Significance of Directivity Effects During the 2011 Lorca
Total Page:16
File Type:pdf, Size:1020Kb
Bull Earthquake Eng https://doi.org/10.1007/s10518-017-0301-9 ORIGINAL RESEARCH PAPER Significance of directivity effects during the 2011 Lorca earthquake in Spain 1 2 Carlos Gordo-Monso´ • Eduardo Miranda Received: 25 August 2017 / Accepted: 17 December 2017 Ó Springer Science+Business Media B.V., part of Springer Nature 2017 Abstract The May 11th 2011, Lorca earthquake in Southeastern Spain was a moderate magnitude event (Mw 5.1) yet it caused nine fatalities, more than 300 injuries and more than 462 million euros in economic loses. Peak ground accelerations as well as response spectral ordinates far exceed expected values from various ground motion prediction models. In par- ticular, spectral ordinates computed from recorded ground motions significantly exceed those in current Spanish probabilistic seismic hazard models, as well as those in the Spanish and European building codes. The objective of this paper is to assess directivity effects on ground motions recorded during the 2011 Lorca earthquake, and to evaluate the significance of these effects in earthquake resistant design on moderate seismic regions. In the first part of this paper, we study the likelihood of the presence of a directivity pulse, by conducting a comparison of different parameters of recorded ground motions to analytical pulses. In the second part, we relate the recorded ground motion and its inelastic displacement spectra to some recent sta- tistical models that try to capture the displacement demand features of earthquakes presenting directivity-pulse characteristics. It is shown that simple analytical pulses are capable of reproducing very well pulse-type near-fault ground motions recorded during the event. It is concluded that directivity effects played a major role in the large impact caused by this relatively small event. Furthermore, directivity effects which are typically ignored, both in probabilistic seismic hazard analysis and in most building codes, may lead to important underestimations of ground motions. & Carlos Gordo-Monso´ [email protected] Eduardo Miranda [email protected] 1 E. T. S. de Ingenieros de Caminos C. y P., Universidad Polite´cnica de Madrid, 28040 Madrid, Spain 2 Civil and Environmental Engineering Department, Stanford University, Stanford, CA 94305, USA 123 Bull Earthquake Eng Keywords Lorca earthquake Á Pulse Á Directivity Á Directionality Á Near-fault Á Inelastic spectra 1 Introduction The Lorca, May 11th 2011, earthquake in South-Eastern Spain was a damaging event, especially considering its relatively low magnitude Mw 5.1 (Lopez-Comino et al. 2012; Caban˜as et al. 2013). The earthquake lead to a death-toll of 9 people, more than 300 people injured, and approximately 462 €M in direct economic losses (Alvarez-Cabal et al. 2013). The epicenter was located approximately 2 km east-northeast of the city of Lorca on the Alhama de Murcia fault in the southeastern seismic region of the Iberian Peninsula. The focal mechanism solution corresponds to a reverse and strike-strike slip faulting mecha- nism with a very shallow crustal depth of approximately 3 km (Caban˜as-Rodriguez et al. 2011), to 4 km (Martinez-Solares et al. 2012). Several source models have been proposed for this earthquake (Gonza´lez et al. 2012; Martı´nez-Dı´az et al. 2012a, b; Rueda et al. 2014; Santoyo 2014), for which a good review and discussion is provided by Moratto et al. (2017). The maximum peak ground acceleration in the event, which was recorded at the Lorca station, was 0.37 g in the N30W component and is more than three times larger than the one specified in the Spanish code for this region (based on a 10% probability of exceedance in 50 years) and is also the largest peak ground acceleration ever recorded in Spain (Caban˜as et al. 2013; Moratto et al. 2017). The city of Lorca has a population of about 92,000 habitants. The earthquake produced the collapse of a 4-story modern reinforced concrete building (Fig. 1a) and the partial collapse of an eighteenth century church (Fig. 1b). However, most of the damage was caused by the fragility of some non-structural elements (Alvarez-Cabal et al. 2013), such as masonry parapets at roof level, masonry infills and unreinforced masonry claddings in buildings (Fig. 2). It is well known that rupture directivity effects can lead to strong pulse-like ground motions (Bertero et al. 1978; Anderson and Bertero 1987; Hall et al. 1995; Iwan 1997; Iwan et al. 1998; Alavi and Krawinkler 2001). Earthquakes produce a series of shear dislocation waves that propagate away from the rupture. The propagation of fault rupture toward a site, at a velocity close to the shear wave velocity, causes most of the seismic energy from the rupture to arrive in the form of a single large pulse of motion that occurs early in the seismic record (Somerville et al. 1997). Furthermore, the radiation pattern of Fig. 1 a Collapsed modern reinforced concrete building. b Partial collapse of the Santiago Church, a masonry building dating from the eighteenth century (from Caban˜as-Rodriguez et al. 2011) 123 Bull Earthquake Eng Fig. 2 a, b Failure of unreinforced masonry cladding during the Lorca 2011 earthquake. c Failure of unreinforced masonry parapets on the roof level, the debris on the ground level are the remainings of the roof parapet completely torn apart. d Partial collapse of a reinforced concrete building (photo a from Caban˜as-Rodriguez et al. 2011, photos b–d from Alvarez-Cabal et al. 2013) the shear dislocation on the fault causes these large pulse-like ground motions that tend to be oriented in the direction perpendicular to the fault plane, leading to more intense ground motions in the normal component than in the parallel component. The potential of these near-fault pulse-like to cause large damage on structures was first recognized by Prof. Bertero and his collaborators as a result of ground motions recorded in the 1971 Mw 6.7 San Fernando Earthquake (Bertero et al. 1978). Not much attention was given to this type of ground motion until the 1994 Mw 6.9 Northridge earthquake occurred when a number of near-fault pulse-like ground motions were recorded and other investigators started working on this topic. For a literature review of work done after the 1994 Northridge earthquake the reader is referred to Alavi and Krawinkler (2001). However, as commented by Makris and Black (2004) most of the attention focused on the peak ground velocity of these pulse-like ground motions and not on the area under the acceleration pulses, referred to as the ‘‘incremental velocity’’ by Bertero et al. (1978), which provides a better characterization of what makes near-fault ground motion particularly destructive. Although near-fault pulse-like records have been included in the development of con- ventional ground motion prediction models (GMPE), formerly referred to as attenuation relationships, they do not properly account for directivity effects because most of them are aimed at predicting the geometric median of the intensity of the two horizontal compo- nents, therefore systematically underestimating the intensity of ground motions affected by directivity in the maximum direction, which is typically the fault-normal component. Furthermore, the standard deviation of most ground motion prediction models is averaged over all distances leading to underestimation of the dispersion, especially near the fault (Abrahamson 2000). The first ground motion prediction model to incorporate directivity effects was not developed until 1997, approximately 35 years after the first attenuation relations for instrumental intensity parameters were developed (Somerville et al. 1997). This first attenuation relation to incorporate directivity effects included two period-de- pendent modification factors: a first one to consider the position of the site relative to the geometry of the fault and its rupture direction, and a second one to amplify or deamplify in the fault-normal or fault-parallel directions, respectively. This first model only included directivity effects for earthquakes with magnitudes equal or larger than 6.5 Recently, such directivity effects have been observed in earthquakes with smaller magnitudes such as the 2004 Mw 6.0 Parkfield, earthquake (Shakal et al. 2005), the 2009 L’Aquila Mw 6.3 earthquake (Chioccarelli and Iervolino 2010) or the 2011 Mw 6.3 123 Bull Earthquake Eng Christchurch earthquake (Bradley et al. 2014), pointing out that these damaging pulse-like ground motions can occur even in earthquakes with magnitudes between 6.0 and 6.5. This attenuation relationship with directivity effects by Somerville et al. (1997) was later modified by Abrahamson (2000) to make the directivity model distance dependent, in order to incorporate it in probabilistic seismic hazard analysis (PSHA). However, his model or more recent ones (e.g., Shahi and Baker 2013) only incorporate directivity effects for magnitudes larger or equal to 6.0. Several researchers (Lopez-Comino et al. 2012; Rueda-Nun˜ez et al. 2012; Rueda et al. 2014; Alguacil et al. 2014; Pro et al. 2014) have conjectured the presence of a directivity pulse in the 2011 Lorca 2011 earthquake. The objective of this paper is to evaluate directivity effects on ground motions recorded during the 2011 Lorca earthquake, and to evaluate the significance of these effects in earthquake resistant design on moderate seismic regions. In the first part of this paper, we study the likelihood of the presence of that pulse, by conducting a comparison of different parameters of recorded ground motions to analytical pulses. In the second part of the paper, we relate the recorded ground motion and its inelastic displacement spectra to some of the most recent statistical models that try to capture the displacement demand features of earthquakes presenting directivity-pulse characteristics. 2 Lorca 2011 earthquake ground motion data There was an accelerograph station in the city of Lorca that recorded two horizontal ground motion signals in the N30W and N60E directions (Fig.