1

An Introduction to Swiss Academic Institutions;

Fatigue Strengthening of Metallic Members using Pre-stressed CFRP

Laminates

Elyas Ghafoori

PhD ETH Zürich

Research Scientist

Empa, Swiss Federal Laboratories for Material Science and Technology,

Dübendorf, Zürich, Switzerland

2

Where is Switzerland?

5,500 KM

3

Universities and Research Institutes

1. Swiss Federal Institute of Technology Zurich, ETH Zurich

www.ethz.ch

2. Swiss Federal Institute of Technology Lasuanne, EPFL

www.epfl.ch

Research Institutes:

1.Swiss Federal Laboratories for Material Science and Technology, EMPA

2.Swiss Federal Research Institute for Forestry, Snow and Landscape, WSL

3.Swiss Federal Institute for Water Resources and Water Pollution Control, Eawag

4.Paul Scherrer Institute, PSI

Source: the 2013-2014 Times Higher

Education World University Rankings'

Engineering and Technology table.

Admission in Swiss Universitys?

4

Empa Overview

Prof. Masoud Motavalli

5

Empa, Structural Engineering Lab.

Research Priorates:

• Structural Dynamics and Adaptive Structures

• Aplication of Advanced Materials in Construction: strengthening of metallic, concrete and

timber structures

It is the biggest structural laboratory in Europe

2008 2014

6

PhD study at ETH Zurich

One university supervisor + co-supervisor(s)

• 12 credits (3-4 course).

• First year, submission of the research plan

• Average PhD duration 4 years

PhD thesis type:

• Paper-based thesis (consisting of at least 3 journal papers)

• Monolotic (traditional method)

PhD defense:

• At least one external co-examiner.

7

Fatigue strengthening of metallic members using pre-

stressed CFRP laminates

8

Outline

Introduction

Fatigue Theory

Laboratory Experiments

Strengthening of a Bridge

Conclusions

Partners & Sponsors

9

Introduction

Market

Europe:

• 22% of bridges are metallic.

• 70% of these bridges are older than 50 years-old.

Switzerland:

• Swiss Federal Railways (SBB) has 6050 railway bridges.

• 25% of metallic bridges older than 80 years-old are riveted ones.

Problems in Metallic Bridges:

• Insufficient fatigue crack safety.

• Need for an upgrade to carry larger loads and more traffic.

Traditional Strengthening Solutions:

• Steel: heavy.

• Bonded CFRP plates: not working for unsmooth surfaces, e.g., rivets.

σmax, h

σmin, h

σa, h

σm,h

σa, h

stresses at root of notch

2 σ

a,h

2 σ

a,h

plasticzone

Fatiguelife

Material properties

Stress amplitude

Midrangestress

0

+

-

a

a

mStre

ss

Time

max min

2m

max min

2a

Sy.. Yield strengthSe.. Fatigue endurance limitSut.. Ultimate tensile strength

Sy

-Sy Sy0

Midrange stress m

Stress amplitude a

Compression Tension

Sy

-Sy Sy0

Midrange stress m

Stress amplitude a

Compression Tension

Sy

Se

-Sy Sy Sut0

Midrange stress m

Stress amplitude a

Compression Tension

Sy

Se

-Sy Sy Sut0

Safe Zone

Midrange stress m

Stress amplitude a

Compression Tension

1

2

3

1: No fatigue crack

2: Fatigue crack may occur

3: Fatigue crack occurs

Sy

Se

-Sy Sy Sut0

Safe Zone

Midrange stress m

Stress amplitude a

Compression Tension

Elyas Ghafoori, et al. SMAR 2015. Antalya

Design criterion for fatigue strengthening of steel girders using bonded CFRP

laminates

10

Fatigue theory

Stre

ss

Time0

+

-

a

a m

Stre

ss

Time0

+

-

a

a

m

Before strengthening (A):

After strengthening (B):

After strengthening (C):

by prestressed CFRP

by increasing stiffness

A

Sy

Se

-Sy Sy Sut0

Safe Zone

Midrange stress m

Stress amplitude a

Compression Tension

A

B

Sy

Se

-Sy Sy Sut0

Safe Zone

Midrange stress m

Stress amplitude a

Compression Tension

A

B

C

Sy

Se

-Sy Sy Sut0

Safe Zone

Midrange stress m

Stress amplitude a

Compression Tension

Stre

ss

Time0

+

-

a

a mSteel beam

F F

Steel beam

F F

Steel beam

F F

UHM-CFRP

Fatigue theory

2

01 2

/ 21

2 4

p aflange p

s s s a p p

L G NhPa hb m Pa

I I A t E A

2 2

2

1 1

4 2

a p aT

a s s p p s s a s s

G bd x Gh hx V x

dx t E I E I E I t E I

01 sinh

2

xa

a p p s s

G N hPax m P e

t E A E I

CFRP laminate

I-Beam

P P

Adhesive

L

Lp a

tf

h tw I-Beam

bp

ta

tp

bf

P

b b

P

Beam

CFRP Plate

σ(x)

σ(x)

τ(x)

τ(x)

Ns(x) Ns(x)+dNs(x)

Ms(x)+dMs(x) Ms(x

)

Vs(x) Vs(x)+dVs(x)

Np(x) Np(x)+dNp(x)

Vp(x)+dVp(x) Vp(x)

dx

Adhesive layer

21 1

4

a p

a s s p p s s

G b h

t E A E A E I

1 22

a

a s s

G hm

t E I

interfacial shear stress along the CFRP plate

stress in beam bottom flange

N force in CFRP plate

N0 the pre-stress level

Ga adhesive shear modulus

Note: Subcripts ‘s’ and ‘p’ refers to the steel and the CFRP plate

x

flange

Fatigue theory

A

B

C

Sy

Se

-Sy Sy Sut0

Safe Zone

Midrange stress m

Stress amplitude a

Compression Tension

1a m

e utS S

2 2/20 m 0

1 1 m2 2

1 1

2 4 2 4

pLp p pa a aa

s e e s s a p p s ut ut s s a p p p f

b b b dhaP h G N haP h G Nm aP m aP e

I S S I A t E A I S S I A t E A nb k

2

01 2

/ 21

2 4

p aflange p

s s s a p p

L G NhPa hb m Pa

I I A t E A

Fatigue theory

IPE 120:

fy = 383 MPa

fu = 462 MPaEs = 199.3 GPa

Adhesive:

Aradite AW 106

Ga= 1.04 GPa

CFRP:

NM= 159 GPaHM= 220 GPaUHM= 440 GPa

0

300

600

900

1200

1500

1800

0 0.1 0.2 0.3 0.4 0.5

Str

ain

in

CF

RP

(m

icro

-str

ain

)

x/Lp

Modeling: 15 kN

Experiment: 15 kN

Modeling: 30 kN

Experiment: 30 kN

Modeling: 51 kN

Experiment: 51 kN

HM CFRP

I-Beam

P P

x Lp

More details in:Ghafoori E., Motavalli M., Zhao X.L., Nussbaumer A., Fontana M. Fatigue design criteria for strengthening metallic beams with bonded CFRP plates. Engineering Structures, 2015. 101: p. 542-557.

Experiments

SpecimenStrengthening

schemeCFRP type

Load

(kN)

No. of cycles to

crack initiation Failure mode

B0 Unstrengthened - 1.7-18 448,000 Crack initiation

B1 Bonded NM (159 Gpa) 1.7-18 1,705,000 Crack initiation

B2 Bonded HM (220 GPa) 1.7-18 2,000,000 Runout

B3 Bonded UHM(440 GPa) 1.7-18 2,000,000 Runout

100

120

140

160

180

200

150 170 190 210 230 250

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

R=0.09

0

50

100

150

200

250

300

350

400

0 50 100 150 200 250 300 350 400 450

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

R=0.09

1

100

120

140

160

180

200

150 170 190 210 230 250

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

R=0.09

1

0

50

100

150

200

250

300

350

400

0 50 100 150 200 250 300 350 400 450

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

1: Experiment: Reference beam

R=0.09

1

2

100

120

140

160

180

200

150 170 190 210 230 250

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

R=0.09

1

2

0

50

100

150

200

250

300

350

400

0 50 100 150 200 250 300 350 400 450

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

1: Experiment: Reference beam

2: Experiment: NM CFRP

R=0.09

1

2

3

100

120

140

160

180

200

150 170 190 210 230 250

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

R=0.09

1

2

3

0

50

100

150

200

250

300

350

400

0 50 100 150 200 250 300 350 400 450

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

1: Experiment: Reference beam

2: Experiment: NM CFRP

3: Model: NM CFRP

R=0.09

1

2

34

100

120

140

160

180

200

150 170 190 210 230 250

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

R=0.09

1

2

34

0

50

100

150

200

250

300

350

400

0 50 100 150 200 250 300 350 400 450

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

1: Experiment: Reference beam

2: Experiment: NM CFRP

3: Model: NM CFRP

4: Experiment: HM CFRP

R=0.09

1

2

34

5

100

120

140

160

180

200

150 170 190 210 230 250

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

R=0.09

1

2

34

5

0

50

100

150

200

250

300

350

400

0 50 100 150 200 250 300 350 400 450

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

1: Experiment: Reference beam

2: Experiment: NM CFRP

3: Model: NM CFRP

4: Experiment: HM CFRP

5: Model: HM CFRP

R=0.09

1

2

34

5

6

100

120

140

160

180

200

150 170 190 210 230 250

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

R=0.09

1

2

34

56

0

50

100

150

200

250

300

350

400

0 50 100 150 200 250 300 350 400 450

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

1: Experiment: Reference beam

2: Experiment: NM CFRP

3: Model: NM CFRP

4: Experiment: HM CFRP

5: Model: HM CFRP

6: Experiment: UHM CFRP

R=0.09

1

2

34

5

6

7

100

120

140

160

180

200

150 170 190 210 230 250

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

R=0.09

1

2

34

567

0

50

100

150

200

250

300

350

400

0 50 100 150 200 250 300 350 400 450

Alt

ernat

ing s

tres

s (M

Pa)

Mean stress (MPa)

1: Experiment: Reference beam 2: Experiment: NM CFRP 3: Model: NM CFRP 4: Experiment: HM CFRP 5: Model: HM CFRP 6: Experiment: UHM CFRP 7: Model: UHM CFRP

R=0.09

I-Beam

P P Experiments

16

Strengthening of a Swiss bridge

The Münchenstein rail disaster on 1891 is historically the worst railway

accident ever in Switzerland.

1891 2013

Münchenstein Bridge is a railway riveted metallic bridge in Switzerland.

The bridge had been built in 1875 by Gustave Eiffel, who built the Eiffel

Tower later in 1889.

17

FE Modeling

ABAQUS FE package

Shell elements

within connection

regions

Beam elements in

non-critical regions

Steel Material:

Elastic material Properties

E=210 GPa, =0.3

Reference

Node

Rigid-body nodal constraints tied

to reference node (plane sections

remain plane assumption) Reference

Node

Beam element connecting two

shell-element profiles

Shell element

nodes

45 m

5 m

6 m

18

FE Modeling

19

Clamp (permanent)

Column (permanent)

Prestressing chair (temporarily)

Rivet

Cross beamStringer

Stringer

1. Applicable to unsmooth surfaces (i.g.,

riveted beams).

2. Fast installation (no gluing & no surface

preparation).

3. Easy to prestress (no hydraulic jacks).

4. No traffic interruptions for bond curing.5. Minimum damage (no hole, glue & grinding).

6. Easy to remove.

7. Adjustable prestressing level (to compensate relaxation).

Cross beam

Cross beam

Cross beam

PUR System

20

Laboratory Experiments

5000

23

Detail of rivet holes in bottom

flange of beam

Dimensions in mm

Cyclic loading

21

Laboratory Experiments

22

Mechanical friction clamp

3 CFRP plates

Adjustable column

Magnetic

strain gauge

Bridge Strengthening

23

Humidity and

temperature sensors

Wireless sensor node

Bridge Strengthening

24

Conclusions

• A prestressed unbonded reinforcement (PUR) system has been developed.

• The main advantages of PUR system are Applicable to unsmooth surfaces (i.g., riveted beams)

Fast installation (no gluing & no surface preparation)

Adjustable prestressing level

• Fatigue theory was introduced to determine the minimum prestressing level to avoid

fatigue crack initiation.

• Laboratory experiments have shown the validity of the proposed method.

• The system was applied on Münchenstein bridge in 2014 and was monitored for one

year.

25

Partners & Sponsors

Swiss Commission for Technology & Innovation (CTI)

Industrial Partners:

– S&P clever reinforcement company AG

– SBB (Swiss Federal Railways)

Research Partners:

– EPFL, Steel Structures Laboratory, Prof. A. Nussbaumer

– ETHZ, Structural Engineering Section, Prof. M. Fontana

Questions/ Discussion