Electronic Theses and Dissertation

Pada 7 desember 2016, gempa bumi mengguncang pidie jaya dengan magnitudo 6,5 mw di kedalaman 15 km mengakibatkan sejumlah infrastruktur dan rumah tinggal mengalami kerusakan. selain itu, sebanyak ratusan orang menjadi korban jiwa. dalam upaya mengurangi risiko seismik tersebut, diperlukan penilaian risiko seismik terhadap rumah tinggal dengan melakukan prediksi kerusakan bangunan rumah melalui kurva kerapuhan. kurva kerapuhan telah diperkenalkan oleh para peneliti sebagai salah satu alat dalam menilai kerusakan dan kerugian selama gempa bumi. tujuan dari penelitian ini adalah mengembangkan kurva kerapuhan empiris dengan model probit, logit, dan complementary loglog sebagai fungsi generalized linear model (glm) untuk struktur kayu, dinding bata terkekang, dan beton bertulang akibat intensitas gempa di pidie jaya, serta pemilihan model kurva kerapuhan terbaik berdasarkan evaluasi ukuran kesesuaian. metode pengembangan kurva kerapuhan dengan mengumpulankan data kerusakan rumah tinggal berdasarkan tipe struktur dan jumlah rumah tinggal di pidie jaya dengan simulasi peak ground acceleration (pga) menggunakan openquake. probabilitas kerusakan rumah ditinjau berdasarkan tiga tingkat kerusakan (damage state) yaitu kerusakan ringan, sedang, dan berat. hasil penelitian menunjukkan pada intensitas gempa rendah, struktur dinding bata terkekang di pidie jaya memiliki probabilitas kerusakan ringan, sedang, dan berat lebih tinggi dibandingkan struktur kayu dan beton bertulang. sedangkan rumah struktur beton bertulang memiliki kerapuhan lebih besar di intensitas gempa tinggi. sementara itu, model logit memberikan probabilitas lebih besar di pga yang tinggi dan model probit merupakan model terbaik berdasarkan ukuran kesesuaian.

On December 7, 2016, a moderate earthquake in Pidie Jaya with a magnitude of 6.5 Mw at a depth of 15 km caused damage to a number of infrastructure and houses. In addition, hundreds of people died. In an effort to reduce the seismic risk, it is necessary to assess the seismic risk of residential houses by predicting the damage to houses using the fragility curve. The fragility curve has been introduced by researchers serve as one of main tools in assessing damage and loss during earthquakes. The purpose of this study is to develop empirical fragility curves using probit, logit, and complementary loglog models as a function of the Generalized Linear Model (GLM) for wooden structures, confined masonry walls, and reinforced concrete due to earthquake intensity in Pidie Jaya and the selection of the best fragility curve model based on goodness of fit evaluation. Development of a fragility curve by collecting data on residential damage based on the type of structure and number of houses in Pidie Jaya with Peak Ground Acceleration (PGA) simulation using OpenQuake. The probability of house damage is reviewed based on three levels of damage which include slight damage, moderate damage, and heavy damage. Based on the results of the study, it was shown that at low earthquake intensity, confined brick wall structures in Pidie Jaya have a higher probability of slight, moderate, and heavy damage than wooden and reinforced concrete structures. Meanwhile, reinforced concrete houses have greater fragility at high earthquake intensity. In addition, the logit model gives a greater probability in high PGA and the probit model is the best model based on goodness of fit measure.