BridgeFuncRecovery 0.1.1

A Python package to model post-earthquake functional recovery of bridges.

BridgeFuncRecovery

This project aims to probabilistically model the post-earthquake functional recovery of bridges. The source codes are programmed in Python.

Reference

[1] Wu, C., Burton, H., Zsarnóczay, A., Chen. S., Xie. Y., Terzić, V., Günay, S., Padgett, J., Mieler, and M., Almufti, I. (2025). Modeling Post-earthquake Functional Recovery of Bridges. Earthquake Spectra.

Prerequisites

Python: version 3.6 or above.

Necessary Python packages: copy, numpy, pandas, os, scipy, sys, shutil, pathlib, re, time, pickle

File Overview

main.py takes user-specified parameters and performs the full analysis.

utils.py provides auxiliary functions that are called from the main script main.py.

result_anlaysis.py helps visualize the output data stored in Results.pkl.

After running main.py or calling run() programmatically, a Results.pkl file is created that stores all model output data

User-specified inputs for run()

The run() function requires the following inputs:

Required Arguments:

• IM_fixed (float): Evaluated intensity measure level.
• num_span (int): Number of bridge spans.
• CompQty (dict): A dictionary of quantities of bridge components. An example is shown below:

"CompQty": {  # dict[str, int]
    'Col':3, 'Seat_ab':2, 'Super':1, 'ColFnd':2, 'AbFnd':2, 'Backwall':2, 
    'Bearing_ab':2,'Key_ab':2, 'ApproSlab':2, 'JointSeal_ab':4,
    'Seat_super':1, 'Bearing_super':0, 'Key_super':0, 'JointSeal_super': 0
}

• WorkerAllo_percrew (dict): A dictionary listing allocated workers that constitute a worker crew to perform repairs for each bridge component. An example is shown below:

"WorkerAllo_percrew": {  # dict[str, int]
    'Col': 3, 'Seat_ab': 4, 'Super': 5, 'ColFnd': 4, 'AbFnd': 4, 'Backwall': 4, 
    'Bearing_ab': 3, 'Key_ab': 3, 'ApproSlab': 5, 'JointSeal_ab': 5,
    'Seat_super':4, 'Bearing_super':2, 'Key_super':4, 'JointSeal_super': 4
}

• Worker_Replace (int): Number of workers to perform bridge replacement.

Optional Arguments:

• num_rlz (int, default = 1000): Number of Monte Carlo realizations.
• w (list of float, default = [0, 0, 1]): A weighing scheme specifying how correlations are considered in bridge component damage sampling. The first entry is correlation shared for all components; the second entry is shared for components in a smaller family, and the third entry is independent portion. [0,0,1] means all components’ damage are sampled independently. [1,0,0] means all damages are perfectly correlated. For more details, see “A model for partially dependent component damage fragilities in seismic risk analysis” (Baker et al. 2023).
• height (float, default = 35): Bridge height.
• num_lanes_before (int, default = 4): Number of available bridge lanes under normal operations.
• ColSuperMatType_dict (dict, default = {'Col': 'concrete', 'Super': 'concrete'}): Bridge materials used in columns and superstructures. Possible entries: ‘Concrete’ or ‘Steel’.
• NumCrew_percomp (dict): A dictionary specifying number of worker crews per component. Default shown below:

"NumCrew_percomp": {  # dict[str, int]
    'Col': 1, 'Seat_ab': 1, 'Super': 1, 'ColFnd': 1, 'AbFnd': 1, 'Backwall': 1,
    'Bearing_ab': 1, 'Key_ab': 1, 'ApproSlab': 1, 'JointSeal_ab': 1,
    'Seat_super': 1, 'Bearing_super': 1, 'Key_super': 1, 'JointSeal_super': 1
}

• WorkHour_repairable (int, default = 8): Worker-hour per day used to repair bridge components.
• WorkHour_replacement (int, default = 8): Worker-hour per day used to replace a bridge.
• num_concrete_pour_replacement (int, default = 1): How many pours of concrete are considered when replacing a bridge.
• dispersion_assigned (float, default = 0.3): Dispersion considered in repair duration sampling.

Example call:

from BridgeFuncRecovery import run

results = run(
    IM_fixed=0.23,
    num_span=2,
    CompQty={'Col':3, 'Seat_ab':2, ...},
    WorkerAllo_percrew={'Col':3, 'Seat_ab':4, ...},
    Worker_Replace=30
)

This returns a results dictionary and saves it to Results.pkl

Analyzing the Results

The functions in result_analysis.py help interpret and visualize output data from the main analysis. Available functions include:

• plot_fs_initial(data)
• plot_fs_reopening(data)
• plot_total_impeding_ccdf(data)
• plot_total_repair_ccdf(data)
• print_impeding_medians(data)
• print_repair_durations(data)
• plot_repair_class_distribution_single(data, comp_name)
• plot_all_repair_class_distributions(data)
• plot_closed_lane_initial(data)
• show_all_results(data)

Example usage:

from BridgeFuncRecovery import run, plot_repair_class_distribution_single

# Run the analysis
results = run(IM_fixed=0.23, num_span=2, CompQty={...}, WorkerAllo_percrew={...}, Worker_Replace=30)

# Plot the Repair Class distribution for columns
plot_repair_class_distribution_single(results, 'Col')

This saves results to Results.pkl and displays a figure visualizing Repair Class (RC) distribution for columns.

Full Example

Below is a full working example with realistic parameter values:

input_dictionary = {
    # Required inputs:
    "IM_fixed": 0.23,  # float
    "num_span": 2,  # int
    "CompQty": {  # dict[str, int]
        'Col':3, 'Seat_ab':2, 'Super':1, 'ColFnd':2, 'AbFnd':2, 'Backwall':2, 
        'Bearing_ab':2,'Key_ab':2, 'ApproSlab':2, 'JointSeal_ab':4,
        'Seat_super':1, 'Bearing_super':0, 'Key_super':0, 'JointSeal_super': 0
    },
    "WorkerAllo_percrew": {  # dict[str, int]
        'Col': 3, 'Seat_ab': 4, 'Super': 5, 'ColFnd': 4, 'AbFnd': 4, 'Backwall': 4, 
        'Bearing_ab': 3, 'Key_ab': 3, 'ApproSlab': 5, 'JointSeal_ab': 5,
        'Seat_super':4, 'Bearing_super':2, 'Key_super':4, 'JointSeal_super': 4
    },
    "Worker_Replace": 30,  # int
    
    # Optional inputs:
    "num_rlz": 100,  # int
    "w": [0, 0, 1],  # list of ints
    "height": 35,  # int, in ft
    "num_lanes_before": 4,  # int
    "ColSuperMatType_dict": {  # dict[str, str]
        'Col': 'steel', 'Super': 'concrete'
    },
    "NumCrew_percomp": {  # dict[str, int]
        'Col': 1, 'Seat_ab': 1, 'Super': 1, 'ColFnd': 1, 'AbFnd': 1, 'Backwall': 1,
        'Bearing_ab': 2, 'Key_ab': 2, 'ApproSlab': 1, 'JointSeal_ab': 1,
        'Seat_super':1, 'Bearing_super':2, 'Key_super':1, 'JointSeal_super': 1
    },
    "WorkHour_repairable": 10,  # int
    "WorkHour_replacement": 10,  # int
    "num_concrete_pour_replacement": 1,  # int
    "dispersion_assigned" = 0.3 # float
}

from BridgeFuncRecovery import run, show_all_results
# Run analysis
results = run(**input_dictionary, return_data=True)

# Show all the results
show_all_results(results)