CHAPTER 5 Effect of long-term high temperature exposure on mechanical
5.3 Results and Discussion
5.3.5 Tensile test of ±45° oriented angle-ply laminates
165
Figure 5.3.33 Compressive Young’s modulus change for 90C as a function of aging time.
166
Figure 5.3.34 Relation among shear stress, strain, and AE measurements for 45T at non-aging.
Figure 5.3.35 Relation among shear stress, strain, and AE measurements for 45T at 1,000 h.
Figure 5.3.36 Relation among shear stress, strain, and AE measurements for 45T at 2,000 h.
0 20 40 60 80 100
0 5 10 15
Stress (MPa)
Strain (%)
Stress Cumulative AE Event
Cumulative AE Energy Amplitude 20
10
0 Cumulative AE Energy×102(V2s)
10 8 6 4 2 0
Amplitude(V)
CumulativeAE Event ×106(Count) 10
5
0
0 20 40 60 80 100
0 5 10 15
Stress (MPa)
Strain (%)
Stress Cumulative AE Event
Cumulative AE Energy Amplitude 20
10
0 Cumulative AE Energy×102(V2s)
10 8 6 4 2 0
Amplitude(V)
CumulativeAE Event ×106(Count) 10
5
0
0 20 40 60 80 100
0 1 2 3 4 5
Stress (MPa)
Strain (%)
Stress Cumulative AE Event
Cumulative AE Energy Amplitude 50
25
0 Cumulative AE Energy×102(V2s)
10 8 6 4 2 0
Amplitude(V)
CumulativeAE Event ×106(Count) 10
5
0
Strain [%]
Strain [%]
Strain [%]
Shear stress [MPa] Shear stress [MPa] Shear stress [MPa] [V2S] [Count] [V]
[V2S] [Count] [V]
[V2S] [Count] [V]
167
Figure 5.3.37 Relation among shear stress, strain, and AE measurements of 45T for 4,000 h.
Figure 5.3.38 Relation among shear stress, strain, and AE measurements of 45T for 8,000 h.
As shown in these figures, shear stress increased linearly with increasing strain for the lower strain. Although the AE signal onset appeared at lower strain value, no variation was observed in the shear strength between non-aging and 1,000 h. As shown in Fig. 4.3.5, the fiber/matrix debonding and resultant matrix crack from exposed top and bottom surface penetrate to -45° layer. Thus the outermost ±45º layers actually were damaged. However the damage was not critical on the shear strength up to 1,000 h. In latter aging periods, the shear strength decreased with increasing aging time. The AE signal initiation for aged specimens became lower compared with that of non-aging
0 20 40 60 80 100
0 1 2 3 4 5
Stress (MPa)
Strain (%)
Stress Cumulative AE Event
Cumulative AE Energy Amplitude 20
10
0 Cumulative AE Energy×102(V2s)
10 8 6 4 2 0
Amplitude(V)
CumulativeAE Event ×106(Count) 10
5
0
0 20 40 60 80 100
0 1 2 3 4 5
Stress (MPa)
Strain (%)
20
10
0 22Cumulative AE Energy ×10(Vs)
0 20 40 60 80 100
0 1 2 3 4 5
Stress (MPa)
Strain (%)
Stress Cumulative AE Event
Cumulative AE Energy Amplitude 20
10
0 Cumulative AE Energy×102(V2s)
10 8 6 4 2 0
Amplitude(V)
CumulativeAE Event ×106(Count) 10
5
0 0 20 40 60 80 100
0 1 2 3 4 5
Stress (MPa)
Strain (%)
Stress Cumulative AE Event
Cumulative AE Energy Amplitude 20
10
0 Cumulative AE Energy×102(V2s)
10 8 6 4 2 0
Amplitude(V)
CumulativeAE Event ×106(Count) 10
5
0 0
20 40 60 80 100
0 1 2 3 4 5
Stress (MPa)
Strain (%)
Stress Cumulative AE Event
Cumulative AE Energy Amplitude 20
10
0 Cumulative AE Energy×102(V2s)
10 8 6 4 2 0
Amplitude(V)
CumulativeAE Event ×106(Count) 10
5
0 0 20 40 60 80 100
0 1 2 3 4 5
Stress (MPa)
Strain (%)
Stress Cumulative AE Event
Cumulative AE Energy Amplitude 20
10
0 Cumulative AE Energy×102(V2s)
10 8 6 4 2 0
Amplitude(V)
CumulativeAE Event ×106(Count) 10
5
0
Strain [%]
Strain [%]
[V2S] [Count] [V]
[V2 S] [Count] [V]
Shear stress [MPa] Shear stress [MPa]
168
specimen. The AE signal onset would correspond to local delamination onset. After 4,000 h aging specimen, larger strength reduction occurred. The strength retention for 4,000 h was approximately 47%. As shown in Fig. 4.3.6 (i), local delamination onset after isothermal aging was observed. Thus the local delamination had a great influence on the strength reduction. After 8,000 h, shear strength retention was approximately 7%.
In this case, the laminate was severely damaged due to local delamination onset in entire the specimen as shown in Fig. 4.3.6 (j). Figure 5.3.39 shows the images for 45T after failure. (a) is non-aging, (b) is 4,000 h, and (c) is 8,000 h, respectively. From these figures, no change in failure mode was observed. The failure occurred with delamination.
(a) Non-aging (b) 4,000 h (c) 8,000 h Figure 5.3.39 Images for 45T after failure.
Figure 5.3.40 shows the shear stress for final failure and critical stress for edge delamination as a function of aging time. The error bar indicates standard deviation. The effect of 180 °C exposure on strength was severe and the strength retention after 1,000 h,
5 mm 5 mm
5 mm
169
2,000 h, 4,000 h, and 8,000 h with non-aging one was 99%, 90%, 47%, and 7%. As for critical stress for local delamination, it decreased with increase aging time up to 8,000 h.
Figure 5.3.40 Shear strength for final failure and critical stress for edge delamination as a function of aging time.
Figure 5.3.41 shows shear modulus change of 45T as a function of aging time. For non-aging specimen, the average shear modulus was obtained as 5.2 GPa. With increasing aging time, shear modulus significantly decreased. It was caused by local delamination.
Figure 5.3.41 Shear modulus change as a function of aging time.
0 2 4 6 8 10
0 2000 4000 6000 8000
Shear Modulus [GPa]
Aging Time [Hours]
0 20 40 60 80 100
0 20 40 60 80 100
0 2000 4000 6000 8000 Cri
tical Stress for Edge Delamination [MPa]
Shear Stress for Final Failure [MPa]
Aging Time [Hours]
Shear Stress for Fainal Failure Crtical Stress for Edge Delamination
170
In general, for multidirectional laminates under tensile loading, edge delamination appear before laminates failure. The edge delamination is resulted from the interlaminar stresses induced at free edges when the laminate is loaded in in-plane. A fracture mechanism of edge delamination in multidirectional laminate had been established by O’Brien [5-6]. Based on the fracture energy consideration, the relationship between the critical energy release rate and the edge delamination onset strain is described as Eq.
(5-3).
Gc = εc2 t (E−E’) / 2 (5-3)
where εc is delamination onset strain, t is the specimen thickness, E is no damaged Young’s modulus, and E’ is delaminated laminate Young’s modulus.
Lee [5-7] found a relation between E and E’ based on test results.
E’ ≈ 0.7 E (5-4)
Thus Eq. (5-3) is described with Eq. (5-4) Gc = 0.15 εc2 t E (5-5)
In this study, εc was defined as an edge delamination onset strain corresponded to critical stress for edge delamination .
Figure 5.3.42 shows the critical energy release rate Gc change as a function of aging time. The Gc significantly decreased in the first 1,000 h. This result suggested that the matrix or fiber/matrix interface where adjacent to exposed free edge surface in the interlaminar layer was severely damaged. The Fiber/matrix debonding and subsequent matrix cracking progressed along thickness direction as shown in Fig. 4.3.9. On the other hand, no matrix crack was observed at 0°/-45° interface layer after 4,000 h as shown in Fig. 4.3.24 (d). Moreover, as mentioned in chapter 2, the fracture toughness of the neat resin did not show such a larger decrease compared with the Gc result.
Therefore the fiber/matrix interface would be largely damaged.
171
Figure 5.3.42 Critical energy release rate Gc change as a function of aging time.
0 200 400 600 800 1,000
0 2000 4000 6000 8000
Critical Energy Release Rate Gc (J/m2)
Aging Time (Hours)
172
3.6 Compressive test of quasi-isotropic laminate
Figures 5.3.43 to 5.3.47 show the relation among compressive stress, strain, and AE measurements of NHC for non-aging, 1,000 h, 2,000 h, 4,000 h, and 8,000 h, respectively. For 8,000 h, strain could not be measured by strain gauge due to the outermost layer peeled just after loading. Thus the strain for 8,000 was obtained as cross head displacement divided by specimen length. Since noise was generated between the compressive test jig and the platen for compressive test, only cumulative energy was analyzed with audible sound and visual observation during testing. As shown in these figures, compressive stress increased linearly with increase strain. The cumulative AE energy sharply increasing point gradually shifted to lower stress level with increasing aging time. After 8,000h, the outermost 45° layer peeled at 45/0° interlaminar layer as soon as the specimen was loaded. From visual observation during testing, the cumulative AE energy increasing stress corresponded to local delamination onset. The strain where the cumulative AE energy rapidly increasing point corresponded to delamination onset.
Figure 5.3.43 Relation among shear stress, strain, and AE measurements of NHC for non-aging.
0 100 200
0 100 200 300 400 500 600
0 0.5 1 1.5
Stress [MPa]
Strain [%]
Stress
Cumulative AE Energy
Cumulative AE Energy [V2s]
173
Figure 5.3.44 Relation among shear stress, strain, and AE measurements of NHC for 1,000 h.
Figure 5.3.45 Relation among shear stress, strain, and AE measurements of NHC for 2,000 h.
Figure 5.3.46 Relation among shear stress, strain, and AE measurements of NHC for 4,000 h.
0 100 200
0 100 200 300 400 500 600
0 0.5 1 1.5
Stress [MPa]
Strain [%]
Stress-Strain Cumulative AE Energy
Cumulative AE Energy [V2s]
0 100 200
0 100 200 300 400 500 600
0 0.5 1 1.5
Stress [MPa]
Strain [%]
Stress
Cumulative AE Energy
Cumulative AE Energy [V2s]
0 100 200
0 100 200 300 400 500 600
0 0.5 1 1.5
Stress [MPa]
Strain [%]
Stress
Cumulative AE Energy
Cumulative AE Energy [V2s]
174
0 100 200
0 100 200 300 400 500 600
0 0.5 1 1.5
Stress [MPa]
Strain [%]
Stress
Cumulative AE Energy
Cumulative AE Energy [V2s]
Figure 5.3.47 Relation among shear stress, strain, and AE measurements of NHC for 8,000 h.
Figure 5.3.48 shows the compressive strength, the critical stress for local delamination and the critical stress for delamination as a function of aging time. The error bar indicates standard deviation. The strength retentions after 1,000 h, 2,000 h, 4,000 h and 8,000 h compared with non-aging one were approximately 89%, 85%, 68%, and 47%, respectively. With decreasing the critical stress for local delamination and critical stress for delamination, the compressive stress decreased. Thus the failure mechanism was basically same irrespective of aging time. For 4,000 h, the cumulative AE energy rapidly increased without the cumulative AE energy stable increasing as shown in Fig. 5.3.46. This result suggested that the local delamination rapidly grew as a delamination due to the damage of interfaces between layers as show in Fig. 4.3.8 (g).
175
Figure 5.3.48 Compressive strength, critical stress for local delamination and critical stress for delamination as a function of aging time.
Figure 5.3.49 shows the NHC specimen images of just before specimen failure. (a) is non-aging, (b) is 1,000 h, (c) is 2,000 h, (d) is 4,000 h, and (e) is 8,000 h, respectively.
For non-aging, the outermost 45° layer peeled before failure. For 1,000 h, outermost 45°
layer or 0° layer peeled just before specimen failure, while some specimens did not peel before specimen failure. For 2,000 h, the outermost 45° layer in both side peeled as a first and the outermost 0° layer peeled in both side following before specimen failure.
For 4,000 h, failure mechanism was same manner with 2,000 h partway. Before final failure onset, the damage of the 0°/-45° interface was observed. The 0° layer peeled in lower stress level compared with that of 2,000h. For 8,000 h, 45° layer peeled in lower stress level and successive 0° layer peeled. Crack also appeared in outermost -45° layer, 90° layer, secondly 45°layer, and secondly 0°/-45° interface layer before specimen failure.
To estimate the effect of the damage progress from exposed top and bottom surface on the NHC strength, laminate theory was used. Experimental data of non-aging for
[MPa]
176
unidirectional laminate were used for calculation. Table 5.3.1 shows mechanical properties of laminates. Obtained compressive Young’s modulus with using these properties for no damage case was 46.0 GPa. The failure strain was 1.2 % which was obtained from non-aging test result. From the observation of damage behavior during loading, outermost two plies for each side were neglected to calculate strength at 2,000 h aging. For the strength estimation at 4,000 h, outermost four plies for each side were neglected for effective cross sectional are. For strength estimation at 8.000 h, outermost six plies for each side were neglected. Figure 5.3.50 shows compressive strength was calculated with the assumption of such a damage progress. The predictions were in good agreement with the experimental results regardless of simple assumption. Thus the damages from exposed top and bottom surface dominated the compressive strength of the quasi-isotropic laminates. For multidirectional laminates in compressive loading, the damage progress from exposed top and bottom surfaces are important to estimate residual strength. For more accurate residual mechanical properties, modeling for damage initiation and progress would be needed.
177
(a) Non-aging (b) 1,000 h
(c) 2,000 h (d) 4,000 h
Figure 5.3.49 NHC specimen image of just before specimen failure (Continued).
1 mm
1 mm
178
(e) 8,000 h
Figure 5.3.49 NHC specimen image of just before specimen failure.
Table 5.3.1 Mechanical properties of laminates
Properties (Unit) Value
Longitudinal Young's moulus for compression (GPa) EL 116 Tranveres Young's modulus for compression (GPa) ET 9.2
In-plane shear modulus (GPa) G 5.2
Longitudinal Poisson's ratio νLT 0.31
Tranverse Poisson's ratio νTL 0.02
Ply thick ness (mm) t 0.14
Ply width (mm) W 25.4
Longitudinal compressive Young’s modulus (GPa) Transverse compressive Young’s modulus (GPa)
1 mm
179
Figure 5.3.50 Compressive strength comparing the assumption of the damage case for each aging time.
5.3.7 Compressive test of quasi-isotropic laminate with hole
To measure strain, an extensometer (Model 632.11F-20; MTS Systems Corporation) was placed around hole of specimen. In order to obtain nominal stress value, load was divided by reduced cross-sectional area around hole.
Figures 5.3.51 to 5.3.54 show the relation among OHC stress, strain, and AE measurements of OHC for non-aging, 2,000 h, 4,000 h, and 8,000 h, respectively. Since noise was generated between the compressive test jig and the platen for compressive test, only cumulative AE energy was analyzed. The OHC strength increased linearly with increasing strain up to 4,000 h aging. For 8,000 h, the stress did not increased linearly due to delamination onset around hole. The cumulative AE energy increasing point shifted to lower stress level with increasing aging time compared with that of non-aging specimen.
180
Figure 5.3.51 Relation among shear stress, strain, and AE measurements of OHC for non-aging.
Figure 5.3.52 Relation among shear stress, strain, and AE measurements of OHC for 2,000 h.
Figure 5.3.53 Relation among shear stress, strain, and AE measurements of OHC for 4,000 h
Cumulative AE Energy [V2 S]
Stress [MPa] Cumulative AE Energy [V2 S]
Stress [MPa] Cumulative AE Energy [V2 S]
Stress [MPa]
Stress [MPa]
Stress [MPa]
Stress [MPa]
181
Figure 5.3.54 Relation among shear stress, strain, and AE measurements of OHC for 8,000h.
Figure 5.3.55 shows OHC specimen image after specimen failure. For non-aging and 2,000 h, failure mode was laterally compressive failure across the center of the hole. For 4,000h, crack appeared around hole before failure. For 8,000h, crack appeared around hole just after loading and the outermost layer peeled as soon as loaded the specimen was loaded similar to NHC.
With decreasing the cumulative AE energy increasing stress, the compressive strength decreased. Form visual observation in testing, a strain where the cumulative AE energy sharply increasing point corresponded to edge delamination and/or delamination onset around hole for 2,000 h, 4,000 h, and 8,000 h.
(a) Non-aging (b) 2,000 h (C) 4,000 h (d) 8,000 h Figure 5.3.55 OHC specimen images after specimen failure.
10 mm Cumulative AE Energy [V2 S]
Stress [MPa]
Stress [MPa]
182
Figure 5.3.56 shows the OHC strength and the critical stress for edge delamination as a function of aging time. The error bar indicates standard deviation. The strength retentions after 2,000 h, 4,000 h and 8,000 h compared with non-aging were approximately 93%, 82%, and 53%, respectively. For 8,000 h, 45° layer peeled in lower stress level and adjacent 0° layer might peel. Because the strength retention of the OHC for 8,000 h was similiar value compared with that of NHC. For OHC specimen the damages from exposed top and bottom surface were also dominated the OHC strength.
Therefore the strength reduction mechanism was same manner as the NHC.
Figure 5.3.56 OHC strength and critical stress for edge delamination as a function of aging time.
Figure 5.3.57 shows the strength ratio compared the NHC strength to the OHC strength as a function of aging time. For non-aging, it seemed that the stress concentration was not large. The strength ratio decreased with increasing up to 4,000 h.
Thus the stress concentration was relaxed by the crack and delamination onset around hole as shown in Fig. 4.3.10.
Critical Stress for Edge Delamination
Critical Stress for Edge Delamination [MPa]
183
Figure 5.3.57 Strength ratio compared NHC to OHC as a function of aging time.
5.3.8 Mechanical properties retention for CFRP
Figure 5.3.58 and Fig. 5.3.59 show the change in strength and modulus due to thermo-oxidative degradation, respectively. For the change in the strength for 90 T and 45T, the larger decrease was confirmed compared with the other specimen. This result suggested that matrix and fiber/matrix interface were severely damaged. For the change in modulus, the figure shows shear modulus for 45T was severely decreased after 8,000 compared with the other specimen. Local delamination as shown in Fig. 4.3.6 (j) affected modulus reduction. The modulus for the other specimen did not show such a large decrease. The result suggested that the effect of thermo-oxidative degradation on modulus for CFRP was basically small.
0.0 0.5 1.0 1.5
0 2000 4000 6000 8000
NHC/OHC Strength Ratio
Aging (Hours)
Aging Time [Hours]
184
Figure 5.3.58 Change in strength due to thermo-oxidative degradation for each aging time.
Figure 5.3.59 Change in modulus due to thermo-oxidative degradation for each aging time.
5.3.9 Relation between strength retention and weight change
Figure 5.3.60 shows the relation between strength retention and weight change for each type of specimen. In this figure, it could be divided the three types of failure mechanism. First was fiber domination type. Second was matrix domination type. Third was delamination domination type. The first type was 0T and 90C. The second type is