Adhesive

Analysis of Kissing Bonds in
Metallic Joints
Felicity Guild* and Choothum Jeenjitkaew†
School of Engineering and Materials Science
Queen Mary, University of London
Department of Materials
Mile End Road
London E1 4NS
*Now
†
at: Department of Mechanical Engineering
Imperial College London
London SW7 2AZ
Now at: Element Hitchin [formerly: Materials Engineering Research Laboratory (MERL)]
Wilbury Way,
Hitchin, Hertfordshire, UK, SG4 0TW
Our Objectives
1.
Produce reliable and repeatable kissing bonds
2.
Investigate changes in their surface chemistry and
morphology
* Zofia Luklinska
3.
Establish the effects of kissing bonds on joint strength
4.
Correlate experimental measurements in terms of bond
strength and local strains for kissing bonds with
numerical predictions
5.
Investigate means of detection of kissing bonds
*C.Jeenjitkaew, Z. Luklinska and F.J. Guild, Morphology and surface chemistry of kissing bonds in
adhesive joints produced by surface contamination, IJAA, 30 (2010) 643-653
C.Jeenjitkaew, Z. Luklinska and F.J. Guild, Morphology and surface chemistry of kissing bonds in
adhesive joints produced by using ElectRelease™ adhesive, J. Adhes. 87 (2011) 291-312
This Presentation
Production of reliable and repeatable kissing bonds
(1) Surface contamination
(2) ElectRelease™ adhesive
Failure mechanisms (Experimental and FEA)
Failure strength
Mode of failure
Local strains
Future detection of kissing bonds?
I Producing reliable and repeatable kissing bonds
(1) Surface contamination
Hardened steel (HDS) - solvent degreasing+SiC papers+solvent degreasing
Al 2014 T6 - CAE (DEF standard)
High temperature cure adhesive - Redux® 319 (Modified epoxy film adhesive)
Room temperature cure adhesive - E3348 (2 part epoxy adhesive)
Set
1
2
3
Al
2014
T6
Al
2014
T6
Al
2014
T6
Al
2014T
6
Al
2014
T6
Al
2014
T6
Al
2014
T6
Adherend
HDS
HDS
HDS
HDS
Al
2014
T6
Adhesive
Redux
319
Redux
319
Redux
319
Redux
319
Redux
319
Redux
319
Redux
319
Redux
319
E3348
E3348
E3348
E3348
Contami
nants
-
PTFE
film
PTFE
spray
Frekote
-
Frekote
Sweat
Cutting
Oil
-
Frekote
Sweat
Cutting
oil
DLJ specimens – contaminated surface in the middle of the joint (25% of effective bonded area)
I Producing reliable and repeatable kissing bonds
(2) Using ElectRelease™ adhesive
HDS - solvent degreasing+SiC papers+solvent degreasing
Electrically debonding adhesive – ElectRelease™ (2 part amine cured epoxy adhesive)
Set
A
Adherend
HDS
HDS
Adhesive
ElectRelease™
ElectRelease™
Electric field
-
10 V DC
3.0
Current (mA)
2.5
+
2.0
1.5
Power
10 V
1.0
Upper adhesive
layer
+
A
+
A
0.5
0.0
0
5
10
15
20
25
Weakened interface
Time (minutes)
DLJ specimens – weakened interface by applying 10 V DC for 25 mins
Double-lap joint
Load end
Clamped end
strain gauges
spacer
spacer
2
1.6
adhesive
3.2
0.3
1
contaminant
11
12.7
12.7
3
3
6.35
31.8
12.7
203.3
All dimensions are in mm and not to scale
25.4
Surface Chemistry : SEM with EDS
• Surface contamination - (HDS+Redux®319)
Control
Frekote
PTFE spray
B
Steel
PTFE spray lubricant
PTFE spray lubricant
Steel
Redux®319
Redux®319
A
AB = contaminated interface
• Migration of PTFE spray is evident
• Similar migration observed for sweat
• Cutting oil appears to be absorbed by adhesive
• Frekote tends to remain at /near the interface
Comparison of Failure Loads
Contaminant
•
•
Effect of Carrier
Effect of Adhesive
Redux® 319
•
•
Effect of PTFE spray/sweat/cutting oil
Effect of PTFE film/Frekote
Frekote
HDS/
Redux 319A
(carrier)
HDS/
Al2014T6/
Redux 319 Redux 319
(no carrier)
Al2014T6/
E3348
Failure mechanisms
Contaminant
(HDS+Redux®319)
Control
PTFE film
Frekote
Inner adherend
Inner adherend
Outer adherend
Outer adherend
Inner adherend
Outer adherend
10 mm
Cohesive failure
Adhesive failure at the
position of PTFE film
and mixed mode at
noncontaminated
area
Adhesive failure at the
position of Frekote
and mixed mode at
noncontaminated
area
Failure of ElectRelease® Bonds
With electric
field
Without electric
field
Cathodic
adherened
Inner
adherend
Anodic
adherends
Outer
adherends
10
mm
10
mm
Cohesive failure
dominates
Value of average failure load reduced by 57%
Adhesive failure at anodic
surfaces
Change in Failure mechanism
Finite Element Model
Material Models
• Adherends simulated as elastic/plastic materials
assuming von Mises yield criterion
– Failure occurred in elastic region for HDS but AL2014 T6 within
plastic region
• Simulation of adhesives more complex
– Isotropic
– Shear behaviour not accurately derived from tension behaviour
(as assumed for von Mises yield criterion)
• Use Exponent Druker-Prager model
– Successful model derived for Redux®319
– Model for ElectRelease® failed to converge – used elastic/plastic
• FEA models for ElectRelease® may not be fully accurate
Finite Element Mesh
Modelling of Contaminant
strain gauges
spacer
spacer
2
1.6
adhesive
3.2
Line of symmetry
0.3
1
contaminant
Frekote modelled as uncoupled surfaces
11
12.7
12.7
3
3
25.4
Remaining adhesive interface modelled
using zero thickness cohesive elements
6.35
31.8
12.7
203.3
All dimensions are in mm and not to scale
Properties of cohesive zone gained from material tests
Fixed arm peel tests (Mode I)
Four point bend end notch flexure test – 4ENF (Mode II)
Assumed Mode II and Mode III parameters identical
The cohesive properties are derived from independent material tests
Modelling of ElectRelease®
strain gauges
spacer
spacer
2
1.6
adhesive
3.2
Line of symmetry
0.3
1
Weakened
interface
11
12.7
3
25.4
3
31.8
Interface modelled using zero thickness
cohesive elements
12.7
203.3
All dimensions are in mm and not to scale
Properties of cohesive zone gained from material tests
Measured before and after application of the current
Fixed arm peel tests (Mode I)
Four point bend end notch flexure test – 4ENF (Mode II)
Assumed Mode II and Mode III parameters identical
The cohesive properties are derived from independent material tests
Comparison of Experimental and
Predicted Values of Failure Load
HDS/Al2014 T6 + Redux® 319
40
Failure Load (KN)
35
30
25
20
Experimental
FEA
15
10
5
0
HDS
HDS
Al
Al
Control Frekote Control Frekote
Comparison of Experimental and
Predicted Values of Failure Load
HDS/ElectRelease™
16
Failure Load (KN)
14
12
10
Experimental
FEA
8
6
4
2
0
Control
After current
Comparison of Values of Local strain
• Surface contamination:
HDS+Redux®319
1 mm in the middle of bonded area
25 mm over bonded area
Extensometer
Strain gauge
25 mm
40
35
35
30
30
25
20
Control_EXP
15
Control_FEA
Load (kN)
Load (kN)
40
25
20
Control_EXP
15
Control_FEA
10
Frekote_EXP
10
Frekote_EXP
5
Frekote_FEA
5
Frekote_FEA
0
0
0
0.0002 0.0004 0.0006 0.0008
0.001
0.0012
Local strain
Good agreement between EXP and FEA
for control and contaminated DLJ.
0
0.001
0.002
0.003
0.004
0.005
Local strain
Good agreement between EXP and FEA
for both control and contaminated DLJ
Comparison of Values of Local strain
• Surface contamination:
Al2014 T6+Redux®319
1 mm in the middle of bonded area
25 mm over bonded area
Extensometer
Strain gauge
25 mm
40
40
35
35
30
Load (kN)
20
Control_EXP
15
Control_FEA
Load (kN)
30
25
25
20
Control_EXP
15
Control_FEA
10
Frekote_EXP
10
Frekote_EXP
5
Frekote_FEA
5
Frekote_FEA
0
0
0
0.0005
0.001
0.0015
Local strain
0.002
0.0025
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
Local strain
Good agreement between EXP and FEA
Good agreement between EXP and FEA
for control DLJ but less good for contaminated DLJ for both control and contaminated DLJ
Comparison of Values of Local strain
• Using ElectRelease™ adhesive:
HDS+ElectRelease™
1 mm in the middle of bonded area
Strain gauge
25 mm
14
12
Load (kN)
10
8
6
Without current_EXP
4
Without current_FEA
2
With current_EXP
With current_FEA
0
0
0.0001
0.0002
0.0003
Local strain
Good agreement between EXP and FEA for control and contaminated DLJ at 1mm gauge length.
Same local stiffness before and after current.
Comparison of DIC Results
Axial (applied) Strain
Control Joint
Contaminated Joint
load
Failure mechanism: Possible Future detection of kissing bonds?
Defect
LE11
0.001
0.0009
0.0008
0.0007
0.0006
0.0005
0.0004
0.0003
0.0002
0.0001
0
-0.0001 0
Applied strain (LE11) at approximately 22-24% of
failure load
Noncontaminated
model
Contaminated model
2
4
6
8
10
12
14
Distance (mm)
Strain profiles across top plane of
adherend above the kissing bond
Lateral Strain (LE22) at approximately 22-24% of
failure load
0
0
2
4
6
8
10
12
14
-0.00005
-0.0001
Presence of kissing bond changes
the strain profiles
LE22
Could this be used as a means of
detection?
-0.00015
-0.0002
Noncontaminated
model
Contaminated model
-0.00025
Distance (mm)
Conclusions
We produced reliable and repeatable kissing bonds
Contamination by Frekote
ElectRelease™ adhesive
We established the changes in surface chemistry and morphology at the
interface/interphase for kissing bonds
We measured significant changes in joint strength and adhesive failure at the interface
for kissing bonds
We successfully modelled kissing bonds using finite element analysis and correlated
the reduction in joint strength with the change in adhesive strength at the interface
We propose that a future method of kissing bond detection could be based on
measurement of strain in the adherends, particularly lateral strain