The document describes experiments using ultrasonic inspections and confocal microscopy to evaluate fatigue damage in fiber reinforced polymer composites. Glass fiber reinforced phenolic and carbon fiber reinforced epoxy composites were tested using 4-point bending fatigue tests. Ultrasound equipment and visualization software were used to scan the composites at various cycle intervals to identify damage development. Confocal microscopy was also used to examine fatigue damage in samples tested to 20,000-26,000 cycles. The results were intended to characterize fatigue behavior and damage progression in the composites.

1. Ultrasonics inspections and confocal
microscopy to evaluate fatigue damage in
fiber reinforced polymer composites
V.G. García, J. Sala, L. Crispí, J.M. Cabrera, A. Istúriz, A. Sàez,
M. Millán, C. Comes, D. Trias
Composites

2. 1. Introduction: Outline
2. Experimental Procedure:
Materials inspected and tested
Ultrasound equipment
Ultrasound visualization software
4-Point bending fatigue tests
3. Results:
Wöhler plot
Ultrasound scans at N=4,000 cycles
Ultrasound scans in at N=26,079 cycles
Confocal microscopy in a GFRP at N=26,079 cycles
Confocal microscopy in a CFRP at N=20,000 cycles
4. Conclusions
2/20

3. 2. Experimental Procedure
Fig. 1. Dimensions of the glass fiber/phenolic resin bars
Fig. 3. Ultrasound inspections every
2,000 cycles.
Until
20,000
Fig. 2. Dimensions carbon fiber/epoxy resin bars cycles.
Fig. 4. 4-Point bending flexural fatigue. 3/20

4. 2. Experimental Procedure: Materials inspected and tested
The Glass Fiber Reinforced Polymer (GFRP) was Isovid G-3
manufactured by Composites Ate.
-Isovid G-3 consists of 200g/m2 plain weave E-glass, and a
modified phenolic resin that enhances flame retardant
characteristics.
-Layers are 0.24mm thick.
-All plies were stacked to match 0º and 90º.
-Composite was high speed milled to reach dimensions.
Fig. 5. GFRP samples.
Fig. 6. GFRP sample after exposure Fig. 7. Woven appearance.
to sun light.
4/20

5. 2. Experimental Procedure: Materials inspected and tested
The Carbon Fiber Reinforced Polymer (CFRP) was manufactured using Hexcel
8552/32%/134/IM7(12K) at the INTA Materials and Structures Department.
-The 8552/32%/134/IM7(12K) pre-impregnated carbon fiber
plies consisted of tows of 12,000 individual fibers of IM7
intermediate modulus carbon.
-The pre-preg contained 32% of 8552, an amine cured,
Fig. 8. CFRP samples. toughened epoxy resin system.
-The nominal ply thickness was 134µm.
-Measured ply thickness was 129µm to 134µm.
Table 1. Stacking sequences and percentages of layers oriented 90º, 0º or ±45º.
% % % % #
Specimens Stacking sequence
90º 0º 45º -45º layers
[(0/±45/02/±45/02/90) /
CFRP-P011 8.8 54.4 18.4 18.4 283
( 02/±45/02/±45/02/ )12]S
90
[(0/90) /
CFRP-P021 9.5 54.7 17.9 17.9 201
( 02/±45/02/±45/0290)9]S
/
Fig. 9. Smooth, rough, and [(0/90/0/±45/02/90) /
machined surfaces of CFRP CFRP-P041 11.9 54.2 16.9 16.9 59
( 02/±45/02/±45/0290)2]S
/
samples.
5/20

6. 2. Experimental Procedure: Ultrasound equipment
-A single crystal longitudinal wave transducer of 1MHz and
13mm diameter was used for most inspections.
-The gain was set at 15dB.
-Specimens CFRP-P041 were inspected using a 5MHz (10mm
diameter) transducer at 5dB.
-Acoustic wave propagation was set at 3275m/s.
-Mapro developed a software, based on LabView, to process
the pulse-echo signals and visualize A-scans, B-scans, C-
scans, plus optional ∫-scans.
Fig. 10. Ultrasound equipment built by Mapro
using a Socomate USPC3100LA card.
Y (Index)
X (scan)
Fig. 11. A pulse-echo inspection in an Fig. 12. Before an inspection the Fig. 13. Inspection of
immersion bath. transducer is positioned at the (0,0) a GFRP specimen.
origin.
6/20

7. 2. Experimental Procedure: Ultrasound visualization software
Fig. 14. Screen image of the visualization software, after loading the ultrasound signal data. 7/20

8. 2. Experimental Procedure: Ultrasound visualization software
Fig. 15. Screen image after: digitally aligning the first peak, positioning the cut off lines, and setting a color range. 8/20

9. 2. Experimental Procedure: Ultrasound visualization software
Interface (I) echo
Back-wall (B) echo
Fig. 16. In this study the A-scans was passed through different algorithms to create alternative scan images. 9/20

10. 2. Experimental Procedure: Ultrasound visualization software
C-scan
minus I & B
∫-scan with I
&B
∫-scan
minus I & B
∫-scan with I
& B minus
mean
∫-scan
minus I & B
minus mean
Fig. 17. A menu in the C-scan label allows processing the A-scan to create different ∫-scans. 10/20

11. 2. Experimental Procedure: 4-Point bending fatigue tests
Flexural Stress
The maximum strength σ0 for
σ = (3FL)/(4Wh2)
the GFRP was 391MPa, and
for the CFRP was 1135MPa.
F→ force,
L → support length,
W → specimen width,
Normalized fatigue stresses for
h → specimen thickness.
L GFRP specimens was set at 0.2
to 0.9 and for the CFRP
F F Fig. 19. GFRP specimen tested until fracture specimens at 0.36 to 0.60.
2 2 to determine σ0.
F L S L F Normalized fatigue stress is
2 4 4 2 defined as σmax / σ0 where σmax
L is the maximum flexural fatigue
Fig. 18. Loading diagram
stress in the outer layers.
according to ASTM D6272-
02 (2008).
L
Fig. 20. CFRP specimen tested until fracture
to determine σ0. 11/20

12. 3. Results: Wöhler plot
600 400
Maximum stress, σmax (MPa)
GFRP-P011-40-1000-38_1/3 GFRP-P031-40-710-16_2/6
CFRP-P011-40-1440-38_1/2
Maximum stress, σmax (MPa)
400 CFRP-P011-40-1440-38_2/2 300 GFRP-P011-40-1000-38_2/3
GFRP-P011-40-1000-38_3/3
GFRP-P031-40-710-16_3/6
GFRP-P031-40-710-16_4/6
CFRP-PO21-40-1040-27_1/2 GFRP-P021-40-1000-27_1/3 GFRP-P031-40-710-16_5/6
200 GFRP-P021-40-1000-27_2/3 GFRP-P031-40-710-16_6/6
200 f = 0,8 Hz CFRP-P021-40-1040-27_2/2 GFRP-P021-40-1000-27_3/3 GFRP-P041-40-307-8_1/2
R=0,1 CFRP-P041-40-307-8_2/2 100 GFRP-P031-40-710-16_1/6 GFRP-P041-40-307-8_2/2
0
0 2 4 6 8 10 12 14 16 18 20 0 0 2 4 6 8 10 12 14 16 18 20
Number of cycles (x1000)
-200 Number of cycles (x1000)
-100
4-point bending tests
-400 -200
-600
-300 4-point bending tests
f = 0,8 Hz
-400 R = 0,1
-800
Fig. 21. Maximum stresses during fatigue every Fig. 22. Maximum stresses during fatigue every
2,000 cycles for the CFRP specimens. 2,000 cycles for the GFRP specimens.
400
Maximum stress, σmax (MPa)
350
300 4-point
bending
250
fatigue tests, 235 MPa
200 R=0,1
150 f=0,8Hz 13,344 cycles
100 Isovid G-3:
50
Plain weave 200g/m2 fiber glass
with flame retardant phenolic resin
0
101 102 103 100 104 105
Number of cycles to failure, N
Fig. 23. Maximum stresses versus cycles to failure of Isovid G-3 GFRP. 12/20

13. 3. Results: Ultrasound scans a GFRP at N=4,000 cycles
Y (Index)
X (scan)
(a) X (scan)
Y (Index)
C-scan of the B echo
(b)
∫-scan with I & B
(c)
∫-scan minus I & B
(d)
∫-scan with I & B minus mean
(e)
∫-scan minus I & B minus mean
(f)
Fig. 24. In (a) specimen GFRP-P031-40-710-16_5/6 that broke at
N=4001 cycles is shown. Images (b) to (f) show different types of
scans performed at N=4000.
13/20

14. 3. Results: Ultrasound scans in a GFRP at N=26,079 cycles
N=26,504
Y (Index) Y (Index)
(a) X (scan)
(b) C-scan of the B echo
(c) ∫-scan with I & B
∫-scan minus I & B
(d)
∫-scan with I & B minus mean
(e)
∫-scan minus I & B minus
mean
(f)
Fig. 25. In (a) specimen GFRP-P031-40-710-16_3/6 that broke at
N=26,504 cycles is shown. Images (b) to (f) show different types of
scans performed at N=26,079.
14/20

15. 3. Results: Ultrasound scans in a GFRP at N=26,079 cycles
N=26,504
(a) X (scan)
Y (Index)
(b) C-scan of the B echo
(c) ∫-scan with I & B
∫-scan minus I & B
(d)
∫-scan with I & B minus mean
(e)
∫-scan minus I & B minus
mean
(f)
Fig. 26. In (a) specimen GFRP-P031-40-710-16_3/6 was cracked open
to reveal the red color of the penetrating liquid. Images (b) to (f) show
different types of scans performed at N=26,079. 15/20

16. 3. Results: Confocal microscopy in a GFRP at N=26,079 cycles
(b)
(a)
(c)
Fig. 27. Image in (a) shows specimen GFRP-P031-40-710-16_3/6 being inspected by a
Sensofar Plμ 2300 confocal microscope, in (b) and (c) two contiguous surface areas show a
crack on the site of maximum bending stress.
16/20

17. 3. Results: Confocal microscopy in a CFRP at N=20,000 cycles
X (scan)
(a)
Y (Index)
(d)
(b)
X (scan)
(c)
Y (Index)
(e) (f)
Fig. 28. Images in (a) and (b) shows specimen CFRP-P021-40-1040-27_1/2, in (c) the C-
scan image, in (d) a photo of the damaged surface, and in (e) and (f) two contiguous confocal
images show a normal surface and a worn surface, respectively.
17/20

18. 3. Results: Confocal microscopy in a CFRP at N=20,000 cycles
(e) (f)
Fig. 28. Images in (a) and (b) shows specimen CFRP-P021-40-1040-27_1/2, in (c) the C-
scan image, in (d) a photo of the damaged surface, and in (e) and (f) two contiguous confocal
images show a normal surface and a worn surface, respectively.
18/20

19. 3. Results: Confocal microscopy in a CFRP at N=20,000 cycles
(f)
(g) (h)
Fig. 29. Image in (f) is again the worn surface of specimen CFRP-P021-40-1040-27_1/2. A
closer observation in shown in (g) and the damage can be quantified in image (h).
19/20

20. 4. Conclusions
-The additional information (i. e. newer spots) provided by the ∫-scans still require
further corroboration with the actual internal damage.
-However, the observations of this study seem to point to the ∫-scan without the I &
B echoes, as the image that can combine the most information both from internal
and superficial damage.
-Ultrasound inspections in the GFRP specimens of this study helped locate
possible surface damage, and confocal microscopy was capable of noticing cracks
on that surface.
-In CFRP specimens the damage was visually evident, but confocal images helped
quantify the depth and type of damage.
20/20

21. Thank you for your attention.