IAG Meeting

MMS8

Interfacial Adhesion Strength

TWI 30th April 2003

Bruce Duncan, NPL Materials Centre

MMS8: Interfacial Adhesion StrengthT1: Adhesion Strengthtesting and design

T2: Surface Characterisation

design methods adhesion tests

keyproperties

characterisationmethods

T3: Correlation between surface characteristics and adhesion

T4: Inclusion of adhesion in design methods manufacture process

development

durability

Industrial Validation Exercises

T5: Dissemination

Task 1: Adhesion Strength Tests

1.Review of adhesion test methods (Dec 01)2.Evaluation of existing tests (Feb 02)3.Development of adhesion test method (Sept 02)4.Evaluation of test method (March 03)

1. Measurement note completed.5.Report on adhesion test (June 03)

Task 2: Surface Characterisation

6.Review of simple surface characterisation techniques (Jan 02)

7.Sample preparation (June 02)8.Develop and evaluate techniques for surface

characterisation (Sept 02)9.Report on surface characterisation methods (Apr 03)

Task 3: Correlation between surface characteristics and adhesion strength

10. Establish experiments and test specimens (Jun 03)11. Surface properties and durability (Feb 04)12. Measurement good practice guide (Jun 04)

NPL – Move to New Laboratories, Summer 2003

Old NPL

New NPL

June/July 2003Decant to new laboratorieSome disruption to work

New NPL

Interfacial Adhesion

• Surface Characterisation• Adhesion Tests

• Pull-Off• Butt Joint• Pull-Out• Bend Tests

• Conclusions

• Review - NPL Report MATC(A)66• Adhesion (mechanical and intrinsic)

• Surface treatments

• Surface characterisation

• Measurement Note MATC(MN)24• Metallic Surfaces

• Measurement Note MATC(MN)25• Polymer Surfaces

Surface Characterisation MethodsSurface Characterisation Methods

Chemical CharacteristicsChemical Characteristics

Electron, Photon and Ion Electron, Photon and Ion SpectroscopiesSpectroscopies•• Auger, XPS, EELS - electron•• EDX, XRF, IR, Raman, (XRD) - photon•• SIMS, RBS - ion

Surface Analysis TechniquesSurface Analysis Techniques

Physical CharacteristicsPhysical Characteristics

Strength, Shape and SizeStrength, Shape and Size•• hardness - indentation•• roughness – gloss, profilometry, AFM, SEM

•• oxide layer thickness – ellipsometry•• oxide layer stability - scratch•• surface energy – contact angle•• adhesion – tensile, flexure

Bulk and Local CompositionBulk and Local Composition•• constituent elements•• constituent compounds•• functional groups•• distribution

SUMMARYSUMMARY

• Materials / case studies

• Anodised aluminium alloy (Alcan)

• Hot dipped galvanised steel (Corus) – oil lubricants

• Corona discharge treated glass-fibre/polypropylene

• Grit blasted and chromic acid etched Al alloy

• Surface characterisation

• Statistical analysis (where valuable)• NPL Measurement Note MATC(MN)24

Evaluation of Surface Characterisation MethodsEvaluation of Surface Characterisation Methods

• Surface treatments

• Mill-finish (a non-treated surface)

• Light clean (lightly degreased to remove oils - most of the surface

structure and oxides still remaining)

• Full clean (an extensive electrolytic acid etch - equivalent of the

optimised system without an anodised layer)

• Full clean + 25 nm barrier film ("optimised" system with a full etch

and a 25 nm barrier anodised film)

• Light clean + 25 nm barrier film (a variant of (4) to give a lower level

of etch)

• Full clean + over-anodised - (25 nm barrier + 75 nm anodised film)

Anodised AA5754Anodised AA5754--O Aluminium AlloyO Aluminium Alloy((AlcanAlcan Banbury)Banbury)

Anodised Aluminium: Nominally 25 nm barrier 75 nm filaments

20nm

Approx. 20 nm barrier

Sputtered platinum

70 nm filaments

©© AlcanAlcan International LtdInternational Ltd

Surface characterisation techniques assessed include:Contact and non-contact profilometry and AFM (surface roughness)

Contact angle (wettability)

Micro-hardness

Gloss/reflectivity

Colorimetry

Ellipsometry

Surface resistivity

Optical stimulated electron emission (OSEE)

Surface Characterisation TechniquesSurface Characterisation Techniques

AFM Surface Map Of AFM Surface Map Of Optimised Anodised Optimised Anodised

TreatmentTreatment

GritGrit--Blasted 5251 Aluminium AlloyBlasted 5251 Aluminium Alloy

Clear differences in Ra and gloss for treated and untreated surfaces

No evidence of intermediate state

Ra and gloss remain constant with increasing level of treatment

Either method could be used to check for untreated regions

Inspection Angle Treatment Time (secs) 20° 60°

0 (untreated) 10 30 60 120

121 ± 6 0.8 ± 0.1 0.8 ± 0.1 0.7 ± 0.1 0.7 ± 0.1

> 200 2.9 ± 0.1 3.1 ± 0.2 3.0 ± 0.1 2.9 ± 0.1

Treatment Time (secs)

Ra

(µm) Profilometry 0 (untreated)

10 30 60 120

0.24 1.83 1.79 1.98 2.08

Roughness Gloss

Chemical Etch Treatments - Key Parameters

• Adherend type • Initial surface treatment • Solution type, pH and concentration of chemical agents • Minimum volume required to treat a surface • Solution temperature • Treatment time • Method of rinsing • Method of drying • Exposure time

Surface Characteristics • Surface roughness / texture • Oxide layer morphology and depth • Surface composition / reactive groups / crystallinity / contamination • Surface energy / tension • Surface hardness / cohesion strength • Surface reflectivity / gloss • Surface resistivity

Vickers MicroVickers Micro--HardnessHardness

Non-significant differences between treated and untreated surface

Non-significant differences between level of chromic acid etching

Chromic Acid Etched (CAE) 5251 Aluminium Alloy

60

62

64

66

68

70

72

0 mins 10 mins 60 mins

Treatment time

Vick

ers

mic

ro-h

ardn

ess

num

ber

Gloss/Colour MeasurementsGloss/Colour MeasurementsCAE 5251 Aluminium AlloyCAE 5251 Aluminium Alloy

Clear differences for treated and untreated surfaces

Clear differences between levels of treatment

Minimal difference between GB only and GB + CAE

350 450 550 650 75050

60

70

80 GB+CAE (60 mins)

GB+CAE (30 mins)

GB+CAE (10 mins)

CAE only (10 mins)

CAE only (30 mins)

Ref

lect

ance

(%)

Wavelength (nm)

Inspection Angle Treatment Time (mins) 20° 60° 0 (untreated)

10 20 30 60

121 ± 6 35.7 ± 1.1 30.6 ± 1.2 15.6 ± 0.3 9.3 ± 0.9

> 200 133 ± 3 102 ± 1 76 ± 1 50 ± 2

Dynamic Contact Angle Dynamic Contact Angle -- ((WilhelmyWilhelmy Plate Method)Plate Method)

speed controlled movable stage

sensitive microbalance

solid plate sample

liquid

draught exclusion case

vibration damping base

Dynamic Method θadv θrec θadv - θrec Mill finish Light clean Full clean

Full clean + 25 nm barrier Light clean + 25 nm barrier

Full clean + 25 nm barrier + 75 nm porous

89.5 / 91.9 84.5 / 84.8 81.6 / 86.1 74.2 / 80.4 53.2 / 65.7 85.5 / 89.1

51.3 / 27.0 0.0 / 0.0 0.0 / 0.0

0.0 / 33.6 36.0 / 28.5

0.0 / 0.0

38.2 / 64.9 84.5 / 84.8 81.6 / 86.1 74.2 / 47.8 17.2 / 37.2 85.5 / 89.1

Sessile Method θstatic Mill finish Light clean Full clean

Full clean + 25 nm barrier Light clean + 25 nm barrier

Full clean + 25 nm barrier + 75 nm porous

96 0 ± 6.7 / 97.0 ± 4.1 98 3 ± 7.3 / 97.2 ± 4.2 53 7 ± 4.3 / 76.8 ± 3.5 32 4 ± 6.2 / 42.1 ± 1.6 83 3 ± 4.2 / 74.1 ± 1.7 85 4 ± 2.8 / 86.7 ± 4.1

• Difference between mill finish and other treatments

• Difference between full clean with and without barrier layer

• Difference between full and light clean surfaces with barrier

layer

• No differences between light clean, full clean and the over-

anodised treatments

Results were inconclusive

Possible to differentiate between a mill finish and surfaces without

a barrier or porous layer

Over anodised surface – highest resistance (1 x 108 ohms)

Light and full clean surfaces - lowest resistance (2-6 x 105 ohms)

Surface Resistance MeasurementsSurface Resistance MeasurementsAnodised Aluminium AlloyAnodised Aluminium Alloy

(Industrial Development Bangor Ltd)(Industrial Development Bangor Ltd)Specimen Surface Resistance

(ohms) Mill finish Poor clean Full clean Full clean + 25 nm barrier Poor clean + 25 nm barrier Full clean + 25 nm barrier + 75 nm porous

5 x 107

2 x 105

6 x 105

6 x 105

1 x 107

1 x 108

Colorimetry showed most promise

Relatively easy to use, non-subjective and suitable for laboratory

trials and on-line inspection

Suitability of Characterisation TechniquesSuitability of Characterisation TechniquesWith Surface TreatmentsWith Surface Treatments

Technique Grit-Blasted CAE Anodised Oil Lubricated

Corona Discharge

Micro-hardness No No No No No Contact angle Yes Yes Yes No Yes Profilometry Contact Laser AFM

Yes Yes Yes

Yes Yes Yes

Yes Yes Yes

No No No

No No No

Gloss/reflectivity Yes Yes possibly No No Colorimetry Yes Yes Yes Yes No Ellipsometry No Yes Yes No No OSEE possibly possibly Yes No possibly Surface Resistivity No No No No No

Design of Bonded Structures

• Finite Element Methods• Joint Stiffness (N/mm)

• Predict from Elastic-Plastic behaviour

• Joint Strength• Analysis of stress• Failure criteria for

materials• Strength of interfaces

Lap joint

r = 1.0r = 0.5

52.5 mm

15 mm

10 mm

3 mm0.5 mm

3 mm

p

r = 1.0r = 0.5

52.5 mm

15 mm

10 mm

3 mm0.5 mm

r = 1.0r = 0.5

52.5 mm

15 mm

10 mm

3 mm0.5 mm

r = 1.0r = 1.0r = 0.5r = 0.5

52.5 mm52.5 mm

15 mm15 mm

10 mm10 mm

3 mm3 mm0.5 mm0.5 mm0.5 mm

3 mm3 mm

pppp

Adhesion TestsNeed

• Uniform, predictable stress distribution• Failure at the interface being studied• Straightforward specimen preparation• Robust test procedures • Thin sheet or thick section?• Data suitable for design calculations• Consider different properties of near surface layers?

Overlap Tests – Lap ShearLocation of failure initiation depends on shape of fillet and adherend

Possibility of altering surface being ‘tested’Mix thick and thin sections

Problems with interpreting stress/strains at corner ‘singularities’ – FEA or closed form models?

bulk adhesive

bulk substrate

Overlap Tests - Interpretation

Complex stress state at failure initiationIs ‘shear’ a critical stress state for adhesion failure?Properties and geometry local to the interface?

RoughnessShapeMaterial properties

interphase

Direct Pull Tests – Rigid Substrates

Effects of test variables:Alignment Adherend profilesFillet shapes Interpretation of results?

Cheap, simple and quick

Pull-Off Test Results – Different Al Surfaces

• Differences apparent between ‘good’ and ‘poor’ surfaces• Failure stress levels considerably lower than bulk tensile strength of adhesive (58 MPa)• Limited alignment control – ‘cleavage forces’

Alignment is critical• Control of specimen during

bonding• Axiality of load application is

controlled through precision alignment fixtures

• Cleavage forces can be eliminated

Tensile Butt JointSevere test – adhesive experiences high levels of tensile and hydrostatic stress

Buttjoint test

Three extensometer arrangement to detect bending

0

2

4

6

8

10

12

14

0 10 20 30 40

Displacement (µm)

Loa

d (k

N)

HKT 1006 - 021203BCLVDT 1LVDT 2LVDT 3

Interpretation of Butt Joint FailuresFEA shows that in a buttjoint with parallel, flat faces the adhesive layer experiences non-uniformity of the stress distributionPeak stress is considerably higher than ‘average’ stress

Improved Specimen – Uniform Stress DistributionButt Joint with ‘ball and cup’ radius

20 mm diameter buttjoints50 mm radius on facesAssume perfect bond between adhesive and adherendsMaximum stress only a few % greater than average

Butt joint – test and FEA results

0

5

10

15

20

0 5 10 15 20 25 30 35 40displacement (um)

reac

tion

forc

e (k

N)

000120AE000118AK990920BC990920BD020404BA020404BB020404BC020404BD020404BG020404BH020510EAFEFE including cteSeries5

data from buttjoints with original adherends (thin lines) compared to ball & cup adherends (bold lines)FEA predictions shown (symbols)

Butt Joint Test

• Results comparable to non-profiled specimen• Tough adhesive – yield stress provides limiting load

• Alignment critical to good quality results• Uniform stress distribution

• Failure cohesive or adhesive?• Minimum adhesive strength• Failure stress values comparable with bulk tensile

strengths

Pullout TestCommonly used for characterisation of matrix-resin interfaces in composites

Average interfacial shear strength calculated as ratio of pullout load divided by fibre surface

Calculation implies that shear stresses are the prevailing loading of the interface stresses are distributed almost uniformly along fibre

Use analytical & FE methods to assess stress distributions

Pull-out TestBased on fibre pull-out test for compositesInvestigate a modified test

using thin strips embedded in the adhesive

0

500

1000

1500

2000

2500

0 100 200 300 400

Time at constant pull speed (s)Lo

ad (N

)

mild steel shim

shim embedded15 mm

shim only

pull-out specimen

Fibre pullout – analytical calculations

Tensile stress in fibre is:

Interfacial shear stress distribution along fibre is:

/r)sinh(nL/r]) - (Lsinh[n

σ σe

e fef

x=

( )( )/rnL2sinh

/r]Lcosh[n nσ τ

e

efei

x−=

0

5

10

15

20

25

30

35

40

0 0.2 0.4 0.6 0.8 1

normalised lengthsh

ear s

tress

Le = 15 mm, F = 1856NLe = 10 mm, F = 1767 NLe = 5 mm, F = 1727N

( )

+

=

rRlnν1E

E n

mf

m2

Fibre pullout –FE analysis

Contour plots at Fkink= 1300N

Tensile stress (S11)

Shear stress (S12)

0

2

4

6

8

10

12

0 0.2 0.4 0.6 0.8 1 1.2normalised distance

shea

r str

ess

(MPa

)

3 mm by 3 mm5 mm by 1 mm

Fibre Pull-Out FE Results - IFSSMild steel/XD4601 fibre pull out

0

10

20

30

40

50

60

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1normalised distance

(0 = fibre end, 1 = exit point)

shea

r str

ess

(MPa

)

5mm embedded length, F=1727N10mm embedded length, F=1767N15mm embedded length, F=1856N

Pull-out results 2-part epoxy

0

20

40

60

80

100

120

140

160

180

200

0 20 40 60 80 100 120 140Time (s)

Load

/leng

th(N

/mm

)

HKT200HKT201HKT202HKT203HKT205 PHKT206 P

Post-curedCohesive failure

RT-curedAdhesive failure

Flexure – ISO 14679

Tested in 3-point bend

Investigate stiffening the joint with a backing

Bend Adhesion Test

• Simple and quick to perform

• Open face allows for ‘accelerated’ ageing.

• Analytical calculation of interfacial shear strength

−

+

−=

0

11

0

21

21

12

2)

GG

(G2b

)h-h b(FEτGGhh

2)hh b(E

2hbEG

21

22

2111

1−

+=

)hh b(EhbEG 121110 −+=

3)hh b(E

3hbEG

31

32

3111

2−

+=

Bend tests – 2-part Epoxy3- point bending tests 3M 5027

0

20

40

60

80

100

120

140

160

0 0.5 1 1.5 2 2.5 3 3.5

Displacement (mm)

Load

(N)

HKT103

HKT116

HKT104

HKT110

HKT115

HKT106

HKT112

HKT114

HKT108

HKT109

HKT113

HKT111

HKT117

post-curedcohesive failureRT-cured

delamination

suspect cartridgedelamination

Results for Oiled Surfaces3M 5027 on Oiled Galvanised Steel

0

0.5

1

1.5

2

2.5

3

3.5

4

Clean Oiled1 Oiled2

surface treatment

shea

r str

engt

h (M

Pa)

RT-cured adhesivePost-cured adhesive

3-point bend results

Adhesion to different surfaces – Treated Aluminium

3M 5027

0123456789

10

MF PC FC PC+25nm FC+25nm FC+25nm+75nm

Surface treatment

Shea

r str

ess

(MPa

)

Effect of surface energy on adhesion

Correlation of contact angle and bond strength

y = -0.0126x + 8.6376R2 = 0.2298

y = -0.0155x + 6.1379R2 = 0.5671

y = -0.0408x + 16.425R2 = 0.4971

0

2

4

6

8

10

12

14

16

18

20

0 20 40 60 80 100 120

sessile drop contact angle

bond

str

engt

h (M

Pa) XD4601, 3PB

3M5027, 3PB

XD4601, pull-off

Linear (3M5027, 3PB)

Linear (XD4601, 3PB)

contact angle measured on original adherends

ColorimetryColorimetry MeasurementsMeasurementsAnodised Aluminium AlloyAnodised Aluminium Alloy

350 450 550 650 7500

10

20

30

40

50

60R

efle

ctan

ce (%

)

Wavelength (nm)

Milled finish Light clean Full clean Full clean + 25 nm barrier Light clean + 25 nm barrier Full clean + 25 nm barrier + 75 nm porous

Appearance and adhesion

Correlation of reflectance and bond strength

y = 0.0489x + 4.4124R2 = 0.7767

y = 0.0271x + 3.1502R2 = 0.3899

y = 0.1017x + 6.546R2 = 0.6982

0

2

4

6

8

10

12

14

16

18

20

30.0 35.0 40.0 45.0 50.0 55.0 60.0 65.0 70.0 75.0 80.0

avg reflectance

bond

str

engt

h (M

Pa)

XD4601, 3PB3M5027, 3PBXD4601, pull-offLinear (3M5027, 3PB)Linear (XD4601, 3PB)Linear (XD4601, pull-off)

Concluding Remarks

• Four Tests Studied• Pull-Off• Butt Tension• Pull-Out• Three point bend

• Distinguish ‘good’ from ‘bad’ surfaces• Methods for calculating interface stress

available• Accuracy in question – analytical vs. FEA

Ongoing Work• Durability and Adhesion – work started

• 6 aluminium surface treatments• XD4601 adhesive• Hot/wet conditioning – 36 days• 3-point bend and pull-off tests• Surface inspection by colorimetry, roughness and

contact angle

• Repeat with:• Oiled galvanised steel and 3M5027• Corona treated GRP and DP460

Thanks to

• Bill Broughton, Greg Dean, Louise Crocker, Elena Arranz, Jeannie Urquhart, Alan Pearce, Richard Mera

• Dow Automotive, 3M – adhesive supply• Alcan, Corus - metals• DTI – funding under the Measurements for Materials

Systems programme

Contact:

Bruce DuncanNPL Materials CentreNational Physical LaboratoryQueen’s Road, TeddingtonMiddlesex, UK, TW11 0LW

Tel: 020 8943 6795Fax: 020 8943 6046Email: bruce.duncan@npl.co.uk