In the event of mega earthquakes, the magnitude of their ground motions may exceed that specified for the design of elevated-girder bridges. Therefore, the responses of the bridges during the earthquakes will go beyond their ultimate state and may collapse in the worst cases. However, the behavior of the elevated-girder bridges beyond the ultimate state has been investigated only a little to say nothing about their collapse behavior. In order to reduce the damage during the mega earthquakes, it is essential to investigate the dynamic collapse of the bridges and develop some method to control the collapse behavior such that the progressive disproportionate collapse may not occur as observed during the 1995 Kobe earthquake. Herein, the dynamic collapse behavior of thin-walled steel piers is investigated as a key element whose failure may trigger the collapse of the entire elevated-girder bridge systems. First, shake table tests were conducted on single pier models in order to observe their dynamic collapse behavior in detail. Second, based on these test results, the accuracy of a FE-shell model was examined in terms of the applicability to the analysis of the dynamic collapse of the thin-walled steel piers. Third, accurate and practical criterions to predict the dynamic collapse behavior of the piers was discussed both theoretically and numerically.

View full abstract

 The multi-directional seismic behavior of rubber bearings installed in elevated-girder bridges was investigated by a bi-directional shake table test on a relatively large 2-span continuous elevated-girder bridge model, together with an advanced FE analysis of a realistic bridge model. The investigation was focused on the behavior of the rubber bearings and their fixing bolts under biaxial shear forces and bending moments that are ignored in the current seismic design. From this investigation, it was shown that the fixing bolts of the rubber bearings under negative reaction forces may fail by the bending moments around the transverse bridge axis before the shear strain of the rubber bearings attains its allowable value of 250%. A similar phenomenon was observed in the 2016 Kumamoto earthquakes.

View full abstract

 The occurrence of heavily corrosion at the bottom of steel girder ends has been reported. For such situation, analytical and experimental studies on corroded steel girder ends have been carried out. However, in the practical cases, countermeasures for the girder ends have tended to be qualitatively judged ever. This study aims to summarize conventional analytical and experimental data for corroded girder ends. After creating graphs, which represent the influence of corrosion on load-carrying capacity of girder ends, their characteristics are investigated. Herein, the authors set parameters as the remaining thickness ratio, the reduction rate of yield strength that is calculated from structural details and so on. Based on this graphs, equations that estimate the reduction of the load-carrying capacity are proposed considering the lower limit of the reduction rate from the case without damage. Furthermore, by showing the relationship between ultimate and yield strengths without damage and reaction forces, information on the judgement for countermeasure of repair is provided.

View full abstract

 A simple method to simulate time histories of daily maximum wind speed is presented, in which the mean-reverting Ornstein-Uhlenbeck process is incorporated. The data observed at a meteorological station in the past are analyzed to clarify the statistical nature of the daily maximum wind speed at the site, and to obtain the values for parameters in the method. Trials are conducted to investigate the characteristics of the method, and the time histories simulated for the daily maximum wind speed in the year of 2014 are compared with the observation data. It is shown that the simulated ones are mostly similar to the observed one whilst in some cases the simulated ones differ from the observed one in time domain as well as in statistics.

View full abstract

 Fatigue tests have been carried out on out-of-plane gusset welded joint specimens under plate-bending compressive cyclic stresses and tensile cyclic stresses in order to clarify the fatigue crack initiation and propagation behavior, and fatigue strength under plate-bending compressive cyclic stresses. In the fatigue tests, the multiple repeated two-step stress tests have been also performed for the purpose of leaving beach marks in a fatigue failure surface in addition to the constant amplitude stress tests. The fatigue test results suggest that the fatigue crack stops propagating under the compressive cyclic stresses after the crack depth reaches 65 to 75% of plate thickness, propagation behavior before it come up to 50% of plate thickness under the compressive cyclic stresses are the same as those under the tensile cyclic stresses, and fatigue crack initiation life in the compressive cyclic stresses is much longer compared with that in tensile cyclic stresses.

View full abstract

 The concept of performance-based seismic design has been introduced in the 2004 revision and 2007 revision of the Japanese design codes for coastal structures and port structures, respectively. However, the performance criteria for coastal structures is given as the allowable settlement at the crest of the seawalls, and there is no discussion on the allowable horizontal displacement. In reality observed in 1995 Hyogoken-Nanbu earthquake disaster, some seawalls lost their function of continuous walls due to uneven horizontal displacement of the walls. Thus, in this paper, the allowable level of horizontal displacement of seawalls is discussed from the observed damage cases, and the performance criteria is proposed. Then, the feasibility of the proposal was discussed in detail with a project of the remediation of sea walls.

View full abstract

 Some studies indicated one possible factor; the spatial difference in site amplification effect of the surface ground should bring on the heavy damage to the limited area in Mashiki town during the 2016 Kumamoto earthquake. In order to investigate the actual soil property of the area for the experimental proof of the above mentioned assumption, we obtained undisturbed volcanic ash soils of the surface ground at the site close to the seismic station recording the M_J 7.3 mainshock in the heavily damaged area by using the triple-tube sampler. The obtained features of physical and dynamic soil properties of the soil core sample based on the laboratory soil testing are summarized as follows; (1) The grain size of the volcanic ash soil varies widely. (2) The soil particle density of sand and gravel is lighter than that of clay. (3) The dynamic properties of the volcanic ash soil do not depend on the soil type and the confining stress, i.e., the depth. (4) The strain dependencies of shear modulus ratio and damping ratio of the volcanic ash soil are similar to the standard ones of sand and clay, respectively. In addition, we investigated the depth distribution of density at the same site by the loggings, and recognized the similarity in density of each soil type around the area.

View full abstract