Objectives of IASC
The International Association for Structural Control (IASC) is a Nonprofit Public Benefit Corporation. The objectives of IASC are the advancement of the science and practice of structural control and monitoring, by means of education, research and application of knowledge. This includes the response of structures to earthquakes, wind and man-made forces.

Membership in IASC
The Association is composed of National Organizations of Structural Control approved for membership by the Board of Directors of IASC. At present, there are national Panels for Structural Control in the USA, Japan, Europe, China, and Korea. Other national Panels are being formed. The “U.S. Panel on Structural Control Research” is the USA member in IASC. The U.S. National Science Foundation supports the operation of the U.S. Panel.

Financial Resources
IASC by itself does not have financial resources; it relies on its national Panels to coordinate their respective research initiatives from their own funds and resources.

Mission Statement for IASC Task Groups
In order to speed the implementation of the general concept of Structural Control worldwide, and in response to the recommendations of the Second International Workshop on Structural Control (2IWSC) held in Hong Kong in 1996, and the Second World Conference on Structural Control (2WCSC) held in Kyoto, Japan in 1998, IASC has established the following Task Groups:
1. Task Group on Benchmark Control Studies
2. Task Group on Health Monitoring
3. Task Group on Codes for Structural Control

Following is an overview of the mission statement for each Task Group, their operational requirements and overall organizational chart.

Organization of IASC Task Groups:
Each Task Group is composed of the three co-chairs plus chairs of the National Committees, as they are formed.

Co-chair A represents Asian countries: China, Japan, Korea, Taiwan,...
Co-chair B represents American countries: Canada, Chile, Mexico, United States,...
Co-chair C represents European countries: France, Germany, Italy, Russia,...

The Co-chairs are responsible for arranging National Committees in countries that wish to participate. The IASC President officially appoints Co-chairs and National Committee members as they are nominated by the Co-chairs. Each National Committee is responsible for obtaining the necessary financial support to cover its organizational and operational activities.

Operational Requirements
Each of the Groups is expected to submit a formal report to the President of IASC on an annual basis summarizing the accomplishment of the Group and outlining future plans. Formal presentations and activities of the planned Groups will be presented at planned IASC International Workshops and World Conferences.

An organizational chart of a typical Task Group is shown in the adjoining diagram.

1. Task Group on Benchmark Control Studies:
The main goal of this Task Group is to initiate the development of a series of representative benchmark problems that can help to focus structural control research for civil infrastructure systems such as buildings and bridges. Simultaneously, experimental benchmark problems should be formulated, including both small and large scale structures.

2. Task Group on Health Monitoring:
The main goal of this Task Group is to prepare plans for a series of international benchmark studies of proposed methodologies for structural health monitoring using response data from full-scale structures in damaged and undamaged states. For the purpose of working towards this goal, the structural health monitoring system is defined as any non-destructive evaluation technology for damage detection, and/or assessment utilizing structural response data.

3. Task Group on Codes for Structural Control:
The main goal of this Task Group is to survey and summarize existing codes and guidelines for base isolation and passive damping control systems, develop a general philosophical basis for the use and design of structural control systems, and recommend guidelines for the design of structural control systems.
Establishment of the U.S. Panel Task Groups:
The U.S. Panel on Structural Control Research, in collaboration with the American Society of Civil Engineering (ASCE), established in 1999 three National Committees that correspond to the three Task Groups listed above. The operation of the three Task Groups is being coordinated in collaboration with the IASC Executive Committee. Initially, the Chair of the ASCE Engineering Mechanics Division Dynamics Committee worked with the Secretary General of IASC to coordinate the establishment of each Task Group from members and friends of the Dynamics Committee of the Engineering Mechanics Division of ASCE who have been active in the respective areas of the proposed Groups. Start-up funds were furnished by ASCE for each Group (to cover the travel/meeting expenses of about five members per Task Group). Depending on the progress of the Task Groups, it is expected that funds to sustain the operation of the Task Groups beyond the start-up phase may be sought from sources outside ASCE.

A brief report on the activities of two of the U.S. Panel Task Groups is given below.

The joint IASC-ASCE Task Group on Benchmark Control Studies, chaired by Prof. Bill Spencer (Univ. of Notre Dame), first met on 14th June, 1999 at the 13th ASCE Engineering Mechanics Conference held at Johns Hopkins University (Baltimore, Maryland). The goal of the task group was to develop a series of structural control benchmark problems which will assist the research community at large in assessing the relative merits of different control strategies
At this initial meeting, several teams were formed to begin or continue development of several benchmark control studies. In total, four problems are under development:

1.  Wind Benchmark problem
2.  Seismic Building Benchmark Problem
3.  Seismic Bridge Benchmark Problem
4.  Base-Isolation Benchmark Problem

Wind Benchmark Problem
Prof. Yang (Univ. of California, Irvine) presented the Wind Benchmark Problem, which he has developed with Prof. Agrawal (City College of New York). The simulation model is of a 76 story structure, shown in Figure. 1, subjected to along- and across-wind forces. The across-wind forces were deemed to be important subsequent to wind-tunnel testing performed at Univ. of Sydney, Australia by Prof. Samali. In addition, the structure is symmetric, so the centers of mass and stiffness are coincident. Thus design of along- and across-wind control strategies can be developed independently. Up to date details of the wind benchmark problem can be found at: www-ce.engr.ccny.cuny.edu/people/faculty/agrawal/benchmark.html.

Seismic Building Benchmark Problem
Dr. Ohtori and Prof. Spencer presented the proposal for a Nonlinear Seismic Benchmark Problem (NSBP). The model structures to be considered are the 3-, 9-, and 20-story SAC buildings designed for the Los Angeles area. The 9-story building is shown in Figure 2. The proposed excitations are various scaled versions of the 1940 El Centro, 1968 Hachinohe, 1994 Northridge, 1995 Kobe earthquake and something from the recent Turkey earthquake. The simulation will be implemented using SIMULINK. For purposes of comparison, researchers should implement their control strategies neglecting sensor and actuator dynamics, however, these may be considered in addition. Details of the seismic building benchmark problem can be found at: www.nd.edu/~quake/benchmarks/bench3def.

Seismic Bridge Benchmark Problem
Prof. Dyke (Univ. of Washington, St. Louis) presented the proposed Seismic Bridge Benchmark Problem, which is based on the Cape Girardeau bridge (Missouri) which is currently under construction (see Figure 3). The proposed control objectives will concentrate on the cable tension, tower base shears, overturning moments, tower compression forces and pounding of the deck against abutments. At this preliminary stage it was decided not to consider different excitations at different supports.

Base-Isolation Benchmark Problem
Prof. Watanabe (Keio University) presented the proposal for a Base-Isolation Benchmark Problem. For the proposed base-isolated structure model, it will be assumed that the superstructure behaves linearly, and that the isolation layer possesses either a combination of linear stiffness and hysteretic damping or bilinear hysteresis. The superstructure will be 6-10 stories, and have a nominal damping ratio of 0-2%. Again the proposed excitations will be scaled versions of several major earthquakes, and will be applied in one horizontal direction. The main evaluation criteria will be peak base drift, base acceleration, structural acceleration, base shear, and interstory drift. Researchers will be encouraged to use any type of control strategy, i.e., passive, semi-active or active. For further information on the activities of the Task Group on Benchmark Studies contact Prof. Bill Spencer at: spencer@nd.edu.
 

The joint IASC-ASCE Task Group on Health Monitoring, chaired by Jim Beck (California Institute of Technology), met for the first time on 13 June 1999, at the 13th ASCE Engineering Mechanics Conference held at Johns Hopkins University (Baltimore, Maryland). The declared goal of the group was to focus and coordinate research in the field of structural health monitoring and system identification. The most direct means of reaching these goals is to propose various benchmark problems which the greater research community could work on and compare levels of performance. Although previous blind benchmark tests (for example, that proposed at the 1997 IMAC conference) generated significant interest, very few researchers actually submitted identification results. Because of the demonstrated hesitance to engage in blind benchmark tests the task group decided to develop benchmark tests where the damage states are known.

At this initial meeting, an immediate plan of action was agreed upon to generate a computer model of an existing test structure at the University of British Columbia (Vancouver, Canada). The test structure, shown in Figure 4, is a 1/3-scale 4-story 2 bay x 2 bay steel-frame structure which can later be physically tested by Prof. Carlos Ventura (UBC) with the help of members of the task group. Details of the physical test structure can be found at: www.civil.auc.dk/i6/forskpr/proj01/testa.html. Initially members of the group would use the computer simulation response of the structure to perform system identification studies. Later these studies can be repeated through the use of physical test data.

The initial task was then to create a computer model of the test structure. Prof. Erik Johnson (University of Southern California) generated a 12 degree-of-freedom (DOF) model (using MATLAB) which was distributed to the members of the task group. Three damage states in all were considered, which involved the removal of cross-bracing at various floors. The model was excited by uncorrelated horizontal wide-band random forces at each floor in one of the axes of symmetry of the building. Therefore, this initial simulation did not excite any torsional modes.

Members of the task group quickly began working on the simulation output, and met again on 28 August, 1999 at the California Institute of Technology to compare identification results and to discuss future modifications to the simulation models. At this meeting Prof. Joel Conte (University of California, Los Angeles) and Prof. Lambros Katafygiotis (Hong Kong University of Science and Technology) each presented their own mathematical models of the physical test structure which both had 120 DOF’s. One of the 120 DOF models will be used by the members of the group to determine the effects of performing identification with an assumed model which is of lower order than the simulated model.

Various details for the generation of future simulation data were agreed to at the Caltech meeting. For example, the time-step used for integration and the time-step at which response data is recorded were to be treated separately. In addition consensus was reached to adopt the Nigam-Jennings modal integration technique which assumes piecewise linear excitation between samples. The realistic levels of measurement noise to be added to the simulated ‘clean’ data and the spectral content of that noise were discussed and arrived at by testing the UBC instruments, which will ultimately be used when physical testing begins.

Data Analysis for the Preliminary Benchmark Problem
At this meeting the members of the task group presented their identification results based on the simulation of the original 12 DOF model. Prof. Beck and graduate student Ivan Au presented results using their ‘MODE-ID’ program which adopts a Bayesian approach to damage detection to optimize system parameters to better predict the structural response. Prof. Dyke (Univ. of Washington, St. Louis) presented results from a method which began locating modal frequencies from autospectra, and then optimizing the stiffness distribution.

Prof. Katafygiotis (Hong Kong University of Science and Technology) presented results based on an approach similar to that of Beck and Au, but which weighted prediction error differently at different floors. Prof. Ventura (Univ. of British Columbia) presented the modal identification of the physical test structure based on response from both hammer testing and single actuator white-noise excitation. Prof. Bernal (Northeastern University) presented identification results based on the Observer Kalman Filter Identification (OKID) technique. Prof. Smyth (Columbia University) presented work by he and Prof. Masri (Univ. of Southern California) which involved the use of direct least-squares approach to minimize equation error. In addition, Prof. Smyth presented work done by Prof. Betti (Columbia University) and graduate student Hilmi Lus using the OKID method to accurately identify the system properties and changes.

The task group has laid out a schedule of successively more complex damage detection problems involving simulation of the test structure response with either the refined 12DOF model or a 120 DOF model. The principal variations among these cases involve: 1) the investigation of modeling error by performing identification with a lower order model than that which generates the data; 2) exciting the structure at only one location as opposed to excitation at each floor; and 3) introducing non-symmetric damage patterns.

The task group will collect the results and create reports for each case. It is hoped that physical tests of the structure at UBC will have begun before the 3IWSC in Paris in June, 1999.

For further information on the work of the Task Group on Health Monitoring contact Prof. Jim Beck at: jimbeck@cco.caltech.edu.