico), 1985 Amplification of ground motion Translated in code of practiceNorthridge (USA), 1994 Failure of steel connections in bridges Importance of ductility in construction realizedKobe (Japan), 1995 Massive foundation failure (soil effects) Lateral spreading is thought to be one of the main

causes. JRA (1996) code modified for design ofbridges

Chi-Chi (Taiwan), 1999 Many bridges collapsed, which were located close tofaults

Importance of proximity to plate boundaries andplates realized

Koceli (Turkey), 1999 Damage to Bolu tunnel due to fault movement.Damage to buildings and bridges

Building conforming to design codes performedwell

Bhuj (India), 2001 Large scale destruction National Program for Earthquake EngineeringEducation (NPEEE) formed and EarthquakeGeotechnical Engineering studies started in India

Sumatra, 2004 Destruction to built environment caused byearthquake and giant tsunami waves

New research started on tsunami warning systems

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–67 53

how past earthquakes are constantly teaching engineersabout new aspects of earthquake engineering and leadingto developments within this field. This critical review is acontinuation of the international, award-winning Mondia-lago project, ‘Improving the Structural Strength underSeismic Loading of Non-Engineered Buildings in the Hima-layan Region’ and will build upon and extend the workcarried out to date. It is in this way that this review aims tocontribute towards the ongoing development of engineerswithin the field of earthquake engineering.

2. Types of URM buildings vulnerable to collapse duringearthquakes

URM buildings can broadly be arranged into threecommon categories: adobe, brick and stone masonry. Eachof these has features and construction methods that aredependent on its geographic location and level of localexpertise. Although cheap and easy to build, all URMbuildings have been observed to be susceptible to earth-quakes, as will be explained in this review.

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–6754

2.1. Adobe buildings

Adobe structures, which consist of sun-dried blockswith mud mortar, have been popular throughout theworld for thousands of years. This is because of the low-cost and availability of construction materials, which canbe soil sourced from nearby fields and the ease ofconstruction with no sophisticated equipment or expertiserequired. A newly-constructed adobe house can cost aslittle as $10/m² in India. This popularity is illustrated by thehigh proportion of adobe housing present in differentdeveloping countries of Latin America, Africa, Asia, theMiddle East and Southern Europe. However, most of thesedeveloping countries are being visited by numerous earth-quakes frequently, as may be well understood from thepictorial representation of the seismically active areas allover the world shown in Fig. 2. Further, the strongpresence of such constructions all over the globe may beseen in Fig. 3. Indeed, 30% of the world0s population lives

Fig. 2. Distribution of seismically active regions ac

Fig. 3. Earth construction distribution across

in adobe dwellings, which accounts for 20% of the world0surban/suburban population [4].

A typical adobe house, shown in Fig. 4, is a single storeyconstruction with foundations that consist of medium tolarge graded rocks with mud mortar in between. The roofis usually composed of timber joists, or even bamboo,overlaid by corrugated metal sheets, clay tiles or thatch,depending on the economy of the region. Adobe structuresare highly prone to collapse during earthquakes causingconsiderable damage and loss of lives, as shown in Table 3.Furthermore, it is estimated that over 75% of earthquakefatalities over the last century were a result of buildingscollapse [6].

2.2. Brick masonry buildings

Bricks are extremely common in house constructionand have been used in this way since as early as 3500 BC.Brick masonry buildings consist of fired brick units bonded

ross the globe, extracted from De Sensi [3].

the globe, extracted from De Sensi [3].

Table 3Effects of earthquakes on adobe buildings [5].

Earthquake Fatalities Adobe buildings damaged or collapsed People affected

El Salvador, 2001 1100 150,000 1.6 millionSouthern Peru, 2001 81 25,000 –

Iran, 2003 26,000 85% of infrastructure destroyed 100,000 left homeless

Fig. 4. Typical adobe structures in Argentina and India, respectively, extracted from Blondet and Garcia [5].

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–67 55

to one another with mortar. The strength of the brickdepends on the purity of clay used and the temperature atwhich it is fired. Mortar strength depends on the quality ofthe bonding agent used as well as the sand to bondingagent ratio. The ability of the mortar to adhere to bricks istremendously important for the wall0s ability to resist in-plane shear cracking during an earthquake, a failure modewhich will be discussed in Section 3 of this review. Brickmasonry housing is also vulnerable to collapse underseismic loading; one of the worst death tolls inflicted froman earthquake in China, 1976 caused the loss of approxi-mately 240,000 lives, mainly owing to the collapse of brickmasonry structures. Fig. 5 shows the severe damage ofURM building during Jabalpur earthquake, 1997.

2.3. Stone masonry buildings

Stone masonry buildings can be found in Mediterra-nean Europe, North Africa, the Middle East, India, Nepaland other parts of Asia. They are a common type ofconstruction in developing regions since they providelow-cost housing due to the local availability of thematerials constituting stone masonry buildings; a newhouse in India costs, on average, between $50–$90/m²[8]. Stone constructions can be separated into 2 differentgroups:

�

Rural housing – These are typically small dwellings thatare 1 storey tall and contain small openings for win-dows and doors. The buildings are usually built asisolated buildings, not sharing any common walls withother houses, and are used by a single family. The ‘khan’house is a common rural dwelling and consists of asingle-storey stone masonry outer frame surrounding atimber inner frame overlaid by thick flat roofs provid-ing good thermal insulation spread throughout ruralIndia and in Nepal also.

�

Urban housing – Houses in urban communities areoften used for varying purposes with commerciallyfunctioning ground-floors underneath domestic hous-ing floors. These buildings frequently share commonwalls.

These buildings tend to perform poorly in earthquakesas shown in Table 4 owing to the low strength of the stoneand mortar used and the lack of adequate wall connec-tions. The quality of local construction is often very lowdue to the lack of skilled engineers involved. For example,in Nepal over 98% of buildings are constructed by theowners following the advice of local craftsmen [9].

3. Common failure mechanisms of URM buildings

Construction methods and technologies are dependenton local conditions and level of engineering expertise aswell as demographic factors. For example, URM structuresare common in North America but there are building codesin place to regulate construction and determine how wellprotected structures are from seismic loading. In contrast,the rapid urbanization of countries such as Nepal hasresulted in a huge rise in demand for quick and cheaphousing without regulation or controls in place to ensurethey are built to survive earthquakes. From past earth-quakes, these types of buildings have been observed todemonstrate common failure mechanisms, which will beexplained in this section. However, junction failure pre-ceded by out-of-plane collapse is the prime type of failurein all the 3 mentioned categories of URM buildings.

3.1. Adobe buildings

The reason adobe housing is so vulnerable to collapseduring earthquakes is due to its heavy walls, which causea greater resultant force on the building caused fromthe lateral movement of the ground. In addition, adobe

Table 4Percentage of masonry buildings that have collapsed in recent earth-quakes [8].

Earthquake Location(magnitude)

% of masonrybuildings thatcollapsed

Year

Athens earthquake, Greece (M5.9) 8 1999Bovec earthquake, Slovenia (M5.6) 2 1998Maharastra earthquake, India (M6.5) 5 1993

(i)

(ii)

(iii)

(iv)

(v)

Fig. 6. Common failure mechanisms for adobe structures including:(i) separation of walls at corners; (ii) diagonal cracking in walls;(iii) separation of roofing from walls; (iv) vertical cracking in walls;(v) out-of-plane wall failure, extracted from CENAPRED [10].

Fig. 7. Shear cracks in unreinforced brick masonry building from the1993 Killari earthquake, extracted from D’Ayala [11].

Fig. 5. Unreinforced brick masonry before and after 1997 Jabalpur earthquake, extracted from Sinha and Brzev [7].

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–6756

structures lacks in ductility and are consequently verybrittle resulting in sudden catastrophic failures underseismic loading. Failure of adobe structures is caused byeither the separation of walls at the corners or separationof roofing from walls or more commonly cracking andsubsequent failure of walls. These modes of failure aredemonstrated in Fig. 6. Though, the disintegration of floors

and roofing can be due to local stress concentration due tomany other reasons. It is true if such buildings aresubjected to wave loading or tsunami loading then watermay cause substantial decrease in strength of adobeblocks. So, the initiation of the failure under such type ofloading may be initiated by failure of such blocks.

3.2. Brick masonry buildings

Some of the common modes of failure for brickmasonry buildings are

�

Failure of corner junction leading to out-of-plane col-lapse is the prime type of failure.

�

Shear cracks in walls, which tend to be initiated at thecorners of openings in wall as can be seen in Fig. 7 afterthe 1993 Killari earthquake.

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–67 57

�

Fig

Figdur

Out-of-plane failure in long spans where the walltopples as a result of poor connections between wallsand at roof-wall interfaces.

�

After walls have collapsed, this can lead to the disin-tegration of floors and roofing and in severe cases, totalcollapse of the building occurs.

These failures illustrate the importance of the followingfeatures of brick masonry buildings:

�

Bond between mortar and bricks – this is the maincharacteristic that will resist in-plane shear failures.

�

Connection between wythes of brick walls – this canprevent out-of-plane toppling.

�

Connection between walls at corners/junctions – thiswill prevent disintegration of the masonry at cornersand junctions, which is a common failure mechanismfor brick buildings as shown in Fig. 8.

�

Connections between walls and floors/roofing – this canhave a significant effect on the safety of the buildingduring seismic activity as the collapse of the roofing andfloors accounts for a high amount of the resultingfatalities.

The above types of failure are also identified duringmany historical earthquakes in Indian sub-continent andoutside India. Kern County (1952), Imperial Valley (1979),South Italy (1980), Bihar (1988), Loma Prieta (1989),

. 9. In-plane failure mechanisms of URM walls: (a) shear failure; (b) sliding fa

. 8. Failure of Junction of two orthogonal walls in URM buildinging 2011 Sikkim earthquake.

Uttarkashi (1991), Killari (1993), Northridge (1994), Kobe(1995), Jabalpur (1997), Chi Chi (1999), Chamoli (1999),Bhuj (2001), Andaman (2002), Sumatra (2004), Kashmir(2005), Pakistan (2005), Sikkim (2006), Sumatra (2007),Durgapur (2008), Haiti (2010), Chile (2010), Sikkim (2011),Christchurch (2011) and Tohoku (2011) are a few to namein this context. Further, details in this regard are availablein the well accepted literature [12–30].

3.3. Stone masonry buildings

During seismic activity, common failure modes thatoccur in stone masonry are

�

ilur

De-lamination: It is typical for a stone masonry houseto have two external walls with loose rubble infillbetween for increased thermal performance but whenthese walls are insufficiently attached to one anotherthrough the use of ‘through’ stones, they disintegrateand crack during the lateral motion induced by anearthquake.

�

In long-span walls, overturning can occur in out-of-plane. � When connections between adjacent walls are of

adequate strength, the in-plane shear resistance ofthe wall is mobilized after which shear cracks develop.The various mechanisms for in-plane failure are shownin Fig. 9.

�

Junction failure leading to out-of-plane collapse is alsoobserved in many past earthquakes.

4. Existing URM retrofitting technologies

4.1. Introduction

For the reasons discussed previously, an increasingamount of research has been carried out investigatingthe retrofitting of existing URM buildings in order toincrease their collapse time under seismic loading andthereby reduce the loss of lives resulting from sudden,catastrophic building collapses. This section will collatethese methods and explore current research into theirperformance.

e; (c) rocking failure; (d) toe crushing, extracted from Macabuag [31].

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–6758

4.2. Surface treatment

4.2.1. General surface treatmentThe surface treatment process involves constructing a steel

or polymer mesh around the building exterior, which is thencoated with a layer of high-strength mortar. Alternatively,shotcrete can be applied to the wall surface. The steelreinforcement can have a reinforcement ratio between 3and 8% (which means the ratio of the weight of steelreinforcement used to that of masonry may vary in a rangeof 3–8%) depending on the loading resistance required.

This method is beneficial for use in developing areas asit is possible to use fly-ash or rice-husk ash to replace up to20% of the mortar, thereby reducing costs. It can also beperformed with unskilled workers. The system helps toconfine the masonry after cracking has occurred and it hasbeen shown to moderately increase the structure0s ulti-mate lateral load resistance. Surface treatment alsoimproves the out-of-plane resistance of masonry buildingsas it increases the height-to-thickness ratio of the wallsand thereby reducing any ‘arching action’. The disadvan-tages of surface treatment are that the lack of ‘breathing’ ofthe existing wall can lead to its deterioration and thearchitectural effects are significant.

4.2.2. Application of shotcreteElGawady et al. [32] carried out tests upon URM walls

and the retrofitted walls. In case of retrofitted walls, the

Table 5Shotcrete testing results [32].

Reference specimen Single-side reinforcem

Initially, flexural cracks formed, progressivelyworsening until rocking started (at 0.7% drift)

At 75% of ultimate lo

Sliding occurred followed by cracking in wall toes Small flexural crackshorizontally in shotcr

Average lateral strength of 35.5 kN Separation of shotcre

Fig. 10. Application of shotcrete to URM wall, extracted from ElGawadyet al. [32].

reinforcement arrangement was varied; in the first test,reinforcement and shotcrete were applied to one side ofthe wall and in the second, the same amount of reinforcingmesh was distributed evenly on both sides of the panel.

The tests were prepared in the following way:

�

en

ad, f

formetete la

4 mm diameter shear dowels were fixed into pre-drilled holes in the wall at approximately 250 mmcenters.

�

The reinforcement mesh was fixed into place in 2different configurations.

�

After wetting the surface of the wall panel, shotcretewas applied as illustrated in Fig. 10.

During testing, lateral loading was applied in increasingincrements up to failure of the wall. The results are summar-ized in Table 5. In both of the retrofit tests, the ultimate lateralload resistance of the walls was increased by a factor ofapproximately 3.6, and though the initial stiffness was unaf-fected by the surface treatment, the stiffness at peak loadingwas increased by a factor of 3. Further, it can be concludedthat the use of shotcrete significantly increases the ultimateload carrying capacity of the retrofitted models. However, thelimitations regarding the use lie with the facts that, use ofshotcrete is time consuming, reduces available spaces, createsdisturbance in occupancy and affects the esthetics.

4.3. Stitching and grout/epoxy injection

This method involves injecting grout or epoxy into thewalls in order to fill any voids or cracks that have formeddue to the deterioration of the building. In addition,existing cracks can be ‘stitched’ together using steel tiesand mortar. These techniques can restore the initial stiff-ness of the wall. When using epoxy injection, the stiffnessincrease has been shown to be less dramatic than theincrease in strength [33]. Further, the popularity of the useof this technique is because of minimal cost, availability ofmaterial and ease of implementation without requiringmuch technical rigor. The most important aspect of its vastuse lies with the fact that it is sustainable. In addition toinitial stiffness, this method is also able to restore theinitial strength of masonry. However, the technique will besuccessful only if the mechanical property of the mix andits physical chemical compatibility with the masonry to beretrofitted is achieved.

t Double-side reinforcement

irst cracks occurred Hair cracks formed at load of 53.8 kN

edand propagated

Visible cracking occurred at 82.9 kN

yer from foundations Flexural cracking followed by rockingbehavior until failure at 126 kNToe and heal heavily damaged

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–67 59

4.4. Re-pointing

In cases where bricks are of good quality but the mortaris poor, the mortar can be replaced to some extent with ahigher strength bonding material. However, Tetley andMadabhushi [34] found that the addition of 2% OrdinaryPortland Cement to the mortar made little or no differenceto the ultimate acceleration resistance. The advantages inthe use of this technique may be enlisted as minimal costand convenience of implementation. On the other hand,this method is not sustainable and the success of thistechnique lies with the compatibility of the new mortarwith that of existing bricks.

4.5. External reinforcement

Various external reinforcements which may be used forimproving the strength are discussed briefly under thefollowing subsections.

4.5.1. Bamboo reinforcementThe current method of reinforcement using bamboo is

to use it as part of a system involving buttresses, ringbeam, internal vertical reinforcement and horizontal inter-nal reinforcement. It has been shown that this systemincreases the collapse time of adobe structure but has littlecapacity to prevent cracking at low intensity groundmotions. It was proposed by Dowling et al. [35] that thesame partnership of ring beam and bamboo reinforcement

Fig. 12. Retrofitting of adobe wall using

Fig. 11. Bamboo reinforced wall with ring beam, extracted from Dowlinget al. [35].

could be used with vertical reinforcement being externallyfixed post-construction (Fig. 11). By installing verticalreinforcement after wall construction, complications suchas alignment of the reinforcement and trimming of thebricks are avoided. Horizontal chicken wire mesh was usedin one of the models alongside the bamboo and ring beam.During testing, all reinforced structures survived up to a100% increase in displacement intensity. Better reinforcedmodels survived up to a 125% increase and one heavilyreinforced model was even up to 400%.

4.5.2. Seismic wallpaper or glass fibre-reinforced polymer(GFRP) reinforcement

Testing of this technique was carried out on severalmodels of URM walls with solid clay bricks and lowstrength mortar [36]. All specimens were retrofitted withvertical composite strips which were bonded on both facesof the wall using epoxy resin. Each wall was then subjectedto out-of-plane loading and displacement via an airbagloading system. It was found that the ultimate flexuralstrength of all retrofitted specimens was increased.

Further, research by Man [37] has shown that GFRPs aremost effective when arranged at 451 to horizontal, asdemonstrated in Fig. 12. Diagonal shear compression testsand out-of-plane bending tests showed that this retrofitdoubled the tensile strength of the wall at yield andincreases its ductility up to 7.5 times its original value.

Many researchers have carried out experiments toinvestigate the efficiency of different types of FRP rehabi-litation techniques for enhancement of the seismic resis-tance of the masonry walls. Some of the majorcontributions in this regard are available in [38–51]. Thesestudies revealed that there could be significant increase inultimate load carrying capacity, energy absorption andductility in case of retrofitted walls as compared withun-retrofitted URM walls. The catastrophic failure patternin URM walls could be improved by a gradual failurepattern by retrofitting the URM walls with FRP compositematerial.

4.6. Strengthening of junction of URM walls by usingL-shaped reinforcement and poly-propylene (PP) band

A recent study (Fig. 13) has been carried out by Duttaet al. [52] to investigate the performance improvement

GFRPs, extracted from Man [37].

Fig. 14. Constructing the post-tensioning strap, extracted from Turer et al. [53].

Table 6Performance of retrofitted specimens compared to unreinforced wall[53].

Horizontalreinforcement

Verticalreinforcement

Horizontal and verticalreinforcement

70% increase infailureacceleration

40% increase infailure acceleration

More than 10% increasein failure acceleration

Fig. 13. Strengthening of junction using L-shaped steel bars and PP band, extracted from Dutta et al. [52]. (a) Junction reinforced with L-shaped steel barsand (b) Encasing junction by PP band.

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–6760

in URM by strengthening the junction, the most vulnerablepart of a masonry wall, using L-shaped reinforcement andPP band. For L-shaped reinforcement, steel bars of dia-meter 8 mmwas used in alternate layers to strengthen thejunction. For the method of strengthening using PP band,walls were encased with horizontal and vertical PP bandsat regular interval depending on the size of the wall. Boththe cases were subjected to unidirectional sinusoidal

shaking. For similar wall models tested in shaking table,it was observed that the URM wall failed at an accelerationof 0.25g, whereas, the failure of the walls strengthenedusing L-shaped reinforcement and PP bands was at 0.60gand 0.80g, respectively. These experiments have clearlyshown about 3 fold increases in strength due to theretrofitting options adopted.

4.7. Post-tensioning using rubber tyres

Post-tensioning of URM walls using scrap rubber tyres,which have an embedded steel mesh, has been proposedas a potential retrofitting solution. The tyres are assembledas shown in Fig. 14. To alleviate stress concentrations atwall corners, half cylindrical wooden logs are placed forthe rubber to mold around. After installation, the rubberstretches and looses its post-tensioning effect, and sothe bolts are tightened after 3 days to ensure enoughcompressive force is applied.

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–67 61

Testing was carried out by Turer et al. [53] on a 1:10scale model using a simple shake table. Testing was carriedout using the models that were unreinforced, horizontallyreinforced, vertically reinforced and reinforced in bothdirections, respectively. The results are shown in Table 6.The addition of post-tensioning improves the failuremechanisms of the models; cracks were less pronouncedand better spread and the brittle failure usually associatedwith these buildings changed into a more ductile responseso that the roof did not collapse. The combined horizontaland vertical reinforcement resulted in a foundation failurerather than in shear.

4.8. Confinement

This building method as shown in Fig. 15, involvesintroducing reinforced tie columns that confine the build-ing walls at all corners and intersections [33]. Europerecommends its use in EC8 and in China; it is used innew construction as well as retrofitting existing buildings.Research carried out by Paikara and Rai [54] has investi-gated the performance of half-scale masonry walls (con-fined using reinforced concrete ties) under cyclic loading.The tests demonstrated that confining wall sections into

Fig. 16. Geo-grid and soft fence mesh, respect

Fig. 15. Reinforced tie columns containing masonry wall panels,extracted from ElGawady et al. [33].

smaller elements improves the in-plane deformability andenergy dissipation of the walls. Actually, the masonrybeing confined by reinforced concrete frame may behaveeffectively as a diagonal compressive strut under lateralloading and may improve the behavior considerably ascompared to what is observed in case of a bare frame(frame containing concrete columns and beams only).

4.9. Mesh reinforcement

4.9.1. Polymer mesh reinforcementTwo types of polymer mesh that have been used to

retrofit URM structures are an industrial geo-grid and aweaker mesh that is usually used as a ‘soft’ fence onconstruction sites as shown in Fig. 16. In a study carriedout by Blondet et al. [55] the mesh was wrapped aroundthe wall and then coated with a mud plaster finish. Threevariations of the geo-grid (100%, 75% and 50% area cover-age) and one soft mesh system (80% area coverage) weretested. Each specimen was dynamically tested using a uni-directional shake table. Table 7 summarizes the results.In fact, while going through Table 7, it may be easy tounderstand the relative performance of soft mesh anddifferent types of geo-grid.

4.9.2. PP packaging strip mesh reinforcementThis method of reinforcement uses PP packaging strips

that can be found with many packaged items. The stripsare intertwined to produce a mesh that is then attached tothe wall by drilling through it and using ties. Testing understatic loading had been carried out by Macabuag [31], inwhich two scaled wall sections were constructed (withone retrofitted) and tested in a diagonal compressionmachine (Fig. 17). It was found that the horizontal stripsprevent separation of bricks on the same row. Verticalbands increase the frictional resistance between rows andconsequently prevent sliding. In conclusion, this methodeffectively improves the shear resistance of test specimensunder static loading but there was a recurring problem ofthe mesh snapping at points of stress concentration suchas the wall corners. A similar test was carried out at TokyoUniversity by Meguro et al. [9] under dynamic loadingconditions. The testing showed how the retrofit improved

ively, extracted from Blondet et al. [55].

Fig. 17. Non-retrofitted and retrofitted wall panel, extracted from Macabuag [31].

Table 7Performance summary of polymer mesh reinforcement [55].

Meshtype

Area covering(%)

Results

Geo-grid 100 Behaved as a rigid body with only small cracks developing beyond deflections of 80 mmOne wall slides from its foundation during deflections of 100 mm without any significant damageBeyond deflections of 120 mm walls showed signs of torsional response, sliding at the base and additional cracking

Geo-grid 75 Vertical cracks at the corners and diagonal cracks on the longitudinal walls developed beyond deflections of 80 mmCracking continued beyond deflections of 130 mmThe wall was able to keep its integrity as the mesh provided displacement control and a means of distributing stresses

Geo-grid �50 (critical) Cracks initiated at deflections of 80 mm yet were much largerDuring shaking at 130 mm deflection, severe structural damage was sustainedThough collapse was averted it was clear that insufficient reinforcement was provided

Soft mesh 80 Small cracking initiated at deflections of 80 mmMajor cracks were observed at deflections of 130 mmBeyond deflections of 130 mm the wall had broken into several large pieces and was only held together by the meshIn places the mesh had deformed or snapped indicating that the amount provided was barely adequate

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–6762

the specimen seismic performance significantly, displayingincreased load resistance and ductility before failure.

Macabuag et al. [56] studied the effect of retrofitting ofURM buildings in Nepal using PP meshing, being subjectedto artificially generated strong shocks. The study con-cluded that, use of PP mesh prevents spoiling out themasonry blocks and thus enables the system to accom-modate more deformation without collapse.

4.9.3. Steel reinforcement in PeruAfter extensive testing at the laboratories of the Ponti-

fical Catholic University of Peru (PUCP), it was decided bythe Regional Seismological Center for South America(CERESIS) to reinforce some adobe buildings with steelwire mesh, between 1994 and 1999 [57]. This steel meshwas applied externally at critical locations of adobe wallssuch as at corners and free ends. This was then coveredwith a layer of mortar. A large earthquake (Mw¼8.0)occurred in Pisco, Peru on August 15, 2007 resulting in519 deaths, the collapse of over 70,000 houses and theserious damage of more than 33,000 houses. However, 5houses in Ica that had been reinforced in 1998 survived theearthquake without suffering any damage. Major crackingor complete collapse occurred in adjacent buildings buteven the partially reinforced buildings performed betterthan the unreinforced brick and adobe buildings of Peru.

Furthermore, in 2001, an earthquake occurred (Mw¼8.4)in South Peru with similar results, shown in Fig. 18.

An experimental study using steel strips has beencarried out by Taghdi et al. [59] to retrofit low-risemasonry walls. It has been reported from the study thatsteel strips are effective in increasing the in-planestrength, ductility and energy dissipation capacity. It hasalso been observed that the anchor bolts along the verticalsteel strips help to eliminate the premature buckling ofmasonry walls.

4.9.4. Comparing various methods of mesh reinforcementExperiments were carried out by Tetley and Madab-

hushi [34] to compare various retrofitting methods includ-ing introducing steel reinforcing bars, a steel mesh cagearound the wall and a similar mesh formed from plasticcarrier-bags (Fig. 19). All tests were made with 1:5 scaleadobe walls consisting of gravel, simulating adobe blocksand mortar made from kaolin clay and sand. The controlwall failed at an acceleration of 0.32g. The steel cagingmethod was tested by constructing a 13 mm�13 mm netmesh and then constructing the wall within it. Mortar wasthen plastered over the steel mesh. The results showedthat there was significant increase in ductility demon-strated by large deformations before failure occurred andan increase of acceleration resistance by a factor of 3.

Fig. 19. (a), (c) and (e) shows various retrofitting methods carried out and (b), (d) and (f) show their effect on test walls, respectively, extracted from Tetleyand Madabhushi [34]. (a) Unreinforced control wall, (b) Failure of adobe wall around reinforced corner section, (c) Steel mesh cage, (d) More containedfailure of mesh-reinforced wall, (e) Plastic carrier-bag net and (f) Increased wall ductility using plastic bag mesh.

Fig. 18. Reinforced and unreinforced neighboring houses after 2001 earthquake, extracted from Bartolomé et al. [58].

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–67 63

The final failure, at an acceleration of 1.02g, was due to aseparation of the mesh from the base.

A similar mesh was installed but using plastic carrier-bags instead of steel. The bags were cut into 20 mm strips,braided together and then used to form a 50 mm�50 mmsquare mesh, which was fixed to the wall using simpletacks. Three courses of masonry were then built on top ofthe mesh to fix it properly, after which plaster usingmortar was applied. Testing showed increased ductilityand tensile strength of the wall. The transverse wall failedat 0.64g and then total in-plane failure occurred at 1.02g(double the control wall resistance).

5. Summary and conclusions

The study begins with broad idea about the casualtiesdue to past earthquakes in about last 100 years discussedthrough presentation of Table 1. Following such discus-sions, historic evolution of various earthquake resistingpractices have discussed and presented in Table 2.The importance of failure of adobe and masonry buildings

is then discussed to show that such buildings form a largepart of the habitat of the ordinary people. Thus, failure ofsuch building may have great impact on human civiliza-tion (as may be well understood from Tables 3 and 4).So, need of improving performance of such buildingsbecomes an issue of great importance to civil and struc-tural engineers involved in infrastructure building forordinary people. With this background discussion, variousways of improving the performance of masonry buildinghave been evaluated, summarizing their features inTables 5–11. Such studies highlighting overall structuralperformance, economic aspects and also the aspect ofsustainability and ease of construction may lead to follow-ing broad conclusive understanding about state-of-the-artregarding seismic performance of masonry building.

(A)

From these assessments it can be concluded that themesh retrofitting techniques and the softer methodsas listed here are the most feasible for application indeveloping countries. They are namely, re-pointing,stitching and grout/epoxy injection, polymer mesh

Table 9Discussion of retrofitting technologies according to economic criteria.

Retrofit method Economic features

Surface treatment (shotcrete) Requires purchase of shot-creting system including pressurized hoses and pump, although initial installationcosts could be offest by retrofitting many houses. Also requires manufacturing of concrete and use of steelreinforcement mesh and the inclusion of at least one experienced worker, making it too expensive forapplication in poor communities

Stitching and grout/epoxy injection Minimal costs as such products are readily available and can be easily appliedRe-pointing Minimal costs as only required the manufacturing of a stronger mortarBamboo reinforcement For 1000 sq. ft house, cost of reinforcing is $225, in comparison to $400 when using steel reinforcementSeismic wallpaper ‘E’-glass costs $2–$4 per kg. Field experiments have shown that retrofitting with a composite material is cost-

effective [60]Post-tensioning (rubber tyres) Combined cost of scrap tyres and connectors is �$0.6/m². This can be further reduced by mass production of

connectorsConfinement Cost-effective for application in new building, costing little in comparison to the overall construction costs. As a

retrofit, requires demolition and reconstruction of wall sections making it uneconomical for this purposePolymer mesh reinforcement Industrial geogrid Mesh cost: $2/m² Application cost: $19/m²

Soft polymer mesh Mesh cost: $0.5/m² Application cost: $4/m²PP strip reinforcement Demo by Tokyo University and JICA (Japanese International Cooperation Agency) in Kashmir following the

major 2005 earthquake, performed a retrofit at an estimated 5% total cost of house. Actual cost, whenimplemented by rural masons is expected to be much lower [31]

Steel mesh cage For 1000 sq. ft house, costs $400 when using steel reinforcementPlastic carrier bag mesh This retrofitting technique would cost very little due to the availability of plastic carrier-bags and the cheap cost

of mortar to apply to the wall surfaceL-Shaped reinforcement Adds only 5–10% to the total cost

Table 8Summary of previous research into URM retrofitting techniques.

Technology Conclusions

Surface treatment using Shotcrete Increased lateral load resistance by factor of 3.6 under static loading as well as ductility of wallStitching and grout/epoxyinjection

Grout or epoxy injected into walls to fill voids or cracks. Also, existing cracks can be ‘stitched’ together using steelties and mortar. Can restore the initial stiffness of the wall

Re-pointing with 2% ordinaryPortland cement

No noticeable improvement in performance under dynamic loading

Bamboo Uncertainty about performance of vertical internal reinforcement at low earthquake intensitySeismic wallpaper Shown to be effective at increasing wall resistance to lateral acceleration. Optimum arrangement of GFRPs was at

451 to horizontal, which doubled tensile strength at yieldPost-tensioning using rubber tyres Combining horizontal and vertical reinforcements improved the lateral acceleration resistance of the masonry

models by 100% and increased their ductilityConfinement Confinement improves in-plane deformability and energy dissipation of the wallsCenter core Placing a grouted core in the center of walls can double ultimate lateral load resistance. Creates zones of varying

stiffness/strengthPolymer mesh Can prevent partial or total collapse of brick-masonry building during earthquake. Placing mesh at critical wall

locations may improve efficiency of methodPolypropylene packaging strips Demonstrated improved shear resistance of brick walls under static loading and significant increase in

performance under dynamic loadingSteel mesh cage Improved lateral acceleration resistance and increased ductility of adobe wall under testing. Proven to be highly

effective in adobe houses, surviving two major earthquakes in Peru with little or no damagePlastic carrier bag mesh Similar performance to steel mesh cageL-Shaped reinforcement Increases the resistance by preventing junction failure

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–6764

reinforcement, polypropylene packaging strap reinfor-cement, steel mesh cage, and L-shaped reinforcement.

(B)

(C)

There are some methods in practice or under researchthat are also suitable for implementation in develop-ing communities but with some disadvantages. Thesemethods include the following:� Bamboo reinforcement – There may be difficulties

in sustainably sourcing good quality bamboo. Therelies a need for a good understanding of the meth-odology of bamboo processing. In fact, passing onsuch skills to local workers could be difficult.

� Plastic carrier bag net – Unless the manufacture ofthe mesh is industrialized, the time taken to

construct it is a barrier for the application of thistechnique.

� Post-tensioning using rubber tyres – The pooresthetic appearance is the main disadvantage forthis method.

The least suitable retrofitting methods researched are� Surface treatment using shotcrete – Difficulties

were experienced during application in the labora-tories that are likely to translate to real worldsituations.

� Seismic wallpaper – Although they significantlyimprove structural performance, the need for qual-ity control in manufacture and application means

(D)

Table 11Discussion of retrofitting technologies according to buildability criteria.

Retrofit method Buildability features

Surface treatment (shotcrete) Application requires trained specialists. Evenwhen applied by experienced workers, there have been difficultiesnoted in the application process such as excessive dust and noise, which would make it very disruptive foroccupants

Stitching and grout/epoxy injection Little technical knowledge required and materials can be easily transported, applied and removedRe-pointing Little technical knowledge required. However, some sort of temporary works may be needed to support the

structureBamboo reinforcement If fixed to wall exterior, method is easily buildable. If horizontal reinforcement is tied on the exterior of the wall,

it will overlap openings causing practical and esthetic problems. Mud bricks surrounding the bamboo will notprovide adequate protection against water intrusion and also makes maintenance/inspection of bamboodifficult. Installation is quick to learn for local builders but they need to understand the key earthquakeengineering concepts involved

Seismic wallpaper Manufacture of glass reinforced fibers must be under strictly controlled conditions. Retrofit is easy to apply andinspect/maintain but removal is extremely difficult

Post-tensioning (rubber tyres) Retrofit is reasonably simple but as straps protrude away from walls, they are difficult to mask with plasterresulting in significant architectural impact

Confinement Application is very intrusive requiring demolition of wall sections. Also, requires the use of tie-beams to beeffective

Polymer mesh reinforcement Industrialgeo-grid

Tough nature of material and lack of flexibility makes applicationand removal of this method difficult

Soft polymer mesh Mesh can be easily deformed so transportation, application and removal are easyPP strip reinforcement Retrofit is simple enough for application by local craftsmen and homeowners without any prior knowledge/

expertise on earthquake engineering [9]Steel mesh cage Steel cage can be attached to a building if specific instructions are given, but this is not ideal and if problems

arise, technical knowledge is required to solve them. Demolition of the steel mesh presents difficultiesPlastic carrier bag net The materials used are light and flexible. However, it takes a very long time to construct the mesh, which is a

barrier for this method unless the mesh manufacture can be industrialized/streamlined in some wayL-shaped reinforcement No special expertise is required for using this

Table 10Discussion of retrofitting technologies according to sustainability criteria.

Retrofit method Sustainability features

Surface treatment (shotcrete) Use of concrete in small quantities (i.e. not taking advantage of thermal mass), is unsustainableStitching and grout/epoxyinjection (re-pointing)

Minimal amount of material required and very ‘light’ application make these methods sustainable

Bamboo reinforcement Provided bamboo is responsibly sourced, this method is sustainable. Requires very little processing and can beeasily disposed of at the end of its design life

Seismic wallpaper GFRPs contain a large amount of embodied energy. Also, there is a danger of harmful fumes being released if thematerials were burned in a fire. However, GFRPs are non corrosive and extremely durable and so have a longdesign life

Post-tensioning (rubber tyres) Considered a sustainable solution as it involves re-using a resource otherwise wasted; 190 million tyres a yearare sent to landfill in the U.S.A. This would reduce the need for large tire fires that pollute the environment [61]

Confinement This is a relatively ‘heavy’ engineering solution and therefore requires materials and methods that areintrinsically unsustainable

Polymer mesh reinforcement Polymer mesh requires use and processing of petrochemicals, which is not sustainable unless sourced fromrecycled or re-used units

PP strip reinforcement There is plenty of scope for re-use or recycling due to the massive volume of consumer goods that are protectedby these packaging strips

Steel mesh cage The manufacture of steel is unsustainable and disposal at the end of the design life will be difficult in adeveloping community

Plastic carrier bag net Approximately 17.5 billion plastic carrier bags are thrown away in Britain every year. Re-using these bags istherefore considered a sustainable solution [62]

L-Shaped reinforcement Use of reinforcement will not have any adverse effect from the view point of sustainability

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–67 65

that widespread retrofitting in poor communitieswill be difficult.

� Confinement – It is more suited to new construc-tion as opposed to retrofitting where demolitionand rebuilding are required.

Corner/junction failure leading to out-of-plane collapse inbrick masonry buildings is the most common type offailure. Intuitively it could be suggested that simultaneousapplication of strengthening mechanisms arresting joint

separation as well as providing better integrity to thewalls, would be the suitable one. One such combinationcould be the use of PP band and horizontal L-shapedreinforcing bars while the other might be with combina-tions of horizontal L-shaped reinforcing bars and wiremesh. L-shaped horizontal reinforcing bars will beeffective in arresting corner/junction failure. However,PP band and wire mesh will be beneficial in improvingthe overall integrity providing some strength against

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–6766

bending resulting in better resistance towards out-of-plane failure. However, such combinations may be easierfor applications without requiring any specialized skills.

This review has outlined the problems posed by earth-quakes for developing counties, particularly with regard tostrike a balance between the economy, safety and qualityof life in areas such as Latin America, Peru, Nepal andIndian subcontinent. Primarily, URM buildings affected byearthquakes were highlighted in this study. These includeadobe, stone and brick masonry buildings. The commonmechanisms of failure of these types of buildings wereexplored and remedial measures with their suitability,ease of application and economic viability as well assustainability has been presented in details. The presentstudy thus, not only becomes a literature helping to choosesuitable existing retrofitting techniques for various adobe,stone and brick masonry buildings but also helps indisseminating the existing state-of-the-art knowledgewithin policy makers, architects, designers and civil engi-neering community in general. It may help in boostingfurther research in improving some of the strengtheningstrategies mentioned here, too.

Acknowledgments:

The first author would like to thank the project stu-dents [Josh Macabuag, Louisa Man, Andrew Smith, ThomasRedman, Kate Humphreys and Luke Hobbs] over the yearsfor contributing to the study. Some of the tests reportedwere carried out at the BLADE [Bristol Lab for AdvancedDynamic Engineering] laboratories.

References

[1] ⟨http://denali.gsfc.nasa.gov/dtam/seismic⟩ [accessed March, 2012].[2] Bhattacharya S, editor. Design of foundations in seismic areas:

principles and applications, NICEE, ISBN: 81-904190-1-3, 2007.[3] De Sensi B.Terracruda B. soil dissemination of earth architecture;

2003. (⟨www.terracruda.com/%20architetturadiffusione.htm⟩)[accessed March, 2012].

[4] Houben H, Guillaud H. Earth construction – a comprehensive guide.ITDG Publishing.1994.

[5] Blondet M, Garcia GVM. Adobe Construction, WHE Housing Report,Catholic University of Peru, Peru; 2003.

[6] Papanikolaou A, Taucer F. Review of non-engineered houses in LatinAmerica with reference to building projects and self-constructionprojects. Eur Lab Struct Assess 2004 (EUR21190EN).

[7] Sinha R, Brzev SN. Unreinforced brick masonry building withreinforced concrete roof slab. World Housing Encyclopedia. 2002.Report 21, ⟨http://www.world-housing.net/⟩ [accessed March, 2012].

[8] Lutman M. Stone Masonry Construction, WHE Housing Report,Slovenian National Building and Civil Engineering Institute, Slove-nia; 2002.

[9] Meguro K, Mayorca P, Sathiparan N, Guragain R, NasrollahzadehNesheli K. PP-Band retrofitting technique: affordable, acceptable andfeasible method for developing countries, Institute of IndustrialScience. University of Tokyo; 2005.

[10] CENAPRED (Centro Nacional de Prevencion de Desastres), Metodosde Refuerzo para la Vivienda Rural de Autoconstruccion, PrimeraJornada Nacional De Simulacros Para La Prevencion de Riesgos,Mexico; 2002.

[11] D’Ayala D. Unreinforced brick masonry construction, WHE housingreport. United Kingdom: University of Bath; 2002.

[12] Jain SK. On better engineering preparedness: lessons from the 1988Bihar earthquake. Earthquake Spectra, EERI 1992;8(3).

[13] Bruneau M. State-of-the-art report on seismic performance ofunreinforced masonry buildings. J Struct Eng, ASCE 1994;120(1):230–51.

[14] Jain SK, Singh RP, Gupta VK, Nagar A. Garhwal earthquake of October20, 1991. EERI Special Earthquake Report, EERI Newsletter February1992; 26(2).

[15] Jain SK, Murty CVR, Chandak N, Seeber L, Jain NK. The September 29,1993, M6.4 Killari, Maharashtra earthquake in central India. EERISpecial Earthquake Report, EERI Newsletter January 1994; 28(1).

[16] Bruneau M. Performance of masonry structures during the 1994 North-ridge (Los Angeles) earthquake. Can J Civ Eng 1995;22:378–402.

[17] Bruneau M, Yoshimura K. Damage due to masonry buildings causedby the 1995 Hyogo-ken Nanbu (Kobe, Japan) earthquake. Can J CivEng 1996;23:797–807.

[18] Jain SK, Murty CVR, Arlekar JN, Rajendran CP, Rajendran K, Sinha R.Chamoli (Himalaya, India) earthquake of 29 March 1999. EERISpecial Earthquake Report, EERI Newsletter July 1999; 33(7).

[19] Jain SK, Murty CVR, Dayal U, Arlekar JN, Chaubey SK. Learning fromearthquakes: a field report on structural and geotechnical damagessustained during the 26 January 2001 M7.9 Bhuj earthquake inWestern India. Department of Civil Engineering, Indian Institute ofTechnology Kanpur, Kanpur 208016, India, 2001.

[20] Rai DC, Murty CVR. Reconnaissance Report on North Andaman(Diglipur) Earthquake of 14 September 2002. Department of CivilEngineering, Indian Institute of Technology Kanpur, Kanpur 208016,India, 2003.

[21] Jain SK, Murty CVR, Rai DC, Malik JN, Sheth AR, Jaiswal A. Effects ofM9 Sumatra earthquake and tsunami of 26 December 2004. Curr Sci2005;88(3):357–9.

[22] National Information Centre of Earthquake Engineering (NICEE)Photo gallery 2005, ⟨http://www.nicee.org/photos.php⟩ [accessed09.03.13].

[23] Earthquake Engineering Research Institute (EERI) Special Earth-quake Report (2006). Learning from earthquakes – The Kashmirearthquake of October 8, 2005: Impacts in Pakistan, ⟨https://www.eeri.org/ lfe/pdf/kashmir_eeri_2nd_report.pdf⟩ [accessed 09.03.13].

[24] Dutta SC, Mukhopadhyay P, Bhattacharya S. Impact of moderateearthquakes in post-Bhuj era: case study of Sikkim 2006 andDurgapur 2008 earthquakes, India. In: Proceedings of 3rd Greece-Japan workshop: seismic design, observation and retrofit of founda-tions, Santorini, Greece 2009: p. 154–170.

[25] Earthquake Engineering Research Institute (EERI). Learning fromearthquakes: The MW 7.0 Haiti earthquake of January 12, 2010:Report #2. EERI Special Earthquake Report May 2010, Okland, CA94612-1934, USA.

[26] Kato H, Tajiri S, Mukai T. Preliminary reconnaissance report of theChile Earthquake 2010. Building Research Institute Japan, July 1,2010.

[27] Moon LM, Dizhur D, Ingham JM, Griffith MC. Seismic performanceof masonry buildings in the Christchurch earthquakes 2010–2011:A progress report. In: Proceedings of the Australian earthquakeengineering society 2012 conference; December 7–9, 2012, GoldCoast, Qld.

[28] Nishiyama I, Okawa I, Fukuyama H, Okuda Y Building damage by the2011 off the Pacific coast of Tohuku earthquake and coping activitiesby NILIM and BRI collaborated with the administration, ⟨www.pwri.go.jp/eng/ujnr/joint/43/paper/4_Nishiyama.pdf⟩.

[29] ⟨http://www.nicee.org⟩ [accessed March, 2013].[30] ⟨http://earthquake.usgs.gov⟩ [accessed March, 2013].[31] Macabuag J. An introduction to modelling and retrofitting of non-

engineered masonry buildings under seismic loading. MEngResearch Report, Oxford University; 2007.

[32] ElGawady MA, Lestuzzi P, Badoux M. Retrofitting of masonry wallsusing shotcrete, 2006 NZSEE Conference, Paper 45.

[33] ElGawady M, Lestuzzi P, Badoux M. A review of conventionalseismic retrofitting techniques for URM, 13 thIB 2 MaC. Holland:Amsterdam; 2004.

[34] Tetley R, Madabhushi G. Vulnerability of adobe buildings underearthquake loading, In: Proceedings of the international conferenceon earthquake geotechnical engineering, Greece 2007; June 25–28:Paper no. 1244.

[35] Dowling D, Samali B, Li J. An improved means of reinforcingadobe walls-external vertical reinforcement, SismoAdobe 2005;May 16–19th: PUCP, Lima, Peru.

[36] Ehsani MR, Saadatmanesh H, Velazquez-Dimas JI. Behavior of retro-fitted URM walls under simulated earthquake loading. J ComposConstr, ASCE 1999;3(3):134–42.

[37] Man L. Improving the earthquake resistance of non-engineeredadobe masonry buildings. Oxford University; 2008.

S. Bhattacharya et al. / International Journal of Disaster Risk Reduction 7 (2014) 51–67 67

[38] Gilstrap JM, Dolan CW. Out-of-plane bending of FRP-reinforcedmasonry walls. Compos Sci Tech 1998;58:1277–84.

[39] Corradi M, Borri A, Vignoli A. Strengthening techniques tested onmasonry structures struck by the Umbria-Marche earthquake of1997–1998. Constr Build Mater 2002;16:229–39.

[40] Kuzik MD, Elwi AE, Cheng JJR. Cyclic flexure tests of masonry wallsreinforced with glass fiber reinforced polymer sheets. J ComposConstr, ASCE 2003;7(1):20–30.

[41] Tan KH, Patoary MKH. Strengthening of masonry walls against out-of-plane loads using fiber-reinforced polymer reinforcement.J Compos Constr, ASCE 2004;8(1):79–87.

[42] Korany Y, Drysdale R. Rehabilitation of masonry walls using unob-strusive FRP techniques for enhanced out-of-plane seismic resis-tance. J Compos Constr, ASCE 2006;10(3):213–22.

[43] El-Dakhakhni WW, Hamid AA, Hakam ZHR, Elgaaly M. Hazardmitigation and strengthening of unreinforced masonry walls usingcomposites. Compos Struct 2006;73:458–77.

[44] Wang Q, Chai Z, Huang Y, Yang Y, Zhang Y. Seismic shear capacity ofbrick masonry wall reinforced by GFRP. Asian J Civil Eng (BuildHousing) 2006;7(6):563–80.

[45] Shrive NG. The use of fiber reinforced polymers to improve seismicresistance of masonry. Constr Build Mater 2006;20:269–77.

[46] Mosallam AS. Out-of-plane flexural behavior of unreinforced redbrick walls strengthened with FRP composites. Compos Part B2007;38:559–74.

[47] Prakash SS, Alagusundaramoorthy P. Load resistance of masonrywallettes and shear triplets retrofitted with GFRP composites. CemConcr Compos 2008;30:745–61.

[48] Willis CR, Yang Q, Seracino R, Griffith MC. Damaged masonry wallsin two-way bending retrofitted with vertical FRP strips. Constr BuildMater 2009;23:1591–604.

[49] Hamed E, Rabinovitch O. Failure characteristics of FRP-strengthenedmasonry walls under out-of-plane loads. Eng Struct 2010;32:2134–45.

[50] Maalej M, Lin VWJ, Nguyen MP, Quek ST. Engineered cementitiouscomposites for effective strengthening of unreinforced masonrywalls. Eng Struct 2010;32:2432–9.

[51] Roca P, Araiza G. Shear response of brick masonry small assemblagesstrengthened with bonded FRP laminates for in-plane reinforce-ment. Constr Build Mater 2010;24:1372–84.

[52] Dutta SC, Mukhopadhyay P, Goswami K. Augmenting strength ofcollapsed unreinforced masonry junctions: the principal damagefeature due to moderate Indian earthquakes. Nat. Hazards Rev., ASCE2012;. http://dx.doi.org/10.1061/(ASCE)NH.1527-6996.0000096.

[53] Turer A, Korkmaz SZ, Korkmaz HH. Performance improvementstudies of masonry houses using elastic post-tensioning straps.Earth Eng Struct Dyn 2007;36:683–705.

[54] Paikara S, Rai DC. Confining masonry using pre-cast R.C. element forenhanced earthquake resistance, in: Proceedings of the 8th U.S.National Conference on Earthquake Engineering, San Francisco,California, USA, April 18–22, 2006.

[55] Blondet M, Torrealva D, Vargas J, Velasquez J, Tarque N. Seismicreinforcement of adobe houses using external polymer mesh,September 2006, Paper no. 632.

[56] Macabuag J, Guragain R, Bhattacharya S. Seismic retriofitting of non-engineered masonry in rural Nepal. In: Proceedings of theICE – Structures and Buildings; 2012. doi: 10.1680/stbu.10.00015.

[57] Bartolomé A.S., Quiun D., Zegarra L. Performance of reinforced adobehouses in Pisco, Peru earthquake. In: Proceedings of the 14thworld conference on earthquake engineering, Beijing, China,October 12–17, 2008.

[58] Bartolomé AS, Quiun D, Zegarra L. Effective system for seismicreinforcement of adobe houses. Proceedings of the 13th WorldConference on Earthquake Engineering, Vancouver, B.C., Canada,August 1–6, 2004, Paper no. 3321.

[59] Taghdi M, Bruneau M, Saatcioglu M. Seismic retrofitting of low-risemasonry and concrete walls using steel strips. J Struct Eng, ASCE2000;126(9):1017–25.

[60] ⟨http://www.netcomposites.com/education.asp?sequence=33⟩[accessed December, 2008].

[61] ⟨http://www.ridelust.com/190-million-tires-thrown-away-each-year-wheres-the-rubberized-asphalt-concrete-rac/⟩ [accessedDecember, 2008].

[62] www.thegreenstoreonline.co.uk/default.aspx/Page/122 [accessedDecember, 2008].