1590

Floor response spectra for seismic design of operational and functional components of concrete buildings in Canada
M. Shooshtari, M. Saatcioglu, N. Naumoski, and S. Foo

Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 75.86.130.80 on 04/19/11 For personal use only.

Abstract: It has been observed during previous earthquakes that the damage to operational and functional components of buildings often result in more injuries, fatalities and property damage than those inflicted by structural damage. Operational and functional components of a building include architectural components, mechanical and electrical equipment, as well as building contents. A rational approach to designing these elements against seismic excitations involves the use of floor design spectra. The development of such design spectra for buildings in Canada constitutes the objective of the paper. This objective was achieved by conducting comprehensive analyses of selected reinforced concrete buildings, with different lateral force resisting systems and building heights, under code compatible earthquake records for an eastern and a western Canadian city. It was observed that the floor response was significantly amplified, especially for buildings with short periods. Generally, the higher floors showed higher amplifications with differences in spectra between the floors being more pronounced in low-rise buildings and shear wall buildings with short fundamental periods. The results provided a large volume of data to generate floor response spectra for the design of operational and functional components of buildings in Canada. The details of the approach and the design spectra are presented in the paper. Key words: earthquakes, floor response spectra, frames, nonlinear analysis, non-structural components, operational and functional components, response spectra. ismes ante rieurs, il a e te observe que les dommages aux composantes ope rationnelles et fonctionnelles sume : Lors de se Re timents re sultent souvent en plus de blessures, de de ce ` s et de dommages a ` la proprie te que ceux cause s par les domdes ba rationnelles et fonctionnelles dun ba timent comprennent les composantes archimages structuraux. Les composantes ope quipements me caniques et e lectriques ainsi que le contenu des ba timents. Une approche rationnelle pour la tecturales, les e le ments pour re sister aux se ismes implique lutilisation de spectres de re ponse au niveau du plancher. conception de ces e veloppement de tels spectres de conception pour les ba timents du Canada. Cet objectif a Lobjectif de cet article est le de te atteint en re alisant des analyses comple ` tes de ba timents en be ton arme se lectionne s comportant diffe rents syste ` mes de e rales et diffe rentes hauteurs de ba timents, selon les enregistrements sismiques compatibles aux renfort contre les forces late te remarque que la re ponse au niveau du plancher e tait codes pour une ville canadienne de lEst et une de lOuest. Il a e e, particulie ` rement pour les ba timents a ` courtes pe riodes. Re ` gle ge ne rale, les planchers plus hauts grandement amplifie de plus grandes amplifications et les diffe rences entre les spectres e taient plus prononce es endans la structure ont montre timents de faible hauteur et dans les ba timents comportant des murs de contreventement a ` tre les planchers dans les ba riodes fondamentales. Les re sultats ont engendre un grand nombre de donne es afin de ge ne rer les spectres de re courtes pe rationnelles et fonctionnelles des ba timents au Caponse au niveau des planchers pour la conception des composantes ope sente les de tails de la me thode et des spectres de conception. nada. Cet article pre ismes, spectres de re ponse au niveau des planchers, cadres, analyse non line aire, composantes non structura s : se Mots-cle rationnelles et fonctionnelles, spectres de re ponse. les, composantes ope daction] [Traduit par la Re

Received 03 February 2009. Revision accepted 12 August 2010. Published on the NRC Research Press Web site at cjce.nrc.ca on 22 October 2010. M. Shooshtari. Department of Civil Engineering, Bu-Ali Sina University, Hamedan, Iran. M. Saatcioglu1 and N. Naumoski. Department of Civil Engineering, the University of Ottawa, Ottawa, ON K1N 6N5, Canada. S. Foo. Public Works & Government Services Canada, Gatineau, QC K1A 0S5, Canada. Written discussion of this article is welcomed and will be received by the Editor until 30 April 2011.
1Corresponding

Introduction
Performance of buildings and their contents during previous earthquakes have clearly indicated the vulnerability of operational and functional components (OFC) of buildings to seismic damage and life safety. Operational and functional components in buildings consist of (i) non-structural and architectural components, such as infill walls, partitions, claddings, parapets, stairways, suspended ceilings, lighting systems etc.; (ii) mechanical and electrical equipment, such as pipes and ducts, escalators, central control panels, transformers, emergency power systems, fire protection systems, machinery, etc.; and (iii) building contents, including book shelves, furniture, file cabinets, storage racks, etc. Poor performance of these components claims more injuries, fatalPublished by NRC Research Press

author (e-mail: Murat.Saatcioglu@uottawa.ca).

Can. J. Civ. Eng. 37: 15901599 (2010)

doi:10.1139/L10-094

Shooshtari et al.

1591

ities, property and financial losses during damaging earthquakes than those inflicted by structural damage. There have been many incidences that a building, which sustained only minor structural damage, was deemed unsafe and unusable as a result of extensive damage to its OFCs. Failure of such components also poses serious problems for search and rescue operations after an earthquake, resulting in additional and unnecessary increases in casualties. Equipment failures and the debris caused by falling objects could critically affect the performance of vital facilities such as emergency command centres, fire and police stations, hospitals, power stations and water supply plants. Furthermore, there may be additional seismic risks associated with fire or leakage of hazardous materials triggered by failures of OFCs. Operational and functional component failures during past earthquakes have been reported extensively in previous earthquake reconnaissance reports, including those by McKevitt et al. (1995), Filitrault et al. (2001) and FEMA74 (1994) with specific emphasis on the subject. Only a brief summary of previous observations from these publications is presented here. The collapse of unreinforced brick parapets and exterior walls during the 1906 San Francisco, 1925 Santa Barbara, and 1933 Long Beach earthquakes resulted in heavy casualties. Similar failures, including the failures of interior fixtures were observed during the 1952 Bakersfield, 1971 San Fernando, 1987 Whittier-Narrows, 1989 Loma Prieta, and 1994 Northridge earthquakes. The 1964 Alaska Earthquake exposed the vulnerability of modern exterior precast wall panels, elevators and suspended ceilings. The 1971 San Fernando Earthquake triggered collapses of metal library shelving, failures of suspended ceilings, light fixtures and ducts for heating, ventilation and air conditioning. The 1989 Loma Prieta Earthquake resulted in the collapse of heavy ceiling plasters and ornamentation, lighting grids and their suspended fixtures. This earthquake also resulted in severe economic losses caused by damaged water supply systems. During the 1994 Northridge Earthquake, several major hospitals had to be evacuated, not because of structural damage, but because of the failure of emergency power systems, air control units, falling ceilings and light fixtures. McKevitt et al. (1995) reported that failure of most OFCs during the Northridge Earthquake could be attributed to lack of seismic restraints. They also reported that approximately 20% of deaths during the same earthquake occurred because of poor performance and failure of OFCs. In Canada, the 1988 Saguenay earthquake, the strongest event in eastern North America recorded within the last 50 years, caused very little structural damage, but resulted in injuries, property damage, and economic loss associated with the failures of OFCs in buildings. Past experiences with OFC performances during previous earthquakes prompted research to develop asesimic design methodologies, tools, guidelines, standards and codes for earthquake resistant OFCs. In the United States, several approaches were developed and published. The International Building Code (ICC-IBC 2006) requires the computation of seismic forces applied on OFCs as well as maximum lateral deflections to ensure survivability of these components. The US Federal Emergency Management Agency (FEMA) published design requirements for OFCs in new (FEMA-302 1997; FEMA-303 1997) and existing buildings (FEMA-273

Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 75.86.130.80 on 04/19/11 For personal use only.

1997; FEMA-274 1997), as well as the seismic evaluation (FEMA 1998) and seismic rehabilitation (FEMA 1992, 1994) of OFCs in buildings. Research work related to this topic was conducted by Sankaranarayanan and Medina (2007) and Medina et al. (2006). The focus of these studies was to determine the effects of various parameters on the floor acceleration response spectra, i.e., the effects of the height of the building structure, the floor level for which the spectra are calculated, the damping value for the response spectra, the fundamental period of the building, and the level of the inelastic behaviour of the building. The structural models used in the analysis consisted of single-bay frames. Additional research was conducted by Uma et al. (2010), Rodriguez et al. (2002), Chaudhuri and Villaverde (2008), and Marsantyo et al. (2000) for developing floor response spectra. A state of the art paper was published by Villaverde (1997) on the topic. In Canada, the Canadian Standards Association (CSA) published the first edition of CSA S832 Guideline for Seismic Risk Reduction of Operational and Functional Components (OFCs) of Buildings in 2001 (CSA-S832 2001). The Standard was subsequently revised in 2006 and was published under the title of Standard for Seismic Risk Reduction of Operational and Functional Components (OFCs) of Buildings (CSA-S832 2006). The CSA 8322006 includes two methods for evaluating the OFC behaviour: (i) prescriptive method and (ii) analytical method. The prescriptive method provides general concepts for design and performance, including the details for fastening OFCs to prevent or minimize seismic movements, but otherwise relies on guidelines published by the industry for specific equipment or component manufactured. In the analytical method, the forces and (or) displacements of OFCs are calculated using a simple method, such as the equivalent static force method, or a refined method involving response spectrum or time history analyses. The refined methods are mandatory for OFCs with a mass greater than 20% of that of the floor or 10% of the total building mass. The current National Building Code of Canada 2005 (NRC 2005) addresses the design of non-structural components against seismic effects by classifying them into 21 categories and by suggesting an empirical approach. Accordingly, these components must be designed to carry a lateral force Vp defined in eq. [1] and applied at the centre of mass. 1 V p 0:3F a Sa 0:2I E Sp W p

where, Wp is the weight of non-structural component, IE is importance factor, Fa is acceleration-based site coefficient and Sa (0.2) is the spectral response acceleration value at 0.2 s. The horizontal force factor Sp is computed by eq. [2], with minimum and maximum values of 0.7 and 4.0, respectively. 2 Sp C p Ar Ax =Rp

where Cp, Ar, Ax and Rp are component, component force amplification, height and component response modification factors where the latter reflects the available ductility in the component. These factors are defined for each of the 21 categories of OFCs specified in the code. For most OFCs, Vp
Published by NRC Research Press

1592

Can. J. Civ. Eng. Vol. 37, 2010

Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 75.86.130.80 on 04/19/11 For personal use only.

is applied in the horizontal direction, except for horizontally cantilevered floors, balconies and other similar elements where the lateral force is to be applied in the vertical direction as an upward and downward force. The code also emphasizes that the lateral deflection of an OFC, computed using an elastic analysis, should be multiplied by Rp/IE to obtain a realistic deflection within the inelastic range of deformations. While empirical methods have been developed in the past and are used currently, a more rational approach may be used for designing OFCs in buildings, utilizing floor design response spectra. The NBCC-2005 reflects the most recent seismic hazard data for building designs in the form of uniform hazard spectra (UHS). However, floor design spectra for OFCs, compatible with NBCC 2005, are currently not available for use in design. The current research program is an effort towards filling this gap and meeting the needs of the design profession. The NBCC-2005 compatible floor response spectra were generated by Saatcioglu et al. (2008) for reinforced concrete frame buildings in Canada, as part of an earlier phase of the current research program. The current phase includes additional concrete frame buildings, as well as shear wall buildings. A total of 180 nonlinear dynamic response history analyses were conducted for this purpose, using 30 codecompatible earthquake records for eastern and western Canada. The analyses provided acceleration time histories for each floor of each building, which were then used to conduct a statistical analysis to obtain floor response spectra at mean and mean plus one standard deviation. Floor design response spectra were then generated for the design of earthquake resistant OFCs. Note that no dynamic interaction between the OFCs and the floors was considered in the development of the floor response spectra, i.e., it was assumed that the masses of the OFCs are substantially smaller than the mass of the supporting floor. The details of the procedure are presented in the following sections.

Buildings selected for analysis

A total of 12 reinforced concrete buildings, designed on the basis of the seismic provisions of NBCC-2005, were selected. The buildings consisted of 5, 10, and 15 storey heights. Both moment resisting frame and shear walls structural systems were considered as lateral force resisting systems. Consequently, three moment resisting frame buildings and three shear wall buildings with three different building heights were considered in Vancouver and Ottawa, separately. The Vancouver buildings represented structures in western Canada, and the Ottawa buildings represented structures in eastern Canada. The frame buildings in Vancouver included cases where the buildings were designed without due considerations given to current drift limitations, and hence includes structures that may be representatives of older buildings built prior to the enactment of modern building codes. Figures 1 through 3 show the floor plan and elevation views of buildings. The floor plans were selected to be symmetrical to minimize the effects of torsion. The analyses were conducted in the short direction of plan. The participation of non-structural elements, such as architectural and masonry enclosures, and (or) partition

walls were not accounted for in seismic analysis. Therefore, the computed fundamental periods of buildings may be longer than those with significant stiffening effects of nonstructural components. The periods were computed with due considerations given to the cracking of concrete, as per CSA Standard A 23.3 requirements (CSA A23.3 2004). Accordingly, the horizontal element (beam) and vertical element (column and wall) stiffness were reduced to 35% and 70% of their elastic, uncracked values, respectively. The buildings in Vancouver were designed for full ductility, whereas those in Ottawa were designed for nominal ductility, with elastic design force levels reduced by appropriate force modification factors. Within the NBCC-2005 requirements, this corresponds to the use of Rd = 4.0 and Ro = 1.7 for frame buildings and Rd = 3.5 and Ro = 1.6 for shear wall buildings in Vancouver; and Rd = 2.5 and Ro = 1.4 for frame buildings and Rd = 2.0 and Ro = 1.4 for shear wall buildings in Ottawa. Consequently, a higher reduction of forces was applied to Vancouver buildings relative to those in Ottawa. However, the design spectral values for Vancouver were proportionately higher than those for Ottawa, resulting in essentially the same seismic design forces and element dimensions for buildings with the same heights in Vancouver and Ottawa. The primary difference in design between the buildings in western and eastern Canada was in the design and detailing of reinforcement, reflecting different levels of ductility. This resulted in the same fundamental period for the same structural type and height of buildings in Vancouver and Ottawa. The fundamental period values were 1.75 s, 3.47 s, and 5.11 s, for 5, 10, and 15-storey frame buildings and 0.44 s, 1.30 s, and 2.34 s for 5, 10, and 15-storey shear wall buildings, respectively. All buildings were assumed to be located on firm soil (Class C in NBCC-2005), without the amplification effects of soft soils. The force levels were based on building periods computed by the empirical expressions suggested in the code, increased to 150% and 200% of the empirically computed values as permitted by the code because of the longer periods computed by dynamic analyses.

Selection of ground motions

Ground motion records for dynamic analysis were selected to match the UHS given by NBCC-2005 for Vancouver and Ottawa. A total of 15 synthetic records, generated by Atkinson and Beresnev (1998), were selected for each city with a probability of occurrence of 2% in 50 years, reflecting different earthquake distance and magnitude relationships. The magnitudes of records selected for Ottawa were 6.0 on the Richter scale for short events and 7.0 for long events; and for Vancouver, 6.5 for short events and 7.2 for long events. The distance for the short events varied between 25 km and 50 km; and for long events between 50 km and 100 km for both cities. Table 1 shows the characteristic of the synthetic records considered. Figure 4 shows the response spectrum for each earthquake record relative to UHS. Each record is labeled to reflect its magnitude and distance. For example, M6-R25-1 indicates a record with magnitude 6 and distance 25 km. The short events are intended to govern in the short period range, whereas the long events are intended to govern in the long period range.
Published by NRC Research Press

Shooshtari et al. Fig. 1. Typical floor plans for selected buildings: (a) moment resisting frame building; (b) shear wall frame building.

1593

Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 75.86.130.80 on 04/19/11 For personal use only.

Fig. 2. Elevation views of interior and exterior frames of moment resisting buildings as well as interior frames of shear wall buildings: (a) 15-storey; (b) 10-storey; (c) 5-storey.

Fig. 3. Elevation views of exterior frames of shear wall buildings: (a) 15-storey; (b) 10-storey; (c) 5-storey.

Dynamic inelastic response history analyses

Computer software DRAIN-2DX (Prakash et al. 1993) was used to carry out nonlinear time history analysis of buildings to generate floor response spectra. DRAIN-2DX is a general purpose dynamic analysis program that utilizes the stiffness method of structural analysis. Dynamic equation of motion is solved numerically through the step-by-step linear acceleration method. Three degrees of freedom were used at each node, consisting of X and Y translations and R rotation about the Z-axis. Flexural yield strengths for all structural elements are defined as part of the parameters that defined the hysteretic model. The element stiffness was defined with due considerations given to concrete cracking. The flexural rigidities were taken as 70% and 35% of those computed based on gross cross-sectional properties of vertical (columns and walls) and horizontal (beams) elements, respectively. Takedas hysteretic model (Takeda et al. 1970) was employed to describe the element stiffness during loading, unloading and reloading under earthquake excitations. Structural mass was assumed to be concentrated at each floor level and was specified at each node. Damping was specified as 5% of critical damping, consisting of stiffness and mass dependent components. The lateral force resisting systems were modeled as series of plane frames in the short direction, linked together with rigid links, ensuring equal horizontal displacements at joints

of the same storey. This implies that the floors were assumed to be infinitely rigid. Each building had 6 frames in the short direction. All the interior frames were identical in strength and stiffness, and hence were lumped together as a single frame, representing total strength and stiffness of all interior frames. Similarly, the two exterior frames having the same properties were lumped together to form a single frame representing total strength and stiffness of the exterior frames. These two model frames were then linked to ensure equal displacements at floor levels. Each element of the model was provided with a flexural spring at each end, allowing it to yield beyond the member yield capacity, while also incorporating the rules of the hysteretic model (stiffness during loading, unloading and reloading beyond the elastic range). The elements were assumed to have sufficient capacity to remain elastic in shear. The model for each structure was subjected to the selected earthquake records to conduct dynamic inelastic response history analysis. The results indicated that the buildings in Ottawa, representing eastern Canada, remained elastic at all times, even though they were designed for inelasticity and moderate levels of ductility. The maximum storey drift ratios remained within 2.3% for frame buildings and 1% for shear wall buildings. The frame buildings in Vancouver developed inelasticity in some columns and beams under the majority of earthquake records. The shear wall building in Vancouver only developed inelasticity in
Published by NRC Research Press

1594 Table 1. The characteristic of 15 records selected for Ottawa and Vancouver. Ottawa Record number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 M 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 R (km) 25 30 30 30 30 40 40 50 50 70 70 70 70 100 100 Peak acceleration (cm/s2) 582.79 422.13 512.36 461.16 430.94 299.81 231.14 498.89 619.76 295.3 280.46 335.54 286.07 237.81 256.25 Scale factor 0.58 0.80 0.80 0.80 0.80 1.26 1.46 0.58 0.42 0.80 0.80 0.80 0.80 0.92 0.70 Vancouver M 6.5 6.5 6.5 6.5 6.5 6.5 6.5 7.2 7.2 7.2 7.2 7.2 7.2 7.2 7.2 R (km) 25 30 30 30 30 40 40 50 50 70 70 70 70 100 100

Can. J. Civ. Eng. Vol. 37, 2010

Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 75.86.130.80 on 04/19/11 For personal use only.

Peak acceleration (cm/s2) 586.84 523.51 526.83 567.54 380.05 335.38 288.34 357.06 366.64 241.55 254.33 225.94 247.14 146.82 117.13

Scale factor 0.72 1.00 1.00 1.00 1.00 1.49 1.45 0.55 0.55 1.00 1.00 1.00 1.00 1.38 1.50

Note: M is Richter magnitude, R is distance and the scale factor is used to match the UHS (Atkinson and Beresnev 1998).

Fig. 4. Spectral accelerations for synthetic records: (a) short events in Ottawa; (b) long events in Ottawa; (c) short events in Vancouver; (d) long events in Vancouver.

the walls, under some of the ground motion records. Figure 5 shows examples of hysteretic response in selected structural elements. The analyses results provided response time histories of floor accelerations for use in generating design floor response spectra.

Floor response spectra

Dynamic analyses of 12 buildings under 15 earthquake records resulted in a total of 180 analyses and 1800 floor re-

sponse spectra for frame and shear wall buildings located in Ottawa and Vancouver, one spectrum for each floor of each building. The spectra for each floor were used to compute the mean spectrum. Floor response spectra were also constructed for mean plus one standard deviation values to develop design floor spectra. Design spectra are commonly specified at mean plus one standard deviation level (Rosenblueth 1980) to ensure that there is a relatively small probability that the response will be above the specified design level. The mean plus one standard deviation level correPublished by NRC Research Press

Shooshtari et al.

1595

Fig. 5. Sample hysteretic relationships for exterior frame elements of the 10-storey building in Vancouver: (a) first storey column, (b) first storey exterior beam.

Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 75.86.130.80 on 04/19/11 For personal use only.

Fig. 6. Floor response spectra for frame buildings in Ottawa mean plus one standard deviation values: (a) 5-storey building; (b) 10-storey building; (c) 15-storey building.

sponds to 84% of all spectral values being below the specified level. This level was deemed appropriate for the development of design spectra. Figures 69 show spectral values for all buildings. The examination of response spectra indicates that there is a progressive increase in response going from the first floor to upper floors. The rate of response amplification is higher in low-rise buildings as compared to companion medium and high-rise buildings. The 5-storey frame buildings showed an amplification of approximately a factor of 4 for the roof response relative to that of the first-storey. However this amplification factor was approximately 3 and 2 for 10storey and 15-storey buildings. It was further observed that the shear wall buildings analyzed developed higher floor amplifications than the companion frame structures having

the same height. This may be explained by the dependence of the amplification factor on the fundamental period. Indeed, the amplification in floor response was higher for short period structures. The response spectra further indicated higher amplifications for buildings in Vancouver, which were subjected to stronger ground motions relative to those in Ottawa. Figures 6 through 9 also include the UHS for respective cities, depicting the amount of amplification observed at each floor relative to ground. In general, the UHS was representative of the floor response spectra for lower stories, though in certain buildings, especially in shear wall buildings, the UHS was above the floor spectra for lower floors. However, it was deemed appropriate to use the UHS as representative of the first floor response spectrum for design
Published by NRC Research Press

1596

Can. J. Civ. Eng. Vol. 37, 2010

Fig. 7. Floor response spectra for frame buildings in Vancouver mean plus one standard deviation values: (a) 5-storey building; (b) 10storey building; (c) 15-storey building.

Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 75.86.130.80 on 04/19/11 For personal use only.

Fig. 8. Floor response spectra for shear wall building in Ottawa mean plus one standard deviation values: (a) 5-storey building; (b) 10storey building; (c) 15-storey building.

Published by NRC Research Press

Shooshtari et al.

1597

Fig. 9. Floor response spectra for shear wall building in Vancouver mean plus one standard deviation values: (a) 5-storey building; (b) 10-storey building; (c) 15-storey building.

Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 75.86.130.80 on 04/19/11 For personal use only.

Fig. 10. Comparisons of proposed roof design response spectra for buildings considered for Vancouver in the current investigation.

Fig. 11. Proposed roof design spectra for buildings with a fundamental period of 1.5 s.

purposes. A relationship was then derived for the amplification of roof response relative to UHS as a function of fundamental period. This relationship is given as 3 C 5:0 0:5T a

spectral values relative to the UHS. The application of the amplification factor C provides roof design spectral accelerations for buildings on firm ground, Sf(T). The amplification factor may vary with different soil conditions. The roof spectra established by the amplification factor specified in eq. [3] were observed to overestimate response in the short period range. Hence a cut-off value was derived as 4 Sf T CST a  BS0:2 for T s < T  2:0 s

where Ta is the fundamental period of building housing the OFC in seconds, computed by an accepted method of mechanics; and C is the response amplification factor for roof

where T is the period of OFC in the building, S(Ta) is the

Published by NRC Research Press

1598

Can. J. Civ. Eng. Vol. 37, 2010

UHS value specified in NBCC-2005 for firm ground, C is the amplification factor computed using eq. [3] and B is a coefficient that defines the maximum cut-off value. B = 1.5 for buildings with lateral force resisting systems (LFRS) consisting of frames, and B = 2.5 for buildings with shear wall LFRS. Ts is equal to 0.04 s for Ottawa, representing eastern Canada, and 0.2 s for Vancouver, representing western Canada. The UHS values in NBCC for very short period structures, corresponding to T < 0.2 s, is constant and equal to the UHS at T = 0.2 s. This is a conservative assumption, which may not have much effect on building design, since there would be very few buildings with peirods of less than T = 0.2 s. However, this is not the case for OFCs. Operational and functional components may be quite rigid, with periods of less than 0.2 s. Therefore, further refinements were introduced to eq. [4] to reflect the computed floor response spectra more accurately. Accordingly, the variation of design spectra was suggested to change linearly between Sf(0.01) = D S(0.2) at T = 0.01 s and Sf(Ts) at T = Ts (where, Ts is defined above as 0.04 s for Ottawa and 0.2 s. for Vancouver). The coefficient D is defined as 2/3 for frame buildings and 1.0 for shear wall buildings. A linear interpolation can be made for in-between floors between the roof design spectrum and the UHS representing the first floor spectrum. However, the design spectral values for inbetween floors need not exceed the value specified for the roof. The UHS is assumed to remain constant beyond 2.0 s. This was found to overestimate structural response in the long period range. Further amplification of floor response in this range was found to be overconservative. Therefore, it is recommended that for T > 2.0 s, the floor response spectra change linearly from Sf(2.0) at T = 2.0 s to the UHS value at T = 5.0 s. It is important to note that the floor spectra developed in this study are intended for determining seismic forces acting on OFCs that are attached to the floors. Also, the floor spectra developed correspond to nonlinear response of the supporting structure under seismic motions compatible with the design spectrum for the location. The proposed floor design spectra are compared in Fig. 10 for the 6 buildings considered for Vancouver. The comparison indicates substantially higher amplification for shear wall buildings, relative to frame buildings, with the amplification being more sensitive to building height in frame buildings. The relative impact of eastern and western seismicity on proposed roof design spectra is illustrated in Fig. 11 for selected frame and shear wall buildings.

Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 75.86.130.80 on 04/19/11 For personal use only.

 Floor-to-floor amplification of spectral accelerations is more pronounced in low-rise buildings and shear wall buildings. The building period appears to play a dominant role on floor amplification. Buildings with longer periods show lower values of amplification, indicating that floor response spectra at different floor levels may be approaching each other for tall frame buildings with long periods. The opposite is true for low-rise shear wall buildings, for which higher amplification of floor response may occur between the first floor and the roof.  Equation [4] developed as part of the current research project can be used for the type of buildings and locations considered in Canada on firm ground, to generate floor response spectra from UHS. The amplification may be scaled down within the short and long period ranges, as UHS in these regions tend to over-estimate structural response.

References
Atkinson, G.M., and Beresnev, I.A. 1998. Compatible ground-motion time histories for new national seismic hazard maps. Canadian Journal of Civil Engineering, 25(2): 305318. doi:10.1139/ cjce-25-2-305. Chaudhuri, S.R., and Villaverde, R. 2008. Effect of building nonlinearity on seismic response of nonstructural components: A parametric study. Journal of Structural Engineering, 134(4): 661670. doi:10.1061/(ASCE)0733-9445(2008)134:4(661). CSA-S832. 2001. Guideline for seismic risk reduction of operational and functional Components (OFCs) of buildings. Canadian Standards Association, Rexdale, Ont., Canada. CSA A23.3. 2004. Design of reinforced concrete buildings. Canadian Standards Association, Rexdale, Ont., Canada. CSA-S832. 2006. Standard for seismic risk reduction of operational and functional components (OFCs) of buildings. Canadian Standards Association, Rexdale, Ont., Canada, pp. 104. FEMA-172. 1992. NEHRP handbook of techniques for the seismic rehabilitation of existing buildings. Federal Emergency Management Agency, Washington, D.C., pp. 208. FEMA-74. 1994. Reducing the risks of nonstructural earthquake damage. Federal Emergency Management Agency, Washington, D.C., pp. 139. FEMA-273. 1997. NEHRP guidelines for the seismic rehabilitation of buildings. Emergency Management Agency, Washington, D.C., pp. 435. FEMA-274. 1997. NEHRP commentary on the guidelines for the seismic rehabilitation of buildings. Emergency Management Agency, Washington, D.C., pp. 48. FEMA-302. 1997. NEHRP recommended provisions for seismic regulations for new buildings and other structures, Part 1: Provisions. Emergency Management Agency, Washington, D.C., pp.335. FEMA-303. 1997. NEHRP recommended provisions for seismic regulations for new buildings and other structures, Part 2: Commentary. Emergency Management Agency, Washington, D.C., pp. 366. FEMA-310. 1998. Handbook for the seismic evaluation of buildings: A prestandard. Emergency Management Agency, Washington, D.C., pp. 288. Filitrault, A., Christopoulos, C., and Stearns, C. 2001. Guidelines, specifications, and seismic performance characterization of nonstructural building components and equipment. PEER Report 2002/05, Pacific Earthquake Engineering Research Center, University of California, Berkeley, Calif.
Published by NRC Research Press

Conclusions
The following conclusions can be derived from the research project presented in this paper:  Floor response spectra show amplified spectral accelerations relative to the ground spectral accelerations as described by UHS.  The amplification of floor spectral accelerations is the highest at the roof level, gradually decreasing towards the first floor. The UHS is representative of lower storey spectra, and may be used approximately as the design spectrum for the first floor of a multistory building.

Shooshtari et al. ICC-IBC. 2006. International Building Code (IBC-2006), International Code Council, U.S.A., p. 679. Marsantyo, R., Shimazu, T., and Araki, H. 2000. Dynamic response of nonstructural systems mounted on floors of buildings. Proceedings of the 12th World Conference on Earthquake Engineering, Paper No. 1872, Auckland, New Zealand. McKevitt, W.E., Timler, P.A.M., and Lo, K.K. 1995. Nonstructural damage from the Northridge earthquake. Canadian Journal of Civil Engineering, 22(2): 428437. doi:10.1139/l95-051. Medina, R.A., Sankaranarayanan, R., and Kingston, K.M. 2006. Floor response spectra for light components mounted on regular moment-resisting frame structures. Engineering Structures, 28(14): 19271940. doi:10.1016/j.engstruct.2006.03.022. National Research Council. 2005. National Building Code of Canada. National Research Council, Ottawa, Canada. Prakash, V., Powell, G.H., and Campbell, S. 1993. DRAIN-2DX: Base program description and user guide. Version 1.10. University of California at Berkeley, Berkeley, Calif. Rodriguez, M.E., Restrepo, J.I., and Carr, A.J. 2002. Earthquakeinduced floor horizontal accelerations in buildings. Earthquake Engineering & Structural Dynamics, 31(3): 693718. doi:10. 1002/eqe.149.

1599 Rosenblueth, E. 1980. Design of earthquake resistant structures. Pentech Press Limited, Plymouth, United Kingdom. Saatcioglu, M., Shooshtari, M., Naumoski, N., and Foo, S. 2008. Development of floor design spectra for operational and functional components of concrete buildings in Canada. Proceedings of the 14th World Conference on Earthquake Engineering, 12 17 October, Beijing, China. Sankaranarayanan, R., and Medina, R.A. 2007. Acceleration response modification factors for non-structural components attached to inelastic moment-resisting frame structures. Earthquake Engineering & Structural Dynamics, 36(14): 2189 2210. doi:10.1002/eqe.724. Takeda, T., Sozen, M.A., and Nielsen, N.N. 1970. Reinforced concrete response to simulated earthquakes. Journal of the Structural Division, 96(ST12): 25572573. Uma, S.R., Zhao, J.X., and King, A.B. 2010. Seismic actions on acceleration sensitive non-structural components in ductile frames. Bulletin of the New Zealand Society for Earthquake Engineering. 43(2): 110125. Villaverde, R. 1997. Seismic design of secondary structures: state of the art. Journal of Structural Engineering, 123(8): 1011 1019. doi:10.1061/(ASCE)0733-9445(1997)123:8(1011).

Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 75.86.130.80 on 04/19/11 For personal use only.

Published by NRC Research Press