Link to Report: Coming Soon
Background :
Temporary bridges are critical lifelines built to ensure continuity of service during major renovation projects of ordinary bridges or following natural disaster emergencies. Differently from ordinary bridges, which are expected to be in service for 75 years, these structures have a service life of 5 years. In a time in which investments in existing bridge maintenance and repair are expected to increase by 58%, from $14.4 billion annually to $22.7 billion annually (ASCE, 2021), it is essential to plan investments on a risk-informed basis. Establishing a methodology to conduct performance-based and cost-effective designs of systems with a short service life is fundamental to properly inform the management of large assets, where overdesigns at a large scale would lead to uneconomical solutions. Nevertheless, to date, a nationwide consensus on the most appropriate hazard level to adopt nationwide for the seismic design of temporary structures is yet to be established.
This project will build upon previous research of the PI supported by the California Department of Transportation (Petrone et al., 2025; Kashizadeh et al., 2025a; Kashizadeh et al., 2025b), which provided recommendations for the design of temporary bridges employing light superstructure in California. This research will substantially broaden the scope, by carrying out suites of risk analyses on a wide range of bridge typologies employed by the Departments of Transportation across the nation, for different site conditions, and levels of seismicity. Collectively, this effort will offer a robust performance-based and risk-informed foundation for updating current design provisions for temporary bridges, an often overlooked yet critical component of resilient transportation networks.
In a broader sense, the methodologies developed through this project will go beyond the design of temporary bridges and be applicable to other short-service-life infrastructure systems, expanding relevance and applicability of this research.
Objectives :
This research will conduct comprehensive probabilistic risk evaluations considering multiple return periods and service life scenarios for temporary bridges. The work will be based on numerical and analytical studies involving the execution of nonlinear time history analyses, the derivation of bridge response statistics, and the use of probability concept to obtain structural fragilities and seismic risk estimates.
The specific objectives of this research include: the creation of a database of numerical nonlinear models of temporary bridges validated against experimental data available in the literature and input ground motions corresponding to a range of selected hazard levels; the generation of bridge response statistics and fragility functions for a set of pre-defined simplified damage states specific to temporary structures; and execution of probabilistic seismic risk analyses for a set of nominal service lives. Altogether, these objectives will provide the U.S. Departments of Transportation with an objective and robust foundation to make an informed decision on the most appropriate hazard level to use for the design of temporary bridges.
Scope :
Task 1 – Selection of temporary bridge typologies & Design.
The objective of this task is to select a set of temporary bridges representative of the inventory of typologies used across different departments of transportation in the nation. A literature review and direct communication with Bridge Engineers will inform this task.
It is envisioned that different types of superstructure will be considered, including lightweight (expected unit weight of 3.8 kip/ft), mid-weight (expected unit weight 6.7 kip/ft), and ordinary-weight deck systems (expected unit weight 8.2 kip/ft). Since the type of superstructure has implications on the type of connection between the superstructure and the bent, appropriate modeling choice will be made to reflect realistic boundary conditions as detailed in the next section. Additionally, different constructive solution employed for the foundation for different types of soil (soft, dense, and stiff) will be considered. Likewise, these will reflect in different modeling assumptions detailed afterward.
Moreover, bridges with a varying number of spans and bent typologies will be considered. encompassing single-column, and multiple-column bents.
Seismic design to determine size and reinforcement in the reinforced concrete columns will be performed by applying the target acceleration to the bridge in both transverse and longitudinal directions using the complete quadratic combination (CQC3). Upon application of all relevant load combinations in the AASHTO (2020), the load combination controlling the seismic demands will be used to finalize the design.
Task 2 – Numerical modeling and validation.
The objective of this task is to create efficient and accurate nonlinear models of the selected temporary bridges and validate their nonlinear response against experimental data. To effectively simulate all temporary bridge typologies selected in Task 1, a reduced-order parametric numerical model of the type represented in Figure 1 will be developed in OpenSees (Zhu et al. 2018). It is a three-dimensional nonlinear model where the columns employ force-based elements and fiber sections, and the deck is modeled with a linear elastic element with inertial properties consistent with the target deck system type. The cyclic response of plain concrete will be modeled using Concrete02 material in Open- Sees (Zhu et al. 2018), which is based on the model by Yassin (1994) and encompasses a nonlinear curve in compression up to the peak strength followed by linear softening, and linear elastic behavior in tension up to cracking followed by linear softening. The properties of the confined core will be derived from the confinement model proposed by Mander et al. (1984). The response of the rebars will be simulated using the Hysteretic material, which defines a post-yielding softening response beyond the peak stress in tension and compression through three-point piecewise functions. The input parameters for capturing the buckling in compression will be defined based on the model proposed by Zong et al. (2014), which utilizes a beam-on-springs model Kashizadeh et al. (2025).
The abutments will be modeled with a system of springs to reproduce the nonlinear response of shear keys, the abutment backfill, bearing pads, and any potential pounding between backwall and deck. The connection between the bent cap or columns and the deck will use a combination of rigid links and springs to simulate the response of the bearing pads. Finally, the foundation will reproduce rotational and translational stiffness as well as the foundation depth with a combination of springs. These models’ features are represented in Figure 1. Since the key nonlinear elements in the seismic response of bridges are the bents, the nonlinear response of the columns and bents (when available) will be validated against experimental data available in the literature. The key response parameters that will be considered for validation purposes are the peak strength and lateral displacement, the initial stiffness and stiffness degradation, the hysteresis energy dissipation, and buckling. Figure 2 reports an example of a validation exercise for the response of a typical reinforced concrete column. It is envisioned that a similar validation exercise will be carried out across at least six columns or bents using the DesignSafe Data Depot and the PEER Structural Performance Database
Figure 1. Bridge parametric model (Rahmani and Petrone, 2025)
Figure 2. Numerical model validation. Comparison between simulated and experimental responses of selected columns from the literature: (a) Lehman and Moehle (1998); (b) Henry and Mahin (1999). Adapted from Kashizadeh et al. (2025).
Task 3 – Selection of hazard levels & input ground motions
The objective of this task is to select a suite of hazard levels, generate the corresponding design spectra, and select the sets of ground motions to use as an input for the bridge structures.
Design spectra will be derived for a range of hazard levels (or return periods) and across different seismic locations in the United States (e.g. West Coast). The USGS seismic hazard maps will be used to this purpose. Current design for temporary bridges is based on 100 years return period, corresponding to a probability of exceedance of 5% in 5 years. Previous research from the PI has demonstrated that a performance-based seismic design carried out for return periods shorter than 100 years is controlled by the AASHTO minimum design requirements when employing light weight superstructures (Petrone et al., 2025). These cases will not be investigated in this study. It is envisioned that (at least) the following return periods will be considered: 200, 500, 975 years.
The spectra so derived will be used to select and scale pairs of ground motions. The initial selection will be informed by the results of seismic hazard disaggregation at the sites of interest, while the final selection will be carried out based on the minimum error in approaching the target spectrum in the bandwidth of interest.
The pairs of motions so obtained and rotated at least twice by 45 degrees will be simultaneously applied to the nonlinear model of the bridge to perform nonlinear time-history analyses.
Task 4 – Generation of fragility function and risk analyses
The objective of this task is to conduct fully nonlinear time-history analyses of the bridge modeled in Task 2 with the ground motions selected in Task 3, generate fragility functions, and finally assess seismic risk.
The pairs of ground motions selected in Task 3 will be applied to the bridge models developed in Task 2 with a uniform excitation loading scheme. The peak lateral displacement of the columns will be selected as the reference engineering demand parameter to assess structural responses. Once all the bridge responses are obtained, they will be used to generate bridge response statistics and fragilities. The structural response assessment will be carried out with reference to an updated and simplified set of damage states, starting from those proposed by Vosooghi and Saiidi (2012). Moreover, strain and stress criteria will be defined to precisely identify the attainment of each damage state in the numerical model. As an example, Figure 3 shows the bridge response statistics for two locations (San Francisco and Los Angeles) and three hazard levels (1, 2 and 3, corresponding to 50, 100 and 200-year return period). Statistics of this type will be subsequently used to create probabilistic demand models for the temporary bridges and generate structural fragilities. Figure 4 provides a representation of the fragility functions for the same case studies of Figure 3.
Figure 3. Bridge response statistics for the hazard-consistent bridges in (a and b) San Francisco and Los Angeles. Adapted from Kashizadeh et al. (2025).
Figure 4. Fragility functions for San Francisco and Los Angeles, and hazard levels 1, 2 and 3 corresponding to 50, 100 and 200-year return period. Adapted from Kashizadeh et al. (2025).
In addition, a suite of probabilistic risk analyses will quantify the expected damage under ‘beyond design’ earthquake scenarios, thus providing a comprehensive assessment of the overall bridge performance.
Finally, the probability of exceeding each defined damage state will be coupled with the seismic hazard at each of the considered sites to calculate seismic risk for different values of the service life, including 5, 10 and 15 years and conduct loss assessments. To this aim, the method proposed by Yoon et al. (2022) will be used.
Task 5 – Development of recommendations for seismic design and detailing
Collectively, Tasks 1 to 4 will provide the basis to perform an objective comparison across different bridges typologies and designs, corresponding to different hazard levels for sites representative of regions with diverse seismicity levels.
These results will be used to develop recommendations for the seismic design of temporary structures. These will include both the target hazard level and seismic detailing that differs from what is currently mandated by AASHTO for ordinary bridges. These outcomes will be collected in a summary report.
Research Team :
Principal Investigator: Dr. Floriana Petrone