Predicting and Reducing Liquefaction-Induced Damage to Houses : The Nihonkai-Chubu Earthquake

A comparison of two practical analysis methods for predicting liquefaction with the recorded liquefaction that occurred during the 1983 Nihonkai-Chubu Earthquake was made. Based on findings of surveys, that included bore and standard penetration tests, grain size analyses, and interviews with residents, comparative calculations were made using two practical methods for predicting liquefaction : the procedure given in the Japan Hrghway Bridge Code and the technique which uses residual pore waterpressure data. The comparison of survey findings with those of the analysis showed a coincidence rate of 23% for the procedure of the Japan Highway Bridge Code and 31% for the method using pore water pressure data at k_<so> = 0.15,the value usually used at present in the two practical methods. When, however, k_<so> = 0.22,the ratio of the maximum acceleration to gravity acceleration, obtained at ground surface during the 1983 Nihonkai-Chubu Earthquake was used, the results of the comparison were 81% for the former procedure and 80% for the latter method. We conclude that the occurrence of liquefaction can be accurately estimated by both these methods, if there is close agreement between the value of k_<so> and the ratio of maximum to gravity acceleration at the ground surface for a future massive earthquake. A practical method for predicting and reducing liquefaction-induced damage to houses also is proposed. It is derived from the relationsip between the thickness of the non-liquefied layer H_1 and that of the liquefied layer H_2 obtained from survey findings, : bore and Japanese standard penetration tests, grain size analyses, and interviews with residents. Which kinds of foundations that had been built before the earthquake, remained intact during the 1983 Nihonkai-Chubu Earthquake also are reported. On the basis of these findings, a practical method for reducing liquefaction-induced damage to houses is proposed.