Journal of Japan Society of Civil Engineers, Ser. A1 (Structural Engineering & Earthquake Engineering (SE/EE))
Online ISSN : 2185-4653
ISSN-L : 2185-4653
JSCE Journal of Earthquake Engineering, Vol.39 (Paper)
A RESTORATION PREDICTION MODEL FOR THE DAMAGED ROAD EMBANKMENTS BY STATISTICAL ANALYSIS OF MULTIPLE EARTHQUAKE DISASTER RECORDS
Atsushi NOMOTOHisakazu SAKAIYoshio KAJITANI
Author information
JOURNAL FREE ACCESS

2020 Volume 76 Issue 4 Pages I_409-I_419

Details
Abstract

 A large number of earthquake-induced damages to road embankments have been reported in Japan. Because a road embankment is a linear structure, the traffic functionality can be lost even if just one damage occurred in a link. For example, an embankment around the Masiki bus stop in the Kyusyu expressway had collapsed during the 2016 Kumamoto earthquake and the traffic regulation had been applied for 13 days. An accurate prediction for repair period of damaged road embankment would inform us reinforcement and recovery strategies of embankment ; therefore, it should mitigate a secondary disaster for human living.

 In this study, we investigated the regulation period of roads as well as the damages to road embankment caused by seven earthquakes which induced severe damages to roads in Japan. We compared with the prediction accuracy of multiple statistical methods adopting five explanatory variables: embankment height, damage range, topography, damage form, and traffic volume. As a result, it is found that extended quantification theory I with box-cox transformation perform best for predicting the recovery period of the damaged road embankments.

Content from these authors
© 2020 by Japan Society of Civil Engineers
Previous article Next article
feedback
Top