Field investigation on the performance of building structures during the April 25, 2015, Gorkha earthquake in Nepal

Engineering Structures 121 (2016) 61–74

Contents lists available at ScienceDirect

Engineering Structures journal homepage: www.elsevier.com/locate/engstruct

Review article

Field investigation on the performance of building structures during the April 25, 2015, Gorkha earthquake in Nepal Keshab Sharma, Lijun Deng ⇑, Carlos Cruz Noguez Dept. of Civil and Environ. Engrg., University of Alberta, Edmonton, Alberta, Canada

a r t i c l e

i n f o

Article history: Received 10 August 2015 Revised 18 March 2016 Accepted 18 April 2016

Keywords: Gorkha Nepal earthquake Field investigation Damage Reinforced concrete buildings Masonry building

a b s t r a c t On April 25, 2015, a major earthquake of moment magnitude Mw 7.8 struck the Gorkha District of Nepal at 11:56 a.m. local time (6:11 a.m. UTC). One major aftershock of Mw 7.3 on May 12, 2015, contributed to the devastation of many villages in mountainous areas nearby the epicenter. The spatial distribution of aftershocks, which extended 150 km to the east of the epicenter, suggests that the rupture propagated from west to east, thus producing severe destruction in Kathmandu, at approximately 80 km southeast of the epicenter. A total of 800,000 buildings were severely damaged or collapsed. A post-earthquake reconnaissance showed that damages in reinforced concrete buildings in urban areas were mostly due to poor construction quality, low concrete strength, non-seismic detailing in beam–column joints, and local site effects. Most of the masonry buildings in the villages nearby main shock epicenter were also affected. This paper presents the recorded accelerograms, acceleration response spectra, and the seismological aspects of the earthquake. Case histories of damaged buildings, the patterns, and the failure mechanisms are discussed in this paper. It is concluded that a majority of the damaged buildings were not designed or constructed properly in accordance with national building codes of Nepal or ACI codes. Ó 2016 Elsevier Ltd. All rights reserved.

Contents 1. 2. 3. 4.

5.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ground motions and response spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Building types in Nepal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Field observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Damages to reinforced concrete buildings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.1. Failure of beam–column joints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.2. Short column effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.3. Short splice length and splices near joints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.4. Insufficient stirrups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.5. Thin column and insufficient longitudinal reinforcement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.6. Alteration of columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.7. Quality of construction materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.8. Damages due to strong beam–weak column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.9. Soft storey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.10. Cracks on infill wall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.11. Pounding effect. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.12. Failure of water tanks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.13. Multi-storey apartment building . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Damages to masonry buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Reinforced concrete buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Masonry buildings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

⇑ Corresponding author. http://dx.doi.org/10.1016/j.engstruct.2016.04.043 0141-0296/Ó 2016 Elsevier Ltd. All rights reserved.

62 63 63 64 64 64 64 64 64 66 66 66 66 68 69 69 69 70 72 73 73 74

62

K. Sharma et al. / Engineering Structures 121 (2016) 61–74

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

1. Introduction Nepal is situated in the center of the Himalayan concave mountain chain, that is about 870 km long in the Northwest by west to Southeast by east (NWW–SEE) and 130–260 km long in the North to South (N–S) direction. The Himalayan chain is one of the most active seismic regions in the world, because of the faulting between the subducting Indian Plate and the overriding Eurasian Plate to the north. The Indian plate converges with the Eurasian plate at a rate of approximately 45 mm/year towards the north– northeast [1–3]. Over the last centuries, earthquakes in 1803, 1833, 1897, 1905, 1934, 1950, 2001, and 2005 in the Himalayan region have resulted in a large number of casualties and caused extensive damages to structures [4,5]. Researchers had warned for decades that Nepal was vulnerable to a deadly earthquake, particularly because of its geology, urbanization, poor construction practice, and a lack of disaster preparedness [6]. The Gorkha Nepal earthquake, of the moment magnitude Mw 7.8, occurred at 11:56 a.m. NST (6:11 a.m. UTC) on April 25, 2015, with the epicenter (N: 28°080 49.200 ; E: 84°420 28.800 ) nearby Baluwa village, about 80 km northwest of Kathmandu, at a focal depth of approximately 15 km [7]. A tremor that lasted for 55 s was felt in Nepal, India, Bhutan, Bangladesh, and China. Two aftershocks of Mw 6.7 and 6.3 were also felt in Nepal within 25 h after

the main shock. Another large aftershock (Mw 7.3) occurred at 12:50 p.m. NST (07:05 a.m. UTC) on May 12 in central Nepal. This earthquake was one of the most powerful seismic events since the 1934 Mw 8.1 Nepal Bihar earthquake [8]. As of August 2, 2015, the earthquakes had caused 8710 deaths and injured 22,493 people in Nepal [9]. Jaishi et al. [10], Chaulagain et al. [11], and Parajuli et al. [12] studied the performance of buildings in Kathmandu Valley based on site-specific ground motions. Shakya et al. [13] reported the performance of the buildings during the September 18, 2011, earthquake of Mw 6.9 in eastern Nepal; however, the behavior of the reinforced concrete or masonry buildings of Nepal subjected to a major earthquake is poorly understood or investigated. For this reason, post-earthquake reconnaissance activities that record the performance of structures and provide case studies are of the same significance as research activities for researchers, engineers, policy-makers, and the society in general of Nepal and other countries. A field reconnaissance was carried out by the authors immediately after the 2015 Gorkha earthquake main shock, and the observations were reported in the present paper. The objective of the field reconnaissance was to record and analyze the causes of the damage patterns observed in the buildings, mainly in Kathmandu and other rural regions. The paper discusses the seismological

Fig. 1. (a) Recorded accelerograms at KATNP for the Mw 7.8 main shock, (b) recorded accelerograms at KATNP for the Mw 7.3 aftershock, (c) 5%-damped acceleration response spectra of the accelerograms at KATNP for the Mw 7.8 main shock and the Mw 7.3 aftershock and comparison with the design spectrum of Kathmandu Valley from NBC-105 [13], and (d) location of the KATNP station and epicenters.

63

K. Sharma et al. / Engineering Structures 121 (2016) 61–74 Table 1 Key information of the earthquake. Station name

Date (m/d/y)

Time (Local)

Depth (km)

NS (cm/s2)

EW (cm/s2)

Vertical (cm/s2)

Latitude (N°)

Longitude (E°)

Region

KATNP KATNP

4/25/15 5/12/15

11:56 12:50

15 15

164 87

158 72

184 75

28.15 27.84

84.71 86.08

Gorkha Kodari

Fig. 2. Existing building types in Nepal: (a) adobe, (b) wooden, (c) stone in mud mortar, (d) brick in mud mortar, (e) stone in cement mortar, (f) brick in cement mortar, (g) non-engineered RC building, and (h) engineered RC building.

aspects of the Gorkha earthquake, describes the classifications of buildings in Nepal, and elaborates on the performance of various building types during the earthquake main shock and aftershocks. The paper offers a number of case studies in earthquake-resisting building design.

mic codes [17]. Details of the characteristics of ground motions recorded during the Gorkha earthquake can be found in Dixit et al. [18], Martin et al. [19] and Galetzka et al. [20].

2. Ground motions and response spectra

A large number of vulnerable buildings exist in Nepal. Fig. 2 shows common building types used in the country. The total number of individual households in Nepal is 5,423,297, while the population is 26,494,504 [21]. The data obtained from the census indicates that brick/stone masonry with mud mortar buildings consist of 44.2% of total buildings, followed by wooden buildings (24.9%) in rural areas. In urban areas, buildings with cementbounded brick/stone (17.6%) and cement concrete (9.9%) are widely used. Description of each building type is summarized in Table 2. More detail is available in Chaulagain et al. [22]. Among these building types, adobe buildings (Fig. 2a) are popular in rural communities in Nepal. Wooden buildings (Fig. 2b) are widely used near the forest areas in Nepal. Stone in mud mortar (Fig. 2c) is widely found in hills and mountain. Brick in mud mortar building (Fig. 2d) is very common in old settlement in Kathmandu Valley and areas accessible by road. Adobe, brick and stone masonry with mud mortar are the most vulnerable to an earthquake due to the large mass and brittle low strength of the materials and because of the lack of proper detailing and maintenance. Public buildings such as schools and hospitals in hilly areas are made of stone masonry with cement mortar (Fig. 2e). Brick in cement mortar (Fig. 2f) is the most common in urban areas accessible by road. Reinforced concrete (RC) moment-resisting frame structures (Fig. 2g and h) are the most common type of construction in public and commercial buildings after the 1988 Udayapur earthquake. Structures of this type include schools, hospitals, governmental offices, hotels, and financial and business centers. With

The capability of the network for recording strong ground motions in Nepal is limited. The U.S. Geological Survey (USGS) has one seismological station at Kanti Path, Kathmandu, and makes the records obtained at this station (KATNP) publicly available. The KATNP station is located at the core of Kathmandu city (Fig. 1d). Critical parameters for the April 25th main shock and the May 12th aftershock are given in Table 1. The EW, NS and UD components of recorded accelerograms are shown in Fig. 1a and b. Long-period components can be identified from the acceleration response spectra (Fig. 1c), which may be the results of the soft sedimentary basin effects [14,15]. The peak ground acceleration (PGA) of the recorded ground motions was 150–170 cm/s2 and 70– 80 cm/s2 for the Mw 7.8 main shock and the Mw 7.3 aftershock, respectively. The PGA recorded at KATNP station did not exceed the PGA estimates with 10% probability of exceedance in 50 years from the recent regional seismic hazard studies by JICA [6] and Ram and Wang [16]; however, it is speculated that the local site effects might have contributed to the significant amplification of the motions in Kathmandu Valley and thus make the effects of the earthquake more influential. The USGS preliminary estimation of the PGA in the epicentral area was about 350 cm/s2 [7]. The 5% damped response spectra of the recorded accelerograms for the Mw 7.8 main shock and the Mw 7.3 aftershock are calculated and shown in Fig. 1c. Fig. 1c also shows design response spectra in Zone A, i.e. the Kathmandu Valley, of Nepal according to the Nepal seis-

3. Building types in Nepal

64

K. Sharma et al. / Engineering Structures 121 (2016) 61–74

Table 2 Brief description of existing building typologies in Nepal. Building type

Description

Figure reference

1. Adobe

Materials made of clay, water, and organic matters with or without wooden frame, wall thickness usually about 350 mm, highly vulnerable to earthquakes Constructed using wooden posts from tree trunks and walls made with wooden planks or woven bamboo with mud plastering Walls made of stone with mud mortar, and a building frame made of wood, flexible floors and roofs, highly vulnerable

Fig. 2a

Unconfined, unreinforced brick masonry in mud mortar with either wood or masonry lintels, and wood-framed floors supporting heavy mud floor and roof slabs, wall thickness 300–450 mm, highly vulnerable Dressed or undressed stones with cement mortar with or without light reinforcement, metal or wooden framed floors/roof, generally iron sheet as a roofing material, vulnerable Low residential building, burned bricks with cement mortar generally with light reinforcement, wall thickness 230 mm (outer) or 115 mm (inner), vulnerable Not designed structurally, RC frame with brick infills, cast-in-place concrete beams and columns with cast-in-place concrete slabs for floors and/or roof, lack of sufficient seismic resistance, vulnerable RC frame with brick infills, cast-in-place concrete beams and columns with cast-in-place concrete slabs, often follow the Nepal building codes and Indian standard codes for seismic design, special detailing to provide ductile performance, relatively safe

Fig. 2d

2. Wooden 3. Stone in mud mortar 4. Brick in mud mortar 5. Stone in cement mortar 6. Brick in cement mortar 7. Non-engineered RC 8. Engineered RC

some exceptions, RC moment-resisting frames are also used in the construction of low residential buildings. Most of the RC buildings prior to the 1988 Udayapur earthquake are non-engineered (Fig. 2g), i.e. not designed by structural engineers, and thus lack sufficient seismic resistance. Engineered RC buildings (Fig. 2h) are relatively new and constructed in big cities of Nepal. 4. Field observations The route and location of investigation sites are shown in Fig. 3 along with the epicenters of the main shock. The Kathmandu Valley and regions near the main shock epicenter were surveyed to study the performance of reinforced concrete and masonry buildings. The Kathmandu Valley is covered with thick lacustrine and fluvial deposits of more than 550 m depth, which is originated from Pliocene to Pleistocene. These unconsolidated deposits were mainly derived from the surrounding hills by the rivers. 4.1. Damages to reinforced concrete buildings RC moment-resisting frame structures are the most common type of construction in urban areas including Kathmandu Valley. According to current Nepali code for the seismic design of buildings [17], capacity and ductile design are the basis for the earthquake design of RC frames, with strong-column-weak-beam systems and column beam-joints being capable of ensuring the formation of plastic hinges at the beam ends. In Nepal, typically RC buildings have burnt clay bricks or stones as wall filling materials. Noticeable features of this type of buildings are: (i) absence of RC or masonry bond beams above doors and windows in private buildings; brickwork is generally supported directly by the wooden frame used for doors/windows; (ii) floating columns in upper storey, (iii) intermediate soft storey in multi-storey buildings, and (iv) poor reinforcement detailing. Structural and construction deficiencies and associated damages to RC buildings observed during field visit are presented the present section. 4.1.1. Failure of beam–column joints Severe damages to beam–column joints were observed frequently during the survey. Fig. 4 shows a few typical cases to illustrate the main reasons for the failure or damage of beam–column joints. A lack of transverse reinforcement was observed in the majority of the damaged columns, an example being shown in Fig. 4a. insufficient stirrup spacing near the beam–column joint was found to be the mechanism of the severe damage shown in Fig. 4b and c, in which the stirrup spacing was larger than

Fig. 2b Fig. 2c

Fig. 2e Fig. 2f Fig. 2g Fig. 2h

175 mm, which is more than two times of maximum spacing (75 mm) specified in Nepal National Building Code [23] and about three times of maximum spacing (58 mm) specified in ACI 318 code [24] for a column with the details observed during the survey. 4.1.2. Short column effect Short column failures were widely observed during the field assessment, such as the cases shown in Fig. 5. Fig. 5a shows the formation of a short column due to intermediate staircase landing in between two floors, while Fig. 5b shows the short column effect due to partial infilled frames to fulfill functional requirement of lighting and ventilation. Adequate detailing, with closer stirrup spacing, is required to prevent short column failures, which may develop due to inappropriate structural arrangements, or from continuous openings at the top of infill walls not foreseen in the design of columns. If short columns cannot be avoided, specific detailing for transverse reinforcement is required in seismic zones [23,24]. 4.1.3. Short splice length and splices near joints Fig. 6a, c, and d shows the collapsed 4-storey residential building, the column of a warehouse, and a 6-storey hotel in Kathmandu Valley. A common failure mechanism observed in these RC buildings was the insufficient splicing of main rebars at or near beam–column joints, which is an undesirable location for splicing [23,24], as shown in Fig. 6b, c, and e. Similarly, existence of short lap splices in tension at plastic hinge zones, lack of end hook angle, and widely spaced stirrups (Fig. 6c) increased the severity of the damage. As per the Nepal National Building Code [23] and ACI 318 code [24], all lap splices in columns in seismic regions must be designed as tension lap splices. For the 300 mm by 300 mm size columns and 16 mm diameter main rebars used in these structures, the lap splice length should be at least 896 mm and 752 mm as per Nepal National Building Code [23] and ACI 318 code [24] respectively. As shown in Fig. 6, the maximum lap splice lengths were found to be about 420 mm, which is 44% shorter than the minimum ACI 318 code [24] requirement. Both Nepal National Building Code [23] and ACI 318 code [24] allow lap splices only at the middle of the member, with adequate transverse reinforcement. The irregular geometry in plan (Fig. 6a) also increased the severity of beam-–column failure. Thus, the required ductility in these RC members cannot be maintained. 4.1.4. Insufficient stirrups It was found that the stirrups in the columns of damaged RC buildings (Figs. 6c, 7a, c and e) were poorly detailed. The length of the hook provided in stirrups was short and had inappropriate 90° end hooks, as shown in Fig. 7b, d, and f. It is observed from

K. Sharma et al. / Engineering Structures 121 (2016) 61–74

65

Fig. 3. Survey routes, locations of investigation sites that are referred in this report and the epicenters of the main shock according to USGS (modified from Google map).

Fig. 4. Performance of selected RC moment-resistant frame buildings: (a) a 6 storey residential building in Kalanki (N: 27.711744, E: 85.283072), (b) Deep Jyoti secondary school in Balaju, Kathmandu (N: 27.737957, E: 85.309560), and (c) commercial building (Civil Bank) in Kalanki, Kathmandu (N: 27.71278, E: 85.28279).

Fig. 5. Shear failure of the column due to short column effect: (a) a short column due to intermediate staircase landing in between two floors (N: 27.771751, E: 85.720258), and (b) short column effect due to partially infilled frames (N: 27.733939, E: 85.309280).

Fig. 7b and f that the concrete covering the ties spalled, thus facilitating the opening of the ties. Fig. 6e also shows the detail of the column in the collapsed hotel building in Gongabu, Kathmandu. Sufficient vertical reinforcement was used, with columns and beams of adequate size. However, only one 2-legged stirrup was

used and the bars were spliced at the joint as shown in Fig. 6e. The most critical failure mode observed in columns was shear failure, shown in Fig. 8. Once diagonal cracking occurs (Fig. 8a), the aggregate interlock in the concrete decreases rapidly, resulting in a very brittle and sudden failure if adequate stirrup reinforcement

66

K. Sharma et al. / Engineering Structures 121 (2016) 61–74

Fig. 6. Short lap splices and spliced at beam–column joint: (a) a 4 storey residential building at Swayambhu (N: 27.689236, E: 85.286871), (b) short lap splice and spliced at joint, (c) short lap spliced in a warehouse at Kalanki (N: 27.713248, E: 85.282875), (c) a 6-storey hotel at Gongabu, Kathmandu (N: 27.738114, E: 85.307635), and (e) inappropriate splice at the joint.

is not provided. Excessive spacing of transverse reinforcement throughout the column, ranging from 150 mm to 300 mm, was observed in most of the damaged columns shown in Figs. 6c, 7 and 8. The maximum spacing of column stirrups as per the Nepal National Building Code [23] and ACI 318 code [24] is about 60 mm and 75 mm respectively for a column with the details observed during the survey. This caused shear failures, buckling of longitudinal rebar and poor confinement of the core concrete as shown Figs. 7b, f and 8b. 4.1.5. Thin column and insufficient longitudinal reinforcement In addition to inadequate stirrups, insufficient column crosssectional dimension was another reason for the column damage of the cases shown in Figs. 8 and 9. Fig. 9a shows a 7-storey building in Kapan, Kathmandu, which overturned and collapsed during the earthquake; the provided reinforcement ratio of this column was less than 1%, which is the minimum requirement according to the ACI 318 code [24]. 4.1.6. Alteration of columns An interesting observation during the field survey was the performance of RC structures that was improperly altered before the Gorkha earthquake by owners. RC moment-resisting buildings in many regions of Nepal have been fitted with rolling shutters (Fig. 10a) at the ground floor for commercial use. When installing rolling shutters, the cover of the RC columns is often removed at a few spots along the column axis, and the main rebars are exposed and welded to the shutter guide, as shown in Fig. 10b and d. Many buildings with rolling shutters were found severely damaged, as shown in Fig. 10b and c, due to the column failure. The installation

of the rolling shutter significantly changed the stiffness and moment capacities of the RC columns by: (1) reducing the crosssectional area of the column; (2) reducing the strength and stiffness of the rebars due to the overheating caused by the welding processes. 4.1.7. Quality of construction materials It is observed that the quality of the concrete used in the damaged or collapsed buildings was poor. Examples of the use of low strength concrete observed in the RC buildings are shown in Figs. 8b and 10b. The concrete from these buildings could be easily crumbled by hand. In addition, cobble-sized aggregates were used in the concrete of these buildings (Figs. 7f and 10b). Fig. 10d shows that a large stone was dropped into freshly-cast concrete to economize on the volume of concrete used also known as the plum concrete. The use of large-diameter aggregates might have significantly decreased the strength of the concrete materials. The aggregates seemed to be poorly (or uniformly) graded, which led to a honeycomb pattern in the cast concrete and resulted in a further reduction in concrete strength. 4.1.8. Damages due to strong beam–weak column Strong beam weak column connections were often observed in the damaged or collapsed RC buildings. Examples are presented in Fig. 7c and e. In general, as a column become weaker compared to the beam, the ductility capacity of the structure is reduced. The post-yield behavior of the weak column strong beam mechanism is highly degraded with significant strength deterioration and little displacement capacity, indicating the vulnerability of structures with weak columns and strong beams. Fig. 11 shows the case study

K. Sharma et al. / Engineering Structures 121 (2016) 61–74

67

Fig. 7. (a) A collapsed building near Kalanki (N: 27.723985, E: 85.283295), (b) opened stirrups near beam column joint and short hook, (c) a departmental store building in Swayambhu, Kathmandu (N: 27.723916, E: 85.283283), (d) a short hook at 90° angle and corrosion of reinforcement, (e) collapsed commercial building in Kuleshwar (N: 27.686799, E: 85.298063), and (f) wide spread stirrups and hook at 90° angle.

Fig. 8. Column failure of residential building in Balkhu, Kathmandu (N: 27.682496, E: 85.295434): (a) bulking of main bars, and (b) close-up view of bulked column and widely spaced stirrups.

68

K. Sharma et al. / Engineering Structures 121 (2016) 61–74

Fig. 9. Overturning of a 7-storey building (N: 27.726948, E: 85.350465): (a) thin column, not adequate reinforcement, short splice length and spliced at beam column joint, and (b) the column section showing a low reinforcement ratio of the column in this case study.

Fig. 10. Performance of buildings altered by roller shutter (N: 27.712782, E: 85.282797): (a) overview, (b) welding of roller shutter frame with main bars, (c) lack of transverse bar, bulking of main bar and large size aggregate, and (d) large-diameter cobbles used in concrete.

of a new 3-storey residential building in Balkhu, Kathmandu. It is seen that severe damage occurred at the top end of the column, whereas minimal damage was observed at the corresponding beam ends. 4.1.9. Soft storey Soft first storey is a typical feature in the RC buildings in Kathmandu Valley because the first stories of the building have been often used as commercial areas, shops, or car parking. There are no partition walls, and often there are no outer walls in the first

storey. Soft storey floor behavior is created involuntarily by building owners due to the elimination or reduction of infill walls at the ground floor level. Soft storeys result in increased displacement demands and place the burden of energy dissipation on the first storey columns. Similar damage pattern as a result of soft storey were found in earthquakes in other areas of the world [25,26]. Fig. 12 shows two multi-storey buildings which suffered softstorey failure at the second and first floor respectively resulting from not considering the change in relative stiffness provided by the infill walls between the first and the upper storey.

K. Sharma et al. / Engineering Structures 121 (2016) 61–74

69

Fig. 11. A tilted residential building in Balkhu (N: 27.682034, E: 85.294746): (a) overview of building, (b) and (c) details of column-beam joint damage.

Fig. 12. Typical soft storey failures (no partition and outer wall on collapsed floor): (a) a 4-storey building in Baphal (N: 27.711732, E: 85.337861) and (b) a 5-storey building in Kalanki, Kathmandu (N: 27.711744, E: 85.283072).

4.1.10. Cracks on infill wall Infill walls made of burnt bricks are common in the earthquakeaffected area. Fig. 13a shows the 6-storey building of Acme Engineering College, Sitapaila, Kathmandu. A close-up view at the first storey (Fig. 13b) shows severe damage at the infill walls. Most of the damage occurred in areas of the wall where openings were close to the corners. Fig. 13c shows a 6-storey commercial building in Sitapaila, Kathmandu; close-up views of infill wall damage at the second and third storey of the building are shown in Fig. 13d, where burnt bricks were used as infill wall material. The damage suffered by the infill walls, which are offset from the column axis to increase the office room size, is clearly non-structural, and results from the low construction quality of the infill walls and its inability in accommodating the large drifts experienced by the building at the second and third storey. 4.1.11. Pounding effect Structural pounding refers to the lateral collisions of adjacent buildings during earthquakes. Pounding occurs when building separations are insufficient to accommodate the relative motions of

adjacent buildings. In Kathmandu Valley, many buildings are constructed closely or in contact with each other and therefore are particularly susceptible to the pounding damages. Two case studies are shown in Fig. 14. Insufficient building separation is often due to the high cost of land and small lot sizes in core cities.

4.1.12. Failure of water tanks It is a common practice to place a water storage tank at the roof of the RC frame residential buildings in Kathmandu Valley, as shown Fig. 15a. The tank is made of reinforced concrete and is usually supported by short concrete columns projecting above the roof level. In general, these water tanks are not accounted for in the building design and are constructed as per the owner instruction and requirement. The storage capacity of the tank depends on the size of the building, but is of the order of 0.5–5 m3. Due to their rigid support system, the tanks have a reduced energy absorption capacity and thus experienced severe damage. Many tanks were filled with water at the time of the earthquake, increasing the inertial forces. Damage of a water storage tank is shown in Fig. 15b.

70

K. Sharma et al. / Engineering Structures 121 (2016) 61–74

Fig. 13. Damages on infill walls: (a) a college building, Sitapaila, Kathmandu (N: 27.708111, E: 85.284382), (b) wall cracks in the building (a), (c) a commercial building in Sitapaila, Kathmandu (N: 27.707545, E: 85.283390), and (d) collapsed infill wall of the building (c).

Fig. 14. Damage on building caused by pounding effect due to closely spaced building in Kathmandu (left-N: 27.73803, E: 85.30751, right-N: 27.73800, E: 85.30756).

4.1.13. Multi-storey apartment building Although newly constructed multi-storey apartment buildings are considered as a well-engineered building type in Nepal, severe cracks were found on many infill walls of this type of buildings. It was found that out of 15 newly constructed multi-storey

apartment buildings in Kathmandu Valley, 11 were severely affected by the earthquake. Fig. 16a shows a 16-storey, apartment complex, Park View Horizon, which exhibited many major cracks (Fig. 16b) along its height. Non-structural burnt bricks were used as the infill material, which led to a significant

K. Sharma et al. / Engineering Structures 121 (2016) 61–74

71

Fig. 15. (a) Common practice of constructing water tank in residential buildings (N: 27.685201, E: 85.296675), and (b) a collapsed water tank supported by short columns above the roof (N: 27.685278, E: 85.296118).

Fig. 16. Damage of tall apartment building in Kathmandu: (a) Park View Horizon apartment (N: 27.739762, E: 85.324410), (b) cracks at the wall of Park View Horizon.

increase in weight. Causes of the major damage in wellengineered multi-storey apartment buildings in Kathmandu Valley may be attributed to the long-period ground motions as shown in the spectra of Fig. 1, which shows a predominant period ranging from 0.4 to 0.6 s. In addition, local site effects may have contributed to extensive damage to the apartment as the apartment is located on a small hill as amplification occurs at and near the top of the hill over a broad range of frequencies [27,28].

The Nepal National Building Code 201 [23] ensures adequate detailing of the reinforcement concrete members and connections to allow for the development of ductility and promoting energy dissipation. However, the survey showed that most of the RC building frames did not perform as intended. Most failures developed at the columns, either through shear failure, or through excessive deformation demands at the ground floor columns in the presence of soft-storey mechanisms, leading to partial to full collapse of the building structure.

Fig. 17. Totally devastated small villages: (a) at Baluwa (near epicenter of Mw 7.8), Gorkha, (N: 28.170005, E: 84.706527), and (b) at Thimi, Bhaktpur (N: 27.735881, E: 85.301887).

72

K. Sharma et al. / Engineering Structures 121 (2016) 61–74

Fig. 18. Cracks on the masonry wall: (a) vertical cracks near the corner (N: 27.757277, E: 85.278215), and (b) diagonal cracks starting from corner of the openings (N: 27.758100, E: 85.285615).

4.2. Damages to masonry buildings Most of the buildings in the rural areas of Nepal were built using materials found in local areas i.e. adobe, bricks, and stones. Many residential and public buildings in Kathmandu built before the 1980s were unreinforced masonry buildings, which were not designed or constructed in accordance with the Nepal National Building Code 203 [29]. Generally, stone-masonry structures in the affected area were made of undressed stones bonded with mud mortar. A few important structures (e.g. schools, hospitals) have dressed stone masonry with mud and cement mortar. In general, masonry buildings suffered damages, primarily because of the low strength mortar used. The masonry in these buildings was unreinforced and the walls were not tied to each other or to the floors and roofs. Earthquake-resistant features such as horizontal ties at various levels and stones at the corners are generally not provided in such constructions. This resulted in the formation of severe cracks near the corners and at the location of openings in such buildings, even when subjected to moderate shaking. A large portion of the unreinforced masonry buildings completely collapsed, or were heavily damaged, near the epicenter of the main shock and aftershock, of Mw 7.3. Fig. 17a shows the overview of a totally devastated small village near the epicenter of the main shock (Gorkha, Nepal). No single standing building was found in this village, with all the roofs of unreinforced

masonry buildings collapsing. Similarly, Fig. 17b shows damaged brick masonry buildings with mud mortar in Thimi, Bhaktpur. Fig. 17 clearly highlights the severity of damage to unreinforced masonry buildings in Nepal. Masonry walls of several buildings in the affected area suffered damages in the form of vertical or inclined shear cracks as shown in Fig. 18. Failures and cracks at the corners as shown in Fig. 18a might have been triggered by a few mechanisms: (1) insufficient connections or shear transfer between the walls and floors may result in a poor integrity that led to the formation of vertical cracks; (2) uneven settlement of the foundation; however, noticeable subsidence was not observed around or nearby the building; (3) out-of-plane bending, caused by seismic loads perpendicular to walls, may also have triggered vertical cracks. Unreinforced masonry buildings are vulnerable to flexural outof-plane failure. If the connection between the walls and floors is not adequately implemented, the whole wall panel, or a significant portion, may overturn due to the lateral seismic excitation. Fig. 19a shows an example of out-of-plane collapse of load-bearing walls. Gable walls are extremely vulnerable to drift and out-of-plane lateral loads because they are largely unsupported infill walls detached to the roof structures. One important mechanism of the damages to unreinforced masonry building was the collapse of gable walls, as shown in Fig. 19b. Failure of gable walls might also trigger the failure of the lateral (shorter) wall as shown in Fig. 20a. Due to of the poor performance, stone masonry gable walls should

Fig. 19. Failure of unreinforced masonry walls: (a) out of plane failure, Durbar High school, Kathmandu (N: 27.707426, E: 85.314096), and (b) gable wall failure near the epicentre of the Mw 7.8 main shock (N: 28.025110, E: 84.587180).

K. Sharma et al. / Engineering Structures 121 (2016) 61–74

73

Fig. 20. Partially collapsed stone masonry wall (N: 28.129978, E: 84.539998): (a) opening in short wall, cracks from opening, and (b) separation of wall vertically due to lack of through stone, cavity in wall, and stock of small boulder on wall.

Fig. 21. Schematics of (a) conventional wall section without through stone, (b) wall section with through stones, and (c) wall plan with through stones.

observed in structural walls, as shown in Figs. 19b and 20b. This was another source of damage because such cavities decrease the flexural and shear strength during the earthquake. The cavities were not quantified during the field study, but previous studies have showed that cavities in structural walls have a detrimental effect on the structural performance [31]. Additionally, thick walls made with small size rubbles (Figs. 19 and 20) and no through (key) stones led to walls splitting into two parts (Fig. 20b). Reinforcing the masonry structures can significantly enhance the seismic performance of the structures. It was observed during the field visit that two reinforcement measures with minimal cost resulted in an efficient method in preventing heavy damages to masonry structures. The first measure uses the through stones (also known as key stones) that was illustrated in Fig. 21; the through stones can provide limited resistance to lateral seismic loads and thus probably prevented the out-of-plan failure of masonry walls. The second measure uses the horizontal ties as shown in Fig. 22 in which the masonry building performed well during the earthquakes. The horizontal ties provided on sill and lintel level integrate the building elements into a single block and prevent the crack formation from the opening. The brick pier of the building in Fig. 22 also increases the resilience to seismic loads. Additionally, proper maintenance, good-quality materials, and good workmanship during construction, and regular maintenance seem to have improved the seismic performance of masonry buildings observed in the field trip. 5. Conclusions

Fig. 22. Performance of brick masonry building with seismic elements in Bhaktpur, Nepal (N: 27.650841, E: 85.448181).

be replaced by lightweight materials, i.e. wooden plank or steel sheet, as recommended in [30]. A large portion of stone masonry structures built with mud mortar completely collapsed or were heavily damaged, as shown in Fig. 20. In Nepali villages, wall construction is mainly done with irregularly shaped stones having smooth surfaces and bonded with mud mortar. Due to the reduced bond strength, lateral loads caused the stones to slide horizontally. Openings on walls further reduce the resistance to seismic loading (Fig. 20a). Cavities were

The aim of this paper is to investigate the damage and collapse mechanisms observed in Nepali buildings during the 2015 Gorkha earthquake that struck Nepal on April 25, 2015. A field reconnaissance was conducted immediately after the main shock. This paper described the seismological aspect of the earthquake and discussed the building types of Nepal. The paper summarized the patterns of damages or failures of various types and analyzed the possible mechanisms of the damages. The following conclusions could be reached based on the reconnaissance. 5.1. Reinforced concrete buildings The primary mechanism of collapse was shear failure caused by the wide spacing of the stirrups, buckling of longitudinal rebar, and the poor confinement of the core concrete. Embedded hook length was not enough for stirrups to be effective. Inadequate crosssectional area of columns was another reason of the damage. Insufficient longitudinal reinforcement (less than 1.0%) in columns,

74

K. Sharma et al. / Engineering Structures 121 (2016) 61–74

short lap splices, and incorrect end hook angle of main bars contributed to the severe damages or collapses of buildings. Low quality concrete was used in the affected area and undermined the seismic performance of reinforced concrete buildings. Failures and damages of reinforced concrete buildings due to the soft stories were observed because of elimination of infill wall due to social, public or commercial needs. Damage and failures related to strong beam weak columns mechanism were observed. Improper modification to structural systems by owners was observed to have detrimental effects. Many houses with rolling shutters at the ground level were severely damaged. The majority of rolling shutters were installed inappropriately, such that the adjacent RC columns were damaged; main reinforcements were exposed and welded with the shutter guide. The causes of the damage of a few multi-storey apartment buildings in Kathmandu Valley may be attributed to the local site effect that results in long-period ground motions. The thick and soft geological deposits in Kathmandu Valley tend to amplify the motions at the ground. 5.2. Masonry buildings The causes of the damage of masonry buildings were observed to be poor construction detailing, poor masonry material properties, irregularly shaped stones having smooth surfaces, weak structural walls, unconfined gable walls, and cracks at the corners of windows and doors. Minimum reinforcement measures using through stones in the walls or using horizontal and vertical bands significantly improved the seismic performance of masonry buildings. Overall, the 2015 Nepal earthquake showed the need of developing structural systems with adequate strength, stiffness, ductility, and redundancy in areas subject to significant earthquake forces. The importance of involving experienced engineers and enforcing code compliance at all phases of building design is highlighted, as most of the structures which failed were nonengineered. All modifications to buildings must be avoided until an engineer may evaluate the impact and feasibility of the change. Special emphasis must be placed in the adequate maintenance and seismic upgrade of structures. Providing structural integrity to floor diaphragms, wall systems, and strengthening of walls (for in-plane and out-of-plane bending) and frames proved useful to reduce vulnerability in both RC and masonry structures. Avoidance of brittle details, such as lap splices and insufficient confinement in columns, effectively prevents sudden and catastrophic failures. Acknowledgements The authors are grateful to the financial support of Natural Sciences & Engineering Research Council of Canada under the Discovery Grants program (RGPIN-2014-04707) and the University of Alberta – Joint Research Labs program. The post-earthquake reconnaissance was partially funded by the Japan Society for the Promotion of Science (JSPS). The authors appreciate the discussions with Prof. T. Kiyota of the University of Tokyo, Dr. K. Goda of the University of Bristol, and H.K. Adhikari of Care Nepal, during the reconnaissance. References [1] Paul J, Buergmann R, Gaur VK, Bilham R, Larson KM, Ananda MB, et al. The motion and active deformation of India. Geophys Res Lett 2001;28(4):647–50. [2] White LT, Lister GS. The collision of India with Asia. J Geodyn 2012;56– 57:7–17.

[3] Copeland P. The when and where of the growth of the Himalaya and the Tibetan Plateau. In: Ruddiman WF, editor. Tectonic up lift and climate change. New York: Plenum Press; 1997. p. 19–40. [4] Chitrakar GR, Pandey MR. Historical earthquakes of Nepal. Bull. Geol Soc Nepal 1986;4:7–8. [5] Bilham R, Gaur VK, Molnar P. Himalayan seismic hazard. Science 2001;293:1442–4. [6] Japan International Cooperation Agency (JICA). The study of earthquake disaster mitigation in the Kathmandu Valley, Kingdom of Nepal. Final Report, 2002; I–IV. [7] United States Geological Survey (USGS). [accessed on 10 June 2015]. [8] Ambraseys NN, Douglas J. Magnitude calibration of north Indian earthquakes. Geophys J Int 2004;159:165–206. [9] Nepal Police; 2015. . [10] Jaishi B, Ren WX, Zong ZH, Maskey PN. Dynamic and seismic performance of old multi-tiered temples in Nepal. Eng Struct 2013;25(14):1827–39. [11] Chaulagain H, Rodrigues H, Jara J, Spacone E, Varum H. Seismic response of current RC buildings in nepal: a comparative analysis of different design/construction. Eng Struct 2013;49:284–94. [12] Parajuli HR, Maskey PN, Taniguchi H. Vulnerability assessment of the old brick masonry buildings. Disaster Mitigation Cultural Heritage Historic Cities 2012;6:61–6. [13] Shakya K, Pant DR, Maharjan M, Wijeyewickrema AC, Maskey PN. Lessons learned from performance of buildings during the September 18, 2011 earthquake in Nepal. Asian J Civil Eng 2012;14(5):719–33. [14] Sakai H, Fujii R, Kuwahara Y. Changes in the depositional system of the PaleoKathmandu Lake caused by uplift of the Nepal Lesser Himalayas. J Asia Earth Sci 2002;20:267–76. [15] Paudyal YR, Bhandary NP, Yatabe R. Seismic microzonation of densely populated area of Kathmandu Valley of Nepal using microtremor observations. J Earthquake Eng 2012;16(8):1208–29. [16] Ram TD, Wang G. Probabilistic seismic hazard analysis in Nepal. Earthquake Eng Eng Vib 2011;12:577–86. [17] Nepal National Building Code (NBC)-105. Nepal national building code for seismic design of buildings in Nepal, ministry of housing and physical planning, Department of Buildings, Kathmandu, Nepal, 1994. [18] Dixit AM, Ringler AT, Sumy DF, Cochran ES, Hough SE, Martin SS, et al. Strongmotion observations of the M 7.8 Gorkha, Nepal, earthquake sequence and development of the N-SHAKE strong-motion network. Seismol Res Lett 2015;86(6):1533–9. [19] Martin SS, Hough SE, Hung C. Ground motions from the 2015 M 7.8 Gorkha, Nepal earthquake constrained by a detailed assessment of macroseismic data. Seismol Res Lett 2015;86(6):1524–32. [20] Galetzka J, Melgar D, Genrich JF, Geng J, Owen S, Lindsey EO, et al. Slip pulse and resonance of Kathmandu basin during the 2015 Mw 7.8 Gorkha earthquake, Nepal, imaged with space geodesy. Science 2015;349 (6252):1091–5. [21] National Population and Housing Census 2011 (National Report). Government of Nepal, National Planning Commission Secretariat, Central Bureau of Statistics, Kathmandu, Nepal. [22] Chaulagain H, Rodrigues H, Scapone E, Varum H. Seismic response of current RC buildings in Kathmandu Valley. Struct Eng Mech 2015;53(4):791–818. [23] Nepal National Building Code (NBC)-201. Nepal National Building Code for Mandatory Rules of Thumb for Reinforced Concrete Buildings with Masonry Infill, Ministry of Housing and Physical Planning, Department of Buildings, Kathmandu, Nepal, 1994. [24] ACI Committee 408. Building code requirement for structural concrete (ACI 318–14) and Commentary (ACI 318R–14). American Concrete Institute, Farmington Hills, MI, 2014. [25] Sezen H, Whittaker AS, Elwood KJ, Mosalam KM. Performance of reinforced concrete buildings during the August 17, 1999 Kocaeli, Turkey earthquake, and seismic design and construction practice in Turkey. Eng Struct 2003;25 (1):103–14. [26] Kaushik HB, Dasgupta K, Sahoo DR, Kharel G. Performance of structures during the Sikkim earthquake of 14 February 2006. Curr Sci 2006;91(4):449–55. [27] Bouchon M, Schultz CA, Tokstz MN. Effect of three dimensional topography on seismic motion. J Geophys Res 1996;101:5835–46. [28] Gazetas G, Kallou PV, Psarropoulos PN. Topography and soil effects in the MsZ5.9 Parnitha Athens earthquake: the case of Adames. Nat Hazards 2002;27:133–69. [29] Nepal National Building Code (NBC)-203. Nepal National Building Code for Guidelines for earthquake resistant building construction: low strength masonry, Ministry of Housing and Physical Planning, Department of Buildings, Kathmandu, Nepal, 1994. [30] Hausler E, Hart TM, Goodell G. Case study: design and construction of confined masonry houses in Indonesia. In: Proc 10th National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 2014. [31] Bayraktar A, Altunısßık AC, Muvafık M. Field investigation of the performance of masonry buildings during the October 23 and November 9, 2011, Van Earthquakes in Turkey. J Perform Constr Facil 2012.