PSI - Issue 54

Koji Uenishi et al. / Procedia Structural Integrity 54 (2024) 67–74 Uenishi / Structural Integrity Procedia 00 (2023) 000–000

72

6

a

b

Frequency [Hz]

1 0.5

0.1

100 50

10 5

300

1000

NS

0

Strong TD detected

NS

]

-300

2

100

300

EW

Still

EW

0

doubtful

10

TD

-300

Velocity [cm/s]

300

1

Acceleration [cm/s

TD

0

-300

0.1

0.01

0.1

1

10

0 102030405060

Time [s]

Period [s]

Fig. 4. (a) Seismological records at the site JMA Uki (latitude 32.6476  N, longitude 130.6842  E) for the Kumamoto earthquake (main shock)

that occurred at 1:25 am JST on April 16, 2016, and (b) the associated Fourier spectra (From the online data provided by JMA,

https://www.data.jma.go.jp/svd/eqev/data/kyoshin/jishin/). This earthquake’s epicentral latitude and longitude are 32.7545  N and 130.7630  E,

with a focal depth of 12 km and M j = 7.3. The maximum accelerations in the NS, EW and UD directions at JMA Uki with the epicentral distance

of 14.2 km are 492.8, 342.6 and 313.9 cm/s 2

. The upgraded seismographs installed after the 1995 quake in Kobe have recorded the governing

vertical shaking in a higher frequency range (modified after Uenishi (2021)).

inhomogeneous inner structure of the Earth at larger scales such as the crust, mantle, and outer and inner cores that

are of the order of tens to hundreds of kilometers. For this purpose, seismographs have been designed to be more

sensitive to waves of longer wavelengths, i.e. lower frequencies under 1 Hz that can globally travel from

hypocenters to seismological stations worldwide over thousands of kilometers and be utilized to search such larger

scale inhomogeneities. Higher-frequency waves, usually contaminated due to scattering, are filtered out of the

seismological records and, as mentioned, excluded from the seismic investigation with the belief that only horizontal

vibrations in a lower frequency range can have some influence on our surroundings. The sensitivity of the

seismographs commonly installed at the time of the 1995 quake in Kobe dropped considerably above 20 Hz and

vibrations over 10 Hz were filtered out in the seismic investigation, and regrettably, strong higher-frequency shaking

was hardly recorded (Fig. 2(b)). Therefore, the vertical shock could not be detected, and people except for those who

experienced the quake and shock by themselves could not (and still cannot) believe the existence and the effect of

the vertical shock (Uenishi, 2021).

Thus, the dynamic structural failures in Kobe were almost disregarded, and there seemed almost no chance to

prove the existence of the vertical shock. However, there was an enlightening series of seismic events in 2016 in

j = 6.5) and the main

Kumamoto, Japan, where two strong earthquakes, the foreshock (focal depth of 11 km and M

j = 7.3; Fig. 4) occurred at 9:26 pm JST on April 14 and at 1:25 am JST on April

shock (focal depth of 12 km and M

16, respectively. Although in a lower frequency range horizontal shaking dominated as usual also for both

Kumamoto quakes, the seismological records near the epicenters clearly indicate the existence of governing vertical

shaking in a higher frequency range, for instance, over 5 Hz according to the seismograms recorded by the upgraded

seismographs in Uki City (Fig. 4(b)). That is, for the earthquakes similar to the 1995 one in Kobe in terms of the

magnitude, focal depth and focal mechanism, the possibly significant higher-frequency vertical shaking could be

undoubtedly detected and recorded with seismographs in a way more akin to our experience near epicenters. Of

course, the precise geological condition in Uki may be unlike that in Kobe, but the governing higher-frequency

vertical shaking must not be overlooked. However, regrettably, in contrast to the sensors for detecting vibrations due

to blasting that are insensitive to lower frequencies but sensitive to higher ones up to 1,000 Hz, even upgraded

seismographs are not sufficiently sensitive to seismic waves of audible 17 or 100 Hz estimated from the Daikai and

Bantaki failures. Anyhow, even when such very high frequencies or real shocks have not been detected by