Generation of strong earthquake accelerograms using the particle filter with richter magnitude, source depth & epicentre distance as model parameters

• Main Author: Taclibon, Patrick B.
• Format: text
• Language: English
• Published: Animo Repository 2013
• Online Access: https://animorepository.dlsu.edu.ph/etd_masteral/4577
• Institution: De La Salle University

Strong earthquakes are arguably the most destructive natural phenomena that structures are destined to face at some point in their existence. Current state-of-the-practice uses the equivalent static base shear procedure based on a 10% probability of exceedance in 50 years as well as response spectrum analysis to safeguard structures against earthquakes. However, due to the increasing use of advanced materials and complex geometry to satisfy aesthetics for both horizontal and vertical structures, interest in using earthquake time histories, either real or simulated, has risen. In countries where lack of earthquake records is present, this type of earthquake effect estimation would be hindered. This paper proposes a framework to simulate earthquake time histories by defining a magnitude, epicenter distance and source depth for which the structure will stand through analyzing earthquake records. For this purpose, Japans earthquake records were chosen as the sample data chiefly due to its accessibility and remarkable documentation. Generation of earthquake time histories is achieved through the use of black box earthquake modeling using Time varying autoregressive moving average model and the particle filter to estimate its parameters. Furthermore, a regression analysis is done to correlate the parameters of the model with the aforementioned earthquake descriptors in order to arrive at a physically significant simulated earthquake time history. The resulting prototype earthquake model was able to simulate quite well, in the time domain, the real earthquake time histories for Richter magnitudes between and including 6.0 and 7.1 with epicenter distance and depth of less than 89 kms. However, there still remains the issue of filtering the simulated signal to mimic the frequency content of the real earthquake especially at very low frequencies.