You are on page 1of 8

Built Heritage 2013 Monitoring Conservation Management

Pushover seismic analysis of masonry buildings with different commercial codes


Stefania Arangio; Francesca Bucchi; Franco Bontempi Sapienza University of Rome, Department of Structural and Geotechnical Engineering, Rome, Italy 1. Introduction The seismic assessment of historical masonry buildings is a complex task because the global behavior of this kind of structures depends on numerous factors, as the behavior of the single walls, the connections between them, the typology and stiffness of the oor (exible or rigid diaphragms), and the strong nonlinearities of the material [Calderoni et al., 1994; Lenza, 1996; Boothby et al., 2006; Bontempi et al., 2012]. Various methods can be applied: they range from equivalent frames approaches to detailed nite element models (FEM). In this work two approaches, implemented in some widely used commercial codes, are discussed and compared. They are the Simplied Analysis Method (SAM) based on the equivalent frames approach, that can be implemented with nite element codes as SAP2000 or Strand7/Straus7, and the macroelement approach implemented by the code 3Muri. In particular, some analyses of a simple three-dimensional building are carried out in order to investigate the inuence of various parameters, such as the quality of the masonry, the type of the oor or the presence of reinforcements. Subsequently the results obtained with the different codes are compared. This investigation is carried out with the purpose of understanding the sensitivity of the various parameters involved in the modeling. 2. Masonry buildings modeling: Simplied Analysis Method and macroelements approach In this work two modeling approaches are considered: the Simplied Analysis Method (SAM) and Gambarotta et al.s macro-element. In the SAM, developed since 1996 by Magenes and Calvi (1997), and then modied by Magenes and Della Fontana (1998), the wall is schematized with an equivalent frame composed by column elements representing the piers; beam elements representing the spandrels; rigid offsets describing the joint panel. The joints are usually considered innitely rigid. SAM is an extension of the well-known POR approach [Tomaevic, 1978] but in the SAM the spandrel is considered deformable; furthermore it can move horizontally (as in POR) and can rotate (not allowed in POR). In order to describe the seismic behavior of masonry structures, it is possible to classify the spandrels in three classes [Calderoni et al., 2011]: (1) weak spandrels that have not tension resistant elements, as steel chains or similar; (2) strut spandrels that have at least one tensile resistant element that allows the setup of a strut and tie resistance mechanism; (3) beam spandrels that are reinforced with two elements, one above and one below the panel. The type of spandrel affects the response of the adjacent piers that can act as: (1) cantilever, in case of weak spandrels, (2) partially coupled, when the spandrel is of strut type, (3) shear type, in case of beam spandrels. Both
773

Built Heritage 2013 Monitoring Conservation Management

the pier and the spandrel are modeled considering an elastic-plastic behavior with a deformation limit; the plasticity is model by using plastic hinges that are activated when a dened threshold is overcame. The other considered approach is based on the use of macro-elements. In different forms they were proposed by various authors, as for example DAsdia and Viskovic (1995), Gambarotta and Lagomarsino (1997), Cali et al. (2012). The idea at the base of the approach is that the wall is described by a set of macroscopic no tensile elements, which represent piers, spandrels and joints. The advantage of the macro-elements approach is represented by a reduction of the degrees of freedom that reduces the computational effort. Gambarotta et al. (1997) proposed a macro-element that works in plane and that was implemented in a commercial code, 3Muri. More technical details are given in the manual of the software (2007). The Gambarotta et al.s macro-element is used to carry out the results presented. 3. Case study: Non-linear static analysis of a three-dimensional masonry structure with commercial codes In the following the seismic analysis of a simple three-dimensional structure is carried out with two commercial codes that implement the different approaches discussed in the previous section, and the results are discussed and compared. The rst code is SAP2000 that has been used to implement the SAM equivalent frame approach, while the second one is 3Muri, based on the macro-element approach. The considered structure is a two-oor building with a square plant. The geometrical characteristics are shown in the right side of Fig.1. The building is entirely built by tuff. Its mechanical properties are schematized in the table on the left side of Fig.1. In order to take into account the uncertainties involved in the assessment of an existing structure, the Italian structural code (in accordance with the Eurocodes), has introduced two important concepts: the knowledge levels (LC in Italian) and the condent factors (CF) ( C8.A.1.A.4; NTC08). Three different levels of knowledge are dened: LC1: limited knowledge; LC2: intermediate knowledge; LC3: high knowledge. The level is assigned according to the available information on geometry, details, materials. According to the assigned level of knowledge, the mechanical properties are reduced by means of an additional partial factor, the so called condence factor (CF). Of course, if the level of knowledge is high, the uncertainties will be less and the condence factor will be small or even unitary; on the other hand if the knowledge is limited the condence factor will be higher. In the considered case it has been assumed an intermediate level of knowledge (LC2), so the corresponding CF is 1.2. The loads on the buildings are given by the weight of the masonry, the dead loads (3 kN/m2), and the variable loads (2 kN/m2). The vertical loads are applied as punctual forces on the upper ends of each piers. A non-linear static analysis has been carried out by applying static forces proportional to the masses. The total seismic load is about 823 kN.
774

Built Heritage 2013 Monitoring Conservation Management

Fig.1 - Left) Mechanical properties of the considered masonry; Right) Shape and faade of the considered masonry building

4.1. Pushover analyses with SAP2000 As said, SAP2000 has been used for modeling the building by applying the SAM approach: the frames that schematize the spandrels and the piers have rectangular sections with the same height and width of the actual elements. External xed restraints have been considered. It has been said that the type of spandrel inuences the behavior of the structure, so both cases of weak and strut spandrels have been considered. An overall view of the SAP2000 models is shown in Figure 2. The characteristics of the different models are discussed in the following. In all the models a reduced Young modulus (E/2) has been considered in order to take into account the reduced performance of the material in cracked conditions. The nonlinearity of the material is considered using plastic hinges,. These are dened in accordance with the Italian Code (NTC08, 7.8.1.5.4). Three types of plastic hinges are used: shear hinges (V type), bending hinges (M type) and rocking hinges (PM type). The ultimate shear and bending values of the hinges depend on the type of spandrels and they have been dened according to the indication of the Italian structural code (NTC08). 4.1.1. Modeling of the structure with weak spandrels - SAP2000 As said, a structure could exhibit a weak spandrel behavior if it has not ten775

Built Heritage 2013 Monitoring Conservation Management


Fig.2 - Three dimensional view of the SAP200 models: a) weak spandrels and no oor; b) oor modeled with shell elements; c) strut spandrels

sile resistant reinforcements or when the oor is very deformable. Actually, the Italian code does not consider specically this category of structures. The behavior of this category is recognized in various literature work as for example in [Calderoni et al., 2011] that suggest to simulate the weak spandrels using plastic hinges with ultimate values much lower than those of the hinges of the piers. In Table 1 the equations used for the calculation of those values are shown, where 0 is the normal average compression, l is the length of the compressive zone. For a more detailed description of the quantities see NTC08 and Calderoni et al. (2011). The walls are considered well connected, so for each couple of adjacent upper ends of the piers a diaphragm constraint has been assigned. Two different models have been developed: in the rst one the wood oor is not explicitly modeled (case a in Fig.2), while in the other one (case b in Fig.2) it is modeled with shell elements with isotropic behavior; the mechanical properties of the actual wood oor have been assigned (E=1300000 kN/m2, v = 0.1, G=591 N/mm2). In Fig.3 three steps of the analysis are shown. At step 2 the rst hinges appear. As expected, they affect the spandrels, due to their weak behavior. At step 4 some of the spandrels hinges reach their ultimate values and there are also some hinges on the piers. At step 8 almost all the elements in XZ plane are plasticized. Notice that the different behavior in the XZ and YZ plane depends on the position of the joists that load directly only two of the four walls.

Table 1 - Values of Mu and Vu in case of existing buildings with weak spandrels

776

Built Heritage 2013 Monitoring Conservation Management

Fig.3 - Three dimensional view of the SAP200 models

4.1.2. Modeling of the structure with strut spandrels - SAP2000 As already said, it is possible to consider a strut spandrel behavior when for example there are tensile resistant elements as chains. In Table 2 there are their ultimate values according to the Italian code. For a more detailed description of the quantities see NTC08. Also in this case the walls are well connected, so diaphragm constraints between the adjacent nodes have been assigned. Two different models have been developed: in the rst one (case c in Fig.2) the wood oor is schematized with a diagonal strut; the strut section is about 10 cm high, which is the height of the materials that offers the compressive strength, and 60 cm wide, which is 10% of strut length. In the second case the oor is modeled with shell elements as done in case of weak spandrels (again E=1300000 kN/m2, v= 0.1, G=591 N/mm2).

Table 2 - Values of Mu and Vu in case of existing buildings with strut spandrels

4.2. Non-linear static analysis with 3Muri Modeling with 3Muri is fast because the software have been created specically for dealing with masonry structures and has a user-friendly interface. The input data are mainly: type of material, level of knowledge, geometry (walls, openings, type of oor). The weak spandrels are simulated by using a type of panel that in the software is called simple while for the strut spandrel it is used a panel with chains. The chain has been calculated in such a way that its strength is comparable with that of the struts used in the SAP2000 model. The oor has been modeled by assigning various characteristics as the type
777

Built Heritage 2013 Monitoring Conservation Management

Fig.4 - Results of the analysis with 3Muri

of joists, their direction, the type of wood, etc. The walls have been considered well connected while the connection between the oor and the walls is considered weak. In Figure 4 the results of the analysis of the structure with weak spandrels are shown. It is possible to notice that almost all the spandrels show plastic bending behavior and that the rupture in the piers, that work as cantilever beams, is due to plastic bending and compression. 4.3. Comparison of the results In Fig.5 the response curves of the non-linear static analyses are compared. The position of the considered control point (#26) is indicated in Fig.2. It is possible to note that the linear behavior is similar with both codes whereas the values of the ultimate shear obtained with SAP2000 are always lower than those obtained with 3Muri. This difference is strongly related to the modeling of the oor (neglecting its presence, modeling with a shell slab or as a rigid diaphragm for example).

Fig.5 - Pushover response curve for different spandrel behaviors and software

778

Built Heritage 2013 Monitoring Conservation Management

These results highlight the importance of the right modeling that can be obtained only with a deep knowledge of the structure. So the test and the analysis that are carried out for increasing the level of knowledge are also fundamental tools for supporting the modeling phase. 5. Conclusions In the rst part of this work, various methods for the seismic assessment of existing masonry buildings have been briey discussed, focusing mainly on the Simplied Analysis Method (SAM), which can be implemented with most of the commercial nite element codes, and the macro-element approach, implemented in the code 3Muri.Then, in the second part, a case study has been carried out by applying two of this methods and considering two different spandrel behaviors. The results of the nonlinear analyses have shown that the assumptions at the base of the models strongly inuence the results and they should be carefully made. Moreover this investigation allows understanding the sensitivity of various parameters, suggesting that some of them signicantly inuence the global behavior and need to be accurately evaluated with specic on-site tests.
Acknowledgments The team www.francobontempi.org from Sapienza University of Rome is gratefully acknowledged. Eng. Penna of S.T.A. Data s.r.l. is acknowledged for the use of the Software 3Muri.This work was partially supported by StroNGER s.r.l. from the fund FILAS - POR FESR LAZIO 2007/2013 - Support for the research spin off. References Arangio S., Panetta G., 2009, Response of buildings to ground movements: considerations about the numerical modeling, Proceedings of COMPDYN 2009, Rhodes, Greece. Bontempi F., Gkoumas K., Arangio S., 2008, Systemic approach for the maintenance of complex structural systems, Structure & Infrastructure Engineering, 4(2), 77-94. Boothby T. E., Atamtrktr H.S., Erdogmus E., 2006, Manual for the Assessment of Load-Bearing Unreinforced Masonry Structures, Report for the US Department of the Interior. Bontempi F., Arangio S., Belliazzi G., Schinco M., 2012, Basi per ladeguamento sismico e ruolo della modellazione strutturale (in Italian), AID Monuments, Perugia, 26-28 May 2012 Bucchi F., 2012, Analisi non lineare di edici in muratura con codici di calcolo commerciale (in Italian), Master thesis in Civil Engineering, Sapienza University of Rome, Advisor Prof. F. Bontempi. Calderoni C., Cattari S., Lagomarsino S., 2009, In-plane strength of unreinforced masonry piers, Earthquake Engineering and Structural Dynamics, 38, 243267. Calderoni B., Cordasco E.A., Lenza P., Pacella G., 2011, A simplied theoretical model for the evaluation of structural behaviour of masonry spandrels, Intern. Journal of Materials and Structural Integrity, 192-214, 5(3). Calderoni B., Ghersi A., Lenza P., 1994, Seismic behavior of masonry buildings, Proceedings of the Masonry Council of Canada, 1994. Calderoni B., Marone P., Pagano M., 1987, Modelli per la verica statica di edici in

779

Built Heritage 2013 Monitoring Conservation Management

muratura in zona sismica (in Italian), Ingegneria Sismica 3, 19-27. Cali I., Marletta M., Pant B., 2012, A new discrete element model for the evaluation of the seismic behavior of unreinforced masonry buildings,Engineering Structures, 40, 327-338. DAsdia P., Viskovic A., 1995, Analyses of a masonry wall subjected to horizontal actions on its plane, employing a non-linear procedure using changing shape nite elements, Computational Methods in Experimental Measurements VII, Brebbia C.A. & Caromagno G.M. Editors, 1995. Gambarotta L., Lagomarsino S., 1997a, Damage models for the seismic response of brick masonry shear walls. Part I: the mortar joint model and its applications, Earthquake Engineering and Structural Dynamics, 26:4, 424- 441. Lenza P., 1996, Nonlinear behavior of masonry buildings under seismic actions, Proceeding of the 11th World Conference on Earthquake Engineering, 1996. Magenes G., Calvi G.M., 1997, In-plane seismic response of brick masonry walls, Earthquake Engineering & Structural Dynamics, 26 (11), 10911112. Magenes G., Della Fontana A., 1998, Simplied non-linear seismic analysis of masonry buildings, 5th International Masonry Conference, Proc. of the British Masonry Society, 8, 190-195. Norme Tecniche delle Costruzioni 2008 (NTC 2008), D.M. 14/01/2008. Parisi F., Augenti N., 2012, Uncertainty in seismic capacity of masonry buildings, Buildings, 2(3), 218-230. Pasticier L., Amadio C., Fragiacomo M., 2008, Non-linear seismic analysis and vulnerability evaluation of a masonry building by means of the SAP2000 V.10 code, Earthquake Engineering and structural dynamics, 37, 467-485. Tomaevic M., 1978, The computer program POR, Report ZMRK, Institute for Testing and Research in Materials and Structures, Ljubljiana, Slovenia. UNI EN 1998-3:2005, Design of structures for earthquake resistance: assessment and retrotting. User Manual of the software 3Muri S.T.A. DATA, version 3.0.5, 2007.

780

You might also like