RICERCA ITALIANA     

Innovative technologies and materials for seismic retrofit of existing structures

Università degli Studi "G. d'Annunzio" Chieti-Pescara

Abstract

The main goal of the proposed research is to assess and enhance methodologies and materials for the design, testing and modeling of the seismic strengthening of existing structures. Both reinforced concrete (RC) (mainly frames, shear wall and bridges) and masonry structures (mainly historical and traditional structures and bridges) will be considered. The main strengthening technique studied will be the application of Fiber Reinforced Polymers (FRP), but other methods, such as the application of external prestressing, will also be considered. The design methods will refer to current and emerging design guidelines in Italy, the European Union and North America. Available design equations will be evaluated and eventually improved. The different research units of the proposed project will use a number of structural elements to be strengthened starting from different levels of damage. The research units will also use state of the art techniques for frame and finite element modeling. The research activities will develop along two interconnected phases: 1) experimental tests on structural members to be loaded in the lab and in situ under monotonic and cyclic loads; b) numerical modeling of structural members and frames under both static and dynamic loads. The proposed project is the continuation of a previously funded two-year research (Cofin 2002). The original research scope is extended here to seismic design and masonry structures.

The EXPERIMENTAL phase focuses on a series of tests on as-built and strengthened structural members. The goal of these tests is to understand the damage and failure phenomena of RC beams, columns, joints, shear walls and masonry walls. Some tests will rely on structural members previously tested and damaged in the lab and now available for repair and retesting. Other research units will use structural members extracted from real buildings that were demolished. Finally, new structural members will be designed according to older seismic and non seismic code provisions and tested in the lab. The specimens will be tested under both monotonic and cyclic loads using and comparing different FRP strengthening techniques. For a full scale, five-story RC wall, repair and strengthening with external prestressing will be used. A series of non destructive tests on RC and masonry bridges still in service completes this phase.

The MODELING phase will use a common computational platform for the nonlinear, static and dynamic analysis of RC frames. All the research units have already worked or used this platform. Starting from the experimental results, new fiber elements will be developed and others will be enhanced to introduce the effects of the shear deformations and of strengthening with FRP. Panel elements to model the infills will be added. Plane and solid finite element analyses will also be carried out, using fracture mechanics-based codes. The FE analyses will study in detail the brittle failure mechanisms that characterize rupture in RC and masonry structures. Finally, based on the results of the experimental tests on RC beams strengthened in flexure and shear with FRP, a blind test will be organized and advertised. The purpose is to assess the validity of different modeling techniques proposed by the research community. Researchers from all other the world will be invited to participate. The results of the blind test will be presented at a workshop to be held at the end of the two year research. <<<

Principal Investigator

Enrico SPACONE Università degli Studi "G. d'Annunzio" CHIETI-PESCARA

Research Objectives

The main objective of the proposed research is the advancement of the state-of-the-art in the field of seismic retrofitting of existing RC and masonry structures. The research will develop along two interconnected, parallel phases. The objectives of the two phases are described hereafter.

EXPERIMENTAL PROGRAM: The experimental program will focus on a series of tests on structural members, as built and strengthened, whose aim is a better understanding of the damage and failure mechanisms under monotonic and cyclic loads. Both RC and masonry specimens will be considered.

As for reinforced concrete, the Brescia, Firenze and Rome 3 units can rely on structural members (beams, columns, joints, shear walls) that have already been tested in the lab or that have been extracted from buildings to be demolished. Such members represent a unique opportunity to study structures designed according to dated safety criteria and in need of retrofitting/strengthening in order to satisfy the modern seismic design criteria. Brescia, Firenze, Rome 1 and Rome 3 will deal with the problem of shear strengthening of RC members with insufficient transverse reinforcement. Brescia will retest an already tested and damaged full scale, five-story structural wall. The aim of the new cyclic tests is the assessment of wall strengthening through external prestressing. The Firenze, Rome 1 and Rome 3 units will join forces in studying the shear strengthening with FRPs of RC beams and columns. Different strengthening methods will be tested and compared. The confinement problem in FRP-wrapped vertical RC members will be studied by Firenze, Rome 1 and Rome 3. Rome 3 will also test infilled frames to study the effects of the infills on the column shear response. The FRP-concrete and FRP-masonry adherence problem will also be studied, as it is central to the assessment of the collapse response of FRP-strengthened RC and masonry elements. Delamination tests will be carried out on RC specimens in Roma 1 and Bologna and on RC and masonry specimens in Chieti. The Chieti unit will also consider the delamination problem with a new composite material, FRCP (Fiber Reinforced Cementitious Matrix, a recently developed material where the polymeric resin is replaced by a cementitious matrix). Chieti and Roma 3 will address the problem of the characterization of the brick infills and of their importance in the response of RC frames mostly under moderate seismic forces. Full and 1:2 portal frames will be tested. Finally, nondestructive tests on RC and masonry bridges will be carried out by the Bologna unit on bridges in service in the Pistoia province.

As for masonry, the Chieti unit will study, through a series of tests, the adherence and delamination between FRP or FRCM and masonry. The tests will be carried out on specimens replicating historical and traditional masonry. Other tests will deal with the characterization of raw earth. Raw earth is attracting increasing interest because of its potential as sustainable material, but its mechanical properties are not well know, particularly under seismic loads.

MODELING PROGRAM: The modeling program will focus on: A) the development, enhancement and application of frame elements for nonlinear frame analysis; B) the study of single structural members by plane and solid finite element analyses aimed at tracing the evolution of the cracking patterns in strengthened RC and masonry members loaded up to failure.

The objectives of the section on frame elements and nonlinear frame analysis are: 1) develop a series of fiber elements for RC beams, beam-columns and joints. The elements will run in a general purpose nonlinear frame analysis environment; 2) apply and calibrate the above models starting from the results of the experimental investigation of Phase 1; 3) conduct a series of parametric studies to extend the results of the experimental tests to entire structures and different strengthening methods. First of all, the force-based fiber model will be enhanced to include more refined uniaxial laws. Shear deformations will also be added in order to predict and describe the shear failure in shear deficient RC members. Frame elements with bond slip between reinforcement (steel or FRP) and concrete will also be considered and applied to the prediction of the failure load of FRP-strengthened beams. The fracture energy values obtained from the experimental tests and from the refined FE analyses will be fundamental for these studies.

As for the plane and solid FE studies, they will be carried out on both RC structural members and on masonry walls. Parametric studies on FRP- or FRCM-strengthened beams will analyze the failure modes, focusing mainly on the evolution of the cracking patterns. These analyses will rely on nonlinear fracture mechanics-based FE codes. As for masonry, new panel elements will be developed that include the effects of the FRP reinforcement on masonry brick and stone walls. New panel elements for raw earth panels will also be developed. The new models will be calibrated starting from the experimental results of Phase 1 of the proposed research.

A "blind test" will be organized for the prediction of the experimental results on RC beams strengthened in flexure and shear with FRP materials. The reference tests will be identified in the initial stages of the research. Researcher from all over the world will be invited to participate. They will be given the geometric, mechanical and load sequence of the tests. The test results and the numerical correlations will be presented at a workshop that will take place at the conclusion of the research program.

Finally, it is important to point out that researchers from the units of the proposed project have recently formed, together with other Italian universities, a study group for the preparation of a pre-code document for the design, construction and testing of strengthening techniques with fiber reinforced polymers. One of the objective of the research proposed here is to provide support to the pre-code document group, in particular for the development of design criteria and for the validation and improvement of existing design equations. <<<

First Results

POINT 1: A preliminary numerical study will allow to assess the effectiveness of the new repair and to study the optimal position of the external cables and their pre-tension. Extremely useful and important results will derive from the cyclic tests on the repaired wall and from the comparison of the response of the repaired and original wall.

POINT 2: The tests on the beams will allow to identify the stress distributions in the FRP sheets across the crack, with the aim of 1) expressing them analytically, b) obtaining a design equation expressing the dependence by the crack geometry and by the placing of the FRP sheet/tissue. Regarding the tests on the columns, the effectiveness of the interventions with FRP with respect to the traditional interventions will be evaluated. For what concerns the study of the frames with and without infill the obtained information will be utilized for a) the validation of the FRP reinforcements in case of large contact stresses transferred by the infill and/or by the dissipative bracings, b) the evaluation of the reliability of the results obtainable through nonlinear analysis methods.

POINT 3: Following the experimental tests, it is expected to assemble a comprehensive data base to characterise and quantify, through appropriate design rules, 1) the reduction of the confinement effect when strengthening reinforced columns having rectangular cross-section, 2) the cyclic behaviour of confined structural members.

POINT 4. The results expected in phase 1 are rules for the definition of an analytical bond relationship with good prediction ability. Furthermore, fracture mechanics parameters will be obtained for different failure modes. The bond characteristics between concrete and FRP and between concrete and FRCM will be compared. The results obtained in this point will be used to obtain design equations and for the numerical analyses of phase II.

POINT 5: A first result of the proposed research will be the definition of a design abacus that classifies the mechanical characteristics of masonry infills typically used in Italy according to geometrical, mechanical and constructive characteristics of the components. The second phase of the research will help define efficient evaluation criteria for existing infilled frames and frames repaired with FRP jackets and/or dissipative bracings

PUNTO 6: The first result expected is an indication of the effectiveness, depending on different fiber orientations, of the FRP (or FRCM) bonding to masonry. For the raw earth masonry, the in situ and lab tests will help define its mechanical characteristics as a function of the percentage of the different components. Important results are expected for assessing the effectiveness of different internal reinforcements, using natural and artificial fibers. Based on the above results, current design rules will be reviewed and improved.POINT 1: Important results will be the new and updated constitutive laws for concrete, steel, FRP and bond and for the cross section (shear law).

POINT 2: The two Timoshenko beam elements with shear deformations at the section or fiber level will be relevant partial results. Modeling of the FRP reinforcement will be added to the element. The fiber 2D concrete model will be simplified.

POINT 3: The Merlin analyses will give information a) on the evolution of the cracking patters that eventually lead to detachment of the FRP reinforcement; b) on the fracture energy values that lead to the best correlation with the experimental results; c) on the geometric and mechanical parameters that affect the type of failure observed. The parametric studies with frame elements will help evaluate and eventually improve current design procedures aimed at predicting delamination of the FRP reinforcement.

POINT 4: An exchange of information and students between the Italian Universities and the University of Colorado will lead to the development and strengthening of coordinated experimental and modeling research programs. The Italian Universities will greatly benefit from the long term experience of the University of Colorado on pseudo-dynamic testing.

POINT 5: A first result will be the development of a method for strengthening infill frames via dissipative bracings. Indications will be given on the how to evaluate the structural safety before strengthening, how to design and apply the bracings. Anther important result will be the development of a three-strut infill model. The experimental results will help optimize the panel geometric and mechanical parameters of the infill model.

POINT 6: The new macro elements "masonry panels" with and without FRP or FRCM will be an important result to be used for different types of masonry.

POINT 7: The results of the numerical analyses are fundamental in order to determine the optimal cable position and their post-tension, and to correctly execute the repair works of the structural wall to be tested. Furthermore, the analyses will allow to evaluate the possibility of using innovative materials, such as high performance and fiber-reinforced concrete, in repairing damaged concrete. The numerical study on caisson foundations will allow to asses the effectiveness of the foundation system of the structural wall. Based on the results of this study, a simplified approach for the design of box foundations wil be developed.

POINT 8: Many international experts on modeling and design of concrete structures will take part in the "blind test". This will give great exposure to the results of the proposed research. <<<

Timescale

24 months

National and international background

The assessment of existing RC and masonry structures under static and dynamic loads is a topic of great research and practical interest. Such interest stems on one side from the different levels of degradation shown by the existing infrastructure, and on the other from the fact that many existing structures do not satisfy the safety criteria of the current seismic design philosophies. Recent studies conducted in Canada estimate that the investments necessary worldwide to rehabilitate the existing infrastructure are around a thousand billion dollars (ISIS Canada 2000).

Recent earthquakes have underlined the seismic vulnerability of existing structures and infrastructures, as shown, for instance, by the catastrophic collapses of buildings, bridges and highway structures (Comartin et al. 1995, Hall 1995). Even when human casualties are low, the economic impact can be extremely heavy, as the Northridge earthquake in the US has demonstrated in 1995. This earthquake caused few victims, but the highest economic losses in history due to an earthquake. Few months ago the catastrophic earthquake that destroyed Bam in Iran has once again shown the seismic vulnerability of monumental structures, which are even more difficult to retrofit. Older buildings are of particular interest in Italy because of the presence of several important historical buildings made of load bearing cooked brick, stone and raw earth masonry. For such buildings it is necessary to investigate methods that can enhance their seismic response without modifying their artistic-historical value and their aesthetics.

Existing RC structures are often seismically deficient because they were designed according to old codes that are inadequate to guarantee the safety levels prescribed by the new generation of seismic codes (Eurocodes 2 and 8, 1998, International Conference of Building Officials 2000, Ordinanza PCM 2003). Most of the observed collapses in the last years are primarily due to insufficient transverse reinforcement (and thus low shear strength), low concrete confinement and inadequate lap splicing. Another problem in existing RC structures is the deterioration due to environmental and human factors. Among the most common causes it is worth citing reinforcement corrosion due to excessive use of salt in winter and freeze/thaw cycles, the subsequent loss of bond between steel and concrete, the use of low-quality materials, the damage caused by accidents. The most important consequences of the above-cited problems are one or more of the following structural deficiencies: inadequate flexural and/or axial strength in buildings and bridges, inadequate shear strength, insufficient ductility, inadequate development length for the steel reinforcement, inadequate concrete confinement, inadequate design against buckling of the longitudinal rebars.

As for masonry structures, they generally have good strength reserve. Structural problems are often due to local deficiencies that do not allow different portions of the structure to collaborate together causing partial or total instability of the structure. It is unusual to find structural elements that are deficient. Seismic strengthening of masonry structures must guarantee that the building behaves as a whole under lateral, cyclic loads. Such behavior can be induced by, first, adding structural connections among the different components and, second, by strengthening strategically important structural elements.

Furthermore, recent years have seen a growing interest in materials and building methods that are eco-compatible. The seismic performance of these materials needs to be entirely studied. One such material is raw earth, whose use is spreading again in several European and American countries (it has been continuously used in several African and Middle East countries) both for new constructions and for the preservation of older structures. The use or raw earth is encouraged by the promising perspective of its eco-sustainability (Forlani, 2001). For raw earth too there exists the problem of how do design new structures and how to preserve older ones, especially when it comes to seismic loads. There is a draft law in Italy (L.64/74) that includes raw earth among the materials that can be used to build in seismic areas, but detailed studies are still necessary to define and control construction methods to build seismic resistant raw earth structures.

The above discussion points out the importance of offering designers, on the one hand, the instruments necessary to both assess the current condition of existing structures and to estimate their capacity to resist seismic design loads that satisfy the safety requirements of current codes and, on the other hand, retrofitting methods for strengthening the same structures if this is deemed necessary. Large investments continue to be made all over the world to analyze the problem and to find solutions that are both economical and structurally feasible. Examples of such investments are found, for instance, in the NEES project funded by the US Congress in 2000 and that will last at least until 2014. The NEES project used 80 million dollars to establish a network of internet-connected laboratories for earthquake engineering tests. This initial phase of the project will end in late 2004. More generally, research in the field of existing RC and masonry structures has moved along different interconnected directions that aim at developing a complete approach towards the diagnostics and assessment of existing structures, the design of eventual rehabilitation techniques and the modeling of the structural behavior before and after strengthening/retrofitting.

Experimental research in the field of RC and masonry structures is essential for the development of design criteria, the calibration of design equations and the validation of analytical models. Various testing campaigns on structural elements and on entire frames have been completed all over the world. Significant examples are the RC buildings and bridge columns tested at ELSA (European Laboratory for Structural Assessment, Ispra, Italy), those performed on masonry structures on the shake table of ENEA-Casaccia, Italy, those carried out in the United States at UC Berkeley, UC San Diego, SUNY Buffalo (just to mention some of the biggest laboratories), in Japan, in Canada, in Australia and in New Zealand. Traditionally, the scopes of the experimental tests on RC structures have been both to achieve a better comprehension of the behavior of existing structures and to validate new and improved construction details for new structures. The enormous problem of the aging civil structure and infrastructure inventory has pushed authorities worldwide to look with ever growing attention at the development and the experimental assessment of repair/strengthening/retrofitting techniques for existing structures. Seismic isolation (whose scope is essentially to change the dynamic properties of a structure) was one of the first methods studies. More recently, the development of various strengthening techniques have attracted more attention. These techniques may use both traditional (concrete and steel) and new materials (fibre reinforced composite materials, or FRP, developed first by the military and aerospace industry, and later extended to civil structures).

The existing literature on retrofitting of RC structures with FRP is extremely vast (a partial list is reported in the reference list) and has fostered the development of some strengthening techniques (strips and plates for strengthening of beams and columns, and jackets for columns and bridge piers, just to mention some of the best known applications). The contribution by the composite materials to the strenth of RC structures can be evaluated through appropriate analytical relations that depend on the problem at hand: bending (e.g., Ritchie et al. 1991, Täljsten 1997a), shear (e.g., Chajes et al. 1995, Norris et al. 1997, Triantafillou 1998), ductility (e.g., Spoelstra and Monti 1999, Monti et al. 2000). These relations aim at predicting the strength of structural elements, by considering the contribution of the composite material. In many cases such analytical relations are accurate only if classical failure mechanisms, such as concrete compressive failure or tension composite failure, develop. However, when other mechanisms related to the bonding between FRP and concrete lead to failure (see for ex. Täljsten 1997b, Malek et al. 1998), the prediction capacity is sensibly reduced.

The lack of a satisfactory comprehension of the collapse phenomena of existing RC structures strengthened with composite materials is confirmed by the fact that, while a great number of experimental and analytical work has been published so far, the available design guidelines (mainly ACI 440 F [ACI 2000] and fib bulletin n.14 [fib 2001]) use in many cases design equations that are extrapolated from empirical data rather than derived from a rational understanding of the mechanical phenomena observed experimentally.

The situation for masonry structures is similar or worse, as the application of FRP materials to masonry is even more recent than for RC structures. In this masonry case the problems appear to be somehow different because, on one hand, the bond behavior is assured by applying the FRP fabrics directly to the stone or cooked brick elements, and on the other hand, there is a bigger difference between the mechanical characteristics of the base material and the reinforcement. Another reason of complexity derives from the geometry of masonry structures, made of bi and tri-dimensional elements rather that line elements. An important issue in retrofitting historical masonry structures is the need to choose techniques compatible with the characteristics and needs of a monument, such as the reversibility of the retrofitting and its compatibility with the building esthetics. Finally, there are new retrofitting techniques, in particular by using FRCM (or Fiber Reinforced Cementitious Materials), which guarantee the transpiration of the masonry structures on which they are applied and thus appear to be quite promising for masonry retrofitting.

Parallel to the experimental research, it is very important to develop simple but accurate analytical models for non-linear analyses up to failure of strengthened and as-built RC and masonry structures subjected to both static and dynamic loads. These analytical models play a fundamental role in emerging design philosophies, based on "performance-based design" (SEAOC 1995, FEMA 1997). According to this philosophy, the response of a structure must be explicitly checked under different load scenarios by controlling that a series of safety criteria (based on limiting values for forces, and above all, displacements) are met. The importance of modeling is confirmed by the efforts undertaken in this field by several US earthquake engineering research centers, first of all the Pacific Earthquake Engineering Center (PEER), which is proposing a new computational platform named OpenSees (Opensees 2002) which primarily uses fiber beam/column elements based on studies started by the coordinator of this proposal during his doctoral research at UC Berkeley (Spacone et al. 1996a,b,c).

The modeling of RC structures with frame elements represents a compromise between computational precision and speed. The last years have seen remarkable progresses in this field, leading to a significant increase in computational speed and to the formulation of frame elements based on rigorous energy principles. In particular, force-based elements have become very popular with respect to classical displacement-based elements. These elements are exact (Spacone et al. 1996a) within the limits of the Euler-Bernoulli theory. The use of such elements requires only one two-node element for each structural element, with a remarkable reduction of the total number of degrees-of-freedom of the structure. The combination with a fiber section model produces an element suitable for application to both beams and columns (Spacone et al. 1996 b, c ). More recent developments introduce slippage between reinforcement and concrete (Spacone and Limkatanyu 2001 and Aprile et al. 2001) and add confinement in circular FRP-strengthened columns (Spoelstra and Monti 1999). The fiber frame element still needs enhancements, in particular a more detailed library of uniaxial constitutive laws (which includes for example buckling of the longitudinal reinforcement or confinement of rectangular sections), and the inclusion of shear deformations and shear failure. Finally, in the case of RC beams strengthened with FRP, great uncertainties still exist on how to select the parameters of the partial bond between concrete and FRP, particularly in cracked zones (Monti and Renzelli 2002, Aprile et al. 2001). These developments need further support from experimental research for a deeper look at the problem.

Finally, plane elements are most commonly used for modeling masonry structures. Non linear programs, such as PEFV (developed by the Research group of the University of Chieti), are used to analyze the behavior of panels and entire masonry walls. PEFV uses finite elements with variable geometry in order to avoid mesh sensitivity. Similarly to OpenSees, PEFV is developed using object oriented programming (in C++) which eases the implementation of new elements, constitutive laws and solution algorithms. There remains the need to develop elements that can model masonry panels strengthened with FRP laid in different configurations. Finally, it is important to develop models for infills to be implemented in nonlinear frame analysis programs in order to consider the effects of the infills on the response of RC frames, especially under serviceability conditions. <<<