Demonstration Video

Fragility curves corresponding to the moderate damage limit state are approximately located in the middle between the slight damage and the collapse limit states’ curves for buildings 1 and 3 along both directions. However, collapse fragility curves of building 2 are close to the moderate damage limit state’s which means that building 2 is expected to collapse soon after exceeding the moderate damage level in both directions.

Column shear-axial failure in existing vulnerable reinforced concrete (RC) frame buildings constructed before the mid 1970's is a major seismic risk. The challenges associated with spatial response and system load redistribution capability at the onset of collapse has not been resolved yet. The acceptance criteria in current seismic rehabilitation provisions are defined at the element level with no due consideration for the system robustness. Four sets of three-dimensional, geographically distributed hybrid simulations (HS) will be conducted using the George E. Brown, Jr. Network for Earthquake Engineering Simulation (NEES) facility at the University of Illinois at Urbana-Champaign to obtain the response up to collapse of a representative three-dimensional structural system subjected to one-directional and tri-axial seismic ground motions. This research will investigate, characterize, model, and derive practical procedures for the consequences of column shear-axial failure on the collapse of existing vulnerable RC structures. The project will develop system-level acceptance criteria and analytical tools for near collapse seismic performance of existing non-ductile RC frame structures. Data from this project will be archived and made available to the public through the NEES Project Warehouse/data repository at http://www.nees.org. This project is a collaborative effort between researchers from Northeastern University and Western Michigan University.

This project utilizes the unique HS capability provided by the NEES facility to simulate near collapse response of existing vulnerable RC buildings through the integration of large scale physical experiments and numerical models. If successful, this research project will shift the philosophy of structural assessment of vulnerable buildings from component-level to system-level evaluation. The application of system-level evaluation methods developed in this project can lead to more efficient and cost-effective rehabilitation methods for existing non-ductile RC buildings against collapse by identifying and prioritizing buildings susceptible to partial/total collapse. Thus, optimal use of limited resources can be made. A multi-level education and outreach program will provide students from K-12, college, and underrepresented groups, as well as high school physics teachers, with the opportunity to participate in project activities. This will include development of three teaching modules suitable for elementary, middle, and high school students covering the basics of earthquake engineering and topics relevant to collapse analysis. This award is part of the National Earthquake Hazards Reduction Program (NEHRP).

Project Duration:                01/2014~06/2014

Research Assistant:   Carlos A. Santana

Project Type:   Master Project

In this work, the description of the ground motions and its implications are shown first. Then, the novel technique known as “Hybrid Simulation” (HS) is described with its overall features, and right next, the explanation of several tools (algorithms) used to decipher the dynamic problem seeing several different approaches and concepts. All these tools are fully HS-compatible and their implementations are portrayed including testing validations.

Cross Laminated Timber (CLT) is a building material that was developed in the 1990’s. Initially it was primarily used in Europe. Slowly it has been gaining popularity as a building material in the United States. However, it is far from being commonly used. It is used as a steel or concrete substitute for framing.

CLT is composed of three or more layers of lumber that are adhered together. The layers are stacked crosswise to increase durability. Figure 1 shows how the layers are arranged. CLT has a number of advantages including the following:

         • Low material weight with comparable strength

         • Seismic performance

         • Versatile

         • Made from renewable resources

         • Low jobsite waste

         • Prefabricated for fast and easy installation

         • Carbon absorption for Green building  

A monotonic test is performed in order to develop the cyclic loading protocol. Then, the cyclic test will be performed. The data collected from this test will complement similar testing being done at Clemson University in South Carolina.

Project Duration:                09/2013~01/2014

Research Assistant:   Mohamed Rusthi

Project Type:   Master Project

Earthquake response of steel moment-resisting frames can be evaluated using currently available finite element software SAP2000 and computational platform RT-Frame2D. The objectives of this master project are; 1) design and analysis of a prototype 3D structure using SAP2000 and conduct compete earthquake analysis according to ASCE 7-10, 2) analyse a representative 2D moment resisting frame from the previous 3D structure and compare the response with the newly developed framework RT-Frame2D, and 3) evaluation of RT-Frame2D in terms of its modeling capabilities, implementation, and output results of the dynamic response. The SAP2000 finite element analysis was performed using a four story steel moment resisting frame building with different earthquake analysis methods provided in ASCE 7-10. And the representative 2D moment resisting frame of the building was analyzed using a newly-developed computational platform RT-Frame2D which uses MATLAB/Simulink environment. Finally, the different modeling options available in RT-Frame2D framework were evaluated by comparing the fundamental period of vibration and dynamic responses such as acceleration, velocity, and displacement. The results when compared with SAP2000 showed that the RT-Frame2D framework is reliable for dynamic analysis. Even though, RT-Frame2D has different modeling options such as flexible connections and panel zone elements, replicating the same models in SAP2000 was not possible. When the responses from flexible connection and panel zone models from RT-Frame2D were compared, the addition of panel zones when there is a flexible connection makes no difference in the responses. On the other hand, models without and with linear panel zone have the same responses.

Project Duration:                01/2013~12/2013

Research Assistant:   Christopher S. Sawyer

Project Type:   Master Project

In order to represent the three-story steel frame structure of the Laboratory of Earthquake Structure Simulation (LESS) numerically and maintain stability, a friction damper was designed and installed to the existing very lightly damped specimen. The effect of the friction damper on the seismic response was evaluated through both numerical simulation and experimental investigation. The damping ratio of the structure was increased from 0.277% to 1.699%. Though the desired realistic damping ratio of 2% to 3% for steel structures was not reached, stability was realized in the numerical models. A validated numerical model of both the third and second story substructures were established in MATLAB/Simulink for future development of versatile real-time hybrid simulations at LESS.  

Description: This project utilizes the unique hybrid simulation capability provided by the NEES facilities to integrate the physical experiments on four sets of three columns of four buildings and the numerical simulation of the remaining reinforced concrete frame structure. The goals of the project are to determine the effects and important characteristics of triaxial as opposed to unidirectional seismic ground motions on column failure and collapse mechanisms, to develop reliable analytical modeling tools and methods for collapse analysis and to develop system level acceptance criteria and procedures for collapse analysis of reinforced concrete structures.

For the distributed hybrid simulation, the prototype structure shown below was substructured into three modules. Module 1 is the left column, which was represented by a cantilever column specimen located at WMU. Module 2 is the beam, which was simulated numerically in OpenSees at HIT. Module 3 is the right column, which was represented by a spring specimen located at HIT. UI-SimCor was used as the computational driver.

The transfer of data between UI-SimCor and OpenSees was achieved through a program included in the UI-SimCor package called Network Interface for Console Applications (NICA). However, communication with the physical substructures is more challenging. This was realized through a series of connections. First, a LabVIEW Virtual Instrument (VI) called Network Interface for Controllers (NICON) was used to send and receive data to and from National Instruments (NI) hardware. NICON was developed by Dr. Oh-Sung Kwon and Viswanath Kammula at the University of Toronto.

At HIT, a NI USB-6259 data acquisition device was physically attached to the computer running UI-SimCor via a USB port. NICON was run on the same computer, allowing data transfer between UI-SimCor and the NI device. The data was then transferred to dSPACE through Bayonet Neill-Concelman (BNC) connectors. dSPACE was then able to directly control the hydraulic actuator and receive feedback.

At WMU, the communication is more intricate. NICON was deployed to a NI PXI-8108 processor, and the internet protocol (IP) address of the processor was specified within UI-SimCor. This established a connection between UI-SimCor at HIT and the NI processor at WMU. Also, since the WMU network firewall blocks incoming connections from outside the network, the WMU virtual private network (VPN) had to be utilized at HIT. The VPN client was installed on the same computer running UI-SimCor, and a session was initiated prior to the distributed hybrid simulation. This allowed the computer to send and receive data to and from the NI processor as if it were directly connected to the WMU network. Next, BNC connectors were used to connect between the NI processor and the SC6000 hydraulic controller at WMU. This SC6000 was able to control the hydraulic actuator and receive feedback.

Project Duration:                09/2012~07/2013

Research Assistant:   Bradley Gerbasich

Project Type:   Master Project

This project entailed two separate parts, each with different configurations, design loads, and considerations. For both Part 1 and Part 2, a SAP2000 model was created to reflect the geometry of the building. The design loads were assigned to the model utilizing ASCE 7-10 to determine the loading according to the applied loads or given building criteria.  Also, the Loading Combinations found in ASCE 7-10 were utilized within the SAP2000 model to develop the design loads for each individual member. For Part 1, the design of the steel members utilized both the Structural Steel Design textbook and the AISC Steel Construction Manual. For Part 2, the design of the steel members utilized both the AISC Steel Construction Manual and the AISC Seismic Design Manual.

Project Duration:                01/2013~05/2013

Research Assistant:   Adnan Sanchez 

Project Type:   Master Project

Real-time pseudo-dynamic hybrid simulation is an experimental method consisting physical testing and numerical simulation with the objective to reproduce dynamic behavior of a structural system.  This project evaluates the compensation techniques necessary for real-time hybrid simulation which include additional error compensation for amplitude mismatch between command and measured, first and second order feedforward predictors to account for inherent actuator delays. The best compensation methods were determined for both linear and nonlinear structural models that would render the least peak error (PE) and the root mean square (RMS).

Project Duration:                09/2011~04/2012 

Research Assistant:   Kevin Phillips 

Project Type:   Master Project

Pseudo-dynamic testing is increasingly being used for testing structures that are subject to seismic loadings.  Due to the limited capacity of available testing facilities and also due to economic reasons, testing is often carried out on scaled down models, rather than full scale structures.  The various aspects that are considered, when selecting scale factors for the similitude laws chosen, are explained.  The question of the practicality of scaled down testing on structures is examined using both open and closed loop pseudo-dynamic testing procedures. It was found that the results obtained from pseudo-dynamic testing of scaled models can be considered identical to full scale responses, when used for practical purposes.

Project Duration:                12/2011~04/2012

Research Assistant:   Roger A. Sanchez M. 

Project Type:   Master Project

Incremental Dynamic Analysis (IDA) is an analysis procedure by which can be obtained the dynamic response of a structural model exited by several seismic loads where increasing intensities are applied to analyze the structure performance from its elastic behavior to inelastic response until collapse. After the execution of several IDA analyses the results can be summarized on IDA curves to have a graphical representation of the structure’s performance. On this project the procedure to perform IDA analysis is detailed step by step using the software SAP2000 where it is explained how to create the model of a 2D steel frame, define load and masses, assign section properties, define nonlinear properties, add earthquake record data to the model, define time history dynamic analysis case with the required configuration to get the nonlinear response of the model and how to display and export the results. The procedure to perform IDA analysis is executed using time history analysis showing how to scale the earthquake data and run the analysis for several scale factors to get the response of the structure from its linear behavior to dynamic instability. IDA curves are then generated with the data obtained from the sets of dynamic analysis using Microsoft Excel.

Project Duration:                02/2011~01/2012

Research Assistant:   Chee Kian Teng

Project Type:   Master Thesis

Structural Health Monitoring (SHM) of bridge has rapidly become one of the main interests in civil engineering field. The state-of-the-art sensor monitoring technology poses as one of the most efficient and accurate SHM methods. The premise is that changes to structural properties caused by damage will change the way a structure responds to ambient motions. Modal analysis algorithms are applied to the collected vibration responses of a bridge structure (i.e. a bridge) and the structural properties can be extracted. Finally, with the extracted modal properties, damage detection methods can be used to detect and identify the damages developed within the structure under investigation. The study evaluates three output-only modal analysis algorithms with a wireless sensor network (WSN) based on their accuracy and efficiency in extracting modal properties. Output-only algorithm requires only the measured system response (i.e. displacement, velocity or acceleration), while the input, which is the quantification of the excitation force, is not required. The algorithms studied are: Stochastic Subspace Identification (SSI), Auto-regressive Moving Average (ARMA) and Fast Fourier Transform (FFT). These algorithms were used to extract modal properties using acceleration responses collected using WSN from two case studies. The case studies analyzed which are a three degree of freedom (DOF) benchmark structure and a highway bridge in Holland, Michigan. Based on the accuracy and consistency of the modal properties extracted using the algorithms, FFT was determined to be the most accurate and consistent, followed by ARMA and SSI. The extracted modal properties of the Holland Bridge were confirmed by modal frequencies obtained from a Finite Element (FE) bridge model. Recommendations on sensor's placement for future bridge SHM-related projects were provided.

Project Duration:                08/2009~06/2011

Research Assistant:   Griffin Enyart

Project Type:   Master Thesis

The hybrid testing method was developed to evaluate the seismic performance of a structural system by physically testing part of the structure, called a physical substructure, while numerically simulating the rest of the structure using a computer model, named as computational substructure. Instead of building full sized structural specimen, hybrid testing allows researchers to build a complex substructure to be tested experimentally while the relatively simple part of the structure being numerically simulated. Recently versatile hybrid testing system was built at Western Michigan University including a seismic simulator (often called shake table), a reaction/loading system and an advanced controller. Such testing system was used to evaluate the seismic performance of a three story structure whose top story installed with motion mitigation devices (i.e. dampers). The physical substructure is the top story with damper that was installed on the shake table and the numerical substructure is the bottom two story simulated in the computer. The boundary motions between the physical and numerical substructure were numerically simulated and applied to the top story using the shake table. Test results as well as the development of the test system is presented. 

Project Duration:                08/2009~01/2010

Research Assistant:   Mark Humiecki

Project Type:   Master Thesis

This study evaluates three damage detection methods with a wireless sensor network based on accuracy and efficiency. The three methods assessed are: Modal Assurance Criterion (MAC); Damage Location Assurance Criterion (DLAC); and a method which utilizes the change of the structures flexibility. The concept of each method is discussed and demonstrated with a numerical example of a three story shear type structural model that was designed and constructed as the benchmark structure used in this study. The wireless sensor network adopted here utilizes the Intel Mote 2 pre-configured with the latest tool suite released by the Illinois Structural Health Monitoring Program (ISHMP). The establishment of the wireless sensing network, from software installation to data analysis, is presented. The three damage detection methods are then experimentally validated by using the wireless sensor network to monitor the benchmark structure model. Experimental results are presented and compared to demonstrate the performance of these three methods. The MAC method, with this simple benchmark structure, was determined to be the most accurate and efficient. The flexibility-based method was found to be the least accurate