Structural Health Monitoring and Rapid Post-Event Assessment of Civil Infrastructures through

Digital Twins

Hamed Ebrahimian, Ph.D., P.E. Senior Engineer, SC Solutions, Inc.

S. Farid Ghahari, Ph.D.Assistant Project ScientistDept. of Civil and Environmental Engineering, UCLA

Ertugrul Taciroglu, Ph.D.Professor and ChairDept. of Civil and Environmental Engineering, UCLA

Why Bother?

2

Aging Infrastructures

ThreatsLimited

Resources

Retrofit &Upgrading

Visual Inspection

3

Post

-Ear

thq

uak

e In

spec

tio

n

• Costly

• Periodic

• Subjective

• No system-level insight

• Hidden damage

Mai

nte

nan

ce

Insp

ecti

on

• # of bridges × inspection time × chaos = ?

• Intensity-based metrics can be inaccurate!

• Inspection complexity

• Hidden damage

• No system-level insight

SHM - NDE

4

Point Monitoring

Load Cell

Optical Strain Gauge

Fiber Optics

Crack Sensor

Modal-Based Linear Damage Identification

NDE

Shortcomings?• Data not information• Limited damage detection, and

quantification capability• Require large number of sensors• Maintenance and installation

cost• No system-level insight

Shortcomings?• Not sensitive to local damage• Based on limiting assumptions

(broad band excitations, etc.) –inaccuracy and lack of robustness

Shortcomings?• Traffic interruption• Cost• Access limitations• No system-level insight

(Source: DOI 10.1115/JRC2014-3782)

Source: FHWA SHRP2 Solutions

Integration of Mechanics-Based Models with Data

5

Forward Simulation

• Model ParameterUncertainties

• Unknown Inputs

Model Updating / training

• Estimate Model Parameters

• Estimate Input Forces

Digital World Real World

Mechanics-based damage diagnosis at refined

spatial resolution

Digital Twins

6

Traffic DataTraffic induced Vibration

Earthquake Data(Output or input-output)

Real Twin Data Mechanics-based Model

Digital Twin

Bayesian Data Assimilation

Bayesian Data Assimilation

7

( )ΨfY =ˆ

Y y=

( )|p Ψ y

Ψ

( )( ) ( )

( )

p pp

p=

y Ψ ΨΨ y

y

−−ΨPΨ ˆ,ˆ

++ΨPΨ ˆ,ˆ

Prior Information

Linear/Nonlinear FE Model

Likelihood Function

Measurement

Bayesian Updating

( )Ψp( )Ψp

Ψ

( ) ( ),p NY = y Ψ 0 R

Digital Twins for Operational Monitoring and Management

8

Acceleration THs

Vehicle Locations

1

1

1

4

Object Tracking / Type Detection

2

3Stochastic Filtering

Predictions

-+

Online Portal

Bayesian FE Model Updating(Integrating Data with Model)

Measurements

Mechanics-based FE Model

Estimate jointly the model parameters and vehicular loads

Online Portal

P1P2

Pn

Digital Twin

Share Information with Stakeholders

Permanent/temporary sensors

Regular camera system (tripod or drone mounted)On-site laptopOffice/server computer

3

1

23

Box-Girder Bridge Case Study

9

• San Roque Canyon Bridge

• Typical prestressed cast-in-place RC box girder

Concrete deterioration & rebar corrosion in Deck

Concrete deterioration in Girders

Loss of prestress force

Verification Study

10

• Verification using numerically simulated data

• Poor condition in

R1

R2

R3

R5

R4

40% Reduction in Top slab 𝑓𝑐′

20% Reduction in Girder 𝑓𝑐′

20% Reduction in Tendon PrestressNo damage in Bottom slab

Accelerometers

R2

• 20 unknown material parameters + Rayleigh damping parameters + Vehicular loads

Top Slab 𝑓𝑐′ @ 5 regions

Girder 𝑓𝑐′ @ 5 regions

Bottom Slab 𝑓𝑐′ @ 5 regions

Section prestressforce @ 5 regions

Verification Study

11

Damage Diagnosis

12

Top Slab

Girders

Bottom Slab

Top Slab 𝑓𝑐′ Girders 𝑓𝑐

′

Tendon Prestress Strain Bottom Slab 𝑓𝑐′Tendons

Digital Twins for Post-Earthquake Assessment

13

Output Acceleration THs

4

Stochastic Filtering

Predictions

-+

Online Portal

Bayesian FE Model Updating(Integrating Data with Model)

Measurements

Mechanics-based FE Model

Estimate jointly the model parameters and foundation input motions (FIMs)

Digital Twin

Share Information with Stakeholders

Permanent Sensors

Box-Girder Bridge Case Study

14

Identifiability Analysis

15

No. Description No. Description

1 Modulus of elasticity of the concrete (deck) 19 Longitudinal soil-foundation stiffness (abutment)

2 compressive strength of the concrete (columns) 20 Longitudinal soil-foundation damping (abutment)

3Initial modulus of elasticity of the concrete

(columns) 21 Transverse soil-foundation stiffness (abutment)

4 Modulus of elasticity of the bearing pad 22 Transverse soil-foundation damping (abutment)

5 Shear stiffness of the bearing pad 23 Vertical soil-foundation stiffness (abutment)

6 Mass of the embankment-abutment 24 Vertical soil-foundation damping (abutment)

7 Vertical soil-foundation stiffness (piers) 25Rotational soil-foundation stiffness about the

longitudinal axis (abutment)

8 Vertical soil-foundation damping (piers) 26Rotational soil-foundation damping about the

longitudinal axis (abutment)

9 Longitudinal soil-foundation stiffness (piers) 27Rotational soil-foundation stiffness about the

vertical axis (abutment)

10 Longitudinal soil-foundation damping (piers) 28Rotational soil-foundation damping about the

vertical axis (abutment)

11 Transverse soil-foundation stiffness (piers) 29Far-field soil-embankment stiffness in the

longitudinal direction

12 Transverse soil-foundation damping (piers) 30Far-field soil-embankment radiation damping in

the longitudinal direction

13Rotational soil-foundation stiffness about the

longitudinal axis (piers) 31Far-field soil-embankment material damping in

the longitudinal direction

14Rotational soil-foundation damping about the

longitudinal axis (piers) 32Initial stiffness of the soil-backwall stiffness in the

longitudinal direction

15Rotational soil-foundation stiffness about the

transverse axis (piers) 33 Ultimate strength of the soil behind the backwall

16Rotational soil-foundation damping about the

transverse axis (piers) 34 Initial stiffness of the shear-key

17Rotational soil-foundation stiffness about the

vertical axis (piers) 35 Mass-proportional damping coefficient

18Rotational soil-foundation damping about the

vertical axis (piers) 36 Stiffness-proportional damping coefficient

Validation Study – Santa Cruz 2018 Eq.

16

Damage Diagnosis

17

Passive Soil Force-Deformation

Concrete Fiber Material Response

Shear Key Force-Deformation

• U.S. DOT - SBIR Program Grant # 6913G618P800109

• Caltrans Contract # 65A0642

• My colleagues at SC Solutions, Inc.

18

Acknowledgements

Thank youfor your attention!