Damage Tolerance and Durability of Adhesively Bonded Composite Structures

Hyonny Kim, Associate Professor, Dept. Structural Engineering, UC San Diego C.T. Sun, Professor, School of Aeronautics & Astronautics Thomas Siegmund, Associate Professor, School of Mechanical Engineering

The 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
Purdue University Joint Advanced Materials and Structures Center of Excellence

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

Purdue University Joint Advanced Materials and Structures Center of Excellence

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
Purdue University Joint Advanced Materials and Structures Center of Excellence

Project I. Adhesive Constitutive Behavior Measurement and Bondline Thickness Dependent Mixed Mode Fracture
Hyonny Kim, Associate Professor, UC San Diego, hyonny@ucsd.edu
Students: Jungmin Lee (PhD May 2006), Richard Khoo (MS Dec 2006), Hee Seok Roh

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

Purdue University Joint Advanced Materials and Structures Center of Excellence

Adhesive Constitutive Behavior in Bonded Joints

choice of constitutive curve is not clear adhesive vs 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 property should be geometry independent establish more direct and simple test method for determining constitutive behavior: bulk-adhesive tensile dogbone t.b.d. new method
5000

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

Average Shear Stress, (psi)

4000

3000

2000

8mil 20mil 40mil

1000

60mil

0 0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

Average Shear Strain,

Modified D5656 Test Specimen - less rotation - laser displ. measurement Purdue University Joint Advanced Materials and Structures Center of Excellence

Localized Damage Evolution Causing Apparent Softening Increasing Time 6

Bulk Adhesive Constitutive Behavior Tensile Dogbone Tests

PTM&W ES6292 epoxy paste adhesive 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

Gage Section
6000

5000

Constitutive Behavior is Strain Rate Dependent

4000

Stress, psi

3000

does not include effects of:

2000 1.5 in/min 0.3 in/min 1000 0.05 in/min

0 0.000

0.005

0.010

0.015

0.020

0.025

adherend constraint on adhesive layer possible material micro-structural differences between thin adhesive layer vs. thick bulk
7

Strain Purdue University Joint Advanced Materials and Structures Center of Excellence

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
Long-Distance Microscope Test Specimen Laser Extensometer

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

Matrix of Completed Tests (all tests at RT ambient): Mode Mix (% mode II) 0 50 75 100 ta = 0.008 in. 4 3 3 4 ta = 0.020 in. 5 3 3 7 ta = 0.040 in. 6 3 3 4 ta = 0.060 in. 4 5 3 6 8

Purdue University Joint Advanced Materials and Structures Center of Excellence

Results Mixed Mode GC Envelope

25.00 8mil 20mil 40mil 60mil

FEA Plastic Zone Predictions Mode I Large Shear Strain Visible

Adherend

MMR 100%

20.00
Crack Growth Process Observation by LD Microscope Adhesive

Small Shear Strain Visible

MMR 50% Adherend Plastic Zone (PZ) Contours at Growth Initiation:

15.00 Gc (lb/in)

10.00

No Shear Strain Visible

40 mil. PZ Moderately Constrained

8 mil. PZ Highly Constrained

MMR 0%

5.00

0.00 0

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

20

40

60

80

100

Mixed-Mode Ratio (%) 60 mil. PZ Unconstrained 9 Purdue University Joint Advanced Materials and Structures Center of Excellence

Summary: Comparison of Shear Strength Test and Fracture Properties

4

Parameter Normalized by Value at ta = 0.008 in.

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

Yield Fail
GIC 3 GC 50% Mode II GC 75% Mode II GIIC

Fracture Properties

StrengthTest Properties
0 0 0.01 0.02 0.03 0.04 0.05 0.06

Bondline Thickness ta (in.) Purdue University Joint Advanced Materials and Structures Center of Excellence

10

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 shear strength 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

Purdue University Joint Advanced Materials and Structures Center of Excellence

11

Project II: Modeling Thickness Effect on Strength

of Adhesive Lap Joint Using CTOA

C.T. Sun, Professor sun@purdue.edu, 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
Purdue University Joint Advanced Materials and Structures Center of Excellence

12

Adhesive Thickness Effect on the Strength Lap Joints

L T T
t

adhesive

Adherend: Aluminum Alloy 7075 Adhesive: HYSOL EA9394

Strength of Joints (kN)

L=3in, l=1in, T=0.125in t=0.008in, 0.01in, 0.02in, 0.06in Experimental result

7 6 5 4 3 2 1 0 0 0.5 1 1.5

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

Joint strength increases as the bondline thickness decreases up to 0.25 mm

Bondline Thickness (mm) Purdue University Joint Advanced Materials and Structures Center of Excellence

13

Fracture Initiation is Mode I Dominant in Lap Joints

B

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)

Stress Decreasing
A

Initial crack mode mixity

4 3.5 Energy Release Rate (J/m) 3 2.5 2 1.5 1 0.5 0 0 0.1 0.2 0.3 0.4 0.5 0.6 Adhesive thickness (mm) GI GII

Thick Layer

Thin Layer

Thin Layer

Thick Layer

Purdue University Joint Advanced Materials and Structures Center of Excellence

14

DCB Test Results

failure modes transition from mode I fracture to interfacial failure as adhesive thickness decreases below a certain level

140 120 100 Load (N) 80 60

Load vs Adhesive Thickness

140 130 120 Failure Load (N) 110 100 90 80 70 60 50 0 0.5 1 1.5 2 2.5 3 3.5 Bondline thickness (mm)
Load (N)

0.95mm
40 20 0 0 2 4 6 8 10 Opening End Displacement (mm)

1.3mm 3.3mm

160 140 120 100 80 60 40 20 0 0 5 10 15 Opening End Displacement (mm)

Sudden Failure

0.4mm 0.44mm 0.85mm

Purdue University Joint Advanced Materials and Structures Center of Excellence

15

Effect of Adhesive Thickness on Failure Mode

Mode I crack propagates in thicker adhesive Transition of failure mode in thinner adhesive

Normal stress at the interface

140 Interfacial Normal Stresses (MPa) 120 100 80 60 40 20 0 0 20 40 60 Applied Load (N) 80 100 120

s22-6mil s22-20mil s22-60mil

Maximum Normal Stress

Purdue University Joint Advanced Materials and Structures Center of Excellence

16

CTOA Criterion for Hysol EA9394

CTOA is independent of adhesive thickness before failure mode change
8 7 CTOA (Degree) 6 5 4 3 2 1 0 0 0.5 1 Adhesive Thickness (mm) 1.5 2

Purdue University Joint Advanced Materials and Structures Center of Excellence

17

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

Purdue University Joint Advanced Materials and Structures Center of Excellence

18

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

Thomas Siegmund, Associate Professor, siegmund@purdue.edu Steffen Brinckmann, Post Doctoral Research Associate Jibin 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
Purdue University Joint Advanced Materials and Structures Center of Excellence

19

Experimental Facilities

Crack Growth Resistance

Speckle Images Force Displacement Record

Environmental Degradation

Force Displacement Record

Displacements and Strain fields

Finite Element Method with Cohesive Zone

Digital Image Correlation

0.021

Stereo-Fractography
MeX
yy

a0

1 mm

-0.0035

Purdue University Joint Advanced Materials and Structures Center of Excellence

20

Computational Modeling
The Cohesive Zone Model:
Describes local energy dissipation during fracture and fatigue Is conveniently coupled to other fields (plasticity, moisture, heat, electrical) Adherent
Adhesive

COD H2O F Global Parameters: Load,(F) Displacement (COD) Force Displacement Environment (H2O) Environment Time Cycles

CZ Elements Diffusion Elements

T T Local Parameters: Traction (T) Separation () Traction-Separation Concentration H2O Concentration C(H2O)
Damage

Finite element model with element model with Cohesive elements & H O transport cohesive elements, moisture 2 transport, and cyclic damage

C(H2O)

Purdue University Joint Advanced Materials and Structures Center of Excellence

21

Monotonic Loading
Digital Elevation Maps
F

Stereo Images Pair Left -- Right

Tn ( ) =

G (CTOD )
0.508mm 1.524mm 3.048mm cz law
150 100 200

100 90 80 70

Fracture Profiles
Some plasticity
z-value (m)
50

Fracture Profiles
400 300

Tn [MPa]

60 50 40 30

200
0

20 10 0 0 0.02 0.04 0.06

-50 -100

100

0
-150

max , ,

n [mm]

-200

-100
-250

-300 250

-200
350 450 550 650 750 850 950

100

200

300

400

500

600

700

800

900

Point Number

Purdue University Joint Advanced Materials and Structures Center of Excellence

22

Fatigue Loading

100000

max = max,0 (1 D), ,

da 0.85 = 245 ( G ) dN

Fatigue Crack Propogation Rate, da/dN (m/cycle)

10000

1000

100

Gmax = GC
FE-CZM Experiment
0.1 1 10 100 1000 2 10000

10

1 0.01

Strain Energy Release Rate Range, G [J/m ]

200 150

100

Fractography
0 500 1000 1500 2000 2500

50 z-value (m)

-50

-100

-150

-200

Purdue University Joint Advanced Materials and Structures Center of Excellence

distance along path (m)

23

Time Dependence
Wedge test with constant loading
Precrack Stable Unstable

2500

Crack Extension(micrometers)

2000

1500

1000

500

0 0 2 4 6 8 10 12

Time at Load (min)

Purdue University Joint Advanced Materials and Structures Center of Excellence

24

Moisture Effects on Joint Fracture

Adhesive

Cohesive Experiment Simulation

Purdue University Joint Advanced Materials and Structures Center of Excellence

25

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

Purdue University Joint Advanced Materials and Structures Center of Excellence

26

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
27

Purdue University Joint Advanced Materials and Structures Center of Excellence

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

Purdue University Joint Advanced Materials and Structures Center of Excellence

28