Damage Tolerance and Durability of Adhesively Bonded Composite Structures

The Joint Advanced Materials and Structures Center of Excellence

Damage Tolerance and Durability of Damage Tolerance and Durability of Adhesively Bonded Composite Adhesively Bonded Composite

StructuresStructuresHyonny Kim, Associate Professor, Dept. Structural Engineering, UHyonny Kim, Associate Professor, Dept. Structural Engineering, UC San DiegoC San Diego

C.T. Sun, Professor, School of Aeronautics & AstronauticsC.T. Sun, Professor, School of Aeronautics & AstronauticsThomas Thomas SiegmundSiegmund, Associate Professor, School of Mechanical Engineering, Associate Professor, School of Mechanical Engineering

2Purdue University – Joint Advanced Materials and Structures Center of Excellence

Damage Tolerance and Durability of Adhesively Bonded Composite Structures

• Motivation and Key Issues– failure prediction of composite adhesive joints remains a difficult problem

• multiple failure modes and complex failure processes• damage initiation and growth influenced by geometry, loading, and environmental

factors such as moisture, temperature, etc.– damage in joints is difficult to detect – must design structures to be tolerant to

reasonably-sized flaws• accurate models are needed to predict failure and assess damage tolerance

• Objectives– investigate physical phenomena and processes leading to failure in adhesively

bonded joints– account for bondline thickness and environmental conditions– develop models describing these phenomena

• Approach:– combined experimental/analytical investigations supporting development of

models


FAA Sponsored Project Information

• Principle Investigators & Researchers– Hyonny Kim (now at UCSD)– C. T. Sun– Thomas Siegmund– Post-Doc: Steffen Brinkmann– Students: Haiyang Qian, Nicholas Girder, Matt Wan

• former students: Jibin Han (Dec 2005), J. Lee (May 2006), T.T. Khoo (Dec. 2006), Hee Seok Roh

• FAA Technical Monitor– Curt Davies

• Industry Participation– ABAQUS


Focus Areas Towards Achieving Objectives:

– Adhesive constitutive behavior for use in bonded joint analyses

– Effect of adhesive thickness on mixed mode fracture of joints

– Effect of bondline thickness on strength of adhesively bonded joints – CTOA approach

– Influence of moisture, cyclic loading and time dependence on joint fracture – Cohesive zone model approach


Hyonny Kim, Associate Professor, UC San Diego, [email protected]: Jungmin Lee (PhD May 2006), Richard Khoo (MS Dec 2006), Hee Seok Roh

Project I. Adhesive Constitutive Behavior Measurement and Bondline Thickness Dependent Mixed Mode Fracture

Objective:– support analysis tools used for design and damage tolerance– use of nonlinear FEA and fracture mechanics based analyses has

become more routine• VCCT and cohesive-zone incorporated into commercial FEA codes

Approach– Accurately measure material property data as crucial ingredients to

increasingly capable and available modeling tools– defining improved methods for constitutive curve measurement– investigate bondline thickness dependent mixed mode fracture

envelope

mailto:[email protected]


0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.350

1000

2000

3000

4000

5000

Aver

age

Shea

r Stre

ss, τ

(psi

)

Average Shear Strain, γ

8mil20mil40mil60mil

• choice of constitutive curve is not clear• adhesive τ vs γ measured by ASTM D5656:

– exhibits strong bond thickness dependency– criticized as being inconsistent at ASTM

Symposium on Joining and Repair of Composites (March 2003), and at FAA Adhesive Joints Workshop (June 2004)

•• true material propertytrue material property

Adhesive Constitutive Behavior in Bonded Joints

should be geometry independent

• establish more direct and simple test method for determining constitutive behavior:

– bulk-adhesive tensile dogbone– t.b.d. new method

Shear Stress vs. Shear Strain Relationship for PTM&W ES6292 Measured by ASTM D5656 Test Method

Increasing Time

Localized Damage Evolution Causing Apparent Softening

Modified D5656 Test Specimen- less rotation- laser displ.

measurement

Displacement Control Test


Bulk Adhesive Constitutive Behavior – Tensile Dogbone Tests

0.000 0.005 0.010 0.015 0.020 0.0250

1000

2000

3000

4000

5000

6000

Stre

ss, p

si

Strain

0.3 in/min

0.05 in/min

1.5 in/min

PTM&W ES6292 epoxy paste adhesive

Constitutive Behavior is Strain Rate Dependent

• bulk dogbone partially successful• main issues:

– premature failure – not measuring entire constitutive curve

• voids always present due to manufacturing method (casting)

• initiation of failure leads to immediate cross-width fracture and thus can not develop significant plastic deformation

– does not include effects of:• adherend constraint on adhesive

layer• possible material micro-structural

differences between thin adhesive layer vs. thick bulk

Gage Section


Bondline Thickness Dependent Mixed Mode Fracture

• motivation:– fracture mechanics is capable tool for dam. tolerance

analysis– need mixed mode strain energy release rate (SERR) data

• approach:– SERR measured for range of bondline thickness to

establish mixed mode fracture envelope database– observed processes occurring at crack tip– use nonlinear FEA to understand bondline effect in

measured data– establish fracture criteria in joints that accounts for

bondline thickness dependent GIC and GIIC

Mode Mix(% mode II)

ta = 0.008 in.

ta = 0.020 in.

ta = 0.040 in.

ta = 0.060 in.

0 4 5 6 450 3 3 3 5

75 3 3 3 3

100 4 7 4 6

Matrix of Completed Tests (all tests at RT ambient):

test specimen details:adherends: 2024-T4 Al alloy, 0.25 x 1.0 x 6.0 in.adhesive: PTM&W ES6292 epoxy paste adhesivebondline thickness range: 0.008 to 0.060 in.

Laser ExtensometerLong-Distance

MicroscopeTest

Specimen


Results – Mixed Mode GC Envelope

0.00

5.00

10.00

15.00

20.00

25.00

0 20 40 60 80 100

Mixed-Mode Ratio (%)

Gc (

lb/in

)

8mil 20mil 40mil 60mil

MMR 0%

MMR 100%

MMR 50%

Large Shear Strain Visible

Small Shear Strain Visible

No Shear Strain Visible

FEA Plastic Zone Predictions – Mode I

8 mil. – PZ Highly Constrained

40 mil. – PZ Moderately Constrained

60 mil. – PZ Unconstrained

Adherend

Adherend

Adhesive

Plastic Zone (PZ) Contours at Growth Initiation:

GC highest for 40 mil joints of 0% to 75% mixed mode ratio

Crack Growth Process Observation by LD Microscope


Summary: Comparison of Shear Strength Test and Fracture Properties

• Fracture properties and shear strength test properties show opposite trend over bondline thickness range 0.008 to 0.06 in.

• Fracture Tests: – GIC and GC at 50% Mode II

optimum for ta = 0.04 in.– GC at 75% Mode II relatively

insensitive to ta– GIIC increasing (could plateau and

go down for higher ta than investigated)

– optimal constraint of plastic zone gives highest GC

• D5656 Shear Strength Tests:– shear yield strength decreasing

for higher ta– shear failure strain decreasing for

higher ta– related to localization of plastic

and failure process zone for higher ta

0 0.01 0.02 0.03 0.04 0.05 0.06Bondline Thickness ta (in.)

0

1

2

3

4

Para

met

er N

orm

aliz

ed b

y Va

lue

at t a

= 0

.008

in.

τYield

γFail

GIC

GC 50% Mode IIGC 75% Mode IIGIIC

Fracture Properties

Strength-Test Properties


Project I: Conclusions to Date &Benefits to Aviation Industry

• Tools and Protocols:– modified shear strength tests: localized damage/fracture

develops for thick bonds – this should be accounted for in data processing and analyses

– dogbone test for constitutive curve partially successful– new specimen is being designed that is easy to test like dogbone

but accounts for confinement of adhesive layer• Data

– strong bondline thickness effect observed for fracture and shearstrength tests

– fracture properties and strength test properties show opposing trends over range of bondline thickness

• Analysis– plastic zone confinement shown via FEA to affect critical SERR

dependency on bondline thickness


Project II: Modeling Thickness Effect on Strength of Adhesive Lap Joint Using CTOA

C.T. Sun, Professor [email protected], School of Aeronautics & Astronautics, Purdue University

Haiyang Qian, Ph.D. Student

Objective – Develop a CTOA fracture criterion to model adhesive thickness-dependent lap joint strength

Approach – Conduct fracture experiments using DCB specimens with various adhesive thicknesses to validate the proposed CTOA approach and to determine the limitation on its applicability with finite element analyses of the experiments

mailto:[email protected]


Adhesive Thickness Effect on the Strength Lap Joints

T

T

l

L

t

adhesive

Adherend: Aluminum Alloy 7075

Adhesive: HYSOL EA9394

Surface Treatment: Semco Pasa-Jell 105 (etching method)

L=3in, l=1in, T=0.125int=0.008in, 0.01in, 0.02in, 0.06in

01

23

45

67

0 0.5 1 1.5

Bondline Thickness (mm)

Stre

ngth

of J

oint

s (k

N)

• Experimental result

• Joint strength increases as the bondlinethickness decreases up to 0.25 mm


Fracture Initiation is Mode I Dominant in Lap Joints

Thin Layer

Thin Layer

Thick Layer

Thick Layer

0

0.5

1

1.5

2

2.5

3

3.5

4

0 0.1 0.2 0.3 0.4 0.5 0.6

Adhesive thickness (mm)

Ener

gy R

elea

se R

ate

(J/m

)

GIGII

Stress Decreasing

A

B

Initial crack mode mixity

•Stress concentration near the joint edge and near the interface• Initial flaw (crack) is under mode I loading• Crack growth is along the interface (red line)


DCB Test Resultsfailure modes transition from mode I fracture to interfacial

failure as adhesive thickness decreases below a certain level

50

60

70

80

90

100

110

120

130

140

0 0.5 1 1.5 2 2.5 3 3.5

Bondline thickness (mm)

Failu

re L

oad

(N)

0

20

40

60

80

100

120

140

0 2 4 6 8 10

Opening End Displacement (mm)

Load

(N)

0.95mm

1.3mm

3.3mm

0

20

40

60

80

100

120

140

160

0 5 10 15

Opening End Displacement (mm)

Load

(N)

0.4mm0.44mm0.85mm

Sudden Failure

Load vs Adhesive Thickness


Effect of Adhesive Thickness on Failure Mode

• Mode I crack propagates in thicker adhesive

•Transition of failure mode in thinneradhesive

0

20

40

60

80

100

120

140

0 20 40 60 80 100 120Applied Load (N)

Inte

rfaci

al N

orm

al S

tress

es (M

Pa) s22-6mil

s22-20mils22-60mil

Maximum Normal Stress

Normal stress at the interface


CTOA Criterion for Hysol EA9394

•CTOA is independent of adhesive thickness before failure mode change

0

1

2

3

4

5

6

7

8

0 0.5 1 1.5 2

Adhesive Thickness (mm)

CTO

A (D

egre

e)


Project II: Conclusions to Date &Benefits to Aviation Industry

• Tools and Protocols:– Critical CTOA concept: CTOA is a fracture criterion that

is independent of adhesive thickness if failure mode remains mode I. This is the case for thicker bondlines

• Data– Critical CTOA data determined in dependence of bond

line thickness • Analysis

– FEA analysis predictions using critical initial CTOA and failure mode transition due to high interfacial stress between adherend and adhesive layer


Project III: Influence of Bondline Thickness, Moisture, Load History

Thomas Siegmund, Associate Professor, [email protected]

Steffen Brinckmann, Post Doctoral Research AssociateJibin Han, (PhD 12/2005)Eric Anderson, Nicolas Girder, Matt Wan (SURF Summer Students)

• Objective:– Develop and employ the cohesive zone model approach to fracture to the

analysis of adhesive joint failure

• Approach:– Crack growth experiments: monotonic, fatigue, time-dependence,

environmental degradation– Models: cohesive zone models in 3D, monotonic, fatigue, coupled for

moisture/load interaction– Image analysis: Digital image correlation for strain fields, quantitative fracture

surface analysis and fracture reconstruction


Crack Growth Resistance Environmental Degradation

Displacements andStrain fields

Force –Displacement

Record

Finite Element Method with

Cohesive Zone


Record

SpeckleImages

Displacements andStrain fields


Record

Finite Element Method with

Cohesive Zone


Record

SpeckleImages

Stereo-FractographyDigital Image CorrelationMeX

a0 Δa 1 mm

0.021

-0.0035

εyy

a0 Δa 1 mma0 Δa 1 mm

0.021

-0.0035

εyy

0.021

-0.0035

εyy

Experimental Facilities


Computational Modeling

• The Cohesive Zone Model:– Describes local energy dissipation during fracture and fatigue– Is conveniently coupled to other fields (plasticity, moisture, heat,

electrical…)

F

F

Global Parameters:• Force (F) – Displacement (COD)• Environment (H2O)

COD

H2O

Δ

T

T

Local Parameters:• Traction (T) – Separation (Δ)• H2O Concentration C(H2O)

C(H2O)

Finite element model withcohesive elements & H2O transport

Adherent

Adhesive

CZ ElementsDiffusion Elements

• Load, Displacement• Environment• Time• Cycles

• Traction-Separation• Concentration• Damage

Finite element model withCohesive elements, moisture transport, and cyclic damage


Monotonic Loading

0

10

20

30

40

50

60

70

80

90

100

0 0.02 0.04 0.06

0.508mm1.524mm3.048mmcz law

Δn [mm]

T n [M

Pa]

0

10

20

30

40

50

60

70

80

90

100

0 0.02 0.04 0.06

0.508mm1.524mm3.048mmcz law

Δn [mm]

T n [M

Pa]

F

F

G( )( )nTCTOD∗

∂Δ =

∂

-200

-100

0

100

200

300

400

0 100 200 300 400 500 600 700 800 900

Point Number

z-va

lue

(µm

)

-300

-250

-200

-150

-100

-50

0

50

100

150

200

250 350 450 550 650 750 850 950

Stereo Images PairLeft -- Right

Digital Elevation Maps

Fracture Profiles Fracture Profiles

Some plasticity

max , ,σ δ Γ


Fatigue Loading

1

10

100

1000

10000

100000

0.01 0.1 1 10 100 1000 10000

Strain Energy Release Rate Range, ΔG [J/m2]

Fatig

ue C

rack

Pro

poga

tion

Rat

e, d

a/dN

(μ

m/c

ycle

)

max CG G=

( )0.85245da GdN

= Δ

-200

-150

-100

-50

0

50

100

150

200

0 500 1000 1500 2000 2500

distance along path (µm)

z-va

lue

(µm

)

max max,0 (1 ), ,Dσ σ δ= − Γ

FE-CZMExperiment

Fractography


Time Dependence

Wedge test with constant loading

0

500

1000

1500

2000

2500

0 2 4 6 8 10 12

Time at Load (min)

Cra

ck E

xten

sion

(mic

rom

eter

s)

Precrack Stable Unstable


Moisture Effects on Joint Fracture

Experiment Simulation

Adhesive

Cohesive


Project III: Conclusions to Date &Benefits to Aviation Industry

• Analysis– Cohesive zone models: fracture – fatigue – rate dependence –

moisture degradation• Tools and Protocols:

– In-situ crack growth– Digital image correlation applied to adhesives– Quantitative fractography– Environmentally assisted crack growth with wedge test– Time dependent crack growth with wedge test

• Data– Preliminary data on fatigue crack growth resistance and moisture

assisted crack growth


A Look Forward

• Benefit to Aviation– in response to increasing use of adhesive bonding

– Analysis Tools: supports sophisticated computation-based design• failure process prediction, including adhesive plasticity• CTOA, VCCT, Cohesive Zone model • now available in commercial codes• simulation tools can reduce time to conduct extensive environmental

degradation tests– Data: addressing important issues of bondline thickness

• quantify phenomena governing why “properties” seemingly depend on bondline thickness

• definition and use of local failure criteria that are not bondline thickness dependent

– Protocols: test methods to obtain fracture and constitutive data• seeking to define simpler tests and remove necessity to collect data as

function of bond thickness• Fractography


A Look Forward

• Future Needs– results to date concentrated on adhesive using metal adherends – future work

needed to investigate other adherend (namely composite) and adhesive types and failure modes: interfacial (a.k.a. adhesion) and mixed interfacial/cohesive failure + composite failure

– investigate combined loading (simultaneous effects of temperature, humidity, cyclic loading) for range of bondline thickness and mode mix ratio

– establish mixed mode fracture criteria that accounts for bondline thickness– integrate aspects of individual crack growth models into cohesive zone approach– development of improved test specimen for constitutive curve measurement– account for localized failure evolution in modeling of shear tests – demonstrate

transferability to joints of generic configuration– use the developed fracture models to find optimized adhesive thicknesses for

different adhesives– develop a embedded crack concept in conjunction with the developed fracture

models to predict general bonded joint strength

Documents

Damage Tolerance and Durability of Adhesively Bonded Composite Structures