eprintid: 37782
rev_number: 34
userid: 9231
dir: disk0/00/03/77/82
datestamp: 2020-01-28 17:03:55
lastmod: 2020-01-28 17:03:55
status_changed: 2020-01-28 17:03:55
type: thesis_degree
metadata_visibility: show
contact_email: ath22@pitt.edu
eprint_status: archive
creators_name: Hirata, Amedeo
creators_name: Stephens, Max
creators_email: ath22@pitt.edu
creators_email: mts90@pitt.edu
creators_id: ath22
creators_id: mts90
title: The Model Builder: A Multi-Fidelity Structural Modeling Tool for Seismic Hazard Analysis
ispublished: unpub
divisions: sch_eng_civilenvironmental
full_text_status: public
keywords: Python, OpenSEES
abstract: Earthquakes affect millions of people across the earth.  Earthquakes cause billions of dollars of damage and hundreds of thousands of deaths per year.  For these reasons that seismic resilience is a field of immense concern for government, stakeholders and researchers.  As the field of earthquake engineering has advanced, higher levels of analysis have been developed, both in terms of fidelity and complexity, capable of capturing complex non-linear structural response.  However, regional assessment frameworks, which estimate structural performance at regional scale, do not commonly use building-level structural modeling to capture structural response. Many of these frameworks instead use a fragility-based approach where seismic intensity measures (such as peak ground acceleration) are related to probabilistic levels of damage.  The fragility functions used in this approach are based on the response of structural archetypes such as a steel moment frames or concrete shear walls. Thus, the response and damage estimations generated using these methodologies are not that of the actual structure in question, but rather that of an archetypical representation of the structure.  While this approach can roughly predict the dominant characteristics of vulnerable structures subjected to seismic hazards, it is difficult to pinpoint critical deficiencies for immediate retrofit or evacuation and inspection.  
The objective of this research is to develop a framework to provide building-specific nonlinear response predictions within regional structural simulation.  First, existing regional response prediction frameworks for natural disasters and structural analysis are introduced. Next, the opensource finite element analysis software OpenSeesPy is introduced along with the multi-fidelity Model Builder tool capable of rapidly developing building-level structural models using the software.  This software uses the OpenSEES framework to create the models, apply ground motions, and assess levels of damage of the structural system.  The Model Builder was then validated using several recorded structural response histories.  Based on the findings here, the Model Builder can be applied to develop models for regional response simulation.
date: 2020-01-28
date_type: published
pages: 125
institution: University of Pittsburgh
refereed: TRUE
etdcommittee_type: thesis_advisor
etdcommittee_type: committee_chair
etdcommittee_type: committee_chair
etdcommittee_name: Stephens, Max
etdcommittee_name: Khazanovich, Lev
etdcommittee_name: Oyler, John
etdcommittee_email: mts90@pitt.edu
etdcommittee_email: Lev.K@pitt.edu
etdcommittee_email: Oyler@pitt.edu
etdcommittee_id: mts90
etdcommittee_id: lek103
etdcommittee_id: oyler1
etd_defense_date: 2019-11-15
etd_approval_date: 2020-01-28
etd_submission_date: 2019-11-13
etd_release_date: 2020-01-28
etd_access_restriction: immediate
etd_patent_pending: FALSE
thesis_type: thesis
degree: MS
citation:   Hirata, Amedeo and Stephens, Max  (2020) The Model Builder: A Multi-Fidelity Structural Modeling Tool for Seismic Hazard Analysis.  Master's Thesis, University of Pittsburgh.    (Unpublished)  
document_url: http://d-scholarship-dev.library.pitt.edu/37782/1/Hirata_Amedeo_Thesis_Model_Builder.pdf