The first derailment of a Shinkansen train occurred during the Mid-Niigata Prefecture Earthquake in 2004. This report describes the survey results of ground damage around approximately 190 pairs of elevated piers spanning 1.5km which included the area where the Shinkansen derailed. In particular, the following phenomena were observed and measured: sunken ground around the piers, clearance between the ground and pier, and the flow of muddy water emerging from the clearances. These phenomena were suspected to be closely associated with the topographical and geological features of the affected area. In fact, the prevalence of damage for each topographical condition was highest in the terrace area, lowest in the lowland area, and intermediate in the alluvial fan area. Because the terrace area is formed of a thick layer of soft sediments (N-value=0〜19), as compared with the lowland area, the depth of the pier foundations was greater in the terrace area than in the lowland area. To investigate the process of derailment of the Shinkansen, the distribution of train's time staying on track (TST) for each track marker was analyzed. The point where the gauge of track had widened (205k960m) was taken as the starting point for TST; at 206k200m the track gauge has widened to greater than 40mm. The longest TST interval (i.e., the time required for the length of the train to pass a given point) coincided with the position of the train during the arrival of the S-wave between S1 and S2, the so-called principal motion. These findings suggest that an important factor in the derailment of the Shinkansen was the strong motion of the thick, low N-value terrace deposit.

View full abstract

Laboratory experiments of live wood cuttings of Pinus thunbergii heated in volcanic ash in an electric furnace were carried to make the regression analyses between heating temperature and organic maturation indices, namely H/C atomic ratio, reflectance and IR absorbance of heated wood cuttings. The statistical equations based on these analyses are T=276×(H/C)_-0.66 for H/C atomic ratio, T=201×(Ro)+264 for reflectance (Ro) and T=-58・ln(R_A)+245 for IR absorption, where T is heating temperature (℃) and R_A is a ratio of (CH2+CH3)/[(CH2+CH3)+(C=C)] in IR absorption, respectively. The emplacement temperatures of the Pleistocene Towada-Ofudo and Towada-Hachinohe pyroclastic flow deposits were estimated from the data of H/C atomic ratios, reflectance and IR absorbance of 27 samples of natural carbonized wood fragments involved in these pyroclastic flow deposits at four localities by means of these three equations, respectively. The correlation coefficients of all data of estimated emplacement temperatures among these three thermometers is 0.73 to 0.85. The differences of estimated emplacement temperatures in each sample locality are less than 72℃ regardless of kinds of thermometers. These differences should be geologically negligible because the difference of estimated temperatures among eight samples from a carbonized trunk in the pyoclastic flow deposit was 69℃ at maximum (Sawada et al. 2000), and the estimated temperature range based on paleomagnetism of lithic fragments involved in a pyroclastic deposit is generally more than 100℃. The application of these three thermometers based on organic maturation of carbonized woods to pyroclastic deposits is more simple and reliable than that of paleomagetic thermometer of lithic fragments, and the reliance of temperature estimation can be easily tested by comparison of data of H/C atomic ratio, reflectance and IR absorbance of carbonized wood.

View full abstract