This chapter discusses the role of experimental methods in earthquake engineering with a brief summary of advantages and challenges of the various test methods used in this field. The objectives and motivations of this research are discussed with an overview of the contents of this book.
Hinweise
The original version of this chapter was revised: See the “Chapter Note” section at the end of this chapter for details. The erratum to this chapter is available at https://doi.org/10.1007/978-981-10-5867-7_6
1.1 Importance of Experimental Testing
Natural hazards are the largest potential source of casualties in inhabited areas. Damage to structures causes not only loss of human lives and disruption of lifelines, but also long-term impact on the local, regional and sometimes national and international economies.
One of the main goals of structural and earthquake engineering is to improve the resilience and performance of structures to protect the lives and safety of occupants and control economic losses under an extremely wide range of operational conditions and hazards. Accordingly, the priorities lie in gaining an understanding of the behavior of various classes of structures under different dynamic load types from the elastic range through to developing collapse mechanisms and failure. However, this poses a major challenge as it requires the prediction, with sufficient confidence, of a structure’s response beyond design level, all the way to the state of complete collapse.
Anzeige
Today, dynamic analysis of complex structures can be efficiently computed utilizing readily available software. The cost of computation has been continuously reduced, and now very complex and detailed numerical simulations are possible on personal computers. However, for many components or materials, nonlinear behavior and failure modes are still not well understood. In such cases, numerical analyses and simulations may not be reliable, since more detailed and complex properties are needed for the critical components to obtain meaningful results. Therefore, laboratory testing remains a necessary tool to improve and validate numerical models over the full range of a structure’s response. With this objective in mind, experimental simulations of structures have been conducted to investigate the capacity and failure behavior of various structural systems and critical components that are difficult to model numerically. Based on these studies, the behavior of different structural systems such as multi-story buildings, bridges, coastal structures and others during extreme events is assessed, to enable the design and construction of safer and more resilient structural systems to mitigate natural hazards.
1.2 Experimental Testing Methods
In order to experimentally evaluate the dynamic response of a structure, several techniques can be used in conducting laboratory tests (Filiatrault et al. 2013). The most common method is quasi-static tests that are usually conducted in order to test the behavior of structural components or full-scale structural systems. In this method, the structure is subjected to pre-defined displacement or force history using hydraulic actuators. Typically, these tests are conducted to investigate the hysteretic behavior and capacity of structural components under a cyclic load. Although these tests are fairly easy and economical, they are limited by the predetermined loading protocol. However, in performance-based design, the focus of all decisions is on the demand requirements, the actual behavior of the structural elements and the level of damage during different intensities of extreme loads. Therefore, the predetermined load protocol is generally inadequate for representing the structural behavior, as the load distribution continuously changes during an actual event.
The most realistic approach is the dynamic testing of the entire structure. For example, the use of earthquake shake tables in seismic research provides the means to excite structures in such a way that they reproduce conditions representative of true earthquake ground motions. However, due to the extremely high cost, complexity and damage to the equipment, experimental testing of even a full-scale single-story structure poses significant challenges. The largest shake table in the world, the Hyogo Earthquake Engineering Research Center of Japan (E-Defense) shake table, is located north of Kobe in Miki City, Japan, with dimensions of 15 m by 20 m and the capacity to support building experiments weighing up to 1200 tonnes, which is sufficient to test a full-scale 6-story building. However, not only are these experiments extremely costly, both in terms of operation of a large-scale shake table and in terms of constructing the entire structure on the shake table, they do not provide the large-scale testing environment for tall buildings or horizontally extended structures such as bridges. The George E. Brown Jr. Network for Earthquake Engineering Simulation (NEES) equipment sites in the USA also provide some shake table facilities, such as the twin shake tables in the Buffalo NEES facility, multi-shake table testing in the Nevada NEES facility and a high-performance outdoor shake table in the UC San Diego NEES facility. However, due to limitations on the size and capacity of shake tables, structures are typically tested on a reduced scale or a highly simplified model is used. In particular, in collapse simulation of structures, the shake table test method is expensive, complicated and dangerous, due to the risk associated with the collapse of a structure on the shake table. In addition, a scaled and simplified model does not necessarily represent the response of a full- or large-scale prototype experiencing severe nonlinear deformation and collapse. Scaled specimens can provide a fair understanding of global behavior, but local behavior may not be simulated accurately. However, this local behavior may play a critical role in determining the performance of a structure, given that initial damage usually occurs on a local level. Certain types of behavior, especially local effects such as bond and shear in reinforced concrete members, crack propagation, welding effects and local buckling in steel structures, are well known to have size effects, which casts doubt on the validity of the shake table tests.
The third method is hybrid simulation, also known as pseudo-dynamic testing (Nakashima et al. 1992). Hybrid simulation is a hybrid procedure that combines classical experimental techniques with online computer simulation for cost-effective, large-scale testing of structures under simulated dynamic loads. This method is often called hybrid (rather than pseudo-dynamic testing) since it combines modeling and experiments and can include real dynamic effects in the experiment. According to a report by the US earthquake engineering community, hybrid simulation capabilities are a major emphasis of the next generation of earthquake engineering research (Dyke et al. 2010). However, this role can be expanded to other loading conditions, such as hydrodynamic loading conditions created by waves, traffic and impact loads due to moving vehicles, aerodynamic loads generated by wind, and blast loads. This wide variety of loading conditions can be simulated by incorporating them into the analytical portion of the hybrid model without changing the physical portions of the experiment.
Anzeige
1.3 Hybrid Simulation Fundamentals
Hybrid simulation provides the best advantages of both computational simulation and experimental techniques: the realism of actual testing for the critical components, together with the flexibility and cost-effectiveness of computer modeling. This method is based on domain decomposition, in which the structure of interest can be divided into multiple parts/substructures. On the one hand are the parts and regions that can be reliably modeled in one or more computers, either because of their simple behavior or because they are not considered critical for the analysis conducted. On the other hand are the parts and regions of most interest that are physically tested in one or more laboratories, either because of their highly nonlinear behavior or because they are critical to the safety and performance of the structure. The parts that are numerically simulated are called the numerical or analytical substructures. The parts that are physically modeled and subjected to loads in the laboratory are called the experimental or physical substructures. The combination and interactions of the all substructures form a hybrid model of the complete structure of interest.
In hybrid testing, the dynamic aspects of the simulation are handled numerically. Therefore, such tests can be viewed as an advanced form of quasi-static testing, where the loading history is determined as the simulation progresses for the structure subjected to a specific dynamic load. The governing equation of the motion is solved similar to pure numerical simulations using a time-stepping integration. The displacement/force demands are then applied to the physical specimen(s), and the resisting forces are measured and fed back to the computation solver to calculate the displacement/force demands corresponding to the next time step.
To illustrate this process for the various types of substructures in hybrid simulation, an example is presented for a multi-story building. Utilizing the hybrid simulation technique, the first-story corner-column, considered the critical element, can be constructed and physically tested in the laboratory, and the remaining parts of the structure, the inertia and damping forces and gravity, dynamic loads and the second-order effects can be reliably modeled in the computer (see Fig. 1.1).
Fig. 1.1
Hybrid simulation technique
×
1.4 Advantages and Challenges of Hybrid Simulation
Hybrid simulation provides several advantages, including the following:
1.
Hybrid simulation mitigates the errors related to the simplification of the theoretical modeling of complex nonlinear structures or subassemblies by testing them physically in the laboratory. It is not a trivial task to accurately model complex nonlinear structures, as it requires a series of assumptions that have to be made to simplify the modeling procedure, which is performed at the cost of simulation accuracy and reliability.
2.
Hybrid simulation reduces construction/fabrication costs and the overall time for testing in the laboratory. Dynamic testing of the entire structure requires the construction/fabrication of a whole structure, which is an expensive and time-consuming process for a physical test. Since the damage essentially starts as a local phenomenon, hybrid simulation allows the physical testing of only the critical portion of the structure, where the damage is expected.
3.
Hybrid simulation reduces uncertainties associated with limited scale of shake table tests by facilitating economical large-scale testing. The size and weight of the physical subassemblies are restricted only by the available laboratory space and the strength of the strong reaction wall/floor. The strength of the specimen is also limited only by the actuator capacities that are available in the laboratory.
4.
Hybrid simulation can be conducted on an extended time scale, typically ranging from 100 to 1000 times slower than actual earthquake duration (Carrion and Spencer 2008). This time modification allows the researcher to carefully observe and track the progress of damage throughout the simulation and thus provides important insights into structural component behavior, especially near collapse. Structural performance, such as the failure pattern and initiation of cracks in a special structural region such as beam–column connections, can be closely investigated.
5.
Hybrid simulation can be conducted locally or geographically distributed, meaning that individual substructures do not need to be within the same facility, but can be linked by either the Internet or other methods of data transfer. Therefore, laboratories with much larger capacities can be used for experimental subassemblies.
Although hybrid simulation has attracted many researchers in the evaluation of the seismic behavior of structures and offers many advantages, it provides new challenges, as follows:
1.
Although hybrid testing is an alternative to shake table testing, the accuracy of the hybrid test is often questioned. To guarantee that the results obtained from a hybrid simulation are valid and reliable, it is important to minimize the contamination of the results by errors. The errors that occur at different stages of a hybrid simulation are modeling errors due to the discretization process, analysis assumptions, numerical errors introduced by the integration and equilibrium solution algorithms, experimental errors generated by the control and transfer systems, and the noise in instrumentation devices and the data acquisition system.
2.
Actions on structures during extreme events such as earthquakes are generally multi-directional and continuously varying, due to the time-dependent nature of the input motion. For instance, variations of the axial loads during a seismic excitation may influence the response of the vertical structural components (e.g., bridge piers and building columns) since the response of such elements when combined with flexural, shear and torsional actions may differ from the cases when they are not subjected to the same axial load changes. Simulation of such highly coupled multi-directional loading conditions using conventional structural testing methods can be expensive, time-consuming and difficult to achieve. As a result, advanced and innovative experimental techniques and control strategies are under development by researchers (Nakata 2007; Wang et al. 2012; Hashemi et al. 2014; Hashemi and Mosqueda 2014).
3.
The experiments should be conducted at a large or full scale to accurately capture the local behavior of the elements. However, the conduct of large-scale experiments may not be feasible, due to the limited resources available in many laboratories, including the number and capability of the actuators available, the dimensions and load capacity of the reaction systems, difficulties in the actuator assemblies and testing configuration in reliably simulating the boundary conditions. Consequently, the specimen may be tested at a small scale or under uni/biaxial loading configurations, which do not necessarily represent the actual action or demand on the structural elements and the corresponding nonlinear response of the prototype system.
4.
Conducting multi-directional loading including gravity load effects requires a mixed-mode control strategy. The application of gravity loads has been mainly considered by researchers using a combination of force-control actuators in the vertical direction that are decoupled from displacement-control actuators in the lateral direction of the specimen (Lynn et al. 1996; Pan et al. 2005; Del Carpio Ramos et al. 2015). In these tests, independent of lateral actuators, only the vertical force-control actuators apply the gravity forces, while under large deformations, lateral actuators have a force component in the vertical direction that needs to be accounted for. Therefore, versatile and generally applicable mixed-mode control algorithms are required to take into account instantaneous and spatial coupling in the control systems.
1.5 Objectives and Outline
This manuscript presents the design details and unique capabilities of the MAST system for the hybrid simulation of large-scale structures subjected to extreme dynamic forces. The testing capabilities advance the current state of technology by allowing accurate simulation of complex time-varying 6-DOF boundary effects on large-scale structural components in mixed load/deformation control modes. Utilizing the MAST system, the developments of new materials and structural systems and the effectiveness of new repair/retrofitting strategies can be reliably evaluated using three-dimensional large-scale quasi-static cyclic or local/geographically distributed hybrid simulation tests. The manuscript is organized as follows:
Chapter 2 presents the technical background and literature review on the development of hybrid simulation and summarizes the work by researchers in the fields of substructuring techniques, integration schemes, continuous and real-time hybrid testing, local and geographically distributed hybrid testing and experimental and numerical errors in hybrid testing.
Chapter 3 describes different components of the state-of-the-art system for hybrid simulation at Swinburne, including the design details of the MAST facility, the reaction systems including the strong wall/floor and the cruciform crosshead, servo-hydraulic actuators and the 6-DOF controller system and hybrid simulation architecture.
Chapter 4 presents the results of a range of experiments, including switched/mixed load/deformation mode quasi-static cyclic and hybrid simulation tests to highlight the unique and powerful capabilities of the MAST system, specifically for the assessment and mitigation of the collapse risk of structures.
Chapter 5 presents a summary of key contributions and concluding remarks. Research areas for further development and study are also briefly discussed.
