The National Science Foundation's (NSF) Tokyo Office periodically receives and disseminates reports on research developments in Japan that are related to the Foundation's mission. NSF-sponsored researchers currently working in Japan prepare many of these reports. These reports present information for use by NSF program managers and policy makers; they are not statements of NSF policy. .
Mr. Jerome Lynch, a Ph.D. student in the Department of Civil and Environmental Engineering at Stanford University in Stanford, California, prepared the following report. Mr. Lynch was a participant in the 1998 Summer Institute sponsored by NSF/NIH/USDA and the Science and Technology Agency of Japan. Dr. Akihiro Kondo of the Kobori Research Complex at Kajima Corporation, hosted Mr. Lynch. Mr. Lynch can be reached via email at: jplynch@stanford.edu
Japan is a nation that has been plagued by earthquakes and other natural disasters such as typhoons since the beginning of time. Such harsh environs have resulted in a robust people who seek to protect themselves from these natural disasters. For well over two decades now, Japanese structural engineers have been a major contributor to active structural control research. Structural control technology, still nascent in its development, is a technology used in civil engineering structures to reduce vibrations of
the structure which have resulted from strong winds or earthquakes. For instance, it is quite evident that almost all structural engineering programs on the university level in Japan have made active structural control research a top priority. Japanese construction companies such as Kajima Corporation in Tokyo have taken the worldwide lead by developing successful structural control systems for buildings they have designed and constructed. Through a unique opportunity provided by The Summer Institute in Japan Program, I spent eight weeks in the summer of 1998 at Kajima Corporation investigating structural control systems that they have developed and are developing. Of all of the Japanese companies developing such control systems for structures, Kajima has been the most successful with five different active control systems fully developed to date which have all been used in actual structures. Such an opportunity can only be found in Japan due to the unfortunate fact that the American building industry has not looked seriously enough at active structural control.
Civil engineering structures located in environments where earthquakes or large wind forces are common will be subjected to serious structural vibrations during their life spans. These vibrations can range from harmless to severe with the later resulting in serious structural damage and potential structural failure. For a moment, imagine a world in which structures have the ability to reduce these vibrations resulting in a structure that is damage proof during earthquakes and strong winds. Even though engineers cannot design this type of building yet, the field is getting closer to attaining this goal. Structural control is one area of current research that looks promising in attaining the goal.
Structural control is defined as a mechanical system that is installed in a structure to reduce structural vibrations during loadings such as strong winds and earthquakes. The purpose of such a structural control system is to enhance the safety as well as improve the habitability of structures during these loading scenarios. The structural control system can be divided into two parts, with the first part being the actual control device and the second the control algorithm. It is common for the structural control system to be classified by its device type resulting in three general control types; passive, active and semi-active. If the system is an active or semi-active structural control system, then the algorithm type can be used to further classify the system; feed forward and feed backward.
A passive control system is one in which structural vibrations are reduced from a passive control device imparting a force upon a structure out of response to the motion of the structure. Passive control has many benefits associated with it. First, no external power is required for the passive device to work. This makes a passive device an economical solution. In addition, the device will generally be smaller in size than an active control device. Furthermore, passive devices have been in existence for well over 50 years and have been thoroughly researched and tested resulting in a reliable product. However, there is a negative aspect to the passive device and that is only a limited amount of control can be attained. Even in light of this fact, they are still considered a very cost-effective solution to controlling structures. Examples of passive control devices include base isolation, tuned mass dampers, viscous dampers, elasto-plastic dampers, metallic yield dampers and friction dampers.
An active control system is a much more complex system than the passive control system. External power is employed to power actuators located in the structure in order to apply forces that can put in or take out energy from the system. In order for the actuators to properly apply the desired forces, sensors need to be placed within the structure in order to measure structural response. These sensors relay response information to a central computer that then uses this information to calculate desired actuator forces. The advantage of an active control system is that the system attains excellent control results. However, there are many drawbacks to using an active control system. They are very expensive systems to design and are expensive to operate due to the large amounts of power they need. Furthermore, they tend to take up more space than passive control devices. Some examples of active control devices include the active mass driver system, the active tuned mass system, and the active-passive composite tuned mass damper.
The last broad category of control is semi-active control. Semi-active control falls between passive and active on the control spectrum. A semi-active control system is similar to an active system in that actuators operating on external power are used but the actuators do not add energy to the structure in anyway. The actuators are used to control or assist a passive control device. The inherent benefit of a semi-active control device is that the actuator used does not require large amounts of external power. Semi-active systems are more aggressive than passive systems and usually obtain control results close to that of an active control system. An active variable stiffness system as well as the active variable damping system are considered semi-active systems.
Active and semi-active control systems can be further categorized as a feed forward system or a feed backward system. The feed backward system is the most common control algorithm type that uses information from sensors measuring the structure's behavior to determine actuator forces. In control theory, such a system is also termed a closed loop system. In a feed forward system, the input disturbance on the structure is measured and is used to determine actuator forces. Again, in control theory, this type of system is an open loop system.
The following is a list of structural control systems that I personally studied and researched during my summer at Kajima Corporation.
AMD - Active Mass Driver System: The active mass driver system was the first control system developed by the Kobori Research Complex in the late 1980's. In principle, it is a mass whose motion (displacement, velocity and acceleration) is controlled by a turn screw actuator. Accelerometers are located through out the structure and provide information to the control computer. The computer then uses an acceleration feed back algorithm to calculate the desired behavior of the active mass. The AMD was designed with the intent of controlling structures during winds and moderate earthquake. For these applications, the mass of the AMD is roughly 1% of the total above ground structural weight. The AMD is always on and uses a large amount of power daily. An AMD system could be used for large earthquakes but would require tremendous amounts of electricity and the required mass would be extremely large.
In 1989, the Kyobashi Seiwa Building was constructed using an AMD system. The building is an extremely narrow building measuring 4m x 33m in its floor plan and is 11 stories tall. Due to its tall slender shape, the building is susceptible to transverse vibrations from winds and earthquakes. To control these vibrations, an AMD system was installed upon the roof of the building. Two masses where used. One was placed very close to the center of gravity to control transverse displacements while a second was placed at the front of the roof to control torsional vibrations. Excellent control results were obtained. In the first mode, 20% damping increase was gained while in the second and third modes 5% damping was gained. This is a significant control result at only 1% of the total construction cost. It should be noted that the system is completely shut down if electricity to the building is lost.
TRIGON - Active Tuned Mass Damper: Shortly after the completion of the AMD system, Kajima sought to develop a better active control system using less external power. The TRIGON system was one such system that resulted. TRIGON is a tuned mass damper whose stroke is enhanced by an actuator. This system is for controlling structures during strong winds and moderate earthquakes. When winds are small, the actuator is not in use that means no external power is used. In this situation, the mass of the TRIGON system acts as a tuned mass damper system. During strong excitations, the actuator is turned on and allowed to act. The mass of the TRIGON system is only 0.25% of the total building weight. The actual TRIGON system is 330 metric tons. The total stroke of the mass is 100 cm. The system is very compact and can easily be installed in a structure.
In 1994, the TRIGON system was installed within the newly completed Shinjuku Park Tower in Tokyo. The tower is 54 stories tall and due to its slender shape is susceptible to transverse and torsional vibrations. The system was installed upon the 39th floor and was installed in one day. The system targets the first mode of vibration with damping increased by 5-10% in this mode. In terms of displacements, a 33% to 50% reduction is seen at the top floor. The total cost of the system is 0.5% of the total construction cost.
DUOX - Active-Passive Composite Tuned Mass Damper: The DUOX system is the second system to result from Kajima seeking to improve upon the AMD system. DUOX is composed of a passive tuned mass damper upon which sits an active mass driver system. The actuator controls the active mass driver portion whose inertial force controls the tuned mass damper which in turn places a force upon the structure used to control. This system is an active control system used for vibrations from winds and moderate earthquakes. The total mass of the DUOX system is 0.5% of the total above ground building weight. The DUOX system can control a structure in the X,Y and torsional directions. Very limited power is needed like the TRIGON system and it is very compact. The active mass driver portion of the system only goes on during excitations whereas the tuned mass damper portion is always active. During power failure, the system behaves simply as a tuned mass damper. The system targets the first mode of vibration of the structure and can increase damping by 15% in that mode.
In 1993, DUOX was installed on the roof of the Ando Nishikicho Building. The building is a perfect square in floor plan and is 14 stories tall. Given the wind direction, the building could experience transverse or torsional displacements. The system was quite successful and reduced vibrations of the structure by 66%. An equivalent AMD system that can achieve similar results would be 10% of the building mass and would be terribly large. This system was so successful, that Kajima decided to place it within the Dowa Kasai Pheonix Tower of Osaka, which was completed in 1995.
AVS - Active Variable Stiffness System: This system is a semi-active control system that is ideal for structures employing braced frames for lateral resistance. In the braced frame, a lock is placed which can be locked and unlocked. In its locked state, the brace is active in resisting lateral deflections. When the lock is unlocked, the brace is unattached from the frame and the brace is not effective in resisting lateral deflections. This system is only for earthquake excitations and employs a feed forward control algorithm. Limited power is needed and it is powered by external power sources. During an earthquake, a tall building having AVS system at every braced frame can turn the locks on and off so as to get the building's stiffness away from the resonant frequency. This is extremely successful during narrow banded disturbances and not as effective during wide banded disturbances. The AVS system was used in the construction of the Shaking Table Laboratory at the Kajima Technical Research Institute.
In conclusion, corporations in Japan such as Kajima have proven that active control can be effective in reducing structural deflections during severe loading scenarios. However, what is lacking in all of these systems is proof that they are cost effective. Perhaps one cause for this hurdle is that the control systems developed to date are too ambitious in the goals they seek to accomplish. This observation has prompted me to pursue research efforts that focus more closely upon structural control systems that are more localized and target only critical areas of the structure. This is in contrast to the control systems discussed above which all seek to control the overall behavior of the entire structure. By concentrating more closely upon a smaller portion of the structure, the control systems would require less energy and be more economical to manufacture and install. The data and experiences I accumulated in Japan during the Summer Institute were instrumental in pushing my research in this direction.