**2.4 Micro-hardness testing**

The three samples have been compared in relation to their microhardness performance based on the reinforcement percentage, the heat treatment conditions and the different manufacturing forming processes. Microhardness of the three composites has been measured in order to get the resistance of the material to indentation, under localized loading conditions. The microhardness test method, according to ASTM E-384, specifies a range of loads using a diamond indenter to make an indentation, which is measured and converted to a hardness value [21, 22].

Measuring the different phases in the micro-level it is quite challenging, as the SiC reinforcement of ≈17μm in size was not easy to measure, due to small indentation mark left when a small load on the carbide is applied. When introducing higher values of load, the indentation was not localized in the carbide but covered some of the matrix area too. The load was finally set to 50 grams in order to obtain valid measurements coming from different areas of the samples: SiC, aluminium matrix, and the overall composite i.e. areas superimposing matrix and reinforcement.

There are many factors influencing the microhardness of a composite material, including the reinforcement percentage, interparticle spacing and also particle size. Moreover, manufacturing forming processes influence material's microhardness behaviour in relation to the reinforcement percentages in the composites.

The cast sample in the as received condition has the highest MMC microhardness, where the rolled 20% SiC with lower percentage of reinforcement has the lowest values. By altering the microstructure with modified T6 (HT-1) heat treatment all values of the three samples show an increase between 20-45% from the initial state (Fig. 8). This shows the effect of the heat treatment in the micro-deformation of the matrix-reinforcement interface due to the presence of precipitates and other phases and oxide layers.

Fig. 8. Microhardness values Vs. Heat treatment cycles for the MMC areas

In the T6 condition it was observed the larger increase in microhardness values from the as received state, ranging from 20% to 90% depending on the reinforcement percentage and

The three samples have been compared in relation to their microhardness performance based on the reinforcement percentage, the heat treatment conditions and the different manufacturing forming processes. Microhardness of the three composites has been measured in order to get the resistance of the material to indentation, under localized loading conditions. The microhardness test method, according to ASTM E-384, specifies a range of loads using a diamond indenter to make an indentation, which is measured and

Measuring the different phases in the micro-level it is quite challenging, as the SiC reinforcement of ≈17μm in size was not easy to measure, due to small indentation mark left when a small load on the carbide is applied. When introducing higher values of load, the indentation was not localized in the carbide but covered some of the matrix area too. The load was finally set to 50 grams in order to obtain valid measurements coming from different areas of the samples: SiC, aluminium matrix, and the overall composite i.e. areas

There are many factors influencing the microhardness of a composite material, including the reinforcement percentage, interparticle spacing and also particle size. Moreover, manufacturing forming processes influence material's microhardness behaviour in relation

The cast sample in the as received condition has the highest MMC microhardness, where the rolled 20% SiC with lower percentage of reinforcement has the lowest values. By altering the microstructure with modified T6 (HT-1) heat treatment all values of the three samples show an increase between 20-45% from the initial state (Fig. 8). This shows the effect of the heat treatment in the micro-deformation of the matrix-reinforcement interface due to the

**Microhardness Vs Heat Treatment-MMC**

MMC-AS RECEIVED MMC-HT1 MMC-T6

Fig. 8. Microhardness values Vs. Heat treatment cycles for the MMC areas

**Heat Treatment Cycles**

In the T6 condition it was observed the larger increase in microhardness values from the as received state, ranging from 20% to 90% depending on the reinforcement percentage and

**2.4 Micro-hardness testing** 

converted to a hardness value [21, 22].

superimposing matrix and reinforcement.

100

120

140

160

180

Hv

200

220

240

260

to the reinforcement percentages in the composites.

presence of precipitates and other phases and oxide layers.

 ROLLED20SIC ROLLED31SIC CAST30SIC

manufacturing process. In particular, in the rolled 20% SiC material the increase in microhardness values is in the order of 90%.

Furthermore, variability in microhardness values was observed when comparing cast and rolled materials with different percentage of SiC. However, this variability varied when samples processed at different heat treatment conditions were compared. Highest variability showed samples in the as received condition, whereas lowest variability showed samples in the T6 condition, with samples in the HT-1 condition in between. This can be explained by the fact that precipitates act as strengthening mechanisms and affect the micromechanical behaviour of the composite material.

In the absence of precipitates (in the as received condition), the volume percentage of SiC and the manufacturing processing play a significant role in micromechanical behaviour of the composite. As precipitates are formed due to heat treatment process they assume the main role in the micromechanical behaviour of the material. In the HT-1 heat treatment condition there is presence of β' precipitates which affect the micromechanical behaviour in a lesser degree than in the case of T6 heat treatment condition where fully grown β precipitates are formed. It becomes clear that after a critical stage, which if related to the formation of β precipitates in the composite the dominant strengthening mechanism is precipitation hardening.

While Figure 8 shows results in areas that include the interface region (where precipitates are concentrated) Figure 9, shows results on microhardness values in the aluminium matrix (where precipitates are dispersed). In Figure 9 there is similar variability for all three materials processing states, as received, HT-1, and T6. This implies that in the matrix material the percentage of the reinforcements, the manufacturing process, as well as the precipitation hardening, are strengthening mechanisms of equal importance.

Fig. 9. Microhardness Vs. Heat treatment cycles for Aluminium areas

Figure 10 shows microhardness measurements obtained from areas around the matrixreinforcement interface in a composite heat treated in the T6 condition. The microhardness

Deformation Characteristics of Aluminium Composites for Structural Applications 359

strain for this temper is considerably higher than for the T6 heat treatment; this may be attributed either to the nucleation of the β` precipitate phases which although not yet visible, may lead to the increase of the plastic deformation through crack deflection mechanisms and/or to annealing which acts competitively to the precipitation leading to the toughening of the composite. However, the T6 heat treatment exhibits the highest strength followed by the HT-1 and the as received state. Finally, as was expected, the "as received" composites behaviour in tension deteriorates with increasing filler concentration. The experiments showed that for the same range of conditions tested, the yield and the ultimate tensile strengths of the SiC/Al composites were mainly controlled by the percentage of reinforcement as well as by the intrinsic yield/tensile strengths of the matrix alloys. The addition of the SiC reinforcement created stress concentrations in the composite, and thus the aluminium alloy could not achieve its potential strength and ductility due to the induced embrittlement. Composites in the as-received condition failed in a brittle manner with increasing percentage of reinforcement. As a result, with increasing reinforcement content, the failure strain of the composites was reduced as shown in Figure 11. From the above postulations it is obvious that the phase that dominates the mechanical behaviour of the composite is the precipitation phase created by age hardening while the

The heat treatment affected the modulus of elasticity of the composites by altering the transition into plastic flow (see Table 2 and Fig. 12). Composites in the T6 condition strained elastically and then passed into a normal decreasing-slope plastic flow. Composites tested in the HT-1 condition exhibit a greater amount of strain than the as received and those heat treated in the T6 condition. The failure strain increasing from about 1.5% strain to about 4% but the greater influence was a sharper slope of the stress-strain curve at the inception of

This increase in elastic proportional strain limit and the steepening of the stress-strain curve were reflected by the higher yield and ultimate tensile strengths observed in the heat-treated composites. The increase in flow stress of composites with each heat-treatable matrix

reinforcement phase plays a secondary role.

Fig. 11. Stress / Strain curves of Al/SiC Composites

plastic flow.

values are higher in the close proximity of the interface. It is observed that cast material has higher values than the rolled material. In the case of rolled material, the microhardness raises as the percentage of reinforcement increases.

Fig. 10. Interfacial microhardness showing measurements obtained from areas close to the matrix- reinforcement interface in the T6 condition

### **2.5 Tensile testing**

Aluminium – SiC particulate composite samples were tested in tension for two different volume fractions, 20% and 31%, in reinforcement [23]. The dog-bone coupons were tested according to ASTM E8-04 in the as received and, following two different heat treatments, modified T6 (HT-1) and T6 heat treatment conditions.. The mechanical properties of the composites are presented in Table 2.


Table 2. The mechanical properties of Al/SiC Composites

The engineering stress/strain curves of the composite are shown in Figure 11. As can be clearly seen in Figure 11, the HT-1 heat treatment improved both the strength and strain to failure than the untreated composites for both volume fractions. Furthermore, the failure

values are higher in the close proximity of the interface. It is observed that cast material has higher values than the rolled material. In the case of rolled material, the microhardness

> ROLLED 20% SiC ROLLED 31% SiC CAST 30% SiC

**Interfacial Microhardness - T6**

0


Fig. 10. Interfacial microhardness showing measurements obtained from areas close to the

Aluminium – SiC particulate composite samples were tested in tension for two different volume fractions, 20% and 31%, in reinforcement [23]. The dog-bone coupons were tested according to ASTM E8-04 in the as received and, following two different heat treatments, modified T6 (HT-1) and T6 heat treatment conditions.. The mechanical properties of the

The engineering stress/strain curves of the composite are shown in Figure 11. As can be clearly seen in Figure 11, the HT-1 heat treatment improved both the strength and strain to failure than the untreated composites for both volume fractions. Furthermore, the failure

50

100

150

200

250

300

raises as the percentage of reinforcement increases.

matrix- reinforcement interface in the T6 condition

Table 2. The mechanical properties of Al/SiC Composites

**Hv**

**2.5 Tensile testing** 

composites are presented in Table 2.

strain for this temper is considerably higher than for the T6 heat treatment; this may be attributed either to the nucleation of the β` precipitate phases which although not yet visible, may lead to the increase of the plastic deformation through crack deflection mechanisms and/or to annealing which acts competitively to the precipitation leading to the toughening of the composite. However, the T6 heat treatment exhibits the highest strength followed by the HT-1 and the as received state. Finally, as was expected, the "as received" composites behaviour in tension deteriorates with increasing filler concentration. The experiments showed that for the same range of conditions tested, the yield and the ultimate tensile strengths of the SiC/Al composites were mainly controlled by the percentage of reinforcement as well as by the intrinsic yield/tensile strengths of the matrix alloys. The addition of the SiC reinforcement created stress concentrations in the composite, and thus the aluminium alloy could not achieve its potential strength and ductility due to the induced embrittlement. Composites in the as-received condition failed in a brittle manner with increasing percentage of reinforcement. As a result, with increasing reinforcement content, the failure strain of the composites was reduced as shown in Figure 11. From the above postulations it is obvious that the phase that dominates the mechanical behaviour of the composite is the precipitation phase created by age hardening while the reinforcement phase plays a secondary role.

Fig. 11. Stress / Strain curves of Al/SiC Composites

The heat treatment affected the modulus of elasticity of the composites by altering the transition into plastic flow (see Table 2 and Fig. 12). Composites in the T6 condition strained elastically and then passed into a normal decreasing-slope plastic flow. Composites tested in the HT-1 condition exhibit a greater amount of strain than the as received and those heat treated in the T6 condition. The failure strain increasing from about 1.5% strain to about 4% but the greater influence was a sharper slope of the stress-strain curve at the inception of plastic flow.

This increase in elastic proportional strain limit and the steepening of the stress-strain curve were reflected by the higher yield and ultimate tensile strengths observed in the heat-treated composites. The increase in flow stress of composites with each heat-treatable matrix

Deformation Characteristics of Aluminium Composites for Structural Applications 361

Only when specific validity criteria are satisfied, the provisional fracture toughness, KQ, can be quoted as the valid plane strain fracture toughness, KIC. The standard used for conducting this experiment, ASTM E399, imposes strict validity criteria to ensure that the plane strain conditions are satisfied during the test. These criteria include checks on the form and shape of the load versus displacement curve, requirements on specimen's size and crack geometry, and the 0.2% proof strength values at the test temperature. Essentially, these conditions are designed to ensure that the plastic zone size associated with the precrack is small enough so that plane strain conditions prevail, and that the linear elastic

Fracture toughness tests were conducted using a servo-hydraulic universal testing machine with data acquisition controller. The system was operated on load control for the fatigue pre-cracking stage, and on position control for the crack opening displacement (COD) testing. The fatigue test for pre-cracking was conducted at a frequency of 1 Hz, at a load ratio R = 0.25 and load range of 3.7 - 4.5 KN according to the materials' ultimate tensile strength. The COD was monitored by a clip gauge attached to the specimen with a testing rate set at 1 mm/min. Moreover, a thermal camera was set for thermographic monitoring of the crack opening displacement. Compact tension (CT) specimens were prepared for fracture toughness tests according to ASTM E399. The thickness B of the specimens was 9.2

Provisional KQ values were calculated according to ASTM E399 standard for all specimens according to Equations (1) and (2), where Pq = Pmax. Load versus displacement curves for Al/SiCp composites and unreinforced aluminium alloys are shown in Fig. 13. Fracture toughness data for Al/SiCp and unreinforced aluminium alloys are summarised in Table 2.

Fig. 13. Load – Displacement curves for Al/SiCp composites subjected to T1, T6 and HT-1

heat treatment conditions and three unreinforced aluminium alloy samples

mm for the MMC, and 5 mm for the unreinforced aluminium alloys.

fracture mechanics approach is applicable.

probably indicated the additive effects of dislocation interaction with both the alloy precipitates and the SiC reinforcement. The combination increased the strain in the matrix by increasing the number of dislocations and requiring higher flow stresses for deformation, resulting in the higher strengths observed. Ductility of SiC/Al composites, as measured by strain to failure, is again a complex interaction of parameters. However, the prime factors affecting these properties are the reinforcement content, heat treatment and precipitation hardening.

Fig. 12. Young's Modulus vs. Processing Conditions curves showing T6 treated composites having the highest modulus

#### **2.6 Fracture toughness KIC testing**

The plane strain fracture toughness test involves the loading to failure of fatigue precracked, notched specimens in tension or in three-point bending. The calculation of a valid toughness value can only be determined after the test is completed, via examination of the load-displacement curve and measurement of the fatigue-crack length. The provisional fracture toughness value, KQ, is first calculated from the following equation:

$$K\_Q = \left(\frac{P\_Q}{BW^{1/2}}\right) \cdot f\left(\frac{a}{W}\right) \tag{3}$$

where PQ is the load corresponding to a defined increment of crack length, B is the specimen's thickness, W is the width of the specimen, and f(α/W) is a geometry dependent factor that relates the compliance of the specimen to the ratio of the crack length and width, expressed as follows:

$$f\left(\frac{a}{\mathcal{W}}\right) = \frac{\left(2 + a \,/\,\mathcal{W}\right)\left(0.86 + 4.64a \,/\,\mathcal{W} - 13.32a^2 \,/\,\mathcal{W}^2 + 14.72a^3 \,/\,\mathcal{W}^3 - 5.6a^4 \,/\,\mathcal{W}^4\right)}{\left(1 - a \,/\,\mathcal{W}\right)^{3/2}} \tag{4}$$

probably indicated the additive effects of dislocation interaction with both the alloy precipitates and the SiC reinforcement. The combination increased the strain in the matrix by increasing the number of dislocations and requiring higher flow stresses for deformation, resulting in the higher strengths observed. Ductility of SiC/Al composites, as measured by strain to failure, is again a complex interaction of parameters. However, the prime factors affecting these properties are the reinforcement content, heat treatment and precipitation

**Al A359 / 20%SiC Composite Al A359 / 31%SiC Composite** 

T6-20% HT-1-20% T1-20% T6-31% HT-1-31% T1-31%

T6-20% HT-1-20% T1-20% T6-31% HT-1-31% T1-31% **Materials - processing Conditions**

Fig. 12. Young's Modulus vs. Processing Conditions curves showing T6 treated composites

The plane strain fracture toughness test involves the loading to failure of fatigue precracked, notched specimens in tension or in three-point bending. The calculation of a valid toughness value can only be determined after the test is completed, via examination of the load-displacement curve and measurement of the fatigue-crack length. The provisional

> 1/2 *Q*

where PQ is the load corresponding to a defined increment of crack length, B is the specimen's thickness, W is the width of the specimen, and f(α/W) is a geometry dependent factor that relates the compliance of the specimen to the ratio of the crack length and width,

3/2

(2 / )(0.86 4.64 / 13.32 / 14.72 / 5.6 / ) (1 / ) *a aW aW a W a W a W <sup>f</sup> <sup>W</sup> a W* ⎛ ⎞ + +− + − <sup>=</sup> ⎜ ⎟

⎝ ⎠ <sup>−</sup> (4)

⎛ ⎞ ⎛ ⎞ = ⋅ ⎜ ⎟ ⎜ ⎟ ⎝ ⎠ ⎝ ⎠ (3)

22 33 44

*<sup>P</sup> <sup>a</sup> K f BW <sup>W</sup>*

fracture toughness value, KQ, is first calculated from the following equation:

*Q*

hardening.

90

having the highest modulus

expressed as follows:

**2.6 Fracture toughness KIC testing** 

95

100

105

**Young's Modulus(GPa)**

110

115

120

Only when specific validity criteria are satisfied, the provisional fracture toughness, KQ, can be quoted as the valid plane strain fracture toughness, KIC. The standard used for conducting this experiment, ASTM E399, imposes strict validity criteria to ensure that the plane strain conditions are satisfied during the test. These criteria include checks on the form and shape of the load versus displacement curve, requirements on specimen's size and crack geometry, and the 0.2% proof strength values at the test temperature. Essentially, these conditions are designed to ensure that the plastic zone size associated with the precrack is small enough so that plane strain conditions prevail, and that the linear elastic fracture mechanics approach is applicable.

Fracture toughness tests were conducted using a servo-hydraulic universal testing machine with data acquisition controller. The system was operated on load control for the fatigue pre-cracking stage, and on position control for the crack opening displacement (COD) testing. The fatigue test for pre-cracking was conducted at a frequency of 1 Hz, at a load ratio R = 0.25 and load range of 3.7 - 4.5 KN according to the materials' ultimate tensile strength. The COD was monitored by a clip gauge attached to the specimen with a testing rate set at 1 mm/min. Moreover, a thermal camera was set for thermographic monitoring of the crack opening displacement. Compact tension (CT) specimens were prepared for fracture toughness tests according to ASTM E399. The thickness B of the specimens was 9.2 mm for the MMC, and 5 mm for the unreinforced aluminium alloys.

Provisional KQ values were calculated according to ASTM E399 standard for all specimens according to Equations (1) and (2), where Pq = Pmax. Load versus displacement curves for Al/SiCp composites and unreinforced aluminium alloys are shown in Fig. 13. Fracture toughness data for Al/SiCp and unreinforced aluminium alloys are summarised in Table 2.

Fig. 13. Load – Displacement curves for Al/SiCp composites subjected to T1, T6 and HT-1 heat treatment conditions and three unreinforced aluminium alloy samples

Deformation Characteristics of Aluminium Composites for Structural Applications 363

A rectangular area on the specimen, located just in front of the initial pre-cracking region,

Fig. 14.a CT specimen showing the selected area for thermographic monitoring

the crack was propagated in-plane throughout the experiment.

improvement of the fracture toughness of the material.

The development of fracture was monitored in that area using infrared thermography. The mean temperature in this area versus time during crack growth was calculated using the recorded thermal imprint. As the specimen was stretched in tension, stresses were accumulating in the specimen, and the temperature increased as a function of time. When the accumulated energy became sufficient to propagate the crack, crack growth was observed, resulting in the stress relief. This corresponded to a peak in the temperature-time curve followed by a sudden decrease in temperature. As shown in Fig. 14b, 14c and 14d this behaviour was recurrent until the failure of the specimen. In these figures the thermographic monitoring of Aluminium 2xxx alloy, Al/SiCp T6 composite, and Al/SiCp HT1 composite samples is presented respectively. The different stages of crack growth for each material up to the final fracture of the specimen can be clearly observed. Just prior to fracture, the plasticity zone was clearly delineated on the specimen's surface as a heated region, which may be readily attributed to local plastic deformation. Furthermore, as seen in all figures,

A comparison of the thermography graphs in Figs. 14b, 14c, and 14d leads to the conclusion that the aluminium alloy exhibited different crack propagation behaviour than the Al/SiCp composites. For the aluminium alloy, the temperature versus time curve in Fig. 5b showed extended plasticity behaviour before final fracture occurred. This behaviour was evidenced by the constant increase in temperature between the temperature picks at the 60th and 140th second (figure 14b). This behaviour may be attributed to the small specimen thickness. However, for the T6 heat treated composite material in Fig. 14c, fracture was more elastic as the multiple temperature peaks indicated a confinement of the plasticity zone. Also, plasticity was formed in a more balanced way regarding the overall fracture process. It was also observed that T6 heat treated composites exhibited fewer picks compared to the HT1 heat treated specimens (Fig. 14d). This was attributed to the presence of a stronger interface in the T6 material as the accumulation of precipitates near the interface, resulted in the

was selected, as shown in Fig. 14a.


\*\*Validity criteria:

1 Excessive crack curvature

2. Thickness criteria not satisfied

3. Excessive plasticity

4. a/W out of range

5. Non-symmetrical crack front

6. In plane crack propagation

Table 2. Fracture toughness data for Al/SiCp and Al alloys and test validity

As is shown in Table 2, Al/SiCp composites exhibited lower provisional KQ values than the reference unreinforced aluminium alloys. In addition, heat treatment processing, and especially T6 treated specimen, had the highest KQ values compared to the other two heat treatment conditions. According to the load-displacement curves in Figure 3, composites clearly showed more brittle behaviour than the unreinforced aluminium alloys. T6 heat treated composites have the highest strength, but the lowest ductility compared to the other materials. Although these results provide some insight regarding the fracture behaviour of the materials examined, specific validity criteria have to be satisfied in order to obtain KIC values.

### **2.7 Examination by infrared thermography**

Nondestructive evaluation techniques are powerful tools for monitoring damage in composite materials [24]. Infrared thermography was used to monitor the plane crack propagation behaviour of particulate-reinforced AMCs [25, 26]. The deformation of solid materials is almost always accompanied by heat release. When the material becomes deformed or is damaged and fractured, a part of the energy necessary to initiate and propagate the damage is transformed in an irreversible way into heat [26]. The heat wave, generated by the thermo-mechanical coupling and the intrinsic dissipated energy during mechanical loading of the sample, is detected by the thermal camera. By using an adapted detector, thermography records the two dimensional ''temperature'' field as it results from the infrared radiation emitted by the object. The principal advantage of infrared thermography is its noncontact, non-destructive character.

Table 2. Fracture toughness data for Al/SiCp and Al alloys and test validity

As is shown in Table 2, Al/SiCp composites exhibited lower provisional KQ values than the reference unreinforced aluminium alloys. In addition, heat treatment processing, and especially T6 treated specimen, had the highest KQ values compared to the other two heat treatment conditions. According to the load-displacement curves in Figure 3, composites clearly showed more brittle behaviour than the unreinforced aluminium alloys. T6 heat treated composites have the highest strength, but the lowest ductility compared to the other materials. Although these results provide some insight regarding the fracture behaviour of the materials examined, specific validity criteria have to be satisfied in order to obtain KIC

Nondestructive evaluation techniques are powerful tools for monitoring damage in composite materials [24]. Infrared thermography was used to monitor the plane crack propagation behaviour of particulate-reinforced AMCs [25, 26]. The deformation of solid materials is almost always accompanied by heat release. When the material becomes deformed or is damaged and fractured, a part of the energy necessary to initiate and propagate the damage is transformed in an irreversible way into heat [26]. The heat wave, generated by the thermo-mechanical coupling and the intrinsic dissipated energy during mechanical loading of the sample, is detected by the thermal camera. By using an adapted detector, thermography records the two dimensional ''temperature'' field as it results from the infrared radiation emitted by the object. The principal advantage of infrared

\*\*Validity criteria:

values.

3. Excessive plasticity 4. a/W out of range

1 Excessive crack curvature 2. Thickness criteria not satisfied

5. Non-symmetrical crack front 6. In plane crack propagation

**2.7 Examination by infrared thermography** 

thermography is its noncontact, non-destructive character.

A rectangular area on the specimen, located just in front of the initial pre-cracking region, was selected, as shown in Fig. 14a.

Fig. 14.a CT specimen showing the selected area for thermographic monitoring

The development of fracture was monitored in that area using infrared thermography. The mean temperature in this area versus time during crack growth was calculated using the recorded thermal imprint. As the specimen was stretched in tension, stresses were accumulating in the specimen, and the temperature increased as a function of time. When the accumulated energy became sufficient to propagate the crack, crack growth was observed, resulting in the stress relief. This corresponded to a peak in the temperature-time curve followed by a sudden decrease in temperature. As shown in Fig. 14b, 14c and 14d this behaviour was recurrent until the failure of the specimen. In these figures the thermographic monitoring of Aluminium 2xxx alloy, Al/SiCp T6 composite, and Al/SiCp HT1 composite samples is presented respectively. The different stages of crack growth for each material up to the final fracture of the specimen can be clearly observed. Just prior to fracture, the plasticity zone was clearly delineated on the specimen's surface as a heated region, which may be readily attributed to local plastic deformation. Furthermore, as seen in all figures, the crack was propagated in-plane throughout the experiment.

A comparison of the thermography graphs in Figs. 14b, 14c, and 14d leads to the conclusion that the aluminium alloy exhibited different crack propagation behaviour than the Al/SiCp composites. For the aluminium alloy, the temperature versus time curve in Fig. 5b showed extended plasticity behaviour before final fracture occurred. This behaviour was evidenced by the constant increase in temperature between the temperature picks at the 60th and 140th second (figure 14b). This behaviour may be attributed to the small specimen thickness. However, for the T6 heat treated composite material in Fig. 14c, fracture was more elastic as the multiple temperature peaks indicated a confinement of the plasticity zone. Also, plasticity was formed in a more balanced way regarding the overall fracture process. It was also observed that T6 heat treated composites exhibited fewer picks compared to the HT1 heat treated specimens (Fig. 14d). This was attributed to the presence of a stronger interface in the T6 material as the accumulation of precipitates near the interface, resulted in the improvement of the fracture toughness of the material.

Deformation Characteristics of Aluminium Composites for Structural Applications 365

(d) Fig. 14. Thermographic monitoring of various CT specimens showing the different stages of crack growth up to the specimen's final fracture: (b) Aluminium 2xxx, (c) Al/SiCp T6

A model proposed by McMahon and Vitek [27] predicts the fracture resistance of a ductile material that fails by an intergranular mechanism. Based on this model, an effective work parameter can be developed to predict fracture strength of an interface at a segregated state using Griffith crack-type arguments. The Griffith's equation, which was derived for elastic body, is applied here because it is assumed that the yielding zone size ahead of the crack is small enough and the fracture is governed by the elastic stress field. The model further assumes that small changes in interfacial energy caused by segregation of impurities at the interface will result in a much larger change in the work of fracture. This is due to the fact that the work of fracture must be provided by a dislocation pile-up mechanism around the advancing crack-tip on the interface. This implies that additional work must be provided to deform the material at the crack-tip in addition to the work needed to overcome the interface energy and to replace it with two surfaces. The definition of interfacial fracture

int

<sup>=</sup> (5)

100 *PE d* ε

π

E is Young's modulus, d is the particle thickness, since it is assumed that cracks of the order of the particle size are present when considering crack propagation through the interface

**2.8 A model for predicting interfacial strengthening behaviour of particulate** 

int

σ

composite, (d) Al/SiCp HT1 composite

**reinforced AMCs** 

strength, σint, is then given by:

where,

#### (b)

Fig. 14. Thermographic monitoring of various CT specimens showing the different stages of crack growth up to the specimen's final fracture: (b) Aluminium 2xxx, (c) Al/SiCp T6 composite, (d) Al/SiCp HT1 composite

### **2.8 A model for predicting interfacial strengthening behaviour of particulate reinforced AMCs**

A model proposed by McMahon and Vitek [27] predicts the fracture resistance of a ductile material that fails by an intergranular mechanism. Based on this model, an effective work parameter can be developed to predict fracture strength of an interface at a segregated state using Griffith crack-type arguments. The Griffith's equation, which was derived for elastic body, is applied here because it is assumed that the yielding zone size ahead of the crack is small enough and the fracture is governed by the elastic stress field. The model further assumes that small changes in interfacial energy caused by segregation of impurities at the interface will result in a much larger change in the work of fracture. This is due to the fact that the work of fracture must be provided by a dislocation pile-up mechanism around the advancing crack-tip on the interface. This implies that additional work must be provided to deform the material at the crack-tip in addition to the work needed to overcome the interface energy and to replace it with two surfaces. The definition of interfacial fracture strength, σint, is then given by:

$$
\sigma\_{\rm int} = \sqrt{\frac{100\varepsilon\_P E\_{\rm int}}{\pi d}}\tag{5}
$$

where,

364 Recent Trends in Processing and Degradation of Aluminium Alloys

(b)

(c)

E is Young's modulus, d is the particle thickness, since it is assumed that cracks of the order of the particle size are present when considering crack propagation through the interface

Deformation Characteristics of Aluminium Composites for Structural Applications 367

particle distribution is similar to FCC structure in metals. The fracture toughness of the

( ) int <sup>2</sup> 2 (1 3 ) *<sup>p</sup> <sup>m</sup> IC m m m mm m*

where KIC, Kp = 3 MPa m-1/2, Km = 35 MPa m-1/2, and Kint is the fracture toughness of the composite, SiC particulates, A359 aluminium alloy matrix, and interface, respectively. Lp and Lm are the stress carrying capabilities of a particulate and the matrix, respectively. On average, for SiC particles and aluminium alloy matrix, Lp ~ Lm ~ 2. The value of Lm = 1 is applicable for clean surfaces. However, due to processing conditions and the physical interaction at the matrix/reinforcement interface the interface contains partially contaminated surfaces, therefore Lm = 2 since it cannot be considered as a "clean surface". Vm and (Vm-V'm) are the area fractions for particle cracking and interface failure, respectively. These area fractions though are not accurately known. However Wang and Zhang [33] found that the ratio of particle cracking over interface failure Vm/ (Vm-V'm) was about 0.13 (= 1.4%/10.7 %)

Young's modulus of matrix has been obtained for A359 aluminium matrix. The particles Ep,

Due to the fact that the difference (*V V f f* ) ′ − is very small, a good approximation is to consider that the Young's modulus of the interface region is close to that of the matrix;

The parameter B describes the modification of the boundary energy by impurities using the

where ε2-ε1 is the difference between the formation energy in the impurity in the bulk and the interface region. It is assumed that the values of the surface energy and the impurity formation energy in the bulk are close, and therefore the numerator in the exponential term depends on the impurity formation energy in the interface region, which is assumed to be

where, εs is the surface energy required forming the impurity atom and εe is the elastic energy involved with inserting an impurity atom into a matrix lattice site. This is given by:

( ) 0.5 <sup>8</sup> <sup>2</sup>

*<sup>G</sup> a a a eV*

*f mi m*

π

*e*

ε −ε

Using Faulkner's approach [36], to the derivation of impurity formation energy,

1.94 3 *S*

ε

0.75 εf, where εf is the formation energy of the impurity in the bulk.

ε

1 2 0.75 *<sup>F</sup> RT RT Be e* ⎛ ⎞⎛ ⎞

 ε

= + −+ + − ′ ′ <sup>+</sup> (9)

( ) ( ) 2/3 2/3 2/3 2/3 <sup>1</sup> *E Ev E V E V V Cp m f i <sup>f</sup> f f* = +− + − ′ ′ (10)

⎜ ⎟⎜ ⎟ ⎝ ⎠⎝ ⎠ = ≅ (11)

εf = εs + εe (12)

=+ − (13)

*<sup>K</sup> K K K V VV VK V*

*p pm m*

*L LL L*

composite can then be written as [32]:

in a SiC particle-reinforced aluminium alloy composite.

matrix Em, and interface Ei shown in equation

Zuchovitsky equations [34, 35], given by:

εS is the surface energy (1.02 J m-2)

*E E i m* ≅ [32].

where,

and the particulate, εp is the energy required to create two fracture surfaces = 2εs – εgb (= εο), with εs, the surface energy, and εgb, the grain boundary energy.

The 100 εp component allows for dislocation interaction and movement ahead of the cracktip in ductile materials. This refers to the work required for a total separation of the lattice planes, which is equal to the area under the force-extension curve.

From equation (6) εp can be estimated if Kint (Interface fracture toughness) and Eint (Interface Young's Modulus) are known.

$$\frac{{K\_{\rm int}}^2}{100E\_{\rm int}} = \varepsilon\_p \left(1 - \frac{ZRT\ln\left(1 - c + Bc\right)}{\varepsilon\_p}\right)^n \tag{6}$$

Where,

Z, describes the density of interface sites which are disordered enough to act as segregation sites (= D ρS), with D the thickness of the interface region, and ρS the density of the interface region (D=300 nm) (ρ= 2.6889 g/cm3 for Aluminium and 3.22 g/cm³ for SiC),

R, is the gas constant (= 8.314472(15) J•K-1•mol-1),

T, is the absolute temperature (= 803.15 K for T6, = 723.15 K for HT1),

c, is the segregate concentration needed to cause embrittlement (= 0.1 for pure aluminium),

B, describes the modification of the boundary energy by impurities using the Zuchovitsky equations,

n, is the work hardening exponent (= 10 for FCC aluminium).

In hard particle reinforced metal matrix composites the stress transfer from the matrix to the particles is mainly controlled by the misfit of the elastic constants between the two phases [28]. To measure the stress transfer to the particle, in an homogeneous material subjected to tensile loading, the stress carrying capability of the particle is defined as the ratio of the normal stress σN to the particle in the loading direction to the macroscopic tensile stress, σT, i.e. the ratio L = σN / σT. By using Eshelby's theory, the stress carrying capability of a spherical inhomogeneity can be written as [29]:

$$L = \frac{9\mathbf{x}(2+3\mathbf{x})}{(1+2\mathbf{x})(8+7\mathbf{x})} \tag{7}$$

where, x = Ei / Em, and Ei and Em are Young's moduli for inhomogeneity and matrix, respectively.

Furthermore, the shear lag model, originally developed by Cox [30] modified by Llorca [31], can be used to estimate the stress carrying capability of a particulate, assuming that the volume fraction of reinforcement is small and the average stress in the matrix is approximately equal to the applied stress:

$$L = 1 + \frac{a}{\sqrt{3}}\tag{8}$$

where 2 *h a <sup>r</sup>* <sup>=</sup> is the aspect ratio of the reinforcement, with *h* and *r* the average length and the average diameter of the particle.

A model has been proposed to estimate the effects of particle volume fraction on fracture toughness in SiC particle-reinforced aluminium alloy matrix composites. This model assumes that SiC particles are uniformly distributed in the matrix and that the pattern of particle distribution is similar to FCC structure in metals. The fracture toughness of the composite can then be written as [32]:

$$K\_{IC} = \frac{K\_p}{L\_p}V\_m' + \frac{2K\_{\text{int}}}{L\_p + L\_m}(V\_m - V\_m') + \frac{K\_m}{L\_m}2V\_m + K\_m(1 - 3V\_m) \tag{9}$$

where KIC, Kp = 3 MPa m-1/2, Km = 35 MPa m-1/2, and Kint is the fracture toughness of the composite, SiC particulates, A359 aluminium alloy matrix, and interface, respectively. Lp and Lm are the stress carrying capabilities of a particulate and the matrix, respectively. On average, for SiC particles and aluminium alloy matrix, Lp ~ Lm ~ 2. The value of Lm = 1 is applicable for clean surfaces. However, due to processing conditions and the physical interaction at the matrix/reinforcement interface the interface contains partially contaminated surfaces, therefore Lm = 2 since it cannot be considered as a "clean surface". Vm and (Vm-V'm) are the area fractions for particle cracking and interface failure, respectively. These area fractions though are not accurately known. However Wang and Zhang [33] found that the ratio of particle cracking over interface failure Vm/ (Vm-V'm) was about 0.13 (= 1.4%/10.7 %) in a SiC particle-reinforced aluminium alloy composite.

Young's modulus of matrix has been obtained for A359 aluminium matrix. The particles Ep, matrix Em, and interface Ei shown in equation

$$E\_{\rm C} = E\_p v\_f^{2/3} + E\_m \left(1 - V\_f'^{2/3}\right) + E\_i \left(V\_f'^{2/3} - V\_f^{2/3}\right) \tag{10}$$

Due to the fact that the difference (*V V f f* ) ′ − is very small, a good approximation is to consider that the Young's modulus of the interface region is close to that of the matrix; *E E i m* ≅ [32].

The parameter B describes the modification of the boundary energy by impurities using the Zuchovitsky equations [34, 35], given by:

$$B = e^{\left(\frac{x\_1 - x\_2}{RT}\right)} \equiv e^{\left(\frac{0.75x\_F}{RT}\right)}\tag{11}$$

where ε2-ε1 is the difference between the formation energy in the impurity in the bulk and the interface region. It is assumed that the values of the surface energy and the impurity formation energy in the bulk are close, and therefore the numerator in the exponential term depends on the impurity formation energy in the interface region, which is assumed to be 0.75 εf, where εf is the formation energy of the impurity in the bulk.

Using Faulkner's approach [36], to the derivation of impurity formation energy,

$$
\varepsilon\_{\ell} = \varepsilon\_{s} + \varepsilon\_{c} \tag{12}
$$

where, εs is the surface energy required forming the impurity atom and εe is the elastic energy involved with inserting an impurity atom into a matrix lattice site. This is given by:

$$
\varepsilon\_f = \frac{0.5\varepsilon\_S}{1.94} + \frac{8\pi G}{3e} a\_m \left(a\_i - a\_m\right)^2 eV \tag{13}
$$

where,

366 Recent Trends in Processing and Degradation of Aluminium Alloys

and the particulate, εp is the energy required to create two fracture surfaces = 2εs – εgb (= εο),

The 100 εp component allows for dislocation interaction and movement ahead of the cracktip in ductile materials. This refers to the work required for a total separation of the lattice

From equation (6) εp can be estimated if Kint (Interface fracture toughness) and Eint

( ) <sup>2</sup>

Z, describes the density of interface sites which are disordered enough to act as segregation sites (= D ρS), with D the thickness of the interface region, and ρS the density of the interface

c, is the segregate concentration needed to cause embrittlement (= 0.1 for pure aluminium), B, describes the modification of the boundary energy by impurities using the Zuchovitsky

In hard particle reinforced metal matrix composites the stress transfer from the matrix to the particles is mainly controlled by the misfit of the elastic constants between the two phases [28]. To measure the stress transfer to the particle, in an homogeneous material subjected to tensile loading, the stress carrying capability of the particle is defined as the ratio of the normal stress σN to the particle in the loading direction to the macroscopic tensile stress, σT, i.e. the ratio L = σN / σT. By using Eshelby's theory, the stress carrying capability of a

> 9 (2 3 ) (1 2 )(8 7 ) *x x <sup>L</sup>*

<sup>+</sup> <sup>=</sup> + +

where, x = Ei / Em, and Ei and Em are Young's moduli for inhomogeneity and matrix,

Furthermore, the shear lag model, originally developed by Cox [30] modified by Llorca [31], can be used to estimate the stress carrying capability of a particulate, assuming that the volume fraction of reinforcement is small and the average stress in the matrix is

1

A model has been proposed to estimate the effects of particle volume fraction on fracture toughness in SiC particle-reinforced aluminium alloy matrix composites. This model assumes that SiC particles are uniformly distributed in the matrix and that the pattern of

3

*<sup>r</sup>* <sup>=</sup> is the aspect ratio of the reinforcement, with *h* and *r* the average length

*x x*

*K RT c Bc*

1

*p*

ε

region (D=300 nm) (ρ= 2.6889 g/cm3 for Aluminium and 3.22 g/cm³ for SiC),

T, is the absolute temperature (= 803.15 K for T6, = 723.15 K for HT1),

n, is the work hardening exponent (= 10 for FCC aluminium).

ln 1

⎛ ⎞ Ζ −+ = − ⎜ ⎟ ⎝ ⎠

*p*

ε

*n*

(6)

(7)

*<sup>a</sup> <sup>L</sup>* = + (8)

with εs, the surface energy, and εgb, the grain boundary energy.

planes, which is equal to the area under the force-extension curve.

int int

*E*

100

R, is the gas constant (= 8.314472(15) J•K-1•mol-1),

spherical inhomogeneity can be written as [29]:

approximately equal to the applied stress:

and the average diameter of the particle.

(Interface Young's Modulus) are known.

Where,

equations,

respectively.

where 2 *h a*

εS is the surface energy (1.02 J m-2)

Deformation Characteristics of Aluminium Composites for Structural Applications 369

in the microstructure of the composite at the vicinity of the interphase area, which results to the composite hardening. The creation of the interphase together with the improved stress transfer may be regarded as the main contributing parameters to the improved mechanical properties of the particulate reinforced composite. The improved static strength is followed by a less spectacular performance in fatigue, with the fatigue limit of the material falling to

**S-N Curve Al/SiC 20%**

0 200000 400000 600000 800000 1000000 **No.Cycles to Failure**

To study the crack growth rate (da/dN) vs. stress intensity range (ΔK) data for aluminium SiCp composites and aluminium 2xxx series specimens, the materials were subjected to cyclic loading. Fatigue crack growth tests were conducted according to the ASTM E647 standard using a servo hydraulic testing machine. The tests were conducted under load control. Compact tension (CT) specimens were prepared for the fatigue crack growth experiments. The fatigue tests for the monolithic aluminium specimens were performed at a standard frequency of 5 Hz. However, a lower frequency of 1 Hz was selected for the composite specimens in order to minimize the effect of sudden failure due to the brittle nature of these materials. The experiments were performed at a load ratio R = 0.25 and maximum load ranges of 3.7 - 4.5 KN, keeping the maximum stress at about 70% of the

The technique used for determining the crack growth rate versus stress intensity range during the cyclic loading tests was based on non-contact monitoring of the crack propagation by lock-in thermography. This new technique deals with mapping the crack growth nondestructively. Lock–in thermography is based on remote full field monitoring of thermal waves generated inside the specimen by cyclic loading that caused an oscillating temperature field in the stationary regime. Lock–in refers to the necessity for monitoring the

the 70% of the UTS.

65

material's ultimate tensile strength.

Fig. 15. S-N Curve of Al/SiC 20% Composite

**2.9.1 Crack growth rate vs. range of stress intensity** 

75

85

**Stress %**

95

e is the electronic charge (1.60217646 \*10^19 Coulomb)

ai is the impurity atomic radius (0.118 nm for Si)

am is the matrix atomic radius (0.143 nm for aluminium)

G is the shear modulus (26 GPa for aluminium)

By performing the calculations the impurity formation energy, εf, for A359 aluminium alloy (Al-Si-Mg) can be determined and then substituted in equation (11) to calculate B.

The micro-mechanics model described above is based on thermodynamics principles and is used to determine the fracture strength of the interface at a segregated state in aluminium matrix composites. This model uses energy considerations to express the fracture toughness of the interface in terms of interfacial critical strain energy release rate and elastic modulus. The interfacial fracture toughness is further expressed as a function of the macroscopic fracture toughness and mechanical properties of the composite, using a toughening mechanism model based on stress transfer mechanism. Mechanical testing is also performed to obtain macroscopic data, such as the fracture strength, elastic modulus and fracture toughness of the composite, which are used as input to the model. Based on the experimental data and the analysis, the interfacial strength is determined for SiC particlereinforced aluminium matrix composites subjected to different heat treatment processing conditions and the results are shown in table 2. It is observed that Kint values are close to the K1c values of the composites. Furthermore, σint values found to be dependent on the heat treatment processing with T6 heat treatment composite obtain the highest interfacial fracture strength.

### **2.9 Fatigue testing and crack growth behaviour**

Tension-tension fatigue tests were conducted using a hydraulic testing machine. The system was operated under load control, applying a harmonic tensile stress with constant amplitude. By specifying the maximum and the minimum stress levels, the other stress parameters could be easily determined. These were the stress range, σr, stress amplitude, σa, mean stress, σm, and fatigue stress ratio, R (=σmin/σmax). Throughout this study, all fatigue tests were carried out at a frequency of 5 Hz and at a stress ratio R = 0.1. Different stress levels between the ultimate tensile strength (UTS) and the fatigue limit were selected, resulting in so-called Wöhler or S-N curves. Tests exceeding 106 cycles without specimen failure were terminated. Specimens that failed in or close to the grips were discarded. The geometry of the samples was the same as those used for the tensile characterisation, i.e. rectangular strips of 12.5mm width, and 1.55mm thickness.

The normalised "S-N" curves of the fatigue response of the Al/SiC composites is shown in Fig. 15. The stress was normalised over the UTS of each material and plotted against the number of cycles to failure. As can be observed, whereas in the untreated T1 condition the composite retains at least 85% of its strength as fatigue strength, the corresponding value for the T6 heat treatment is falling to the 70% of UTS. The HT1 heat treatment is exhibiting an intermediate behaviour, with its fatigue strength falling to 75% of the corresponding UTS value. It can be concluded that aggressive heat treatment reduces the damage tolerance of the composites.

A direct comparison of the fatigue performance of the composite with the corresponding quasi static performance in tension reveals that the T6 heat treatment improved the strength of the composite. This can be attributed to a dominant mechanism related to the changes in the microstructure of the composite. This mechanism relates to the precipitations appearing

By performing the calculations the impurity formation energy, εf, for A359 aluminium alloy

The micro-mechanics model described above is based on thermodynamics principles and is used to determine the fracture strength of the interface at a segregated state in aluminium matrix composites. This model uses energy considerations to express the fracture toughness of the interface in terms of interfacial critical strain energy release rate and elastic modulus. The interfacial fracture toughness is further expressed as a function of the macroscopic fracture toughness and mechanical properties of the composite, using a toughening mechanism model based on stress transfer mechanism. Mechanical testing is also performed to obtain macroscopic data, such as the fracture strength, elastic modulus and fracture toughness of the composite, which are used as input to the model. Based on the experimental data and the analysis, the interfacial strength is determined for SiC particlereinforced aluminium matrix composites subjected to different heat treatment processing conditions and the results are shown in table 2. It is observed that Kint values are close to the K1c values of the composites. Furthermore, σint values found to be dependent on the heat treatment processing with T6 heat treatment composite obtain the highest interfacial

Tension-tension fatigue tests were conducted using a hydraulic testing machine. The system was operated under load control, applying a harmonic tensile stress with constant amplitude. By specifying the maximum and the minimum stress levels, the other stress parameters could be easily determined. These were the stress range, σr, stress amplitude, σa, mean stress, σm, and fatigue stress ratio, R (=σmin/σmax). Throughout this study, all fatigue tests were carried out at a frequency of 5 Hz and at a stress ratio R = 0.1. Different stress levels between the ultimate tensile strength (UTS) and the fatigue limit were selected, resulting in so-called Wöhler or S-N curves. Tests exceeding 106 cycles without specimen failure were terminated. Specimens that failed in or close to the grips were discarded. The geometry of the samples was the same as those used for the tensile characterisation, i.e.

The normalised "S-N" curves of the fatigue response of the Al/SiC composites is shown in Fig. 15. The stress was normalised over the UTS of each material and plotted against the number of cycles to failure. As can be observed, whereas in the untreated T1 condition the composite retains at least 85% of its strength as fatigue strength, the corresponding value for the T6 heat treatment is falling to the 70% of UTS. The HT1 heat treatment is exhibiting an intermediate behaviour, with its fatigue strength falling to 75% of the corresponding UTS value. It can be concluded that aggressive heat treatment reduces the damage tolerance of

A direct comparison of the fatigue performance of the composite with the corresponding quasi static performance in tension reveals that the T6 heat treatment improved the strength of the composite. This can be attributed to a dominant mechanism related to the changes in the microstructure of the composite. This mechanism relates to the precipitations appearing

(Al-Si-Mg) can be determined and then substituted in equation (11) to calculate B.

e is the electronic charge (1.60217646 \*10^19 Coulomb) ai is the impurity atomic radius (0.118 nm for Si)

am is the matrix atomic radius (0.143 nm for aluminium)

G is the shear modulus (26 GPa for aluminium)

**2.9 Fatigue testing and crack growth behaviour** 

rectangular strips of 12.5mm width, and 1.55mm thickness.

fracture strength.

the composites.

in the microstructure of the composite at the vicinity of the interphase area, which results to the composite hardening. The creation of the interphase together with the improved stress transfer may be regarded as the main contributing parameters to the improved mechanical properties of the particulate reinforced composite. The improved static strength is followed by a less spectacular performance in fatigue, with the fatigue limit of the material falling to the 70% of the UTS.

#### **S-N Curve Al/SiC 20%**

Fig. 15. S-N Curve of Al/SiC 20% Composite

### **2.9.1 Crack growth rate vs. range of stress intensity**

To study the crack growth rate (da/dN) vs. stress intensity range (ΔK) data for aluminium SiCp composites and aluminium 2xxx series specimens, the materials were subjected to cyclic loading. Fatigue crack growth tests were conducted according to the ASTM E647 standard using a servo hydraulic testing machine. The tests were conducted under load control. Compact tension (CT) specimens were prepared for the fatigue crack growth experiments. The fatigue tests for the monolithic aluminium specimens were performed at a standard frequency of 5 Hz. However, a lower frequency of 1 Hz was selected for the composite specimens in order to minimize the effect of sudden failure due to the brittle nature of these materials. The experiments were performed at a load ratio R = 0.25 and maximum load ranges of 3.7 - 4.5 KN, keeping the maximum stress at about 70% of the material's ultimate tensile strength.

The technique used for determining the crack growth rate versus stress intensity range during the cyclic loading tests was based on non-contact monitoring of the crack propagation by lock-in thermography. This new technique deals with mapping the crack growth nondestructively. Lock–in thermography is based on remote full field monitoring of thermal waves generated inside the specimen by cyclic loading that caused an oscillating temperature field in the stationary regime. Lock–in refers to the necessity for monitoring the

Deformation Characteristics of Aluminium Composites for Structural Applications 371

Fig. 16. da/dN vs. ΔK plots of Al/SiCp composite and monolithic aluminum 2xxx specimens

From the stress maxima versus fatigue cycles curves, for each reference line, shown in Figure 17, the crack lengths versus the number of fatigue cycles were determined for A359/SiCP composites in all three different thermal treatment conditions: T1, T6, and HT-1 (Fig. 18). As it is shown in Figure 18, the crack growth rate was found to be quite linear for all heat treatments. Also, there is a small change in the linear slope for the HT-1 heat treated sample, showing increased ductility, which indicates that more time (i.e. cycles) is needed for the crack to grow in this case. For the T6 heat treatment, the results depict a brittle behaviour, as the crack starts to grow earlier than in the other two cases, supporting

The stress intensity range was further calculated by the data shown in Figure 17. ΔK values have been estimated from the stress maxima versus fatigue cycles curves for each reference line, shown in Figure 17. Each of the four lines provides a stress intensity range and a da/dN value. The data obtained using lock-in thermography, shown in Figures 18 and 19, were correlated with crack growth rate values obtained by the conventional compliance method and calculations based on the Paris law. Furthermore, the da/dN vs. ΔK curves

**2.9.2 Estimation of da/dN vs. ΔK using thermography and compliance methods**  Using the procedure described above based on thermographic mapping, the local stress versus time was measured for the T6 heat treated Al/SiCp along each of the four reference lines placed in front of the CT specimen's notch. The maximum value of stress versus the number of fatigue cycles was then plotted for those four lines (Figure 17). As expected, Figure 17 shows that the local stress, monitored at the location of each line, increases as the crack is approaching that line, then attains a maximum when the crack tip is crossing the line. Finally, after the crack has crossed the line, the local stress measured at the location of the line decreases. This is also expected, since the stress values shown in Figure 17 are stress maxima from all the locations along the particular line. At the exact position on a line where

the crack has just crossed, the local stress is null as expected.

evidence of brittle fracture.

exact time dependence between the output signal (thermal wave) and the reference input signal (fatigue cycle). This is done with a lock–in amplifier so that both phase and magnitude images become available.

The detection system included an infrared camera. The camera was connected with the lockin amplifier, which was then connected to the main servo hydraulic controller. Therefore, synchronization of the frequency through the lock-in amplifier and the mechanical testing machine could be achieved and lock-in images and data capture during the fatigue testing were enabled.

The camera was firstly set at a distance close to the specimen, in order to have the best possible image capture. Then, the fatigue pre-cracking started while synchronizing, at the same frequency, fatigue cycles and infrared camera through the lock-in amplifier.

In order to determine the crack growth rate and calculate the stress intensity factor using thermographic mapping of the material undergoing fatigue a simple procedure was used:


Four lines of the same length, equally spaced at a distance of 1 mm, were set on the thermal images of the CT specimen at a distance in front of the specimen's notch.

In Figure 16, the crack growth rate for the heat treated composite specimens and the reference aluminium alloy samples are plotted on a logarithmic scale as a function of the stress intensity range. The results showed that the heat treatment processing influences crack growth behaviour of the composite materials. Specimens subjected to T6 heat treatment condition exhibited the highest crack growth rate vs. stress intensity range slope compared to the other composite systems. Moreover, the crack growth rate vs. stress intensity range line of specimens subjected to T6 heat treatment was shifted towards higher ΔΚ values compared to that from specimens subjected the other two heat treatment conditions. This implies that in order to attain the same crack growth rate, higher stress intensity factor is required for specimens subjected to T6 condition compared to those subjected to T1 and HT-1 conditions. The need for higher stresses for a crack to propagate reveals the material's microstructural strength, where micro-mechanisms such as precipitation hardening promote high stress concentrations at the crack tip, resulting in the toughening of the crack path. The above postulations agree with previews results, where the T6 heat treated composites showed superior strength but the lowest ductility compared to T1 or HT-1 heat treated specimens. Results, shown in Figure 16, indicate that at intermediate values of crack growth rate (10-2 to 10-5 mm/cycle) the Al/SiCp composites have fracture properties comparable to those of the unreinforced matrix alloys. It is obvious that in these composites crack propagation rate seems more balanced and takes more time than the aluminium alloy obtaining crack growth rate values from around 10-1 to 10-4 mm/cycle.

exact time dependence between the output signal (thermal wave) and the reference input signal (fatigue cycle). This is done with a lock–in amplifier so that both phase and

The detection system included an infrared camera. The camera was connected with the lockin amplifier, which was then connected to the main servo hydraulic controller. Therefore, synchronization of the frequency through the lock-in amplifier and the mechanical testing machine could be achieved and lock-in images and data capture during the fatigue testing

The camera was firstly set at a distance close to the specimen, in order to have the best possible image capture. Then, the fatigue pre-cracking started while synchronizing, at the

In order to determine the crack growth rate and calculate the stress intensity factor using thermographic mapping of the material undergoing fatigue a simple procedure was used: a. The distribution of temperature and stresses at the surface of the specimen was monitored during the test. Therefore, thermal images were obtained as a function of

b. The stresses were evaluated in a post-processing mode, along a series of equally spaced reference lines of the same length, set in front of the crack-starting notch. The idea was that the stress monitored at the location of a line versus time (or fatigue cycles) would exhibit an increase while the crack approaches the line, then attain a maximum when the crack tip was on the line. Due to the fact that the crack growth path could not be predicted and was not expected to follow a straight line in front of the notch, the stresses were monitored along a series of lines of a certain length, instead of a series of equally spaced points in front of the notch. The exact path of the crack could be easily

determined by looking at the stress maxima along each of these reference lines. Four lines of the same length, equally spaced at a distance of 1 mm, were set on the thermal

In Figure 16, the crack growth rate for the heat treated composite specimens and the reference aluminium alloy samples are plotted on a logarithmic scale as a function of the stress intensity range. The results showed that the heat treatment processing influences crack growth behaviour of the composite materials. Specimens subjected to T6 heat treatment condition exhibited the highest crack growth rate vs. stress intensity range slope compared to the other composite systems. Moreover, the crack growth rate vs. stress intensity range line of specimens subjected to T6 heat treatment was shifted towards higher ΔΚ values compared to that from specimens subjected the other two heat treatment conditions. This implies that in order to attain the same crack growth rate, higher stress intensity factor is required for specimens subjected to T6 condition compared to those subjected to T1 and HT-1 conditions. The need for higher stresses for a crack to propagate reveals the material's microstructural strength, where micro-mechanisms such as precipitation hardening promote high stress concentrations at the crack tip, resulting in the toughening of the crack path. The above postulations agree with previews results, where the T6 heat treated composites showed superior strength but the lowest ductility compared to T1 or HT-1 heat treated specimens. Results, shown in Figure 16, indicate that at intermediate values of crack growth rate (10-2 to 10-5 mm/cycle) the Al/SiCp composites have fracture properties comparable to those of the unreinforced matrix alloys. It is obvious that in these composites crack propagation rate seems more balanced and takes more time than the aluminium alloy obtaining crack growth rate values from around 10-1 to 10-4 mm/cycle.

images of the CT specimen at a distance in front of the specimen's notch.

same frequency, fatigue cycles and infrared camera through the lock-in amplifier.

magnitude images become available.

time and saved in the form of a movie.

were enabled.

Fig. 16. da/dN vs. ΔK plots of Al/SiCp composite and monolithic aluminum 2xxx specimens

### **2.9.2 Estimation of da/dN vs. ΔK using thermography and compliance methods**

Using the procedure described above based on thermographic mapping, the local stress versus time was measured for the T6 heat treated Al/SiCp along each of the four reference lines placed in front of the CT specimen's notch. The maximum value of stress versus the number of fatigue cycles was then plotted for those four lines (Figure 17). As expected, Figure 17 shows that the local stress, monitored at the location of each line, increases as the crack is approaching that line, then attains a maximum when the crack tip is crossing the line. Finally, after the crack has crossed the line, the local stress measured at the location of the line decreases. This is also expected, since the stress values shown in Figure 17 are stress maxima from all the locations along the particular line. At the exact position on a line where the crack has just crossed, the local stress is null as expected.

From the stress maxima versus fatigue cycles curves, for each reference line, shown in Figure 17, the crack lengths versus the number of fatigue cycles were determined for A359/SiCP composites in all three different thermal treatment conditions: T1, T6, and HT-1 (Fig. 18). As it is shown in Figure 18, the crack growth rate was found to be quite linear for all heat treatments. Also, there is a small change in the linear slope for the HT-1 heat treated sample, showing increased ductility, which indicates that more time (i.e. cycles) is needed for the crack to grow in this case. For the T6 heat treatment, the results depict a brittle behaviour, as the crack starts to grow earlier than in the other two cases, supporting evidence of brittle fracture.

The stress intensity range was further calculated by the data shown in Figure 17. ΔK values have been estimated from the stress maxima versus fatigue cycles curves for each reference line, shown in Figure 17. Each of the four lines provides a stress intensity range and a da/dN value. The data obtained using lock-in thermography, shown in Figures 18 and 19, were correlated with crack growth rate values obtained by the conventional compliance method and calculations based on the Paris law. Furthermore, the da/dN vs. ΔK curves

Deformation Characteristics of Aluminium Composites for Structural Applications 373

21

20

19

Crack a (mm)

composite

1.602

Crack growth Rate, da/dN (mm/cycle)

1.603

18

17

16

T1 thermography

HT1 thermography

N Cycles 1500 2000 2500 3000

5 15 25 ∆k, MPa √m

Fig. 20. da/dN vs. ΔK for Al/SiCp specimens - Thermography vs. Compliance method

T6 thermography

Fig. 19. Crack growth rate determined by compliance vs. thermography for A359/SiCp

Thermography T1 Compliance T1

> T1 MMC T6 MMC HT1 MMC T6 Thermography T1 Thermography HT1 Thermography

steaming from the compliance method were plotted in the same graph, for comparison purposes, with those obtained using lock-in thermography (Fig. 20). It can be seen in Figure 20 that the two different methods are in agreement, demonstrating that lock-in thermography is a credible nondestructive method for noncontact evaluation of the fracture behaviour of materials.

Fig. 17. T6 al/SiCp stress maxima along the four reference lines vs. number of fatigue cycles

Fig. 18. Crack length vs. cycles obtained from lock-in thermography data for A359/SiCp composites subjected to three different heat treatment conditions

steaming from the compliance method were plotted in the same graph, for comparison purposes, with those obtained using lock-in thermography (Fig. 20). It can be seen in Figure 20 that the two different methods are in agreement, demonstrating that lock-in thermography is a credible nondestructive method for noncontact evaluation of the fracture

Line C

Line D

0 5000 10000 15000 20000 25000 N Cycles

0 1000 2000 3000 4000 5000 N Cycles

Fig. 18. Crack length vs. cycles obtained from lock-in thermography data for A359/SiCp

y = 0.0024x + 12.915

y = 0.0011x + 14.774

T6 MMC HT1 MMC T1 MMC

Fig. 17. T6 al/SiCp stress maxima along the four reference lines vs. number of fatigue cycles

Line B

Line A Line B Line C Line D

behaviour of materials.

100

90

80

Line A

y = 0.0021x + 15.234

composites subjected to three different heat treatment conditions

Stress Max along the line (MPa)

21

20

19

18

Ceack lenght a (mm)

17

16

70

60

50

Fig. 19. Crack growth rate determined by compliance vs. thermography for A359/SiCp composite

Fig. 20. da/dN vs. ΔK for Al/SiCp specimens - Thermography vs. Compliance method

Deformation Characteristics of Aluminium Composites for Structural Applications 375

The results of fatigue testing of aluminium honeycomb sandwich panel both in air and in 3.5% sodium chloride solution, are plotted in Figure 22. A total of twenty cyclic deformation tests were conducted in fully tension-tension at a constant frequency of 2HZ, which is equivalent to two cycles per second. The results of fatigue testing of metallic foam sandwich

*Summary of Crack Propagation*

Long Air Tran Air Long Env Tran Env

> Long Air Tran Air Long Env Tran Env

1.0E+03 1.0E+04 1.0E+05 1.0E+06 1.0E+07 Number of Cycles to Failure

*Summary of Fatigue Crack Propagation*

1.0E+03 1.0E+04 1.0E+05 1.0E+06 1.0E+07 1.0E+08 Cycles to Failure

Cyclic deformation data reveals that honeycomb sandwich panel samples do produce consistent results with acceptable repeatability of results even though the honeycomb core is not a conventional structure due to its complex geometry, but because of its homogeneity, it does compare well to the consistent results we would expect from a less complex conventional aluminium solid sample. The results, also, revealed that samples taken from a longitudinal direction constantly have a longer life expectancy, of approximately 40%, then those samples taken from a transverse direction regardless of environmental exposure.

panel both in air and in 3.5% sodium chloride solution, are plotted in Figure 23.

**3.2 Results and discussion** 

0 0.2 0.4 0.6 0.8 1

0 0.2 0.4 0.6 0.8 1

Fig. 23. Metallic foam fatigue data

**3.2.1 Aluminium honeycomb** 

Load (KN

Fig. 22. Aluminium honeycomb fatigue data

Load (KN
