impact of earthquaker duration on bridge performance

David H. Sanders MAE Presentation

11/9/2017 1

Impact of Earthquake Duration on Bridge Performance

1

David H. Sanders

University Foundation Professor

Department of Civil and Environmental Engineering

University of Nevada, Reno

2

• Chile (2015, 2014 and 2010), Japan (2011), China (2008),and Indonesia (2004) earthquakes are reminders of theimportance of the effect of ground motion duration onstructural response.

• Long fault ruptures including subduction zones

• Chile Earthquake Ruptured over ~ 500 kmDuration ~ 20-90 seconds

• Tohoku Earthquake Fault size ~ 500 km x 210 km

Duration ~ 90-270 seconds

Earthquakes typically last less than 30seconds

Why Long Duration?

• California


11/9/2017 2

Ground Motion Duration Definitions

3

Arias Intensity=∫a2.dt

• Significant Duration (5‐95% of the Arias Intensity)

Recommended by Jack Baker

and Greg Deierlein (2012)

Vu and Saiidi (2005) Rinaldi -1/3 scale

• Design Code: AASHTO

• L/D= 4.5 (flexural control)

Target Displacement demands for our tests

(50%)

4

9.8 in.

Specimen Design/Pre-test Analysis


11/9/2017 3

Specimen Design/Pre-test Analysis

5

- OpenSees model was used to simulate Vu and Saiidi’s Column

- Selection of the motions was based on this model (the displacementdemands to be around half the capacity)

-80 long-duration ground motions from Japan 2011 and Chile 2010were used in the pre-test analysis

6

Damage Prediction Before TestingModified Park-Ang damage index was used to quantify the damage

•δ max = Maximum displacement demand during the ground motion

•δ u =Ultimate displacement capacity (taken 9.8 in. from Vu and Saiidi’s test)

•β = Constant (taken 0.15 for concrete structures)

•Eh = Hysteretic energy

•Fy = Yield force


11/9/2017 4

Modified Park-Ang damage index was used to quantify the damage

Data from past shake-table and cyclic load tests on seismically designed bridge columns (about 25 models) were used to correlate the damage index with different damage states.

Experimental Fragility Curves

0

10

20

30

40

50

60

70

80

90

100

0 0.5 1 1.5 2 2.5 3 3.5

Pro

babi

lity

of e

xcee

ding

a

dam

age

stat

e (%

)

DI (Park and Ang.)

Minor Spalling

Extensive Spalling

Exposed Reinforcement

Bar Buckling

Bar Fracture

‐Example:Run #, DI=1.5

1.5 7

40% Bar Buckling

0% Bar Fracture

90% Exposed RFT.

Damage Prediction Before Testing

8

JapanLong-dur.

Loma PrietaShort-dur.

JapanLong-dur.

0

0.5

1

1.5

2

2.5

3

0 1 2 3 4

Spe

ctra

l Acc

eler

atio

n (g

)

Period (sec.)

Crescent City, CA (2475 years)

FKSH20 N-S (Original)

FKSH20 N-S (Modified)

Column LD-J1(Long-Duration)

Japan

0

0.5

1

1.5

2

2.5

3

3.5

0 1 2 3 4

Spe

ctra

l Acc

eler

atio

n (g

)

Period (sec.)

Seattle, WAPort Angeles, WANewport, ORCrescent City, CAEureka, CA

2475 yr return period

Selected

Spectral Matching

0

0.5

1

1.5

2

2.5

0 1 2 3 4

Spec

tral

Acc

eler

atio

n (g

)

Period (sec.)

Crescent City, CA (2475 years)

MYG006 E-W (Original)

Initial Period

Shake Table Tests


11/9/2017 5

9

JapanLong-dur.

Loma PrietaShort-dur.

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

0 50 100 150 200 250

Acc

eler

atio

n (g

)

Time (sec.)

Ds (5-95%) = 88 sec.

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

0 50 100 150 200 250

Acc

eler

atio

n (g

)

Time (sec.)

Ds (5-95%) = 9 sec. 0

0.5

1

1.5

2

2.5

0 1 2 3 4

Acc

eler

atio

n (g

)

Period (sec.)

Japan

Short

Design codes

These motions are the same

JapanLong-dur.

Shake Table Tests

10

Axial Load

Mass Rig System Lateral

Load cell

Axial Load

Rigid frame

String pots

Shake table

Specimen

Shake Table Tests


11/9/2017 6

11

Mass Rig

Shake Table

Rigid Frame

Axial Load

Shake Table Tests

12

100% of GM

+

AfterShock +

125% of GM+

150% of GM+

etc… (Until Failure)

Loading Protocol


11/9/2017 7

13

Max. Disp.= 4.5” Max. Disp.= 3.88”

South

• 4.4” spalling • Spirals Exposed

North


South

• Cracks (max width= 0.4mm)

North

• 4.5” spalling • No RFT. Exposed

Column 1(Japan- Long Dur.)

Column 2(Short-duration)

100 % of the Ground Motion

Max. Disp.= 4.7”

South

North

• Minor spalling • No RFT. Exposed

Column 3(Japan – Long Dur.)


14


South

North

South

North



Max. Disp.= 7.38”

South

North


• 8.5” spalling • 4 Bars fractured

• 6.4” spalling • Core Damage

• 4.5” spalling • Spirals exposed

• 4.5” spalling• Spirals exposed

• 8.0” spalling • 3 Bars buckled

• 5” spalling

• 1 Bar fractured



11/9/2017 8

15


South

North

South

North



Max. Disp.= 7.38”

South

North


• 8.5” spalling • 4 Bars fractured

• 6.4” spalling • Core Damage

• 4.5” spalling • Spirals exposed

• 4.5” spalling• Spirals exposed

• 8.0” spalling • 3 Bars buckled

• 5” spalling

• 1 Bar fractured

-4

-2

0

2

4

6

8

350 400 450 500 550

Dis

plac

emen

t (i

n.)

+ 7.38 in.

- 2.3 in.

Bar FractureA Column 3


16


11/9/2017 9

17




Max. Disp.= 7.3”

South

North

• 6” spalling • Spirals exposed

• 9” spalling• Spirals exposed

Not

Ap

pli

cab

le

Bar

s F

ract

ure

d a

t 12

5%Max. Disp.= 4.98”

Not

Ap

pli

cab

le

Bar

s F

ract

ure

d a

t 12

5%

Max. Disp.= 7.38”


18




Max. Disp.= 9.22”

Not

Ap

pli

cab

le

Bar

s F

ract

ure

d a

t 12

5%

Max. Disp.= 4.98”

Not

Ap

pli

cab

le

Bar

s F

ract

ure

d a

t 12

5%

Max. Disp.= 7.38”

South

North

• 1 bar fractured• 2 bars buckled

• 4 bars buckled



11/9/2017 10

19

Force Displacement

20

Short-Duration

Long-Duration

(Japan-1)

Long-Duration

(Japan 2)

Strain History


11/9/2017 11

21

Column LD-J1 (Japan-125%)

Column SD-L (Short-175%)

Spectral Acceleration at Final Damage State

22

‐Concrete01 with STC

‐Steel02

‐Wehbe’s method for bond slip

0

20

40

60

80

100

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5

Pro

bab

ilit

y of

exc

eed

ing

a da

mag

e st

ate

(%)

DI (Park and Ang.)

Minor Spalling (M.S.)

Extensive Spalling (E.S.)

Exposed Reinforcement (E.R.)

Bar Buckling (B.B.)

Bar Fracture (B.F.)

50 % probability

Post-Test Analysis


11/9/2017 12

23

Incremental Dynamic Analysis (IDA)

0

10

20

30

40

50

60

70

80

90

100

0 5 10 15

Prob

abili

ty o

f C

olla

pse

(%)

Could be any INDEX

OpenSees Model #

Long-Duration Short-Duration Set A

- Example on how the comparative collapse fragility curves look like

- Each point represents a ground motion that was scaled until failure

Post-Test Analysis

672 total motions that were scaled until collapse

24

Maximum Displacement

Comparative Collapse Analysis(Collapse Fragility Curves)

Post-Test Analysis


11/9/2017 13

25

Spectral AccelerationComparative Collapse Analysis

(Collapse Fragility Curves)

Post-Test Analysis

Conclusions

26

1) Ground motion duration has a significant effect on the collapse capacity of bridge columns.

3) A significant reduction in the spectral accelerations at collapse in case of long duration motions with respect to the short duration motions.

2) A significant reduction in the displacement capacity was observed in case of long duration motions compared to the short duration motions for both the experimental and analytical studies.

Approximate reduction of about 25%

Approximate reduction of about 20%

4) Ground motion duration is an important parameter when selecting ground motions for nonlinear analysis of structures.


11/9/2017 14

Rapid ConstructionBy using precast components to accelerate the onsite construction

I5, Grand Mound Bridge Replacement Project, 2012 (Khaleghi et al. 2012)

27

By using pretensioned rocking columns to reduce the residualdisplacements and to mitigate earthquake damage

Resilient Bridge Using Conventional Materials

Rei

nfor

cing

Ste

el

Ste

el T

ubes

Str

ands

Gro

ut

Con

cret

e

28Faster Construction and Better Seismic Performance-You Can Have Both

Shape Memory Alloys (SMA)Engineered Cementious Composites (ECC)

Resiliency


11/9/2017 15

System Development

29

Damage Mitigation1- By confining the end of the columns to reduce the damage toconcrete

2- By debonding the longitudinal reinforcement at critical zonesto control the fracture of the longitudinal bars

Con

fini

ng T

ube

Deb

onde

d B

ars

30


11/9/2017 16

+ =

Moment-Rotation

Strand Rebar Total

31

System Development

Moment-Rotation

Strand Rebar Total

+ =

32

System Development


11/9/2017 17

Moment-Rotation

Strand Rebar Total

+ =

33

System Development

Moment-Rotation

Strand Rebar Total

+ =

34

System Development


11/9/2017 18

Moment-Rotation

Strand Rebar Total

+ =

35

System Development

Moment-Rotation

Strand Rebar Total

+ =

36

System Development


11/9/2017 19

Moment-Rotation

Strand Rebar Total

+ =

37

System Development

38

Bridge Description


11/9/2017 20

39

Column Description

40

Bridge Construction


11/9/2017 21

1- N-S component Century City Country Club North (CCN) station 1994 Northridge California Earthquake

2-N-S components Sylmar- OliveView Med. Center (SYL)

1994 Northridge California Earthquake

3-N-S component Takatori (TAK) record

1995 Kobe, Japan Earthquake

Motion Record PGA (g) Design Level (%)

14A CCN 0.25 33%

14B1 SYL 0.20 -

14B2 SYL 0.40 -

14C TAK 0.25 -

15 CCN 0.50 67%

16 CCN 0.75 100%

17 CCN 1.00 133%

18 CCN 1.33 177%

19 CCN 1.66 221%

20A CCN 0.75 100%

20B SYL 0.843 -

21A TAK 0.40 -

21B TAK 0.611 -

21C TAK 0.80 -41

Ground Motions

42

Ground Motions


11/9/2017 22

Conventional BridgeMotion 19 (221%DE)

Resilient vs Conventional

43

Resilient BridgeMotion 19 (221%DE) 44



11/9/2017 23

Behavior Comparison (Mantawy et al. 2016)

45


Damage Comparison (Mantawy et al. 2016)

46



11/9/2017 24

Damage Evaluation

47

Bar fracture occurred but after the design level earthquake. (133%)

Fracture of Reinforcing Bars

Motion 17 18 19 20A 20B 21A 21B 21C Total

# of Fractures

1 24 29 5 5 0 1 5 70

Method 1:Fracture Estimation by Acoustic Emissions

48


11/9/2017 25


1- Bar buckling in the columns was inhibited by a confining tube detail;

2- Strains in the reinforcement were distributed over deliberately debonded length;

3- All 72 of the column’s longitudinal reinforcement fractured during testing; and

4- Because the system deformed by rocking, the strain response histories could be calculated from the measured rotations at columns’ ends after the deformation capacities of the strain gauges were exceeded.

Method 2:Fracture Estimation by Connection Rotation

49

Method 2: Fracture Estimation by Connection Rotation

50



11/9/2017 26

Numerical Model Development

51

Method 3:Fracture Estimation by Low Cycle Fatigue Model

52

Numerical Model Development


11/9/2017 27

Numerical Model Assessment

53

Fracture Excluded Motion 20B (SLY 0.834 g)

54

Fracture Included Motion 20B (SLY 0.834 g)



11/9/2017 28

Total Base Shear Envelope

55

COM Displacement during Motion 19 (221%DE)


v v

56

Number of Fractured Bars



11/9/2017 29

Parametric Studies

Using CCN Motion 125%, 150%, and 175% DE

Bent Height and Column Diameter

Steel Ratio, Strand Ratio and Initial Pretensioning

Bar Size - #6, #7, #8

Unbonded Length of conventional reinforcement

What is the impact of configuration on facture and can the unbonded length of

conventional reinforcement delay fracture?

Impact of Debonding

58

Debonding the longitudinal steel at the rocking interfaces in the column, using 24 times bar diameter was effective to delay the fracture of longitudinal bars during the Design Level Earthquake while keeping residual displacement low.


11/9/2017 30

Conclusions

1- The precast system enables the bridge bents to be built faster in the field than conventional construction.

2- The system used conventional technologies while mitigating damage such as:

• Confining steel tubes at the ends of the columns mitigate the damage of concrete and eliminate the buckling of longitudinal reinforcement.

• Debonding delayed yielding and bar facture.

• Rocking system was effective at keeping residual displacement low.

59

The impact of earthquake duration can be significant through

increased strain history,decreased displacement capacity,

and increased chance of bar fracture.

60

Conclusions


11/9/2017 31

Thank you!

This is great honor and very much appreciated.

61