Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
Z. JENDLIZ. JENDLI*, J. FITOUSSI*, F. MERAGHNI** et D. *, J. FITOUSSI*, F. MERAGHNI** et D. BAPTISTE*.BAPTISTE*.
*LM3 UMR CNRS 8006. ENSAM Paris.
**LMPF-JE 2381. ENSAM Châlons en Champagne.
Micromechanical analysis Micromechanical analysis of strain rate effect on damage of strain rate effect on damage
evolution evolution in discontinuous fibre reinforced in discontinuous fibre reinforced
compositescomposites
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
advantages advantages of SMCof SMC composite materialcomposite material
good strength/weight ratio manufacturing process devoted to large series high-energy dissipation with a diffuse damage.
Passengers safety
Structures lightweight
Manufacturing and productivity.
Interaction
Manufacturing ↔ mechanical behaviour
CONTEXTCONTEXT
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
Industrial requestsIndustrial requests
Overall mechanical behaviour prediction under dynamicOverall mechanical behaviour prediction under dynamic loading.loading.
Inaptitude of current dynamic behaviour laws for the structuresInaptitude of current dynamic behaviour laws for the structures calculation in composite materialscalculation in composite materials Insufficiency of the phenomenological approaches currently used.Insufficiency of the phenomenological approaches currently used.
Development of multi-scaleDevelopment of multi-scale approaches.approaches.
Material microstructure integrationMaterial microstructure integration
Physical description of damage mechanisms.Physical description of damage mechanisms.
CONTEXTCONTEXT
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
Behaviour stages
Damage threshold and accumulation
Experimental analysis (at micro and Experimental analysis (at micro and macroscopic scales) macroscopic scales)
AIMSAIMS
Visco-elastic
Visco-damage.
Behaviour modelling using a multi-scale approachBehaviour modelling using a multi-scale approach
Analysing effect in the Analysing effect in the SMC-R SMC-R damagedamage( ( ) )..
( = 10-4 – 400 s-1 )
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
Interrupted high-speed tensile tests.Interrupted high-speed tensile tests.
In-situIn-situ tensile tests. tensile tests.
SEM observations.SEM observations.
Damage mechanisms experimental investigationDamage mechanisms experimental investigation
MODELING COMPOSITEMODELING COMPOSITE DYNAMIC BEHAVIOURDYNAMIC BEHAVIOUR
Introduction of sIntroduction of strain rate effect in a micromechanical modeltrain rate effect in a micromechanical model
Prediction of the Prediction of the dynamicdynamic m mechanicalechanical behaviour and its interaction with behaviour and its interaction with
the composite microstructure.the composite microstructure.
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
MATERIALMATERIAL
SMC-R26 compositeSMC-R26 composite
Sheet Molding Compound-Random.glass E/polyester, discontinuous fibres 26%.
Hydraulic high speed tensile test Hydraulic high speed tensile test machinemachine
20 m/s, piezo-electric load cell 50 kN
Performed strain ratesPerformed strain rates
-1-1-4s400s 10
TESTS PARAMETERSTESTS PARAMETERS
EXPERIMENTAL EXPERIMENTAL INVESTIGATIONSINVESTIGATIONS
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
Strain rates effects on the Strain rates effects on the SMC-R mechanical SMC-R mechanical
characteristicscharacteristics
The composite macroscopic response is widely The composite macroscopic response is widely affected byaffected by
Minor variation of the anelastic slope as a function of the strain rateMinor variation of the anelastic slope as a function of the strain rate
Insensitivity of the Young’s modulus to strain rate increaseInsensitivity of the Young’s modulus to strain rate increase..
( ).
0 0.5 1 1.5 2 2.5
150 s-1124 s-123 s-110 s-10.022 s-1
0
40
80
120
160
(
MP
a)
(%)
8 103
9 103
1 104
1.1 104
1.2 104
1.3 104
1.4 104
1.5 104
10-2 10-1 100 101 102 103
E (
MP
a)
d/dt (s-1)
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
Strain rates effects on the Strain rates effects on the SMC-R mechanical SMC-R mechanical
characteristicscharacteristics
20
30
40
50
60
70
80
10-4 10-3 10-2 10-1 100 101 102 103
th
resh
old
(MPa
)d/dt (s-1)
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
10-4 10-3 10-2 10-1 100 101 102 103
d/dt (s-1)
th
resh
old
(%)
The first non-linearity The first non-linearity Damage Damage thresholdthreshold
is considerably delayed in term of strain and is considerably delayed in term of strain and stress. stress.
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
Strain rates effects on the Strain rates effects on the SMC-R mechanical SMC-R mechanical
characteristicscharacteristics
80
90
100
110
120
130
140
150
160
10-4 10-3 10-2 10-1 100 101 102 103
ul
timat
e (M
Pa)
d/dt (s-1)
0
0.5
1
1.5
2
2.5
10-4 10-3 10-2 10-1 100 101 102 103
d/dt (s-1)
ul
timat
e (M
Pa)
Behaviour accommodation when increases the Behaviour accommodation when increases the strain ratestrain rate :
Steady rise of the ultimate strain (38 %)
Maximum stress increases considerably.
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
Damage analysisDamage analysisINTERRUPTED DYNAMIC INTERRUPTED DYNAMIC
TENSILE TESTTENSILE TEST
Correspondence between ligament and
force
Ligament (mm)
Force (N)
7 822
10,5 1059
12 1089
13,5 1314
14,5 1511
15 1635
15,5 1721
16 1842
The specimen geometry is a bar with dimension 36*6,5*2,7 mm3
The fuses material: PMMA (Poly Methyl Methacrylate) Fragile elastic behavior.
Specimen/fuseIntermediate fixing
Ligament
SMC-R specimen
Mechanical fuse
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
Macroscopic damage analysis Macroscopic damage analysis
Macroscopic damage vs. total strain for three strain rate values01
E
ED
D
Damage initiation is considerably delayed in terms of strain thresholdsDamage initiation is considerably delayed in terms of strain thresholds
Critical damage level : insensitive to the strain rate effect.Critical damage level : insensitive to the strain rate effect.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0 0.5 1 1.5 2
20.5 s-18 s-13 s-1D
mac
rosc
opic
(%)
D_critical
0 0.5 1 1.5 2 2.50
20
40
60
80
100
120
Interrupted high-speed tensile tests
Strain (%)
Str
ess
(MP
a)
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
Strain rate Strain rate increase:increase:
Delayed damage Delayed damage thresholdthreshold
Decreased damage Decreased damage growth speed.growth speed.
Microscopic analysis results Microscopic analysis results corroboratecorroborate those those obtained at the macroscopic level.obtained at the macroscopic level.
Fibre-matrix interface damage descriptionFibre-matrix interface damage description
d_micro = fv_debonded / fv d_micro = fv_debonded / fv total. total.
Overall behaviour Overall behaviour accommodationaccommodation
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0 0.5 1 1.5 2
0.0002 s-18 s-120.5 s-1
d m
icro
scop
ic
(%)
_ult
Microscopic damage analysis Microscopic damage analysis
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
Global dGlobal damage growth in term of micro-cracks lengthamage growth in term of micro-cracks length(including matrix and interface damage)(including matrix and interface damage)
Delayed damage thresholdDelayed damage threshold Decreased damage growth speed.Decreased damage growth speed.
0 100
1 103
2 103
3 103
4 103
0.2 0.6 1 1.4 1.8
0.0002 s-18 s-120.5 s-1
L (
µm
)
(%)
Microscopic damage analysis Microscopic damage analysis
l1
l2
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
CONCLUSIONCONCLUSION
Strain rate effects :Strain rate effects :
Insensitivity of the material elastic properties.
Delayed damage threshold.
Decreased damage growth speed.
Accommodation of the overall behaviour leading
to an increase of the ultimate characteristics.
Integration of the experimental findings to set up physical Integration of the experimental findings to set up physical damage modelling.damage modelling.
Prediction of Prediction of elastic visco-damagedelastic visco-damaged behaviour. behaviour.
Viscosity effects Viscosity effects on the fibres-matrix on the fibres-matrix interfaces damage.interfaces damage.
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
THE ENDTHE END
☺☺
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
Experimental investigationExperimental investigationSpecimen optimised Specimen optimised geometry :geometry :
L1= 6 mm, L2 = 80 mm, L3 = 30 mm, R = 7 mm, e=3
mm
Evolution of the longitudinal stress for a test calculated at 200/sEvolution of the longitudinal stress for a test calculated at 200/s
homogeneoushomogeneous
constantconstant
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
Contrainte= f(Déformation) à 1m/s
0
20
40
60
80
100
120
0 0,5 1 1,5 2 2,5
déformation (%)
Cont
rain
te (M
Pa)
7mm
10,5mm
13,5mm
14,5mm
15,5mm
15mm
1ms à rupture
Experimental Experimental MethodologyMethodology
Analysis on the two scales of materialAnalysis on the two scales of material
Macroscopic analysisMacroscopic analysis Microscopic analysisMicroscopic analysis
Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003Composites Testing and Model Identification Châlons-en-Champagne, 28 January 2003
MULTI –SCALESMULTI –SCALESMODELLINGMODELLING
MatériauMatériauHomogène Homogène EquivalentEquivalent( )
.
MatrixMatrixPolymer or Polymer or
MetallicMetallic
FibresFibresArchitecture, Architecture,
Geometry, Geometry, quantity, ...quantity, ...
damagedamageMicro discontinuitiesMicro discontinuities
Mori Tanaka’s approachMori Tanaka’s approachEschelby inclusion theory Eschelby inclusion theory
Behaviour law Visco-elastic
.
PROCESSPROCESS Probabilistic approachesProbabilistic approachesWeibull, Monte Carlo, ...Weibull, Monte Carlo, ...
Local criterionLocal criterion
Pr =1 - exp[(0202m
(0 0 m) = f ( ).
Experimental investigationExperimental investigationInterrupted dynamic testsInterrupted dynamic testsSEM and ultrasonic testsSEM and ultrasonic tests
Damage growth D=f (Damage growth D=f (