Com

ptes

t_20

04_G

agel

_ 1

Andreas Gagel, Dirk Lange, and Karl Schulte

Technische Universität Hamburg-Harburg (TUHH), GermanyPolymers and Composites Section (5-09)

http://www.tuhh.de/kvweb

gagel@tuhh.de

On the Relation between Crack Densities, Stiffness Degradation,and Surface Temperature Distribution

of Tensile Fatigue Loaded Glass-Fibre Non-Crimp-Fabric Reinforced Epoxy

Comptest 2004, September 21st to 23rd, Bristol, GB, 2004

Title

Com

ptes

t_20

04_G

agel

_ 2

• Motivation

• Experimental- Quasi-static Tensile Test

- Fatigue Test

- Crack Density

- Mechanical Degradation

- Thermography

• Consideration of Relations- Load / Crack Density

- Crack Density / Degradation

- Temperature / Degradation

- Temperature / Failure

- Temperature / Crack Density

• Summary and Conclusion

Content

Com

ptes

t_20

04_G

agel

_ 3

• Large, lightweight, and economical structures with textile reinforced polymers

• High potential of non-crimp-fabric (NCF) reinforced polymers

• Complex damage and failure behaviour

• Understanding the relation between load, damage, degradation and failure

Motivation

Com

ptes

t_20

04_G

agel

_ 4

Mechanical TestingMaterial0°

45°90°

-45°

• E-glass fibre non-crimp fabric• Epoxy L135i/H137 for RTM at RT• [0°(48%),45°(23%),90°(6%),-45°(23%)]am

• Specimens DIN EN-ISO 527-4

• Quasi-static tensile test

• Constant amplitude fatigue

• Stress/strain analysis

Optical analysis Thermography

20 mm

45°-crack

90°-crack

5.2

mm

+Load

Material and Experimental

Com

ptes

t_20

04_G

agel

_ 5

0

1000

2000

3000

4000

5000

6000

0 0,5 1 1,5 2 2,5

45° CD [1/m]

Cra

ck D

ensi

ty[1

/m]

Strain [%]

0

1000

2000

3000

4000

5000

6000

0 0,5 1 1,5 2 2,5

45° CD [1/m]

Cra

ck D

ensi

ty[1

/m]

Strain [%]

0,8

0,85

0,9

0,95

1

1,05

0 500 1000 1500 2000 2500 3000 3500 4000

Rel. Stiffness, FEA [1]Rel. Stiffness [1]

Rel

. Stif

fnes

s[1

]

+/- 45° CD [1/m]

0,8

0,85

0,9

0,95

1

1,05

0 500 1000 1500 2000 2500 3000 3500 4000

Rel. Stiffness, FEA [1]Rel. Stiffness [1]

Rel

. Stif

fnes

s[1

]

+/- 45° CD [1/m]

• Development of relative stiffness E/E0

measured and FEA prediction*

• About linear relation between stiffness

and crack density* A. Gagel, C. Müller, and K. Schulte, ECCM11, Rhode 2004

• Quasi-static tensile strain increments

• Optical measurement of crack density

• About linear increase with applied strain

• Onset of cracks between 0.2% and 0.3%

Crack Density and Relative Stiffnessunder Quasi-Static Tension

Com

ptes

t_20

04_G

agel

_ 6

• S-N curve

• Force controlled

• Tension – tension fatigue (R=0.1)

• Test frequency 5 Hz

• Interrupted after load increments

of 1000 cycles each

• Visual determination of crack density

by transmitting light microscopy

0

1000

2000

3000

4000

5000

6000

7000

8000

0 1000 2000 3000 4000 5000 6000

CD +/-45° [1/m]

Cra

ck D

ensi

ty +

/-45°

[1/m

]

Load cycle number [1]

σmax = 350MPaR = 0.1f = 5Hz

0

100

200

300

400

500

0 1 2 3 4 5 6 7

y = 611,13 - 70,603x R= 0,91201

Max

. stre

ss [M

Pa]

log(Nf ) [1]

0

100

200

300

400

500

0 1 2 3 4 5 6 7

y = 611,13 - 70,603x R= 0,91201

Max

. stre

ss [M

Pa]

log(Nf ) [1]

σmax = 350MPaR = 0.1f = 5Hz

Fatigue Test

Com

ptes

t_20

04_G

agel

_ 7

• Heterogeneous distribution while fatigue loaded

• Characteristic zig-zag pattern

• Final failure location early detectable* (n/Nf<0.1)* A. Gagel, B. Fiedler, and K. Schulte, 9th European-Japanese Symposium on

Composite Materials, Hamburg, 2004f=10Hz, σmax=0.6 σut, R=0.1

0

1

2

3

4

5

0 0,2 0,4 0,6 0,8 1

dTho

tspo

t-d

T sam

ple

[K]

n/Nf[1]

0

1

2

3

4

5

0 0,2 0,4 0,6 0,8 1

dTho

tspo

t-d

T sam

ple

[K]

n/Nf[1]

σmax = 350MPaR = 0.1f = 10Hz

f

0

10

20

30

40

50

60

70

0 0,2 0,4 0,6 0,8 1

dT(hot spot) [K]dT(sample) [K]

dT[K

]

n/N [1]

0

10

20

30

40

50

60

70

0 0,2 0,4 0,6 0,8 1

dT(hot spot) [K]dT(sample) [K]

dT[K

]

n/N [1]f

0

10

20

30

40

50

60

70

0 0,2 0,4 0,6 0,8 1

dT(hot spot) [K]dT(sample) [K]

dT[K

]

n/N [1]

0

10

20

30

40

50

60

70

0 0,2 0,4 0,6 0,8 1

dT(hot spot) [K]dT(sample) [K]

dT[K

]

n/N [1]

σmax = 350MPaR = 0.1f = 10Hz

Surface Temperature Distribution

Com

ptes

t_20

04_G

agel

_ 8

0,75

0,8

0,85

0,9

0,95

1

0

20

40

60

80

100

0 1000 2000 3000 4000 5000 6000 7000 8000

Edyn/Edyn0 [1] DTloc [K]DTglob [K]

Edy

n/Edy

n0 [1

] T-T0 [K

]

Cycles [N]

Stage I

Stage II

Stage III

dT(hot spot)dT(sample)

0,75

0,8

0,85

0,9

0,95

1

0

20

40

60

80

100

0 1000 2000 3000 4000 5000 6000 7000 8000

Edyn/Edyn0 [1] DTloc [K]DTglob [K]

Edy

n/Edy

n0 [1

] T-T0 [K

]

Cycles [N]

Stage I

Stage II

Stage III

dT(hot spot)dT(sample)

Load cycle number [1]

• Initial, steady and final stage of both stiffness decay and temperature increase

• Mean and hot spot temperature develop proportional until final failure occurs

• Stiffness develops inverse to samples mean temperature

Temperature and StiffnessEvolution while Fatigue

Com

ptes

t_20

04_G

agel

_ 9

• Temperature measurements along sample axis

• Existence of relative (but uncritical) hot (L02) and relative cold (L01) areas

• Temperature distribution does not change significantly with load increments

• Temperature difference about 10% of the samples mean temperature increase

f = 5Hz,σmax = 0.6 σut

R = 0.1

Interrupted FatigueDistribution of Surface Temperature

0

5

10

15

0 0,2 0,4 0,6 0,8 1

T(cold area) - T(amb) [K]T(hot area) - T(amb) [K]T(sample) - T(amb) [K]

T (col

d ar

ea) -

T (am

b) [K

]

n/Nf [1]

0

0,4

0,8

1,2

1,6

0 0,2 0,4 0,6 0,8 1

T (hot

are

a) -

T (col

d ar

ea) [K

]

n/Nf [1]

Temperature differencebetween hot – cold area

Com

ptes

t_20

04_G

agel

_ 10

Cold area

20mm20mm

n=1000 n=2000

2mm

2mm

n=3000 n=4000 n=5000

Hot area

• Defect free sample – no disturbances, voids, etc. visually detectable

• Relative hot and relative cold areas form while fatigue

• No obvious differences detectable with respect to damage types or distribution

Interrupted FatigueMeasurement of Crack Density

Com

ptes

t_20

04_G

agel

_ 11

0

1000

2000

3000

4000

5000

6000

7000

8000

0 1000 2000 3000 4000 5000 6000

CD +45°, hot area [1/m]CD +45°, cold area [1/m]CD -45°, hot area [1/m]CD -45°, cold area [1/m]CD +/-45°, hot area [1/m]CD +/-45°, cold area [1/m]

Cra

ck D

ensi

ty +

/-45°

[1/m

]

Load Cycle Number [1]

+/-45° cracks

+45° cracks

-45° cracks

• Same crack density measured in hot and cold areas

• Slight, but not significant difference in +45° and –45° crack densities

• Temperature distribution is not caused in local differences of the crack density

Crack Density in Hot and Cold Areas

Com

ptes

t_20

04_G

agel

_ 12

Dark field microscopy, staining of inter fibre failures with eosin in alcohol

• Diagonal cracks are evenly distributed in + 45° and -45° degree layers

• In hot and cold areas no differences in crack distribution were found

• Temperature distribution is not caused in spatial crack distribution

Spatial Crack Distribution

Com

ptes

t_20

04_G

agel

_ 13

Conclusions

1. For GF-NCF-EP under mechanical load the stiffness loss can be related to and

modelled by the crack density.

2. The location of the fatal failure in fatigue loaded GF-NCF-EP can be located via

thermography in an very early stage of the fatigue life but

3. The location of the fatal failure can not be related to the crack density.

⇒ Stiffness degradation and failure of GF-NCF-EP under fatigue load seem to be

governed by different mechanisms.

Summary

• Onset of inter fibre failure in GF-NCF-EP at low strains between 0.2 and 0.3%

• Three-staged surface temperature increase and stiffness decrease

• Characteristic surface temperature pattern at fatigue load

• Surface temperature and crack density distribution can not be correlated

• The stiffness decrease can be correlated to crack density

Summary and Conclusion