SESOC Journal

RETROFIT TECHNIQUES FOR SEISMIC IMPROVEMENT OF URM BUILDINGS

By: Cass GooDWin1, Garry Tonks2 and Jason inGhaM3
Tribute: Dr Garry Tonks died suddenly at home on 21 April 2009. Garry taught architecture at the University of Auckland for 18 years, was a past president of the New Zealand Timber Design Society, and most of all was a friend and mentor. abSTRaCT There are many complex considerations when seeking to seismically retrofit unreinforced masonry (URM) buildings. Strength requirements need to be evaluated, and the intrinsic strength of the existing materials should be maximised before adding any additional structure. Descriptions of current common techniques for improving the seismic response of URM buildings are outlined, and comments made as to their general appropriateness with regard to architectural and heritage principles. These techniques are explained and critiqued using examples to illustrate their merits or lack thereof. INTroDUCTIoN As highlighted by the 2010 Darfield earthquake, many older buildings, particularly those constructed from unreinforced masonry (URM), now require seismic improvement in order to withstand earthquake forces. There are many approaches and techniques to seismic improvement, the most common of which are outlined below and analysed in terms of the impact of each on the heritage and architectural value of the building. This review is intended as a continuation of an earlier article (Goodwin et al. 2009) which gave background information to furnish designers with information prior to considering seismic retrofit. This information included a simple process for determining the basic heritage value of a building and some guidelines for good building conservation practice. A heritage building is more than simply a collection of building materials. As buildings age, they become more permanent parts of local culture. They become associated with notable people; with the memories of the public; and places where important historic events occurred. They become unique, as buildings are simply no longer made the same way, and time provides them with characteristics that could never be designed into a new building1. Old buildings can become heritage, where they become important to the public rather than just to the owner and occupants; a recent example of this is the birdcage hotel in Figure 1. There is no line on one side of which lies heritage, and on the other not heritage; all old buildings will have some heritage value. The difficulty is in determining the degree of heritage value, and what is important and what is less important about a building.

Such as the patina of age - how a building looks old.

PAPER CLASS & TYPE: REFEREED

1 2

University of Auckland, BE, B Arch (Hons), M Arch (Hons) University of Auckland, B Arch, PhD, ANZCA University of Auckland Associate Professor, BE (Hons) PhD, MBA Journal of the Structural Engineering Society New Zealand Inc.

30

SESOC Journal

Figure 1. The Birdcage hotel prior to seismic improvement and being moved. (Dizhur) Once it has been accepted that seismic improvement of a building is required, it is necessary to determine the extent of required intervention and the appropriate methods for that particular building. Generally, an analysis is undertaken in stages, beginning with assessing the architectural character of the building, then addressing the basic known inadequacies of URM construction (Russell & Ingham 2010), then determining what types, if any, of additional retrofit systems are required. Additional requirements should be met by structural elements which are sensitively designed and in keeping with the architectural character of the building, of the least extent, and reversible if possible (Goodwin et al. 2009). Often, several or many different techniques will be used within one building, and it is vital to have an appreciation of how each will impact upon the architectural and historic aspects of the building, and the ways in which these impacts can be minimised. 1.0 OBJECTIVES

The objectives of seismic improvement are twofold; to protect the occupants from injury or death, and to protect the building itself from damage due to deficiencies in its design. The protection of occupants is facilitated by protection of the building, so the building itself is the focus here. The designer needs to consider the building as a functional but potentially deficient system which can be modified by various degrees in order to improve its performance. 1.1 Strengthening Requirements

Territorial Authorities require most buildings with strength of less than 33% of current requirements to be upgraded under The Building Act 2004. It is recommended that building performance be improved to at least 67% of current code requirement in AS/NZS 1170.0-5: 2004 (NZSEE 2006). Buildings between 33% and 67% are considered to be earthquake risks and retrofit should be seriously considered. Buildings over 67% are generally considered to be satisfactory, although are still up to 5 times more likely to suffer significant damage in an earthquake than a building meeting AS/NZS 1170 requirements (NZSEE 2006). Many unstrengthened (and some previously strengthened) URM buildings provide less than 67% of the strength of AS/NZS 1170 requirements, and many fall below the 33% minimum level, particularly in more seismically active parts of New Zealand (Russell & Ingham 2010). 1.2 Heritage and Conservation

When dealing with buildings with notable heritage characteristics, there are several objectives to be considered, particularly when undertaking a seismic retrofit (Robinson & Bowman 2000). These objectives are not rigid rules, but guidelines for achieving a sensitive solution, so there is always some leeway for alternative approaches and creativity. If the spirit of these guidelines is adhered to, generally a good outcome will be reached2. 1.2.1 Knowledge of important characteristics of the building A quick reading of the architectural character of the building will allow its key features to be identified, which can then be worked with to arrive at an appropriate solution (Goodwin et al. 2009). Seismic improvement systems should neither detract from the character or quality of the building, nor interfere overmuch with individual character-defining elements. When the important elements and aspects are known, care may then be taken to preserve these as much as possible. 1.2.2 All work should involve minimal intrusion Only add as much as is necessary and change as little as possible (Robinson & Bowman 2000). This will also aid in the economic impact of the retrofit work. This guideline has some degree of flexibility in that extra work may sometimes result in a better overall solution, so sound judgement is necessary. Heritage buildings represent a piece of history, so any changes may obscure or alter the ways that this history can be understood.

These guidelines are a heavily summarised reading of the ICOMOS New Zealand Charter for the Conservation of Places of Cultural Heritage Value. Refer to http://www.icomos.org.nz/NZ_Charter.pdf for further information.

Volume 24 No. 1 April 2011

31

SESOC Journal 1.2.3 Retrofit work should be reversible and efficient It must be accepted that current technology is imperfect, and that project requirements will almost never allow for the ideal solution to be implemented. Reversibility allows for the possibility of improved methods in the future, as well as allowing for the potential fallibility of the designer (Robinson & Bowman 2000). As confident as any designer might be that their solution is as good as possible, this may prove to be otherwise, and when dealing with something irreplaceable, it is prudent to err on the side of caution. 1.2.4 Retrofits must respect integrity and character Heritage buildings are called heritage because of the way that they are before they are retrofitted. The act of seismic improvement should not detract from the building as a whole, alter the way it is appreciated, or greatly damage it (Robinson & Bowman 2000). The job of the designer is to ensure that the building remains as part of our heritage, not to alter it to be something different altogether. Considering the above points, in conjunction with sound engineering judgement, should assist in the design of an appropriate strengthening system. Generally, the answers to individual questions are obvious, provided that the cues given by the building itself are noted. 2.0 INITIAL CONSIDERATIONS AND STABILISATION

Once a building has been identified as a seismic risk, a more thorough investigation is required. It makes sense at this stage to establish a detailed understanding of the buildings intrinsic seismic resistance, and assess whether simple measures which address basic inadequacies could be undertaken to give significant increases in strength (Russell, 2010). Maximising the existing strength will ensure that the existing building fabric is retained and utilised as much as possible and will minimise the extent of additional measures. Various techniques for addressing these basic issues prior to the addition of any large supplementary structure are outlined below. Unmodified URM buildings usually have a number of critical structural deficiencies which make them prone to earthquake forces, but these can often be addressed without significant alteration to the building fabric, resulting in a relatively large increase in strength (Robinson & Bowman 2000). The overarching problem is that URM buildings were simply not designed for earthquake loads, and while they can be made to perform adequately in an earthquake, they lack a basic degree of connection to allow all parts of the building to act together as one. 2.1 URM Material Stabilisation

URM deteriorates in the environment over time. Occasionally this deterioration will result in local failures and cracking which affect the overall effectiveness of the building. Various external forces can also cause cracking and damage in URM, such as dampness, subsidence, earthquakes, and impacts. Deterioration similar to that shown in fig. 2 can often be remedied by reinstatement and repointing of mortar3, but sometimes more substantial measures are required. There are various techniques for the repair of cracks, securing of lintels, and reinstatement of damage. Bonding agents such as grout or epoxy can be injected into the mortar; there are also several metal-based inserts such as

Figure 2. Severely degraded brick and mortar due to moisture. (Dizhur) shaped dowels, or reinforcing bars, which can be used to reinstate and strengthen the brickwork (Croci 1998). The visual impact of reinstatement and strengthening can be minimal if done carefully, and the result is potentially far superior to a cracked and broken faade. Such measures are often irreversible however, and care needs to be taken with colour matching and the concealment of holes drilled for inserting rods. Lintels and arches will sometimes require strengthening, particularly when these are constructed from URM. One of the best ways to achieve this is by using drilled and inserted rods which are grouted or epoxied into place. These rods provide the requisite tensile strength to the structural element while having little visual impact.

Lime mortars should always be repointed with new lime mortars. Mixing lime and Portland cement mortars can cause numerous problems.

32

Journal of the Structural Engineering Society New Zealand Inc.

SESOC Journal 2.2 Falling Hazards and Parapets

URM buildings will often feature numerous decorative elements built with brick and plaster which are important parts of its architectural character, such as parapets, chimneys, gable walls, and other, smaller, decorative features. In the past, some buildings had these elements removed wholesale, rather than being strengthened or secured. Parapets and chimneys are usually the first parts of a building to fail in an earthquake due to their low bending strength and high imposed accelerations (FEMA 547). Parapets in particular are simple to strengthen. Generally a steel section running along the length of the parapet which is fixed back to the roof structure behind is a suitable technique, if a little crude. The back of a parapet is not often seen, which makes the visual impact of this method low, and the steel section is bolted to the URM, which also has good potential for reversibility. Chimneys are more complex but can usually be strengthened using a combination of fixing them to the building diaphragms at each level and either strengthening the projecting portion or bracing it back to the roof structure with steel members in a similar way to parapets, or fixing steel sections to the sides to provide flexural strength. The technique shown in fig. 3 may be appropriate where the exterior has been plastered. Chimneys contribute to the architectural form of a building and often help define its roofscape, and as such should be preserved if possible. Other elements such as decorative plaster features on the face of a wall can be effectively fixed with a single bolted connection. Less secure elements, such as

Figure 3. Parapets strengthened using near-surface mounted FRP strips. (Dizhur) plaster finials or balusters, can be secured with a single epoxied bolt connected to a strand of stainless steel wire, to mitigate falling hazards. However, more complex strengthening work may be appropriate in some cases. 2.3 Floor and Roof Connection Upgrades

The most problematic deficiency in URM construction is inadequate connection of diaphragms to walls (FEMA 547), as failure of these connections can potentially lead to global collapse of the building. The addition of a network of small ties can substantially increase the strength of the building by fixing the walls to the floor and roof diaphragms (Robinson & Bowman 2000). These ties need to resist two actions; shear from the diaphragms trying to slide across the walls, and tension from the diaphragm and wall trying to separate. If these ties are missing, the walls will be acting in cantilever

Figure 4. An extreme case in the 2010 Darfield earthquake where inadequate connections have resulted in wall collapse. (Dizhur) from ground level under lateral loads, and floors and roofs are far more likely to be dislodged from their supports, which is the most common mode of failure for URM buildings in an earthquake. This is shown in fig. 4. Walls are not only secured by floors and roofs however; lateral support is also provided by cross walls, columns, and buttresses, and the influence of these elements should be taken into account. The thickness and length of walls both affect their performance; lower level walls in URM buildings are usually several leaves4 thick and can have significant out-of-plane resistance. Similarly, the shear response of these walls can be substantial, particularly for walls with few or no penetrations (Robinson & Bowman 2000). Horizontal diaphragms formed by floors or roofs can distribute loads to other load-resisting elements and tie buildings together. The connections of these systems are often inadequate or non-existent, and require upgrading before they can

A row of bricks at a certain level. A wall 3 bricks thick is known as a 3 leaf wall.

Volume 24 No. 1 April 2011

33

SESOC Journal provide any positive effects. If the resistance of the building still falls below the critical level after upgrading these connections, then additional strengthening will need to be undertaken. The most common form of flooring in URM buildings is timber, although concrete is also encountered (Russell 2010). Timber floors usually consist of tongue and groove planking nailed to joists which are supported on URM walls, or on bearers which are supported on internal columns. The direction of the joists affects the way in which these ties can be connected and there are various strategies for fixing in each direction. These strategies usually involve adding new timber chords and blocks and a number of steel fasteners to ensure adequate fixity between the diaphragm material, the chord, and the URM wall, although where access is limited this system may be substituted for a steel angle at floor or ceiling level. If done sensitively and without excessive visible use of new timber and fasteners, this method can have minimal visible impact. Consideration should be given to the type of bolt used, as tensile bolts often require a plate on the outside wall of the building, which results in exterior visual impact as shown in

Figure 5. An assortment of diaphragm fixings. (Dizhur) fig. 5. Systems have now been developed which forego this requirement, but require the fixing of bolts into the brick with epoxy. Through-bolting has good reversibility, but visual impact, and epoxying has no visual impact but is irreversible, so a balance must be struck. 2.4 Outer Layer Fixing

The outer faade layer of a cavity wall is problematic as it is particularly susceptible to failure by peeling off outwards. The steel ties which were commonly installed to connect this layer to the more robust wall behind are often missing or have rusted out, requiring attention during retrofits (Russell et al 2006). One approach to this problem has been to fill the cavity with grout and reinforcing steel, which has the dual benefits of bonding the outer layer to the main wall and also forming a reasonably strong shear wall which is hidden from view. However, this approach fails to consider the purpose of the cavity. Lime mortars are porous whereas concrete is not, which means that any moisture which finds its way inside the building for whatever reason will remain trapped there. This has resulted in extensive mould and the subsequent destruction of at least one heritage building in Auckland, making this practice hazardous and ill-advised. While a filled cavity may seem like an excellent strengthening solution, it is vital that the drainage functionality of the cavity is fully considered beforehand. The current preferred approach is to use a series of proprietary ties at regular centres which are drilled through the face layer and are epoxied into the structure behind, the fixing of which is shown in fig. 6. This is effectively a retrofit of the steel ties which ought to have

Figure 6. Spiral threaded rods being installed to secure an outer layer of bricks. (Dizhur)

34

Journal of the Structural Engineering Society New Zealand Inc.

SESOC Journal been present. The visual impact of these ties is minimal, although care needs to be taken when concealing drilled holes. These ties are irreversible, but their presence is visually negligible. 3.0 ENHANCING ExISTING WALLS

Seismic improvement should be designed with an appreciation of the architectural character in mind. The locations for strengthening elements will often present themselves; many strengthening elements will be able to be located within hidden or secondary spaces where they will have little architectural or historic impact. However, structure will often be required immediately adjacent to the street faade(s), in primary circulation spaces, lobbies, stairways, corridors, and large public spaces (Robinson & Bowman 2000). In these situations, particular care needs to be taken to ensure that the strengthening technique is appropriate to the architecture and heritage value of the building. Often it is imprudent to add elements in addition to the URM itself for reasons of visual impact. This section explains ways of strengthening the existing URM without significantly altering its thickness or aesthetic appeal. 3.1 Out-of-Plane Strengthening

URM walls are weak when subjected to forces other than compression. Even when fully secured to floors at each level, out-of-plane forces can cause significant bending. This bending is governed by the ratio of the height between levels of support to the thickness of the wall (Rutherford & Chekene 1990). Some walls are of sufficient thickness or have cross-walls or buttresses which enable them to withstand these out-of-plane forces without modification, however many walls will require seismic improvement. There are a number of approaches to combat this problem. 3.1.1 Inter-Floor Wall Supports A series of vertical steel sections can be bolted to the inside face of the wall at sufficient spacing to ensure that the width of wall between supports is capable of resisting the out-of-plane forces. These sections act in bending to transfer wall loads to the adjacent floor diaphragms, essentially breaking up a large planar wall into a number of buttressed segments. This is a simple method which may be appropriate in, for example, an industrial building, where visible steel bolted to the walls is in keeping with the character of the building, or in other buildings where the steel can be made to be architecturally appealing. In some other situations it may be less appropriate but less intrusive than other techniques. If there is existing internal framing with space behind for these columns, and no historic material is lost during installation, then it is a perfectly acceptable method. Steel sections generally fix to the historic material with bolts only, which allows a high degree of reversibility. In the past, rather than only supporting the URM walls for out-of-plane actions, these systems have been conceived as a method to support the floors in the event that the walls fail and collapse (Cattanach et al. 2006). In all URM buildings and in heritage buildings in particular, this is a defeatist approach to design. To design for the inevitability of the building collapsing rather than attempting to prevent its collapse is to deny the building itself of any inherent value, which in most cases is patently untrue. A technique in a similar vein to vertical steel is to provide a horizontal steel member at the mid-height of the wall and brace this with diagonal struts up to the floor or ceiling diaphragm above, as shown in fig. 7. This technique might be more suitable than vertical members if there is a cornice part way up the wall which needs to be conserved, or which can be used to disguise the steelwork. However care needs to be taken to ensure than the struts are visually unobtrusive. Both of these techniques can also be undertaken with the steel substituted with concrete, where this is more appropriate visually, or less commonly with timber. Steel struts can also be recessed within the width of the wall. Recessing the members results in an irrecoverable loss of material and may result in other complications such as cracking, although recesses may be preferable if used beneath a plastered surface, as there it will not affect the interior space. Concrete will be more massive than steel and therefore more intrusive. Also, once cast, concrete is difficult to remove without significant damage, particularly from a porous and naturally coloured material like clay brick. In-situ concrete is a fairly permanent measure, so any activity

Figure 7. Struts from the floor above to improve out-of-plane performance. (Dizhur)

Volume 24 No. 1 April 2011

35

SESOC Journal which requires concrete to be cast against brick should be given careful thought before being undertaken. As so often happens in conservation work, there are a number of measures available, all of which have positive and negative aspects. 3.1.2 Post-Tensioning and Other Core Reinforcement Post-tensioning is an extremely effective way of providing out-of-plane strength in URM walls. The process involves drilling vertical cores through the middle of a URM wall then inserting steel rods into them. The rods ay or may not be set in grout, and are then tensioned, which provides an additional compressive force in the wall. This loading modifies the stress behaviour of the URM in bending (i.e. the result of out-of-plane loading). Instead of bending instantly causing tensile forces; to which URM has little resistance, the wall remains in compression (Ismail et al 2010). This modification of the material properties also results in an increase in the shear strength of the wall, making post-tensioning an attractive strengthening solution. Post-tensioning has very little visual impact, although its installation may be unsuitable in some buildings, as access is needed to the top of the wall, and walls need to be of a certain minimum thickness. Drilling cores involves some loss of historic material from the holes, though compared to some methods this is a very minor impact. A partially completed example is shown in fig. 8. If the bars are fully grouted in place, post-tensioning is essentially irreversible, although this does not necessarily have to be done. The presence of post-tensioning bars is not likely to result in any negative effects to the historic material should their function no longer be required, provided care is taken with all core reinforcement to ensure that it is adequately protected from corrosion. This problem can be completely avoided by using plastic coated steel or FRP bars.

Figure 8. Post-tensioning bars used in the Birdcage hotel with concrete load spreaders cut into the brick. (Dizhur) There are other methods of core reinforcement; the most common being non-stressed bars set in grout, where the steel only becomes stressed when the wall is loaded laterally. The visual impact and reversibility of these methods are exactly the same as for fully grouted post- tensioning, though they are less effective structurally.

3.1.3 FRP and Others There are a number of other methods, such as strips of fibre reinforced polymer (FRP) fitted into vertical saw cuts in URM. This is known as near surface mounting (NSM). NSM is a relatively recent technique which involves epoxying FRP into saw cuts in the surface of the URM and covering the cut with a grout mixed with brick dust (Dizhur et al. 2011). This technique would have some visual impact in naked brick, but little if done within an existing grout line, and none if installed in plastered walls being repainted. This technique can be a particularly effective and non-intrusive method of strengthening, although the finishing of this system is noticeable and work needs to be done to conceal the inserts. Steel strapping as shown in fig. 9 can also be an effective technique. 3.2 In-Plane Strengthening

Most URM walls are required to transfer some degree of shear loading along their length. If a building has insufficient shear capacity in a particular direction, capacity of existing walls can be increased instead of inserting additional structure. There are various methods for achieving this strength increase which generally involve the application of an additional layer of material bonded to the surface of URM to increase

Figure 9. Simple and visually interesting in-plane strengthening. (Dizhur)

36 Journal of the Structural Engineering Society New Zealand Inc.

SESOC Journal its strength, although there are some measures which involve altering the wall itself, such as post-tensioning, as described above. Most of these measures involve a plane of extra independent structure being applied over the surface of the URM, effectively forming new shear walls, which are described below. The presence of openings in a shear wall renders that section less stiff than the surrounding full height walls, meaning that the wall above and below, or between closely spaced openings, will likely be the first areas to fail in the event of an earthquake. Infilling the openings will eliminate this problem by making the wall continuous, and has been advocated as a valid solution in the past. Problems with altering the character of the building and matching brick and mortar colours mean that this should only be used as a last resort and even then preferably not in visible areas. Infilling openings is likely to be somewhat reversible if done with brick, but not completely, and visual impact will depend on the location. If in-filled with concrete, the work will be less reversible and the ductile behaviour of the wall may be affected due to incompatible stiffnesses. Localised steel cross bracing near openings is another technique which can prove effective, but again this system is likely to be highly visible and should only be undertaken when it does not detract from the character of the building. 4.0
NEW BUILDING STRENGTHENING SYSTEMS

URM walls will often be unsuitable for strength upgrades, or are not of sufficient extent for these to provide enough strength. Taller buildings in particular will frequently require additional strengthening elements to supplement their inherent strength, as their facades often consist of a series of piers with openings between without enough planar surfaces to form effective shear walls. As shown in fig. 10, these are susceptible to earthquake damage. This section describes the more common techniques which can be utilised.

Figure 10. Pier damage between openings in the Darfield Earthquake. (Dizhur) 4.1 Moment Frames

Moment frames are a common method of gaining additional horizontal resistance which can also be used as a local strengthening solution. The advantage of this system is that it is comprised of beams and columns, so is fully customisable, and there is space between the vertical and horizontal elements. Moment frames allow full visual and physical access between each side of the frame, and minimal spatial disruption. In building faades with numerous openings, some form of moment frame can often be fitted to the masonry piers on the inside or outside (or both) depending on the effect on the architectural character. Moment frames can be a particularly effective solution, especially where the frame is tailored to the character of the building. Care needs to be taken with steel frames in particular to ensure stiffness compatibility with the existing structure (Robinson & Bowman 2000). Steel is a ductile material, but URM is not, meaning that under earthquake loads the added stiffness of the steel might not come into effect until a load is reached where the URM has already been extensively cracked. Moment frames can be an excellent strengthening technique, either to supplement an existing wall or as a new, stand alone element. If a steel frame is erected against an existing wall where weakness exists, the frame needs to be fixed either directly to the URM using bolted connections into the wall or to the diaphragm. Installing concrete frames is a more complex undertaking, as these will often be constructed by thickening existing piers, although a concrete frame which is separate from the existing structure is possible. In both situations it is important that architectural character is retained, and historic material conserved. Some considerate and artful design strategies may need to be undertaken to achieve this. Steel moment frames have a high degree of reversibility, as again they rely on mechanical connections and

Figure 11. A concrete moment frame inside the faade of a large URM building in Wellington. (Dunning Thornton Consultants)
Volume 24 No. 1 April 2011 37

SESOC Journal relatively small ties to connect to the existing structure. Concrete frames are generally far less reversible, but can sometimes be better concealed when this is a requirement. Fig. 11 shows a large new moment frame which is expressed as a new element. Some recent buildings have very effectively used precast concrete load-resisting elements which are separate from the URM walls, solving the problem of reversibility (Cattanach et al. 2006). 4.2 Braced Frames

Braced frames are available in various configurations: concentric, tension only concentric, eccentric, and K

bracing. The key functional difference between braced frames and moment frames is that due to the diagonal braces, braced frames prevent physical continuity between spaces on either side of the frame. Braced frames are also generally constructed from steel rather than concrete, and are much more rigid than moment frames.

Figure 12. Eccentric bracing in a walkway. (Dunning Thornton Consultants)

Figure 13. An eccentrically braced core. (Dunning Thornton Consultants) Braced frames are a very efficient method of transferring horizontal forces but have significant setbacks. Their use in faade walls is usually precluded by the presence of windows, as diagonal braces crossing window openings are generally considered to be poor design. It is also difficult to get a braced frame to conform to an existing architectural character; however they can be used to very good effect within secondary spaces, and can be made to fit architecturally in some situations with careful consideration. Figs. 12 and 13 show braced frames in use. Generally speaking, steel braced frames have a good degree of reversibility and can provide excellent strengthening when used appropriately. 4.3 Shear Walls

Shear walls are used to increase the strength of existing URM or are added as new elements. Materials which resist shear loads can be added to the surface of the URM; these might include gypsum plasterboard, particle board, plywood, or plate steel (Robinson & Bowman 2000), and are generally fixed to the URM with bolts via a supplementary structure. This means that the surface of the URM is generally covered and may interfere with decorative elements on walls and openings, although this can be worked around with stronger materials such as plate or strap steel. They can also increase the thickness of the wall, which is not particularly desirable as it can reduce the scale and area of the interior. For these reasons shear walls can be visually detrimental if used indiscriminately. Stand alone shear walls, which are independent of URM walls, can be introduced, although these can be detrimental for similar reasons. Despite these negatives, shear walls are a practical and efficient method for strengthening and are commonly used. All of these materials can be easily removed in the future, which makes them good solutions for shear walls in two to three storey buildings with moderate horizontal loads. During the 1980s, Shotcrete shear walls were a common strengthening technique. This involves spraying concrete onto the surface of a URM wall to essentially cast a new wall against the existing wall, as shown in fig. 14. This technique provides plenty of strength to the wall, both in-plane and out-of-plane; however it is now largely regarded as unacceptable unless absolutely necessary. It causes a significant increase in wall thickness and it is very difficult to remove the concrete, and even more so to restore the wall behind to any semblance of its character prior to concreting. Furthermore, its installation generally requires the building to be gutted, which results in the loss of much heritage material and creates an essentially new interior (Robinson & Bowman 2000). Another technique for forming shear walls is the addition of fibre reinforced polymer sheets is in fig. 15. These do not require the

same invasive installation as Shotcrete walls, but are however generally equally

38

Journal of the Structural Engineering Society New Zealand Inc.

SESOC Journal

Figure 14. Shotcrete walls on the rear of the birdcage hotel. (Dizhur)

Figure 15. FRP applied to a wall. (Dunning Thornton Consultants) permanent, and have potentially limited application, although new technology may soon change this. If it is possible to provide out-of-plane strength using FRP inserts, coupled with an FRP surface layer for shear, this solution could be far superior to Shotcrete from an architectural perspective. An important consideration with the use of sheets of FRP is that it is impermeable, which can lead to problems with water trapped within the building resulting in damp and mould issues, and potential de-bonding of the epoxy. 4.4 Diaphragms

Diaphragms are useful because they provide a layer through which lateral forces can be distributed from their source to remote resisting elements, and also act to bind the whole building together at each level. A building which acts as one rigid body rather than a number of flexible panels is far more likely to survive an earthquake. Tying floors to the outer walls is generally required regardless to ensure that joists are not dislodged (Robinson & Bowman 2000).

Figure 16. It is possible to make diaphragm improvements visibly interesting, as with this recently installed floor. (Dizhur) Floor diaphragms consist of three main elements; chords, sheathing material, and supplementary structure. To form a diaphragm in a typical URM building, chords need to be established, and mechanical fastenings added to take shear and tensile loads (Rutherford & Chekene 1990). Several secondary fastenings between the chord and the floor or roof may also be required depending on the technique used. Some tensile ties will penetrate to the outside of the building and others will be drilled and epoxied in place. Existing historic sheathing may prove inadequate and require strengthening or an additional layer of more rigid material. The main visual impact of forming diaphragms is if new sheathing is required. Historic flooring material is often a significant contributor to the character of a place and ought to be retained in view whenever possible. If the existing sheathing is inadequate, a ceiling diaphragm below, or stiffening the existing material might be preferable to covering it. Another approach is to remove the existing sheathing and install a structural layer beneath it. This exercise requires extreme care; firstly because existing sheathing, particularly tongue and groove, is very easily damaged during removal; and secondly, care needs to be taken to restore the boards in the correct order. Ties to the outside of walls may require metal load spreaders which visually impact the exterior. Many New Zealand buildings display these, and they seem to have become somewhat accepted as part of the strengthening process. Nevertheless, care needs to be taken when considering their visual impact and invisible solutions may be preferable. Much of the additional required work can be hidden within the floor space, but if this is exposed or the connections are extensive, special attention will be required to preserve the visual character of the inter-floor space.

Volume 24 No. 1 April 2011

39

SESOC Journal Diaphragms which are formed using mechanical connections have a high degree of reversibility; where ties are epoxied into walls, there is less reversibility, but minimal visual intrusion. Additional sheathing may damage or alter the nature of the historic timber below, making it less desirable as a solution, though this can be mitigated in cases such as that in fig. 16 by an interesting replacement. Occasionally, pouring concrete over an existing timber floor is considered. This solution can greatly increase the stiffness of the building, but in turn increases its weight and therefore the forces acting upon it. Further, it completely changes the material of the floor and is not a reversible action, because even if it can be removed, the concrete would essentially destroy the character of the underlying timber. This procedure is therefore not recommended except in exceptional circumstances. Roof diaphragms where the structure is exposed are slightly different, as the inclusion of a plywood diaphragm above timber sarking is generally acceptable if this area can be accessed, for example if the roofing is being replaced. This installation can also help to protect the sarking beneath. Roofs with suspended ceilings can be made to accommodate cross bracing, struts, and more innovative solutions, as they can be hidden within the ceiling space. In instances where the roof provides little diaphragm action, or the forming of a diaphragm is uneconomic or impossible, a horizontal load resisting member at the level of the top of the walls can be used to provide stability to the walls under out-of-plane loads. However, this member needs to be fixed to stiff elements at regular intervals to transfer horizontal loads, and these stiff elements may need to be introduced to the building if other structure cannot perform this task. 4.5 Foundations

Foundations in existing URM elements are usually adequate (FEMA 547). However, when transferring horizontal loads, new strengthening elements also usually induce significant concentrated vertical loads. This extra loading, combined with the added weight of additional structure, means that often the foundations in URM buildings will need to be upgraded in order to deal with these new loads. There are various techniques for meeting these new requirements, which will usually include enlarging the existing foundations by casting a new concrete footing beside the old and tying these footings together with drilled reinforcing rods. Existing foundations may also need to be tied together with ground beams to increase stiffness. Foundations themselves are not generally considered to have a great deal of heritage value; however the ground level floor often will. If the building has a basement or subfloor, strengthening foundations is made much easier, but care needs to be taken when considering foundation extensions to reduce the damage to the existing above ground material as much as possible. Some buildings are also located in historic areas and excavations beneath will require the input of an archaeologist. 5.0 REDUCTION OF FORCES

Another approach to seismic improvement of URM stems from its weight. Seismic actions are directly proportional to the mass of the building, so if mass is reduced, so are the forces acting upon the building. A lighter building requires less lateral strength and therefore less additional strengthening. Reducing the mass of a building may seem at face value

to be a sensible approach; however past experience has shown this to not be so. The mass must be removed from somewhere, and the higher up the mass is, the stronger the forces upon it and the more difficult it is to strengthen, so the top of the building is the first place which has been looked at. Historically this has led to the ad-hoc removal of decorative elements such as parapets, gables, chimneys, and occasionally whole towers (Robinson & Bowman 2000). These elements will almost always significantly contribute to heritage value and character, and their retention is essential to preserving this. Indeed, it is often desirable to replace these features if they have been removed from buildings and still exist. While reduction of weight may be achieved in more minor ways, such as removal of internal URM partitions or the removal of plant loads, the wholesale removal of decorative elements is strongly discouraged. 5.1 Base Isolation

A more recent approach to reducing forces acting on a building is to introduce a damping layer between the building and the ground. This process is known as base isolation, and works by separating the mass of the building from the lateral loads induced by ground movement with a layer of flexible material. Base isolation is the most effective but also the most radical strengthening technique, as it requires the building to be physically separated from its foundations by a layer of flexible material, and the foundations and the ground floor will usually need to be extensively strengthened. However, the strengthening required to the building above this level is minimised, making base isolation an attractive option for particularly important or valuable buildings (Robinson & Bowman 2000), or where the contents of the building are valuable; for example, hospitals or museums. Visual impact is minimised as most of the work takes place at ground level or below, although it is generally not reversible. As it is often below ground and largely invisible, this work may be of little consequence. The main problem with base isolation is

40

Journal of the Structural Engineering Society New Zealand Inc.

SESOC Journal that currently it is often prohibitively expensive to install in retrospect; however as new technology emerges this may change. There are other methods for modifying the behaviour of the building. One which has seldom been used in New Zealand is the installation of energy dissipaters or dampers. These devices resemble large hydraulic cylinders and work in combination with a steel frame to allow some deflection, but provide an increasing level of resistance. The downside of systems such as these is that they are expensive and difficult to conceal, as the dissipater works on the diagonal in place of a brace. Damping is most relevant to flexible buildings, however, and will not be effective in most URM buildings. 5.2 Torsion

Torsion can be a significant contributor to the total force in a structural system. It is caused by asymmetry in the mass and structural stiffness of the building. If the centre of mass and the centre of stiffness of a building are aligned, then there is no torsional effect, however this coincidental alignment is almost never the case as most buildings are not symmetrical. Buildings with one or more weak walls with extensive openings, such as shop-front faades, and one or more stiff walls, such as party walls, are particularly susceptible to torsional loads. In an earthquake the loads are focused through the centre of mass, as mass is the primary generator of seismic loads. The building resists these loads through its centre of stiffness, so if the centre of mass is offset from the centre of stiffness, then a rotational force is induced in addition to the basic horizontal force. In extreme cases the rotational force can be greater than the actual linear earthquake forces. The main way to combat excessive torsional response is to make the stiffness of the building more uniformly distributed. This exercise may involve stiffening weak walls with frames or bracing, or in more extreme cases involve placing vertical saw cuts periodically along the length of a stiff wall to reduce its stiffness. However, the architectural impact of each of these practices must be considered. The stiffening of elements involves the same considerations as the strengthening practices above, while saw-cutting requires a consideration of cut location and visibility within the structure of the building. Torsional forces must be considered in the retrofit of all asymmetrical buildings, particularly those which have one or more walls that are significantly weaker or stronger than others. 6.0 APPROACHES TO SEISMIC IMPROVEMENT

There are numerous approaches to improving heritage buildings and it is very important that proposed retrofit work is considered and reviewed prior to installation to prevent well-intentioned but insensitive solutions. It is generally accepted that the basic inadequacies of URM construction should be addressed in every case, which includes upgrading connections between the walls and the floors, fixing outer leaves of cavity walls, and securing parapets and chimneys. Addressing these very simple problems will ensure that the building will behave as predicted, and allows the designer to focus on any additional requirements.

The target strength of the building needs to be decided beforehand. In many cases it is accepted that strengthening to AS/NZS 1170 requirements is simply unrealistic and uneconomic, but the NZSEE recommended 67% of current requirements should be aimed for if possible. The approach taken should always refer back to the main principles of dealing with heritage as outlined in section 1.2 of this article. There are sufficient seismic improvement techniques currently available for all retrofits to be reversible and sympathetic to architectural and heritage value. Many approaches and options are available, and all solutions will result in some impact on the building, so the type of strengthening system used needs to be very carefully considered. As illustrated in fig. 17, the way a component looks, how it connects, its impact on existing material and spaces, and its harmony with the historic aesthetic are all important considerations. Furthermore, strengthening technology is constantly improving, and it is vital that current best practice is understood. 6.1 Improvement Type

Generally, the architecture and layout of the building will inform the overall approach to seismic improvement as well as individual strengthening solutions. The impacts of the proposed improvement systems need

Figure 17. A roof bracing system which clashes with the existing structure, resulting in a confused ceiling space. (Dizhur)
Volume 24 No. 1 April 2011 41

SESOC Journal to be considered and the simplest or most structurally efficient system may not always be appropriate. The designer needs to ask how a strengthening element will impact the building overall: visually and functionally; if it will detract from the heritage of the building; involve excessive removal of historic material; and its degree of reversibility. A wall with many openings is generally going to be unsuitable for use as a shear wall or for use with diagonal bracing, but might suit a series of moment frames, whereas an ornate and decorative exposed brick wall or complex forms, as in fig. 18 and fig. 19, might not suit a moment frame, but post-tensioning bars could be appropriate. An industrial building might already have exposed steel structure, so some additional steel, if suitably differentiated from the original structure, could be an excellent solution. 6.2 Placement

Placement of new structure is also often informed by the layout of the building. It is desirable for strengthening elements to be as close as possible to the source of the loads in order to minimise additional work as in fig. 20. It is generally undesirable to subdivide interior spaces particularly primary spaces, into smaller secondary spaces. The practice of placing intrusive bracing in walkways or in the middle of rooms is discouraged as it forms a barrier; physical or implied, between volumes.

Figure 18. The ornate faade of this building made exterior strengthening difficult. (Dizhur)

Figure 19. The solution was to innovatively strengthen the interior, which allowed retention of its most important visual elements. (Dizhur)

Figure 20. Openings in this Christchurch caf allowed placement of frames which compliments the space. (Dizhur)

Figure 21. Steel structure installed in an intrusive way which is visually distracting. (Dizhur)

42

Journal of the Structural Engineering Society New Zealand Inc.

SESOC Journal 6.3 Visibility

It is not always desirable to conceal new structure as if it were part of the building; the new work becomes part of the history of the building and it may be preferable to expose the structure if it is done in an architecturally sensitive way. It is very easy for an ill-considered yet simple structure to negatively affect a building, as in fig. 21. A particularly important heritage building will need careful treatment overall, and it may be possible and desirable to hide all new improvement work, allowing the building to retain as much of its character as possible. In other cases, with industrial or less ornate or important buildings, it is sometimes appropriate to celebrate strengthening work by exposing it and making it into a feature of the improved building as in fig. 22 and fig 23. These two schools of thought may seem to be conflicting, or even mutually exclusive, but careful aesthetic consideration can potentially make either approach appropriate for any given case.

Figure 22. No attempt has been made to conceal this new shear wall, rather the aesthetic is celebrated by the rest of the architecture. (Dizhur)

Figure 23. Excessive floor and wall ties somehow work with this faade. (Dizhur)

CoNClUSIoNS Seismic improvement is a complex art. Balancing strength requirements with visual impact, economy, reversibility, and architectural expression is a delicate act, but one which is possible in most situations, and one which ought to be the aim of all retrofits. Sometimes the design process will involve choosing the lesser of many evils, but with skill and care, a good solution can always be reached. There are myriad approaches and options for the seismic retrofit of URM buildings. The first step which must always be taken is to establish the buildings intrinsic resistance and utilise this to the greatest degree possible. In the likely event that this proves to be inadequate, this resistance should be enhanced by sensitively upgrading connections and other basic URM inadequacies using some of the methods described in this article, or by creating new techniques. A basic reading of the architectural character of the building will often assist in deciding on an appropriate approach, and an appreciation of which parts of the building are visually important will inform which improvement techniques will allow a sensitive solution. Most solutions will result in some visual impact to the building; invisible solutions are generally the safest option for visual impact, but these are often precluded by some other consideration. Of visible solutions, the best option will work with the existing material if not to blend in, then to complement the architectural character.

Volume 24 No. 1 April 2011

43

SESOC Journal It is emphasised that the most important attribute that a designer can have is creativity; conservation principles exist to inform the type of action that will result in a good solution from the perspective of conservation of heritage. Stringent adherence to these principles will not necessarily result in a solution that is good from an architectural or usability perspective and the best solution will work with the existing material to complement the architectural character with a minimum of impact. refereNCeS Cattanach, A.G., G.W. Alley, and A.W. Thornton. (2006) Appropriateness of Seismic Strengthening Interventions in Heritage Buildings: A framework for Appraisal Paper 30 Presented at the 2008 New Zealand Society for Earthquake Engineering Conference: Napier, NZ. Charleson, A.W. (2008) Seismic Design Boston; London: Elsevier/Architectural Press. for Architects: Outwitting the Quake. Amsterdam;

Charleson, A.W., Preston, J. and Taylor, M., (2001) Architectural Expression of Seismic Strengthening. Earthquake Spectra, 17(3), 417-426. Croci, G. (1998) The Conservation and Structural Restoration of Architectural Heritage. Southampton, U.K.: Computational Mechanics Publications/WIT Press. Derakhshan, H., Dizhur, D., Lumantarna, R., Cuthbert, J., Griffith, M. C., Ingham, J. M. (2010) In-field simulated seismic testing of as-built and retrofitted unreinforced masonry partition walls of the William Weir House in Wellington, Journal of the Structural Engineering Society of New Zealand, ISSN 01142879, 23, 1, April, 51-61.

Dizhur, D., Derakhshan, H., Griffith, M. C., Ingham, J. M. (2011) In-Situ Testing of a Low Intervention NSM Seismic Strengthening Technique for Historical URM Buildings, International Journal of Materials and Structural Integrity (IJMSI) (submitted May 2010). Feilden, B.M. (2003) Conservation of historic buildings. Oxford, U.K.: Architectural Press.

FEMA, (2006) FEMA 547 Techniques for the Seismic Rehabilitation of Existing Buildings. Washington D.C.: FEMA. FEMA. (2004) FEMA 398 Incremental Seismic Rehabilitation of Multifamily Apartment Buildings. Washington D.C.: FEMA. FEMA. (2004) FEMA 399 Incremental Seismic Rehabilitation of Retail Buildings. Washington D.C.: FEMA. FEMA. (2003), FEMA 397 Incremental Seismic Rehabilitation of Office Buildings. Washington D.C.: FEMA.

Gatley, J. (2008) Salvation by works. Architecture NZ, 2: 48-58. Goodwin, C. O. (2008) Architectural Considerations in the Seismic Retrofit of Unreinforced Masonry Heritage Buildings in New Zealand M.Arch Thesis, University of Auckland: Auckland, NZ. Goodwin, C. O., Tonks, G. and Ingham, J.M. (2009). Identifying Heritage Value in URM Buildings. Journal of the Structural Engineering Society New Zealand, 22(2), 16-28. Holman, S. (2000) Guidelines for developing heritage buildings. Wellington, NZ: New Zealand Historic Places Trust. ICOMOS NZ. (1993) Charter for the Conservation of Places of Cultural Heritage Value International Council on Monuments and Sites New Zealand.

Ismail, N., Laursen, P. T., Ingham, J. M. (2010) Seismic performance of face loaded unreinforced masonry walls retrofitted using post-tensioning. AJSE. Look, D. W., T. Wong, and S.R. Augustus. (1997) Preservation Briefs 41 The Seismic Retrofit of Historic Buildings; Keeping Preservation in the Forefront. Washington D.C.: U.S. Department of the Interior, National Parks Service. Mack, R. C., FAIA, and J. P. Speweik. (1997) Preservation Briefs 2 Repointing Martar Joints in Historic Masonry Buildings. Washington D.C.: U.S. Department of the Interior, National Parks Service, 1997. Mahmood, H., Ingham, J. M. (2011) Diagonal Compression Testing of FRP-retrofitted Unreinforced Clay Brick Masonry Wallettes, ASCE Journal of Composites for Construction (submitted May 2010). McLean, R. (2010) Seismic Strengthening Improving the Structural Performance of Heritage Buildings Wellington, NZ: New Zealand Historic Places Trust.

44

Journal of the Structural Engineering Society New Zealand Inc.

SESOC Journal NZSEE (New Zealand Society for Earthquake Engineering) (2006). Assessment and Improvement of the Structural Performance of Buildings in Earthquakes: Prioritisation, Initial Evaluation, Detailed Assessment, Improved Measures In Recommendations of a NZSEE Study Group on Earthquake Risk Buildings. Wellington, NZ: NZSEE Robinson, L. and I. Bowman. (2000) Guidelines for Earthquake Strengthening. Wellington, NZ: New Zealand Historic Places Trust. Russell, A.P. (2010), Seismic Assessment of In- Plane Unreinforced Masonry Walls in New Zealand, PhD Thesis, University of Auckland: Auckland, NZ. Russell, A. P., Ingham, J. M. (2010) Prevalence of New Zealands Unreinforced Masonry Buildings. Bulletin of the New Zealand Society for Earthquake Engineering (submitted January 2010). Russell, A. P., and J.M. Ingham. (2008) Trends in the Architectural Characterisation of Unreinforced Masonry in New Zealand. Paper presented at the 14th International Brick and Block Masonry Conference (14IBMAC): Sydney, Australia. Russell, A., J. Ingham, and M. Griffith. (2006) Comparing New Zealands Unreinforced Masonry

Details with those Encountered in Other Seismically Active Countries. Paper presented at the 7th International Masonry Conference: London, U.K., 30 October - 1 November. Rutherford & Chekene. (1990). Seismic Retrofit Alternatives for San Franciscos Unreinforced Masonry Buildings: Estimates of Construction Cost & Seismic Damage. San Francisco, CA. City and County of San Francisco Department of City Planning (prepared by Rutherford & Chekene, Consulting Engineers). Schofield, H. A., J.M. Ingham, and S. Pampanin. (2006) Critical earthquake risk detailing in New Zealands multi-storey building stock: understanding and improving the current perception. Paper 39 presented at the 2008 New Zealand Society for Earthquake Engineering Conference: Napier, NZ. Taylor, M., Preston, J., and Charleson, A. (2002) Moments of Resistance. Sydney: Arcadia Press. Trapeznik, A. (2000) Common Ground? Heritage and Public Places in New Zealand. Dunedin, NZ: University of Otago Press. Weeks, K. D. (1997) Preservation Briefs 14 New Exterior Additions to Historic Buildings: Preservation Concerns. Washington D.C.: U.S. Department of the Interior, National Parks Service.

round of golf anyone?

A priest, a doctor, and an engineer were waiting one morning for a particularly slow group of golfers. The engineer fumed, "What's with those guys? We must have been waiting for fifteen minutes!" The doctor chimed in, "I don't know, but I've never seen such inept golf!" The priest said, "Here comes the green-keeper. Let's have a word with him." He said, "Hello, George. What's wrong with that group ahead of us? They're rather slow, aren't they?" The green-keeper replied, "Oh, yes. That's a group of blind firemen. They lost their sight saving our clubhouse from a fire last year, so we always let them play for free anytime." The group fell silent for a moment. The priest said, "That's so sad. I think I will say a special prayer for them tonight." The doctor said, "Good idea. I'm going to contact my ophthalmologist colleague and see if

there's anything she can do for them." The engineer said, "Why can't they play at night?"
Volume 24 No. 1 April 2011 45