experimental and numerical study to improve damage...

CRC 2nd Annual Meeting Feb. 1-3, 2017

The University of North Carolina at Chapel Hill

CRC 2nd Annual Meeting: February 1 – 3, 2017, Chapel Hill, NC

Experimental and Numerical Study to Improve Damage and Loss Estimation due to Overland Wave and Surge Hazards on Near-Coast Structures

John van de LindtColorado State University

Daniel CoxOregon State University

Kevin Cueto, Diego Delgado, Trung Do, Ben Hunter, Tori Johnson, Hyoungsu Park, Willliam Short



Project Overview

Objective 1: Quantify surge/wave forces on near-coast structures and develop new predictive equations.

Objective 2: Develop the conditional probabilities (fragilities) for exceeding key thresholds.

Objective 3: Illustrate next-generation risk-informed design.

Introduction



Existing FEMA Damage Predictions

Introduction

Case 1: •Freeboard=-2.7 m •Damage: Interior water damage, minimal structural damage•Significant Wave Height: 0.99m

Case 2: •Freeboard=-1.5 m •Damage: 100%•Significant Wave Height: 1.37 m

Predicted Damage: 72% Predicted Damage 60%



Technical Approach

Task 1: Hydraulic model test program at OSU (Yr 1) and data analysis (Yr 1, 2).

Task 2: Numerical model program at CSU. Verification (Yr 1) and fragility development (Yr 1, 2).

Task 3: Develop performance based design examples to illustrate methodology for engineering practice (Yr 2).

Introduction



Task 1: Hydraulic model test program

Research Accomplishments

Hurricane Ike 2008 / Bolivar Peninsula, TX

Observed surge and wave action:Offshore significant wave height: 5.8 m

Storm surge: 4.3 m

Inundation: 4.2-4.7 m

Estimated overland waves: 1.9 m

CRC 2nd Annual Meeting Feb. 1-3, 2017The University of North Carolina at Chapel Hill

Geometric scale 1:10Wave height: 0.10 < H < 0.50 mInundation: 0.40 mSpecimen dimensions: 1.02 x 1.02 x 0.61 m

Froude similitude 1:3.16Wave period: 2.5 < T < 5.0 s

Simplifying Assumptions• No substructure

• No sediments, scour

• No debris

• No currents

First comprehensive measurements of wave forces on elevated residential structures







Gulf of Mexico Barrier island Bay



Research AccomplishmentsTable 1. Experimental wave conditions 1

Exp. TMA REG TRAN

HS

(m)

TP

(s)

h

(m)

Dur.

(min) 𝐻 (m)

𝑇 (s)

h

(m)

Dur.

(min)

A

(m)

TR

(s)

h

(m)

ts

(s)

X1 0.10 3.72 2.15 40.0 0.10 4.10 2.15 4.00 0.51 36.4 2.00 10.0

X2 0.19 3.86 2.15 40.0 0.21 4.10 2.15 4.00 0.34 51.0 2.00 15.0

X3 0.29 4.10 2.15 40.0 0.29 4.10 2.15 4.00 0.28 87.2 2.00 20.0

X4 0.40 4.10 2.15 40.0 0.40 4.10 2.15 4.00 0.21 109 2.00 25.0

X5 0.50 3.86 2.15 40.0 0.50 4.10 2.15 4.00 0.18 117 2.00 30.0

X6 0.16 2.52 2.15 25.0 0.16 2.52 2.15 2.50 0.16 120 2.00 35.0

X7 0.21 2.98 2.15 30.0 0.23 2.98 2.15 3.00 0.14 154 2.00 40.0

X8 0.25 3.28 2.15 35.0 0.26 3.64 2.15 3.50 0.13 162 2.00 45.0

X9 0.34 4.68 2.15 45.0 0.35 4.68 2.15 4.50

X10 0.39 5.04 2.15 50.0 0.42 5.04 2.15 5.00

2

- Data are hosted on NSF DesignSafe-CI for public access




a (m)Air Gap

cases

REG

TMA

TRAN

a0 -0.40 -0.25

a1 -0.30 -0.15

a2 -0.20 -0.05

a3 -0.10 0.05

a4 -0.05 0.10

a5 0.00 0.15

a6 0.05 0.20

a7 0.10 0.25

a8 0.20 0.35

a9 0.28 0.43SWL

a




15

Effects of Air Gap on Horizontal, Vertical Forces

- Horizontal force increases as air gap decreases.

- The maximum vertical force is found when the air gap is zero.

- In some cases vertical force can exceed horizontal force

- There are limited provision in FEMA 55 or ASCE 7-16 to estimate these forces

Horizontal Force

VerticalForce



Contributed to October 2016feature article on CRC page

Participation in SUMREX program Kevin Cueto; Diego Delgado7-week summer program at HWRL

Thank you, Josh!

Education and Outreach




• Model verification using existing data - Tsunami loads on wood-frame wall at full scale test by Linton et al. (2013)- Uplift forces on a large scale bridge superstructure by Bradner et al. (2011)- Elevated Structure Impact test at OSU (2016)




Calibrating model using wood-frame wall at full scale test


v1= v (z,t)

Wave maker

Wood wall

7.3m 0.5m2.5m 25.4 m

Tsunami wave

1:12

2.3

6 m

3.66 m

3.58 m

2.4

4 m

Location 1Location 2

3.6 m

1.8

5 m

2.3

6 m

3.66 m

3.58 m

2.4

4 m

Load cells (Supports)

Displacement (Location 3)Flume botom

Flume wall 0.3

3 m

1.8

5 m

3.58 m

2.4

4 m

0.06

5 m

0.3

3 m

a) b)

c)

Location 3

Location 3

d)

2.3

6 m

7.0

m

3.66 m

35.7 m

sea water materialEmpty materialbottom face: v3=0

flow-out

face, p=0

side face, v2=0

flow-in face, v1=v(z,t)

top facep=0

Sketch of transverse wood wall in flume

Numerical model of wood wall

v1= v (z,t)

Wave maker

Wood wall

7.3m 0.5m2.5m 25.4 m

Tsunami wave

1:12

2.3

6 m

3.66 m

3.58 m

2.4

4 m


3.6 m

1.8

5 m

2.3

6 m

3.66 m

3.58 m

2.4

4 m



Flume wall 0.3

3 m

1.8

5 m

3.58 m

2.4

4 m

0.06

5 m

0.3

3 m

a) b)

c)

Location 3

Location 3

d)

2.3

6 m

7.0

m

3.66 m

35.7 m


flow-out

face, p=0

side face, v2=0


top facep=0

v1= v (z,t)

Wave maker

Wood wall

7.3m 0.5m2.5m 25.4 m

Tsunami wave

1:12

2.3

6 m

3.66 m

3.58 m

2.4

4 m


3.6 m

1.8

5 m

2.3

6 m

3.66 m

3.58 m

2.4

4 m



Flume wall 0.3

3 m

1.8

5 m

3.58 m

2.4

4 m

0.06

5 m

0.3

3 m

a) b)

c)

Location 3

Location 3

d)

2.3

6 m

7.0

m

3.66 m

35.7 m


flow-out

face, p=0

side face, v2=0


top facep=0

Numerical of wave flume and wood wall



Comparing numerical results and tested data


Compare deep water wave height at location 1 (a), running up wave height(b) and velocity (c)

In front of the wall (location 2)

Compare flux (d), total horizontal force (e) at location 2, and deflection at mid span of the wall

(location 3)



Model verification with uplift loads


Compare result for total uplift on bridge between model and test

Analysis of failure mechanics: total uplift exceed the self-weight of bridge superstructure causing failure




Elevated Structure Impact test at OSU (2016)

Numerical model of flume and elevated structure in wave flume

Locations of pressure gauges

Locations of wave gauges



Video showing wave impact on elevated structure




Comparing wave heights at wage gausses





Comparing results at pressure gauges




• Fragility development

• 3 Building archetypes are selected from 6 residential wood building archetypes of the hurricane wind project

• Set up numerical model and collect total uplift and shear as well as force on components such as doors, windows, and walls

• Establish damage states based on damage of components such as door, windows, and nails connection of wood walls.

• Generate fragility surfaces based on both significant wave heights and flood levels


Archetype 1 23 x 50 ft (7x15m), Rectangle, 1-story

Archetype 2 38 x 52 ft

(12x16m), 2-story

Archetype 3 38 x 65 ft

(12x20m), 2-story



Building models


• The buildings are modeled in ANSYS Fluent

• Piles rising from 0, 1m, 2m, and 3m, from the ground

Archetype 1 With 3-m elevated pile from ground

• TMA spectrum for hurricane waves with 𝐻𝑠 = 1,2 , and 3 m, which can cover up to 12msignificant wave height in deep water, and wavepeak period, 𝑇𝑝, from 8s to 14s

• Surge (SWL, food) levels ℎ = 1, 2, and 3m



Example of wave impact on building archetype 1


• Piles rising 1m from the ground

• Significant wave height = 1m

• Flood level = 1m



Pressure measured location and distribution at one wave impact event


PG1

PG2

PG3

PG4

PG5

PG6

PG7

PG3

PG5

PG7



Table of combinations for each building archetype


Pile Elevation (m)

Surge level (Hs, m)Significant wave height

(m)

Archetype 1

0

1

1

2

3

2

1

2

3

3

1

2

3

1

1

1

2

3

2

1

2

3

3

1

2

3

Pile Elevation (m)

Surge level (Hs, m)

Significant wave height (m)

Archetype 1

2

1

1

2

3

2

1

2

3

3

1

2

3

3

1

1

2

3

2

1

2

3

3

1

2

3



Define of fragility curve for coastal structures


A fragility, 𝐹𝑟 , can be expressed as

𝐹𝑟(𝑥) = 𝑃[𝑄 > 𝑅|𝑥]

For elevated structures subjected to storm surge:

-Q = loading from the model for everycombination of (𝐻𝑠, 𝑇𝑝)and/or surge levels, 𝑆

-R= Resistant/ capacity

-Hazard intensities, x = Significant wave height(𝐻𝑠, 𝑇𝑝), Surge levels, 𝑆.

For different time duration

𝑃𝑓 𝑖𝑛 𝑋 ℎ𝑜𝑢𝑟𝑠 = 1 − 1 − 𝑃𝑓 𝑖𝑛 𝑌 ℎ𝑜𝑢𝑟𝑠𝑋𝑌

DamageState

Window/door failure

Wall failure Floor failure

0 (no damage) <1% No No

1 (Minor damage)

>1% and <5%

No No

2 (Moderate) >5% and <25%

>5% and <25% >5% and <25%

3 (Severe damage)

>25% and <50%

>25% and <50% >25% and <50%

4 (Destruction) >50% >50% >50%



Component Resistance Values Used to Model Residential Buildings (Hazus 2.1- Hurricane)


Component Distribution Parameters

Window on 1 story

Weibull C = 54.49psf, k = 4.7

Window on 2 story

Weibull C = 38.7psf, k = 4.8

Entry door Normal Mean=50psf, COV=0.2

Toe-nail Normal Mean=415lb, COV=0.25




Fragilities surfaces for building subjected to wave and surge

Probability of failure base on two environmental variables given some damage states

-Significant wave height-Surge/flood level

Example: when significant wave height = 2.0m, surge level = 2.5m => Probability of failure due to Damage State 1 = 19% when building was raised 1 m from the ground

Task 3: Develop performance based design examples to illustrate methodology for engineering practice (Yr 2).

Application to Galveston, TX

• Damage to residential housing• Crystal Beach, TX• Hurricane Ike (2008)

Crystal Beach, TX

Applications

Applications

Current Flood Maps for Crystal Beach Online Mapping Tool: similar elevations

Applications

Study area: 394 houses.Use appropriate FIRM and distance from shore to estimate house elevation

Date of Construction

Applications

Approximate elevation, LCM

Applications

Overlay building stock with hazard:

• Hs max, from ADCIRC simulations by Bret Webb, U. South Alabama. via NIST project

• Houses are colored by the age classification. • Fragility curve selection in Tomiczek et al.

Applications

Applications

Results: Likelihood of Failure (Collapse Limit State)

Applications

Likelihood of Failure (Collapse Limit State)Increase elevation of all structures by 1 m

Potential for decision support tool



End User Engagement

We plan to work with the following people involved in the End-User Transition: • Eric Berman, HAZUS Program Manager at FEMA HQ • Ed Laatch, FEMA Building Science Division• Chad Berginnis, ASFPM Executive Director and CRC Advisory Board Member• Doug Bellomo, USACE Institute for Water Resources• Christina Lindemer, Coastal Engineer, FEMA Risk Analysis Branch, Atlanta GA

Revisions to FEMA 55 Coastal Construction Manual and ASCE 7-16• Workshop July 19+20, 2017 at OSU to develop research roadmap and

implementation plan • Bill Coulbourne, FEMA-55 CCM committee lead• Chris Jones, Chair of ASCE 7 Flood Load Subcommittee• Gary Chock, Chair of ASCE 7 Tsunami Subcommittee



Proposed Follow-on Work

Tracking modeling uncertainties

3 October 2016

43

1. Building Data2. Energy/EPN3. Water4. Transportation5. Communication

Fragility FunctionsExpected Damage

Kennedy et al., 2011

Hazard Data(ADCIRC/SWAN)




• Comparison of new physics-based fragilities to empirical and Hazus fragilities

• Application to NJ coast (post-Sandy)




Mitigation measures for existing and new construction: Examples of mitigation measures to increase building performance for wind (a – d) from FEMA’s Coastal Construction Manual. (e) Example of testing elevated bridge at HWRL.




Debris Impact: (a) Flood-borne debris hazards (FEMA 2000), (b) log-structure impact at 1:1 scale and (c) shipping container-column impact at 1:25 scale.




Prototype-scale testing for breakaway walls: (a) New construction of elevated home with breakaway wall on first floor. (b) Example of prototype-scale testing at HWRL of wave loads on subassembly and (c) structural failure.



Thank you

Daniel Cox ([email protected]), John van de Lindt, Kevin Cueto, Diego Delgado, Trung Do, Ben Hunter, Tori Johnson, Pedro Lomonaco, Hyoungsu Park, William Short

experimental and numerical study to improve damage...

Documents