Toshiro Hayashikawa

Load-carrying capacity of steel-to-timber joints with a pretensioned bolt

Previous experimental studies reported that bolt pretensioning greatly increases the initial stiffness and load-carrying capacity of bolted joints. It is also a matter of great importance to structural designers to understand the effect of pretension on the load-carrying capacities of bolted joints, and this study presents an extended yield model that considers the fastener’s pretension force. In the extended yield model, the load-carrying capacity was defined as the load at a slip of 15 mm. The ultimate fastener bending angle at the yielded cross section equivalent to this joint slip, which was affected by the fastener’s axial force, was iteratively evaluated in numerical analyses. The introduction of bolt pretensioning largely increased the joint slip resistance at initial loading, but it decreased the ultimate fastener bending angle. This decrease of fastener bending angle resulted in a relatively low stiffness hardening (or secondary stiffness), which is caused by secondary axial forces associated with embedment of steel plates into the wood member. Prediction was verified by the tests of 36 steel-to-timber joints under three different pretension forces and two loading directions relative to the grain. Some of the observed load-carrying capacities of the joints, particularly in loading perpendicular to the grain, however, were not as high as those expected by the numerical analyses considering the given pretension forces.

Bearing properties of Shorea obtusa beneath a laterally loaded bolt

Empirical equations to determine the bearing strength have been proposed by many researchers and design standards. Because these equations have been developed mainly based on test results of softwood species, it is a matter of great importance (to ASEAN structural engineers) to verify the applicability of these equations for tropical hardwood species, which are commonly used in many ASEAN countries. In this study, wood specimens of Shorea obtusa (a tropical hardwood species) were used and the bearing test under full-hole confi guration was carried out for fi ve different loading angles to the grain. The bearing stress-embedment curve obtained from the test was approximated by a linear elastic-plastic diagram indicating the initial and fi nal stiffness of the curve. Testing showed that the average bearing strength parallel to the grain was 7.25% lower than the prediction given in Eurocode 5. The bearing strength perpendicular to the grain evaluated based on bearing load at initial cracking was substantially different from any predictions given by previous studies or design standards. It was also found that the bearing strength and initial stiffness from the bearing stress-embedment curve for loading at intermediate angles to the grain could be satisfactorily predicted with Hankinson’s formula.

Effects of pretension in bolts on hysteretic responses of moment-carrying timber joints

The adoption of a concept similar to the prestressing technique used in laminated wood decks of bridge structures might increase the initial stiffness or ultimate resistance of dowel-type timber joints by applying pretension to their bolts. This study investigated the effect of pretension in bolts on hysteretic responses and ultimate properties of moment-carrying timber joints with steel side plates. A pretension of 20 kN that yielded a prestress level of 1600 kPa or about 90% of the allowable long-term end-bearing strength of spruce species was applied to the bolts of prestressed joints. The superiority of the prestressed joint over the non-pre-stressed joint was proved by very high hysteretic damping, equivalent viscous damping ratio, and cyclic stiffness. At any given rotation level, hysteretic damping reduction and moment resistance decrement due to continuously reversed loads were found to be small because bolt pretensioning minimized the pinching effect. This study showed that the hysteresis loop of the prestressed joint can be obtained by adding the frictional hysteresis loop due to pretension force into the hysteresis loop of the non-pre-stressed joint. Despite a great increase of initial stiffness, only slight increments in ductility coefficient and ultimate moment resistance were found in the prestressed joint.

Soil-structure interaction modeling effects on seismic response of cable-stayed bridge tower

A nonlinear dynamic analysis, including soil-structure interaction, is developed to estimate the seismic response characteristics and to predict the earthquake response of cable-stayed bridge towers with spread foundation. An incremental iterative finite element technique is adopted for a more realistic dynamic analysis of nonlinear soil-foundation-superstructure interaction system under great-earthquake ground motion. Two different approaches to model soil foundation interaction are considered: nonlinear Winkler soil foundation model and linear lumped-parameter soil model. The numerical results show that the simplified lumped-parameter-model analysis provides a good prediction for the peak response, but it overestimates the acceleration response and underestimates the uplift force at the anchor between superstructure and pier. The soil bearing stress beneath the footing base is dramatically increased due to footing base uplift. The predominant contribution to the vertical response at footing base resulted from the massive foundation rocking rather than from the vertical excitation.

One-year stress relaxation of timber joints assembled with pretensioned bolts

In our previous study, great increases of hysteretic damping and initial slip resistance of timber joints were attained by applying axial pretension to the steel fasteners. To evaluate the effectiveness of this method, 1-year stress-relaxation measurement was carried out. Nine prestressed joints were prepared and three of them were restressed after 3 and then 6 months after the initial prestressing. All joints were exposed to indoor conditions, and relaxation of the pretension was regularly measured from time-dependent decreases of axial strain of the bolts. After measurement, the joints were subjected to cyclic and monotonic loading tests until failure. The average ratio of residual stress to the initial prestress after 1 year was about 0.23 and 0.66, respectively, for joints without restressing and those with restressing. A simulated stress-relaxation curve developed from the four-element relaxation model predicted 3% of the initial stress after 5 years. Without a regular restressing program, the initial prestressing effect therefore must be considered negligible. However, about 20% of the pre-stress level can be reasonably assumed if restressing is carried out annually. This small residual stress was found to introduce suffi cient frictional damping to signifi cantly increase the equivalent viscous damping ratio of the joints.

Effects of bottom bracings on torsional dynamic characteristics of horizontally curved twin I-girder bridges with different curvatures

Curved twin I-girder bridges (CTIGBs) have low torsional stiffness that makes them vulnerable to dynamic loads. This study investigates the effects of bottom bracings on the torsional dynamic characteristics of CTIGBs. Five types of bottom bracings are designed to investigate their effects on the dynamic characteristics of CTIGBs with different curvatures under free and forced vibrations. To perform numerical investigations, three-dimensional (3-D) finite element (FE) bridge and vehicle models are established using commercial ANSYS code, and then a vehicle-bridge interaction analysis approach is proposed. Road roughness profiles generated from power spectral density and cross spectral functions are also taken into account in the analyses. The numerical results show that torsional frequencies increase significantly after providing bottom bracings, and the increasing rate depends on the type of bottom bracings and their locations of installation. Bottom bracings can act as load transmitting members from one main girder to the others. Large negative bearing forces that have occurred in bridges with small radii of curvatures can be remarkably reduced by providing bottom bracing systems. It is found that the performances of several bottom bracing systems are effective in improving the torsional dynamic characteristics of the bridges in this study.

Control Strategy for Seismic Pounding Mitigation of Bridge Structures

Damages of adjacent bridge structures due to relative responses such as poundings and unseating have been observed in many earthquakes. The isolators in bridge structures are effective in mitigating the induced seismic forces, however, the deck displacement becomes excessively large when subjected to a ground motion with unexpected characteristics, therefore increase the possibility of pounding; contribute to the unseating of bridge decks and subsequent collapse. An analytical model of expansion joints that takes account of the interaction between adjacent segments of the bridge deck and the effect of impact and restrainers is developed for nonlinear time history analysis. The numerical results show that pounding between adjacent bridge segments could amplify the relative displacement, resulting in the requirement of a longer seat width to support the deck. Pounding results in a transfer of large lateral force from a deck to the other, consequently, results in significant changes in the global response of the participating structural systems. So it is effective to provide a shock absorber between adjacent decks for the mitigation of pounding effect.

Effectiveness of Stoppers and Shock Absorber Devices in Mitigating Earthquake Damage to Curved Viaducts with Steel Bearing Supports

Viaducts with steel bearings are subject to damage during earthquakes. Damage to conventional steel bearings could cause viaduct decks to collapse. This study attempts to prevent large displacements by installing steel stoppers at both sides of roller bearings. Shock absorber devices (SADs) are also installed between roller bearings and stoppers to reduce pounding forces. The study then explores the effectiveness of different stopper gaps and SADs installed between roller bearings and stoppers in mitigating viaduct damage. The effectiveness of stoppers without and with SADs is also investigated. The pounding forces between roller bearings and stoppers, the displacements on the tops of piers, the displacements of superstructures, and the bending moment–curvature relationship at pier bases are evaluated accordingly. Numerical results reveal that earthquake damage to viaducts with small stopper gaps is not significant. The application of SADs is also found to significantly mitigate viaduct damage.

Prediction and mitigation analysis of ground vibration caused by running high-speed trains on rigid-frame viaducts

In this study a 3D numerical analysis approach is developed to predict the ground vibration around rigid-frame viaducts induced by running high-speed trains. The train-bridge-ground interaction system is divided into two subsystems: the train-bridge interaction and the soil-structure interaction. First, the analytical program to simulate bridge vibration with consideration of train-bridge interaction is developed to obtain the vibration reaction forces at the pier bottoms. The highspeed train is described by a multi-DOFs vibration system and the rigid-frame viaduct is modeled with 3D beam elements. Second, applying these vibration reaction forces as input external excitations, the ground vibration is simulated by using a general-purpose program that includes soil-structure interaction effects. The validity of the analytical procedure is confirmed by comparing analytical and experimental results. The characteristics of high-speed train-induced vibrations, including the location of predominant vibration, are clarified. Based on this information a proposed vibration countermeasure using steel strut and new barrier is found effective in reducing train-induced vibrations and it satisfies environmental vibration requirements. The vibration screening efficiency is evaluated by reduction VAL based on 1/3 octave band spectral analysis.

Assessment of modal pushover analysis procedure for seismic evaluation of buckling–restrained braced frames

Buckling-restrained braced frame (BRBF) has become a viable choice of lateral-force resisting system for earthquake resistant buildings as its hysteretic behaviour is non-degrading and much hysteretic energy can be dissipated. Meanwhile, an improved pushover analysis procedure, called Modal Pushover Analysis (MPA), has been proposed by Chopra and Goel [(2002), Earthquake spectra, 20(1), 255–271] to include contributions of higher ‘modes’ in estimating seismic demands of buildings. The MPA procedure was demonstrated to provide reasonably accurate seismic demands compared with non-linear response history analysis (NL-RHA) when analysing moment-resisting frames. This research further examined the bias and accuracy of MPA procedure when it is applied to analyse 3-, 6-, 10- and 14-storey BRBF buildings due to two sets of strong ground motions having 2% and 10% probability of being exceeded in 50 years. The MPA estimates are compared with ‘exact’ results from NL-RHA. It was found that the peak roof displacements can be estimated by responses of equivalent single-degree-of-freedom systems and MPA procedure can accurately predict the maximum inter-storey drift over all storeys. However, the bias of MPA in estimating peak storey drifts at certain stories can be significant.