Associate Professor University of British Columbia
Datasets Supporting the 2026 NEHRP Functional Recovery Design Provisions
Co-Author: Kristen Blowes (University of Colorado Boulder)
Abstract: This presentation introduces a comprehensive dataset (PRJ-5893 on DesignSafe-CI) developed to support the 2026 NEHRP functional recovery design provisions. The dataset includes engineering demand parameters, component fragilities, safety-critical structural damage indicators, and recovery-time data produced in the development of prescriptive functional recovery design requirements for both structural and nonstructural systems. We provide an overview of the dataset structure, content, and curation approach, highlighting its value for calibrating and validating recovery-based computational models. This work is particularly timely given the forthcoming integration of the ATC-138 functional recovery framework into the SimCenter’s Performance-Based Engineering (PBE) tool. By providing relationships between damage to structural and nonstructural components and observed recovery outcomes, this dataset offers a critical benchmark for evaluating, refining, and ultimately operationalizing ATC-138 within the PBE workflow.
Research Structural Engineer National Institute of Standards and Technology
NED: A Nonstructural Element Database for more robust and transparent performance-based seismic deign and functional recovery analysis
Co-Authors: Siamak Sattar (NIST) and Adam Zsarnoczay (Stanford University)
Abstract: Observations since the 2011 Christchurch, New Zealand earthquake, and subsequent earthquakes around the globe have revealed two important findings that are shaping structural engineering research and design: (1) there is a pressing need for functional recovery design solutions to support long-term community resilience, and (2) damage to nonstructural components is often the leading cause of loss of building function after earthquakes. However, for practitioners and researchers utilizing the FEMA P-58 methodology for building performance and recovery assessments, the available nonstructural fragility data is often outdated, static, and often difficult to verify. To bridge this gap, the U.S. National Institute of Standards and Technology (NIST) developed the Nonstructural Element Database (NED), a public repository that fundamentally restructures and extends the existing nonstructural fragility data landscape.
Moving beyond static spreadsheets, NED employs a relational SQL architecture designed specifically to address the scalability limits of legacy databases. Through an extensive literature review, we have curated over 2,000 experimental test records, explicitly mapping raw experimental data to specific seismic fragility functions. This structure allows for greater model transparency and ensures the database is extensible. As new testing or nonstructural performance observations are collected, they can be seamlessly uploaded and integrated into NED to refine or generate new fragility functions. By providing this centralized, open-source framework, NED improves modeling confidence of nonstructural damage outcomes generated in performance-based assessments and provides the technical infrastructure necessary to evolve FEMA P-58 for future resilience-based design.
Professor Auburn University
Leveraging WE-UQ for Generating Wind Fragility Functions for Performance-Based Engineering of Low-Rise Buildings
Co-Authors: Prethesha Alagusundaramoorthy (Auburn University) and Adam Zsarnoczay (Stanford University)
Abstract: Fragility functions are a core element of Performance-Based Engineering for all hazards. Methods for generating seismic fragility functions are well-established, and extensive libraries exist. For wind hazards however, fragility libraries for common building types and components are growing, but are still relatively incomplete, and further, there is often a lack of clarity and standardization in how the fragility functions were generated. This presentation will explore utilizing the WE-UQ tool and other SimCenter tools within a workflow for generating new fragility functions. A comprehensive framework will be presented, and the implications of various simplifications to the approach that are commonly used will be demonstrated.
PhD Student The University of British Columbia
Sensitivity of Safety-Critical Structural Damage Estimates to Modeling Assumptions in Recovery-based Seismic Design
Co-Authors: Huy Pham (The University of British Columbia) and Carlos Molina Hutt (The University of British Columbia)
Abstract: Recent developments in recovery-based seismic design have defined and studied safety-critical structural damage, i.e., damage that must be repaired before safe reoccupancy of a building, as a performance metric to design structural systems for functional recovery performance. Estimating this metric requires mapping component-level damage states to safety classes, which are then evaluated through fault-tree analyses to determine the likelihood of safety-critical damage at the story, system, and directional levels based on the distribution and severity of damage throughout the structure. This study presents a framework to systematically examine how different modeling assumptions influence these results. The computational workflow leverages the SimCenter Pelicun/PBE tool to perform the FEMA P-58 component damage and loss assessment, coupled with ATC-138 fault trees to estimate the probability of safety-critical structural damage. Sensitivity of outcomes due to variations in (1) damage state to safety class mapping, (2) assumptions about damage correlation, (3) component counting rules, and (4) fault-tree thresholds are evaluated within this computational framework. Results are demonstrated using an archetype eight-story reinforced concrete moment frame. The findings provide insight into how modeling choices within commonly used performance-based earthquake engineering tools influence estimates of safety-critical structural damage and identify key factors that contribute most to this outcome.
PhD Student Univeristy of Michigan, Ann Arbor
A Computational Framework for Two-Way Coupling Between Progressive Building Envelope Damage and Building Aerodynamic Loads
Co-Author: Seymour MJ Spence (University of Michigan)
Abstract: Building performance during extreme winds is governed not only by the structural system but critically by the envelope system, as severe envelope damage can compromise structural integrity and disrupt the immediate functionality of buildings. Existing assessments of building envelope performance under extreme wind events often assume that while external pressures can damage envelope components, subsequent aerodynamic loads remain unaffected. In reality, envelope damage is closely linked to local peak pressures, and progressive loss of envelopes can redistribute loads onto remaining components, altering both external and internal wind pressure fields. To address this limitation, this work develops a computational framework that couples large-eddy simulation with a building envelope fragility model, capturing two-way interactions between progressive envelope damage and evolving wind pressure fields. Dynamic openings and background leakage are explicitly modeled, with envelope components modelled by dynamic baffle elements, which are removed once the net pressures exceed the component-specific capacities sampled from the fragility functions. A two-story building is employed as a case study to demonstrate the framework. The proposed framework advances performance-based wind engineering by enabling more accurate predictions of the wind pressure field associated with progressive envelope failure, and identifying conditions where two-way coupling between envelope failure and aerodynamic loading becomes essential.
Postdoctoral Scholar University of British Columbia
Recovery Time and Seismic Performance of Life-Safety and Functional Recovery - Designed Buildings across Hazard Intensity Levels
Co-Authors: Ziyi Wang (University of British Columbia) and Carlos Molina Hutt (University of British Columbia)
Abstract: The 2026 National Earthquake Hazards Reduction Program (NEHRP) Recommended Provisions introduce prescriptive design requirements aimed at achieving functional recovery performance in buildings following earthquakes. These requirements supplement conventional life-safety design by introducing requirements imposed at the risk-targeted Functional Recovery Earthquake (FRER), alongside conventional checks at the Design Earthquake (DE). This study investigates the seismic performance of an 8-story reinforced concrete moment frame building designed to meet life-safety requirements alone with that of an otherwise identical building designed to satisfy both life-safety and functional recovery criteria. Structural design and nonlinear response analysis are performed using the HBRisk Structural Response Prediction Engine for earthquake intensities with return periods ranging from 14 to 4975-years. Probabilistic loss estimation is conducted using the SimCenter Pelicun tool, which translates engineering demand parameters into component damage states based on the FEMA P-58 methodology. Recovery time estimates are calculated using the ATC-138 methodology, which is planned for integration into future versions of the SimCenter PBE tool. By presenting the performance across intensity levels, this study demonstrates how the proposed functional recovery provisions perform at the FRER level and benchmarks its performance at design-level hazards.
PhD Student Stanford University
Functional recovery of mass timber buildings with post-tensioned rocking walls
Co-Author: Barbara Simpson (Stanford University)
Abstract: Post-tensioned mass timber rocking wall lateral force-resisting systems are expected to reduce time to re-occupancy and probability of demolition after seismic events by mitigating the formation of story mechanisms and re-centering after extreme shaking. To characterize the seismic resilience and of mass timber buildings with post-tensioned rocking walls relative to traditional, ductile structural systems, a comparative performance-based earthquake engineering (PBEE) assessment was conducted for [i] a mass timber building with post-tensioned rocking walls employing buckling-restrained boundary elements (BRBs) as energy dissipators, and [ii] a six-story steel framed structure with buckling-restrained braced frames (BRBF). Six-story archetype commercial buildings were designed for a hypothetical site in Seattle, WA, and nonlinear numerical finite element models were defined using OpenSees to support the analysis. The PBEE assessment was conducted using the SimCenter’s PBE tool and Pelicun engine to characterize expected damage and expected downtime for each archetype building. Results were disaggregated to identify how contributions to the total expected loss and downtime may vary between systems, including contributions from collapse cases, non-collapse cases, damage to drift-sensitive non-structural components, and damage to acceleration-sensitive nonstructural components. The results of the study highlight potential trade-offs in seismic performance between a re-centering, elastic-spine lateral force-resisting system and a traditional, ductile lateral force-resisting system.
Postdoctoral Scholar UC Berkeley
Redefining Casualty Estimation Inside Buildings
Co-Author: Luis Ceferino (UC Berkeley)
Abstract: During earthquakes, casualties inside buildings result from a combination of structural and non-structural vulnerabilities, architectural layout, occupant demographics, and timing. Conventional seismic performance-based frameworks use empirical casualty rates linked to structural damage states, overlooking injuries caused by non-structural damage. While recent approaches have begun to focus on this gap, they currently assume uniform exposure of occupants to damage across the entire building floor. Such assumptions lead to inaccurate injury estimations and limit the analysis of harmful potential in non-structural components. As a result, such inaccurate estimates can lead to ineffective risk mitigation, including specific non-structural regulations and human response education. Building on existing performance-based frameworks, we propose a new approach for quantifying earthquake injuries at the room interior level. Injury consequences are treated as a probabilistic outcome estimated by modeling interactions between occupants and physical damage as independent variables. The model integrates occupancy modes into the existing performance-based frameworks and accounts for the spatial variability of non-structural hazards. To illustrate applicability, we used particular hourly occupancy profiles alongside California residential layouts to assess casualties from hazardous non-structural components (e.g., toppling bookcases, falling items). Our model suggests that casualty tolls are primarily driven by the presence of specific components, household composition, and activities, rather than floor area or just the time of day, as suggested by conventional HAZUS and FEMA P-58 methodologies. These findings underscore the importance of room-level exposure modeling, offering enhanced metrics for hospital surge estimation and occupant safety strategies in building design.