For crack prediction in permanent mold casting of lead-free bronze, elasto-viscoplastic and elasto-plastic thermal stress analyses were performed and compared with the casting tests. The elastoviscoplastic model was constructed by the method which enables to determine rheological properties without steady state stress in experiments. Maximum deviatoric principal (only by the elastoviscoplastic analysis) and hydrostatic stress were in reasonable agreement with the locations of cold cracking and hot tearing in the casting tests, respectively. It was suggested that considering the theological properties of semi-solid alloys was important especially to predict cracks during and after solidification

Fluxless bonding of plateless Cu-Cu substrates at processing temperature lower than 250 A degrees C and low pressure of 0.1 MPa was achieved by transient liquid phase sintering (TLPS) of mixed Cu nanoparticles and Sn-Bi eutectic powders. The effects of mixture composition, and sintering temperature on the shear strength, microstructure, and remelting temperature were investigated. Lowering the sintering temperature of Cu mixed with 65 weight percentage of Sn-Bi (Cu-65SnBi) resulted in decreased shear strength, however, at 200 A degrees C sintering temperature, the obtained highest shear strength was more than 20 MPa. It was found that it is essential to use Cu nanoparticles to accelerate the consumption so that no initial Sn-Bi phases remained after processing. The liquid phase generated at approximately 196 A degrees C during sintering from the reaction between newly formed Cu6Sn5 and Bi-phase was expected to facilitate the densification and strengthening of the joints. Although this newly generated liquid phase was known to solidify as hypereutectic Sn-Bi, by controlling the sintering temperature at 200 A degrees C, the remelting event at 139 A degrees C was not observed by differential scanning calorimetry. It is assumed that the proportion of solidified Sn-Bi eutectic phases in Cu-65SnBi that was sintered at 200 A degrees C were significantly small, hence, when reheated at 150 A degrees C, the obtained shear strength was equivalent to that at room temperature.

For crack prediction in permanent mold casting of lead-free bronze, elasto-viscoplastic and elasto-plastic thermal stress analyses were performed and compared with the casting tests. The elastoviscoplastic model was constructed by the method which enables to determine rheological properties without steady state stress in experiments. Maximum deviatoric principal (only by the elastoviscoplastic analysis) and hydrostatic stress were in reasonable agreement with the locations of cold cracking and hot tearing in the casting tests, respectively. It was suggested that considering the theological properties of semi-solid alloys was important especially to predict cracks during and after solidification

Bonding silicon carbide/gallium nitride (SiC/GaN) based power modules, particularly epoxy-molded modules to heat-substrate and/or heat sink, requires low processing temperature preferably lower than 250 degrees C, and low pressure as low as 0.1 MPa to prevent damage to the modules. In addition, due to the impracticality of depositing metal-plating to the epoxy-molded module, bonding of plateless Cu-substrates is in great demand. Furthermore, post-processing residual flux cleaning, which is costly and unfavorable to industry need to be avoided as possible by opting out the usage of flux. Up to authors' knowledge, our study is the first to fulfill all requirements stated above. Transient liquid phase sintering (TLPS) of mixed fluxless Sn-Bi (tin-bismuth) eutectic alloy and Ag (silver) particles was applied to bond plateless Cu-Cu substrates. Sintering temperature of 250 degrees C, sintering pressure of 0.02 MPa, and reducing environment were applied during processing. The effects of addition amount of Sn-Bi and sintering holding time to the shear strength and microstructure were investigated. The remelting temperature after sintering was also examined. Shear strength of 30 wt% added Sn-Bi was over than 20 MPa, which qualify the requirement of MIL-STD-883K, and larger than conventional Pbbased solder. Formation of intermetallic compounds are thought to strengthen the interface and matrix. Remelting temperature shifted from eutectic temperature of Sn-Bi to approximately 262 degrees C, allowing the application of modules at higher operating temperature than the processing temperature.

Bonding silicon carbide/gallium nitride (SiC/GaN) based power modules, particularly epoxy-molded modules to heat-substrate and/or heat sink, requires low processing temperature preferably lower than 250 degrees C, and low pressure as low as 0.1 MPa to prevent damage to the modules. In addition, due to the impracticality of depositing metal-plating to the epoxy-molded module, bonding of plateless Cu-substrates is in great demand. Furthermore, post-processing residual flux cleaning, which is costly and unfavorable to industry need to be avoided as possible by opting out the usage of flux. Up to authors' knowledge, our study is the first to fulfill all requirements stated above. Transient liquid phase sintering (TLPS) of mixed fluxless Sn-Bi (tin-bismuth) eutectic alloy and Ag (silver) particles was applied to bond plateless Cu-Cu substrates. Sintering temperature of 250 degrees C, sintering pressure of 0.02 MPa, and reducing environment were applied during processing. The effects of addition amount of Sn-Bi and sintering holding time to the shear strength and microstructure were investigated. The remelting temperature after sintering was also examined. Shear strength of 30 wt% added Sn-Bi was over than 20 MPa, which qualify the requirement of MIL-STD-883K, and larger than conventional Pbbased solder. Formation of intermetallic compounds are thought to strengthen the interface and matrix. Remelting temperature shifted from eutectic temperature of Sn-Bi to approximately 262 degrees C, allowing the application of modules at higher operating temperature than the processing temperature.

Recovery behavior (recovery) and strain-rate dependence of the stress-strain curve (strain-rate dependence) are incorporated into constitutive equations of alloys to predict residual stress and thermal stress during casting. Nevertheless, few studies have systematically investigated the effects of these metallurgical phenomena on the prediction accuracy of thermal stress in a casting. This study compares the thermal stress analysis results with in situ thermal stress measurement results of an Al-Si-Cu specimen during casting. The results underscore the importance for the alloy constitutive equation of incorporating strain-rate dependence to predict thermal stress that develops at high temperatures where the alloy shows strong strain-rate dependence of the stress-strain curve. However, the prediction accuracy of the thermal stress developed at low temperatures did not improve by considering the strain-rate dependence. Incorporating recovery into the constitutive equation improved the accuracy of the simulated thermal stress at low temperatures. Results of comparison implied that the constitutive equation should include strain-rate dependence to simulate defects that develop from thermal stress at high temperatures, such as hot tearing and hot cracking. Recovery should be incorporated into the alloy constitutive equation to predict the casting residual stress and deformation caused by the thermal stress developed mainly in the low temperature range. (C) The Minerals, Metals & Materials Society and ASM International 2017

© 2017 JSPS 191st Committee on Innovative Interface Bonding Technology. A low temperature and low pressure fluxless bonding of plateless Cu-Cu substrates has been achieved by transient liquid phase sintering of Ag and Sn-Bi eutectic powder mixture in a formic acid reducing environment. The effects of Sn-Bi addition amount and sintering temperature to the shear strength and microstructure were investigated. Remelting temperature of the sintered paste was also examined. Shear strength of 30 weight percentage added Sn-Bi that was sintered at 250°C was over than 20 MPa. The microstructure varied with the Sn-Bi addition amount, however, mainly consisted of Ag solid solution and/or Ag-Sn intermetallic compounds (IMCs), Bi-rich phase and Cu-Sn IMCs. No remelting event at Sn-Bi eutectic temperature was observed and the remelting temperature shifted to approximately 262°C, implying the possibility for higher operation temperature although the processing was performed at lower temperature.

Recovery behavior (recovery) and strain-rate dependence of the stress-strain curve (strain-rate dependence) are incorporated into constitutive equations of alloys to predict residual stress and thermal stress during casting. Nevertheless, few studies have systematically investigated the effects of these metallurgical phenomena on the prediction accuracy of thermal stress in a casting. This study compares the thermal stress analysis results with in situ thermal stress measurement results of an Al-Si-Cu specimen during casting. The results underscore the importance for the alloy constitutive equation of incorporating strain-rate dependence to predict thermal stress that develops at high temperatures where the alloy shows strong strain-rate dependence of the stress-strain curve. However, the prediction accuracy of the thermal stress developed at low temperatures did not improve by considering the strain-rate dependence. Incorporating recovery into the constitutive equation improved the accuracy of the simulated thermal stress at low temperatures. Results of comparison implied that the constitutive equation should include strain-rate dependence to simulate defects that develop from thermal stress at high temperatures, such as hot tearing and hot cracking. Recovery should be incorporated into the alloy constitutive equation to predict the casting residual stress and deformation caused by the thermal stress developed mainly in the low temperature range. (C) The Minerals, Metals & Materials Society and ASM International 2017

This paper proposes a new method for obtaining the rheological properties of solidifying alloys in the brittle temperature range (BTR). In that range, alloys show not only rheological, but also brittle behavior. Conventional methods to obtain rheological properties require steady state stress with ductility. Therefore, rheological properties of BTR alloys are unobtainable, or are otherwise including the effects of microscopic damage. The method proposed in this paper uses the stress-strain relation derived from the Maxwell model assuming that strain hardening is negligible in solid-liquid coexistence states. By removing the plastic strain term, the creep strain rate in Norton's law is derived by the total strain rate and stress rate without the steady state stress condition. Consequently, the stress exponent n and material constant A of Norton's law can be obtained even for alloys in the BTR. We applied this method to both tensile process before crack initiation and stress relaxation process. According to the Maxwell model, couples of the properties (n and A) obtained in both processes must be equal. Therefore, the difference can validate the obtained properties. From tensile and stress relaxation tests, we obtained the properties of solidifying Al-5 wt pct Mg alloy. We validated results by examining the difference. This report is the first to provide a method to obtain the rheological properties of BTR alloy without damage.

This paper proposes a new method for obtaining the rheological properties of solidifying alloys in the brittle temperature range (BTR). In that range, alloys show not only rheological, but also brittle behavior. Conventional methods to obtain rheological properties require steady state stress with ductility. Therefore, rheological properties of BTR alloys are unobtainable, or are otherwise including the effects of microscopic damage. The method proposed in this paper uses the stress-strain relation derived from the Maxwell model assuming that strain hardening is negligible in solid-liquid coexistence states. By removing the plastic strain term, the creep strain rate in Norton's law is derived by the total strain rate and stress rate without the steady state stress condition. Consequently, the stress exponent n and material constant A of Norton's law can be obtained even for alloys in the BTR. We applied this method to both tensile process before crack initiation and stress relaxation process. According to the Maxwell model, couples of the properties (n and A) obtained in both processes must be equal. Therefore, the difference can validate the obtained properties. From tensile and stress relaxation tests, we obtained the properties of solidifying Al-5 wt pct Mg alloy. We validated results by examining the difference. This report is the first to provide a method to obtain the rheological properties of BTR alloy without damage.

<p>  It is known that hot tearing tendency increases with the increase in the cooling rate of casting during solidification. As the direct control factors of hot tearing, the cooling rate dependences of the temperature field of the casting and mechanical properties of alloys in the semi-solid state have been suggested. However, quantitative evaluation has yet to be carried out to clarify which parameters are more important for predicting hot tearing tendency.</p><p>  In this study, through thermal stress analysis using the cooling rate-dependent temperature fields of casting and cooling rate-dependent creep parameters in the semi-solid state, hot tearing tendency was predicted for an Al-Mg alloy during solidification. For the prediction, the maximum principal creep strain accumulated during the solidification was used as an indicator of hot tearing tendencies. Then, the hot tearing tendencies were compared with experimental results. As a result, in the cooling rate range of this study which corresponded to gravity die casting, it was found that temperature fields are relatively more critical for predicting hot tearing tendency than the creep parameters.</p>

&lt;p&gt;To predict hot tearing in direct chill (DC) casting and shape casting of aluminum alloys using thermal stress analysis, cooling rate dependence of viscoplastic properties in a partially solidified state is indispensable. Based on viscoplastic properties determined from experiments, this study develops a method to predict the temperature dependence of viscoplastic properties at an arbitrary cooling rate through the Clyne–Kurz microsegregation model. For validation of the developed method, tensile tests were performed on Al–5 mass%Mg alloy in a partially solidified state at three cooling rates. Results show good agreement between the predicted values and experimentally obtained values, which demonstrates that the developed method is effective for predicting the cooling rate dependence of viscoplastic properties.&lt;/p&gt;&lt;p&gt; &lt;/p&gt;&lt;p&gt;This Paper was Originally Published in Japanese in J. JILM &lt;b&gt;67&lt;/b&gt; (2017) 214–221.&lt;/p&gt;

<p>To predict hot tearing in direct chill (DC) casting and shape casting of aluminum alloys using thermal stress analysis, cooling rate dependence of viscoplastic properties in a partially solidified state is indispensable. Based on viscoplastic properties determined from experiments, this study develops a method to predict the temperature dependence of viscoplastic properties at an arbitrary cooling rate through the Clyne–Kurz microsegregation model. For validation of the developed method, tensile tests were performed on Al–5 mass%Mg alloy in a partially solidified state at three cooling rates. Results show good agreement between the predicted values and experimentally obtained values, which demonstrates that the developed method is effective for predicting the cooling rate dependence of viscoplastic properties.</p>

To predict hot tearing in direct chill (DC) casting and shape casting of aluminum alloys using thermal stress analysis, cooling rate dependence of viscoplastic properties in a partially solidified state is indispensable. Based on viscoplastic properties determined from experiments, this study develops a method to predict the temperature dependence of viscoplastic properties at an arbitrary cooling rate through the Clyne-Kurz microsegregation model. For validation of the developed method, tensile tests were performed on Al-5 mass%Mg alloy in a partially solidified state at three cooling rates. Results show good agreement between the predicted values and experimentally obtained values, which demonstrates that the developed method is effective for predicting the cooling rate dependence of viscoplastic properties.

To predict hot tearing in direct chill (DC) casting and shape casting of aluminum alloys using thermal stress analysis, cooling rate dependence of viscoplastic properties in a partially solidified state is indispensable. Based on viscoplastic properties determined from experiments, this study develops a method to predict the temperature dependence of viscoplastic properties at an arbitrary cooling rate through the Clyne-Kurz microsegregation model. For validation of the developed method, tensile tests were performed on Al-5 mass%Mg alloy in a partially solidified state at three cooling rates. Results show good agreement between the predicted values and experimentally obtained values, which demonstrates that the developed method is effective for predicting the cooling rate dependence of viscoplastic properties.

Consideration of recovery behavior in the constitutive equation of alloy is crucially important for accurate thermal stress analysis of the casting process. However, previously reported empirical equations present difficulties when including a coefficient that determines the ratio of the recovery (or strain-hardening) strain component because the alloy recovery behavior has not been investigated systematically. First, this study systematically revealed effects of the temperature and amount of inelastic strain at 200-440 degrees C on the increase of the initial yield stress of the room temperature (RT) in an Al-Si-Cu high-pressure die casting alloy. Secondly, from those results, an initial yield stress- pre-strain diagram was proposed, demonstrating the effect of the amount of the inelastic strain given at each temperature on the increase of the initial yield stress of RT. Finally, this study newly defined a temperature-dependent contribution ratio of strain hardening to parameterize and duplicate the alloy recovery behavior based on the initial yield stress - pre-strain diagram for the constitutive equation of the alloy. Using this ratio as the coefficient to determine the proportion of the strain contributing the strain hardening for the constitutive equation, thermal stress analysis can produce more accurate duplication of the experimentally determined recovery behavior. Thereby, the residual stress and deformation of the component in the casting process can be predicted more accurately than when using conventional empirical constitutive equations.

Consideration of recovery behavior in the constitutive equation of alloy is crucially important for accurate thermal stress analysis of the casting process. However, previously reported empirical equations present difficulties when including a coefficient that determines the ratio of the recovery (or strain-hardening) strain component because the alloy recovery behavior has not been investigated systematically. First, this study systematically revealed effects of the temperature and amount of inelastic strain at 200-440 degrees C on the increase of the initial yield stress of the room temperature (RT) in an Al-Si-Cu high-pressure die casting alloy. Secondly, from those results, an initial yield stress- pre-strain diagram was proposed, demonstrating the effect of the amount of the inelastic strain given at each temperature on the increase of the initial yield stress of RT. Finally, this study newly defined a temperature-dependent contribution ratio of strain hardening to parameterize and duplicate the alloy recovery behavior based on the initial yield stress - pre-strain diagram for the constitutive equation of the alloy. Using this ratio as the coefficient to determine the proportion of the strain contributing the strain hardening for the constitutive equation, thermal stress analysis can produce more accurate duplication of the experimentally determined recovery behavior. Thereby, the residual stress and deformation of the component in the casting process can be predicted more accurately than when using conventional empirical constitutive equations.

Accurate simulation of residual stress and deformation is necessary to optimize the design and lifetime of casting components. Therefore, the recovery and strain-rate dependence of the stress-strain curve have been incorporated into empirical constitutive equations to improve the thermal stress analysis accuracy. Nevertheless, these equations present several difficulties related to the determination of material constants and their physical bases. This study suggested an empirical elasto-plastic-creep constitutive equation incorporating these phenomena. To determine the material parameters used in this constitutive equation, this study investigated tensile test methods to obtain stress-strain curves that most closely resemble those during or immediately after casting for the Al-Si-Cu high-pressure die-casting alloy JIS ADC 12 (A383.0), which exhibits natural aging. Results show that solution heat treatment with subsequent cooling to the test temperature should be applied to obtain stress-strain curves used for the thermal stress analysis of high-pressure die casting process of this alloy. The yield stresses obtained using the conventional heating method were 50-64 pct higher than those of the method described above. Therefore, the conventional method is expected to overestimate the overestimation of the predicted residual stress in die castings. Evaluation of the developed equation revealed that it can represent alloy recovery and strain-rate dependence.

Accurate simulation of residual stress and deformation is necessary to optimize the design and lifetime of casting components. Therefore, the recovery and strain-rate dependence of the stress-strain curve have been incorporated into empirical constitutive equations to improve the thermal stress analysis accuracy. Nevertheless, these equations present several difficulties related to the determination of material constants and their physical bases. This study suggested an empirical elasto-plastic-creep constitutive equation incorporating these phenomena. To determine the material parameters used in this constitutive equation, this study investigated tensile test methods to obtain stress-strain curves that most closely resemble those during or immediately after casting for the Al-Si-Cu high-pressure die-casting alloy JIS ADC 12 (A383.0), which exhibits natural aging. Results show that solution heat treatment with subsequent cooling to the test temperature should be applied to obtain stress-strain curves used for the thermal stress analysis of high-pressure die casting process of this alloy. The yield stresses obtained using the conventional heating method were 50-64 pct higher than those of the method described above. Therefore, the conventional method is expected to overestimate the overestimation of the predicted residual stress in die castings. Evaluation of the developed equation revealed that it can represent alloy recovery and strain-rate dependence.

After JIS FC300 (almost identical to ASTM 45) gray cast iron was cast in a furan sand mold,both the load on the casting from the sand mold and contraction of the casting were measured dynamically from the beginning of solidification to 200 degrees C. During solidification, the cast iron casting received compressive load from the sand mold attributable to expansion of the casting. After solidification, the load on the casting and the contraction of the casting increased linearly with cooling. During A1 transformation, the load on the casting was relieved because of the casting expansion. After the A1 transformation, the load on the casting and the contraction of the casting increased linearly with cooling again. When the load on the casting reached approximately 13 kN, the load stayed constant, probably because of the sand mold fracture. (C) 2016 Elsevier B.V. All rights reserved.

After JIS FC300 (almost identical to ASTM 45) gray cast iron was cast in a furan sand mold,both the load on the casting from the sand mold and contraction of the casting were measured dynamically from the beginning of solidification to 200 degrees C. During solidification, the cast iron casting received compressive load from the sand mold attributable to expansion of the casting. After solidification, the load on the casting and the contraction of the casting increased linearly with cooling. During A1 transformation, the load on the casting was relieved because of the casting expansion. After the A1 transformation, the load on the casting and the contraction of the casting increased linearly with cooling again. When the load on the casting reached approximately 13 kN, the load stayed constant, probably because of the sand mold fracture. (C) 2016 Elsevier B.V. All rights reserved.

This study investigated a mechanism of reducing solidification cracking susceptibility by grain refinement. Previously, the occurrence of grain boundary sliding (GBS) is implied as one mechanism, which is known for the characteristics of superplastic deformation.
To verify GBS occurrence during solidification of an Al-5 wt%Mg alloy, the effects of refinement on the true stress-true strain curve and the effective power-law coefficient n(eff) (the reciprocal of the strain rate sensitivity) were investigated in a partially solidified state. Furthermore, the contribution of GBS to deformation was examined using the index xi(GBS) defined in the research field of superplasticity. The results are following.
(1) The grain refined structure elongation is greater than that in the coarse grain structure regardless of the solid fraction. Greater elongation is expected to result from both lower value of the coefficient n(eff) and the higher contribution of the GBS, which provide more uniform deformation.
(2) Maximum elongation of up to 20% was found at 480 degrees C (f(s): 0.939) in the grain refined structure. Based on the knowledge of superplastic behavior, the values of both n(eff)approximate to 12 and xi(GBS)approximate to 70% at the temperature indicates superplastic-like deformation during the solidification, reducing solidification cracking susceptibility by grain refinement. (c) 2016 Elsevier B.V. All rights reserved.

Ultrasonic treatment (UST) of molten metals produces a fine grain microstructure. Several mechanisms of grain refining by UST have been suggested; however, experimental verification has not yet established the actual mechanism. In this study, UST was applied to Al-4 wt pct Si molten alloy (1) above the liquidus temperature, (2) during undercooling before recalescence, (3) during recalescence, and (4) after recalescence. After UST treatment, the average grain sizes of the solidified samples were measured, and the effects of UST were evaluated. In the case of (1), the temperatures of the crucible, ultrasonic horn, and the atmosphere of the molten alloy were also confirmed above the liquidus temperature, and a finer grain structure was obtained. This result reveals that UST promotes the nonequilibrium nucleation mechanism. UST during (2) and (3) also caused finer grain microstructures, whereas (4) did not. UST during undercooling hastened nucleation. However, the previously reported mechanism of the breaking dendrite was not activated by UST after recalescence. (C) The Minerals, Metals & Materials Society and ASM International 2016

Ultrasonic treatment (UST) of molten metals produces a fine grain microstructure. Several mechanisms of grain refining by UST have been suggested; however, experimental verification has not yet established the actual mechanism. In this study, UST was applied to Al-4 wt pct Si molten alloy (1) above the liquidus temperature, (2) during undercooling before recalescence, (3) during recalescence, and (4) after recalescence. After UST treatment, the average grain sizes of the solidified samples were measured, and the effects of UST were evaluated. In the case of (1), the temperatures of the crucible, ultrasonic horn, and the atmosphere of the molten alloy were also confirmed above the liquidus temperature, and a finer grain structure was obtained. This result reveals that UST promotes the nonequilibrium nucleation mechanism. UST during (2) and (3) also caused finer grain microstructures, whereas (4) did not. UST during undercooling hastened nucleation. However, the previously reported mechanism of the breaking dendrite was not activated by UST after recalescence. (C) The Minerals, Metals & Materials Society and ASM International 2016

For predicting solidification cracking by thermal stress analysis, the mechanical properties in the partially solidified state based on the experimental results are the best hope. However, the Young's modulus has never been investigated for copper alloys. In this study, stress-strain curves of a Cu-Zn alloy in the partially solidified state for various solid fractions were obtained using a specially developed horizontal tensile test device. Furthermore, by removing the load during the tensile test, the spring-back (elastic behavior) was observed and the Young's modulus was obtained.

For predicting solidification cracking by thermal stress analysis, the mechanical properties in the partially solidified state based on the experimental results are the best hope. However, the Young's modulus has never been investigated for copper alloys. In this study, stress-strain curves of a Cu-Zn alloy in the partially solidified state for various solid fractions were obtained using a specially developed horizontal tensile test device. Furthermore, by removing the load during the tensile test, the spring-back (elastic behavior) was observed and the Young's modulus was obtained.

This study investigated the effect of Mn adding on the T5 and T6 age-hardening behaviors of Al-(9.0-10.0)%Si-0.3%Mg (equivalent to A360 alloy) and Al-10% Si-2%Cu-0.3%Mg (equivalent to A383 alloy) die casting alloys using a hardness measurement, an electron probe micro analyzer (EPMA) and thermodynamic software. When 0.4% Mn was added, the hardness of the T5 heat-treated Al-Si-Mg die casting alloy increases because the Mn addition suppresses the formation of the pi-Al8FeMg3Si6 phase during solidification and distributes Mg into the a solid solution. This mechanism named as pi-phase suppression mechanism promotes precipitation strengthening. In the T6 heat treatment, the hardness of the alloys hardly increased when Mn was added. In the Al-Si-Cu-Mg alloys, the hardening by the 0.5% Mn addition was negligible compared to that of the Al-Si-Mg alloys regardless of T5 or T6 treatment. Therefore, the hardening by Mn addition is the characteristic phenomenon in the T5 heat-treated Al-Si-Mg high-pressure die casting alloy.

This study investigated the effect of Mn adding on the T5 and T6 age-hardening behaviors of Al-(9.0-10.0)%Si-0.3%Mg (equivalent to A360 alloy) and Al-10% Si-2%Cu-0.3%Mg (equivalent to A383 alloy) die casting alloys using a hardness measurement, an electron probe micro analyzer (EPMA) and thermodynamic software. When 0.4% Mn was added, the hardness of the T5 heat-treated Al-Si-Mg die casting alloy increases because the Mn addition suppresses the formation of the pi-Al8FeMg3Si6 phase during solidification and distributes Mg into the a solid solution. This mechanism named as pi-phase suppression mechanism promotes precipitation strengthening. In the T6 heat treatment, the hardness of the alloys hardly increased when Mn was added. In the Al-Si-Cu-Mg alloys, the hardening by the 0.5% Mn addition was negligible compared to that of the Al-Si-Mg alloys regardless of T5 or T6 treatment. Therefore, the hardening by Mn addition is the characteristic phenomenon in the T5 heat-treated Al-Si-Mg high-pressure die casting alloy.

This study investigated the effect of Mn adding on the T5 and T6 age-hardening behaviors of Al-(9.0-10.0)%Si-0.3%Mg (equivalent to A360 alloy) and Al-10% Si-2%Cu-0.3%Mg (equivalent to A383 alloy) die casting alloys using a hardness measurement, an electron probe micro analyzer (EPMA) and thermodynamic software. When 0.4% Mn was added, the hardness of the T5 heat-treated Al-Si-Mg die casting alloy increases because the Mn addition suppresses the formation of the pi-Al8FeMg3Si6 phase during solidification and distributes Mg into the a solid solution. This mechanism named as pi-phase suppression mechanism promotes precipitation strengthening. In the T6 heat treatment, the hardness of the alloys hardly increased when Mn was added. In the Al-Si-Cu-Mg alloys, the hardening by the 0.5% Mn addition was negligible compared to that of the Al-Si-Mg alloys regardless of T5 or T6 treatment. Therefore, the hardening by Mn addition is the characteristic phenomenon in the T5 heat-treated Al-Si-Mg high-pressure die casting alloy.

Pulsating low-cycle fatigue processes, up to the present, have been divided into three states: the transient state, steady state, and accelerating state of ratcheting. In our previous work, we suggested that fatigue behavior of pulsating fatigue process should be classified into five stages in which the plastic strain amplitude and the ratcheting strain rate are plotted on the X and Y axis, respectively. In this study, at the condition of R=-0.3 (partially pulsating fatigue), the change in the plastic strain amplitude and ratcheting strain rate for each cycle to failure was examined on AISI 1025 carbon steel. The dislocation substructure was examined using transmission electron microscopy (TEM) for each stage, except for stage I. It was also demonstrated that the fatigue process can be divided into five stages: stage I corresponds to the un-pinning of dislocations from the Cottrell atmosphere and propagation of the Luders band. Stage II corresponds to the restriction of dislocation movement by dislocation tangles. Stage III corresponds to the formation of dislocation cells. Stage IV corresponds to the promotion of the to-and-fro (back-and-forth) motion of dislocations by a re-arrangement of the dislocations in the cells. Stage V corresponds to the release of dislocation movement by the collapse of dislocation cells. (C) 2015 Elsevier B.V. All rights reserved.

Pulsating low-cycle fatigue processes, up to the present, have been divided into three states: the transient state, steady state, and accelerating state of ratcheting. In our previous work, we suggested that fatigue behavior of pulsating fatigue process should be classified into five stages in which the plastic strain amplitude and the ratcheting strain rate are plotted on the X and Y axis, respectively. In this study, at the condition of R=-0.3 (partially pulsating fatigue), the change in the plastic strain amplitude and ratcheting strain rate for each cycle to failure was examined on AISI 1025 carbon steel. The dislocation substructure was examined using transmission electron microscopy (TEM) for each stage, except for stage I. It was also demonstrated that the fatigue process can be divided into five stages: stage I corresponds to the un-pinning of dislocations from the Cottrell atmosphere and propagation of the Luders band. Stage II corresponds to the restriction of dislocation movement by dislocation tangles. Stage III corresponds to the formation of dislocation cells. Stage IV corresponds to the promotion of the to-and-fro (back-and-forth) motion of dislocations by a re-arrangement of the dislocations in the cells. Stage V corresponds to the release of dislocation movement by the collapse of dislocation cells. (C) 2015 Elsevier B.V. All rights reserved.

Predicting hot tearing during direct chill casting using thermal stress analysis requires constitutive equations in both semi-solid state and below the solidus of the alloy. However, numerous difficulties have been hindered constitutive equations used heretofore for hot tearing predictions. (1) Testing methods for obtaining material constants were inappropriate. First, the elastic strain reversibility was unconfirmed. Second, a flat distribution of temperature in the specimen gauge length was not guaranteed. Third, strain was measured not from local strain but from cross-head displacement. Fourth, the melt-back phenomenon was unavoidable in test during partial remelting because of homogenization of the segregation structure. (2) Temperature dependence of the strain-rate sensitivity of stress was not considered. (3) Some material constants were inferred, not obtained experimentally. This study developed elasto-viscoplastic constitutive equations (Hooke's and viscoplastic Norton-Hoff laws) for partially solidified state and below the solidus. To obtain material constants experimentally, two tensile tests for which issue (1) was addressed were conducted using Al-5 mass%Mg alloy. They were a tensile test after partial solidification and high-temperature tensile test with high-frequency induction coil. After the temperature dependence of elastic and viscoplastic properties was investigated, material constants were obtained and were compared with those obtained using earlier testing methods.

Predicting hot tearing during direct chill casting using thermal stress analysis requires constitutive equations in both semi-solid state and below the solidus of the alloy. However, numerous difficulties have been hindered constitutive equations used heretofore for hot tearing predictions. (1) Testing methods for obtaining material constants were inappropriate. First, the elastic strain reversibility was unconfirmed. Second, a flat distribution of temperature in the specimen gauge length was not guaranteed. Third, strain was measured not from local strain but from cross-head displacement. Fourth, the melt-back phenomenon was unavoidable in test during partial remelting because of homogenization of the segregation structure. (2) Temperature dependence of the strain-rate sensitivity of stress was not considered. (3) Some material constants were inferred, not obtained experimentally. This study developed elasto-viscoplastic constitutive equations (Hooke&#039;s and viscoplastic Norton-Hoff laws) for partially solidified state and below the solidus. To obtain material constants experimentally, two tensile tests for which issue (1) was addressed were conducted using Al-5 mass%Mg alloy. They

Predicting hot tearing during direct chill casting using thermal stress analysis requires constitutive equations in both semi-solid state and below the solidus of the alloy. However, numerous difficulties have been hindered constitutive equations used heretofore for hot tearing predictions. (1) Testing methods for obtaining material constants were inappropriate. First, the elastic strain reversibility was unconfirmed. Second, a flat distribution of temperature in the specimen gauge length was not guaranteed. Third, strain was measured not from local strain but from cross-head displacement. Fourth, the melt-back phenomenon was unavoidable in test during partial remelting because of homogenization of the segregation structure. (2) Temperature dependence of the strain-rate sensitivity of stress was not considered. (3) Some material constants were inferred, not obtained experimentally. This study developed elasto-viscoplastic constitutive equations (Hooke's and viscoplastic Norton-Hoff laws) for partially solidified state and below the solidus. To obtain material constants experimentally, two tensile tests for which issue (1) was addressed were conducted using Al-5 mass%Mg alloy. They were a tensile test after partial solidification and high-temperature tensile test with high-frequency induction coil. After the temperature dependence of elastic and viscoplastic properties was investigated, material constants were obtained and were compared with those obtained using earlier testing methods.

In previous studies, the relationship between stacking fault energy (SFE) and the fatigue process has been examined under the fully reversed loading in which ratcheting does not occur. In this study, under the fully repeated loading, for pure Ni (128 mJ/m(2)), pure Cu (40 mJ/m(2)) and Cu-35 wt% Zn (11.5 mJ/m(2)) with different SFEs, the effect of fatigue damage and ratcheting damage on fatigue life was investigated. It was observed that for pure Ni, compared with pure Cu and Cu-35 wt% Zn, ratcheting damage is dominant. For Cu-35 wt% Zn, compared with pure Ni and pure Cu, fatigue damage is dominant. For pure Cu, both ratcheting damage and fatigue damage are dominant (mixed type). Thus, it was observed for the first time that high SFE tends to enhance ratcheting damage and low SFE tends to promote fatigue damage under the fully repeated loading. (C) 2014 Elsevier B.V. All rights reserved.

The residual stresses in the thick part of the stress lattice shape casting consist of the residual stress due to the temperature differential between the thick part and the thin part and the residual stress due to the temperature differential in the radial direction of the thick part. In this study, the gray cast iron stress lattice shape castings were cast and both types of the residual stresses were separately measured. Thermal stress analyses based on the casting experiment were conducted. Next, the measurements in this study were compared with both types of the simulated residual stresses. The thermal stress analyses estimated the residual stress due to the temperature difference in the radial direction of the thick part to be significantly higher than the measurement, although the residual stress due to the temperature difference between the thick part and the thin part was successfully predicted within a 10 pct error. Thus, this study suggested the introduction of the mechanical melting temperature, above which the very low yield stress is applied conveniently to describe the losses of the deformation resistance of the casting, to more accurately predict the residual stress due to the temperature difference in the radial direction of the thick part. From the verification of the suggested model, this study demonstrated that the conventional elasto-plastic model must introduce the mechanical melting temperature to predict the residual stress due to the temperature difference in the radial direction of the thick part and thus the overall residual stress in the stress lattice. (C) The Minerals, Metals & Materials Society and ASM International 2013

The residual stresses in the thick part of the stress lattice shape casting consist of the residual stress due to the temperature differential between the thick part and the thin part and the residual stress due to the temperature differential in the radial direction of the thick part. In this study, the gray cast iron stress lattice shape castings were cast and both types of the residual stresses were separately measured. Thermal stress analyses based on the casting experiment were conducted. Next, the measurements in this study were compared with both types of the simulated residual stresses. The thermal stress analyses estimated the residual stress due to the temperature difference in the radial direction of the thick part to be significantly higher than the measurement, although the residual stress due to the temperature difference between the thick part and the thin part was successfully predicted within a 10 pct error. Thus, this study suggested the introduction of the mechanical melting temperature, above which the very low yield stress is applied conveniently to describe the losses of the deformation resistance of the casting, to more accurately predict the residual stress due to the temperature difference in the radial direction of the thick part. From the verification of the suggested model, this study demonstrated that the conventional elasto-plastic model must introduce the mechanical melting temperature to predict the residual stress due to the temperature difference in the radial direction of the thick part and thus the overall residual stress in the stress lattice. (C) The Minerals, Metals & Materials Society and ASM International 2013

Various forms of damage to the bridge foundation structure in the revetment along riverbanks and sea coasts caused by liquefaction had been observed during past earthquakes. Several studies on liquefaction using physical model tests and numerical analysis have been conducted in recent years. However, few studies have investigated the seismic behavior of the foundation in a revetment with a slope. In strong earthquakes, the sloped ground is expected to be unstable, and lateral spreading of the ground may occur simultaneously with the loss of soil strength in the liquefaction layer. Moreover, in the seismic design specification (JRA-2002) of the bridge, the liquefaction verification of the foundation is stipulated for a flat ground but not for a sloped ground. Therefore, the effect of the lateral pressure of the liquefaction layer on the foundation in the revetment must be investigated further. This study aims to investigate the dynamic behavior of a steel pipe sheet pile (SPSP) foundation of a cable-stayed bridge and its effect on the performance of the superstructure in the revetment with a slope. A 1-G shaking table test with a scale of 1:60 was conducted on a flat model and a slope model of 15°. In addition, 2-D numerical modeling was applied in an effective stress analysis method that was used on a multi-spring model and cocktail glass model. The differences in the dynamic responses between the two models clearly illustrate the significant effect of the ground slope on the seismic behavior of the SPSP foundation and superstructure.

In the pulsating fatigue process, the relationship between cyclic softening (hardening) and ratcheting should be clarified from the perspective of the practical use of the mechanical parts. However, most previous reports have been limited to predicting of ratcheting curves using constitutive equations. Consequently, there are few qualitative discussions regarding the relationship between cyclic softening (hardening) and ratcheting for changes in the number of cycles. In this study, the plastic strain amplitude, which represents the fatigue damage, is plotted on the X axis and the ratcheting strain rate is plotted on the Y axis, cycle by cycle, to investigate the fatigue and ratcheting damage simultaneously. This figure is known as the SH curve after Dr. Susumu Horibe. As an example, the pulsating fatigue processes of JIS S25C (AISI 1025) with three different carbide shape types are analyzed under the conditions that the engineering stress amplitudes are over their lower yield points. Using SH curves it is shown that regardless of the shape of the carbide, the fatigue behavior should be divided into five stages. Stage I corresponds to the un-pinning of dislocations from the Cottrell atmosphere. Stage II corresponds to the propagation of the Luders band. Stage III corresponds to an increase in short range dislocation movement during the formation of cellular structures due to multiple slip locations. Stage IV corresponds to the fracture of the cellular structure due to an increase in the true stress. Stage V corresponds to crack initiation and propagation. The plastic strain amplitude increases with the number of cycles, while the ratcheting strain rate decreases, especially in stage III; this phenomenon has never been reported previously. At this stage, the microstructure was observed by TEM. It is also determined that in this case, the maximum stress is over the lower yield point, and the ratcheting strain rate is dominant over the fatigue life. (C) 2013 Elsevier B.V. All rights reserved.

In the pulsating fatigue process, the relationship between cyclic softening (hardening) and ratcheting should be clarified from the perspective of the practical use of the mechanical parts. However, most previous reports have been limited to predicting of ratcheting curves using constitutive equations. Consequently, there are few qualitative discussions regarding the relationship between cyclic softening (hardening) and ratcheting for changes in the number of cycles. In this study, the plastic strain amplitude, which represents the fatigue damage, is plotted on the X axis and the ratcheting strain rate is plotted on the Y axis, cycle by cycle, to investigate the fatigue and ratcheting damage simultaneously. This figure is known as the SH curve after Dr. Susumu Horibe. As an example, the pulsating fatigue processes of JIS S25C (AISI 1025) with three different carbide shape types are analyzed under the conditions that the engineering stress amplitudes are over their lower yield points. Using SH curves it is shown that regardless of the shape of the carbide, the fatigue behavior should be divided into five stages. Stage I corresponds to the un-pinning of dislocations from the Cottrell atmosphere. Stage II corresponds to the propagation of the Luders band. Stage III corresponds to an increase in short range dislocation movement during the formation of cellular structures due to multiple slip locations. Stage IV corresponds to the fracture of the cellular structure due to an increase in the true stress. Stage V corresponds to crack initiation and propagation. The plastic strain amplitude increases with the number of cycles, while the ratcheting strain rate decreases, especially in stage III; this phenomenon has never been reported previously. At this stage, the microstructure was observed by TEM. It is also determined that in this case, the maximum stress is over the lower yield point, and the ratcheting strain rate is dominant over the fatigue life. (C) 2013 Elsevier B.V. All rights reserved.

The restraint exerted on a casting by a furan sand mold on the casting and the contraction of the casting during cooling was dynamically and simultaneously measured using a device that we developed. The measurements were compared during cooling with thermal stress analyses. The thermal stress analyses were based on the representative mechanical models for the furan sand mold, i.e., the elastic and elasto-plastic models used in previous studies. The comparison demonstrated that the elasto-plastic model simulates the restraint force more accurately than the elastic model. In the thermal stress analysis, it was important to describe the development of inelastic deformation and the fracture of the sand mold. However, the simulated restraint force was still twice as large as the measured force even in the elastoplastic model. This error is most likely attributable to using the temperature-independent mechanical properties of the furan sand mold and the mechanical model of the casting alloy, which neglected the viscoplasticity at high temperature in the thermal stress analysis. (C) 2013 Elsevier B.V. All rights reserved.

The restraint exerted on a casting by a furan sand mold on the casting and the contraction of the casting during cooling was dynamically and simultaneously measured using a device that we developed. The measurements were compared during cooling with thermal stress analyses. The thermal stress analyses were based on the representative mechanical models for the furan sand mold, i.e., the elastic and elasto-plastic models used in previous studies. The comparison demonstrated that the elasto-plastic model simulates the restraint force more accurately than the elastic model. In the thermal stress analysis, it was important to describe the development of inelastic deformation and the fracture of the sand mold. However, the simulated restraint force was still twice as large as the measured force even in the elastoplastic model. This error is most likely attributable to using the temperature-independent mechanical properties of the furan sand mold and the mechanical model of the casting alloy, which neglected the viscoplasticity at high temperature in the thermal stress analysis. (C) 2013 Elsevier B.V. All rights reserved.

For S45C carbon steel with three heat treatments, pulsating tension tests were carried out under controlled stress condition. The effect of compressive pre-strain, Bauschinger effect and controlling factors on fatigue life were examined. The following results were obtained. The decrease of pulsating fatigue life was caused by three factors, i.e. increase of the total plastic strain amplitude, cyclic plastic strain amplitude and cyclic creep rate. The increase of Bauschinger strain led to the increase of the value of the three factors. Then it resulted in shorter fatigue life. In order to examine the cyclic stress conditions to the fatigue life, the stress ratio was defined as the ratio of maximum cyclic stress to yield stress of each material. Bauschinger strain was occurred when the stress ratio was under the threshold for each material. As the result, cyclic plastic strain amplitude and cyclic creep rate were increased and then pulsating fatigue life was reduced.

It is well known that creep at room temperature remarkably occurs in alpha-Ti alloys. Cyclic creep also appears. Moreover, Ti alloys have different crystal structures and various mechanical properties. Although the cyclic creep behavior and the cyclic softening behavior have been studied, integrated effect of them on fatigue life has not been clarified yet. In this study, fatigue behavior under stress ratio R=0, on Ti alloys of three crystal structures, is investigated. Influence of the cyclic creep and fatigue damage, and influence of the crystal structure to fatigue life are examined. As the results, it is found that as to the alpha-Ti, creep is the dominant damage factor. On the other hand, as to the beta-Ti alloy, fatigue is the dominant damage factor. As to the alpha + beta-Ti alloy, both creep and fatigue are the dominant damage factors (intermediate type). Furthermore, it is found that the area ratio and grain diameter of alpha phase, the spacing between different phases, should influence on cyclic creep strain rate.

For magnesium alloy ZK60 with different heat treatments, controlled plastic strain amplitude fatigue tests were conducted. Effects of twinning and aging treatments on fatigue process were investigated by analyzing the second derivative of hysteresis half-loop. It is found that during compression the convex peak appears at the strain where twinning begins. It is also found that during tension the convex peak appears at the strain where detwinning begins and the concave peak appears at the strain where slip begins to be predominant. Analyzing the strain at the convex peak in compression at the plastic strain amplitude of 6 x 10(-3), for solution treated material and over-aged material twinning hardens with increasing cycle. On the other hand, for T6 material twinning softens from 5 cycle to N-f/5 cycle and then hardens from N-f/5 cycle to N-f/2 cycle. It was suggested that residual twins cause twinning hardening. It was also suggested that cut and dissolution of rod-like precipitates should cause twinning softening. Due to different states of precipitates of materials, the twinning softening appeared for only T6 material at the plastic strain amplitude of 6 x 10(-3). Moreover, the softening were not found for T6 material under the plastic strain amplitude of 4 x 10(-3).

It is well known that creep at room temperature remarkably occurs in alpha-Ti alloys. Cyclic creep also appears. Moreover, Ti alloys have different crystal structures and various mechanical properties. Although the cyclic creep behavior and the cyclic softening behavior have been studied, integrated effect of them on fatigue life has not been clarified yet. In this study, fatigue behavior under stress ratio R=0, on Ti alloys of three crystal structures, is investigated. Influence of the cyclic creep and fatigue damage, and influence of the crystal structure to fatigue life are examined. As the results, it is found that as to the alpha-Ti, creep is the dominant damage factor. On the other hand, as to the beta-Ti alloy, fatigue is the dominant damage factor. As to the alpha + beta-Ti alloy, both creep and fatigue are the dominant damage factors (intermediate type). Furthermore, it is found that the area ratio and grain diameter of alpha phase, the spacing between different phases, should influence on cyclic creep strain rate.

For magnesium alloy ZK60 with different heat treatments, controlled plastic strain amplitude fatigue tests were conducted. Effects of twinning and aging treatments on fatigue process were investigated by analyzing the second derivative of hysteresis half-loop. It is found that during compression the convex peak appears at the strain where twinning begins. It is also found that during tension the convex peak appears at the strain where detwinning begins and the concave peak appears at the strain where slip begins to be predominant. Analyzing the strain at the convex peak in compression at the plastic strain amplitude of 6 x 10(-3), for solution treated material and over-aged material twinning hardens with increasing cycle. On the other hand, for T6 material twinning softens from 5 cycle to N-f/5 cycle and then hardens from N-f/5 cycle to N-f/2 cycle. It was suggested that residual twins cause twinning hardening. It was also suggested that cut and dissolution of rod-like precipitates should cause twinning softening. Due to different states of precipitates of materials, the twinning softening appeared for only T6 material at the plastic strain amplitude of 6 x 10(-3). Moreover, the softening were not found for T6 material under the plastic strain amplitude of 4 x 10(-3).

In a previous study, the hypothesis that dislocation arrangements do not change while Masing behavior occurs in materials was proposed. However, by validating the hypothesis via TEM observation, the authors incidentally found that dislocation structures are not always the same during Masing behavior. Therefore, the aim of this work was to understand the relationship between Masing behavior and dislocation structures using AISI 1025 under stress-controlled fatigue testing. In addition, to understand the more systematical relationship between Masing behavior and dislocation structures, three stress ratios, R=-1, -0.3 and 0, were applied, and each cyclic stress-strain curve (CSSC) was obtained for the first time. To investigate the Masing behavior, a new analytical method was proposed due to the difficulty of evaluating hysteresis loops with mean stress using previous methods. Based on the results, the dislocation structures are observed for the first time to determine whether the dislocation structures are same when Masing behavior occurs. As a result, the dislocation structures that exhibited Masing behavior were categorized into two types: (i) fully homogeneous dislocation structures and (ii) dislocation structures with a dual microstructure composed of cell and loop-patch structures, each of which exhibits a different volume fraction. The relationship between the CSSCs under stress ratios of R=-0.3 and 0 and the dislocation structures was examined. There are B and C regions in the CSSCs under stress ratios of R=-0.3 and 0. It was found that each dislocation structures shows loop-patch and cell structures. At the end of region C, a collapsed cell structure was observed. (C) 2013 Elsevier B.V. All rights reserved.

For magnesium alloy ZK60 with different heat treatments, controlled plastic strain amplitude fatigue tests were conducted. Effects of twinning and aging treatments on fatigue process were investigated by analyzing the second derivative of hysteresis half-loop. It is found that during compression the convex peak appears at the strain where twinning begins. It is also found that during tension the convex peak appears at the strain where detwinning begins and the concave peak appears at the strain where slip begins to be predominant. Analyzing the strain at the convex peak in compression at the plastic strain amplitude of 6 x 10(-3), for solution treated material and over-aged material twinning hardens with increasing cycle. On the other hand, for T6 material twinning softens from 5 cycle to N-f/5 cycle and then hardens from N-f/5 cycle to N-f/2 cycle. It was suggested that residual twins cause twinning hardening. It was also suggested that cut and dissolution of rod-like precipitates should cause twinning softening. Due to different states of precipitates of materials, the twinning softening appeared for only T6 material at the plastic strain amplitude of 6 x 10(-3). Moreover, the softening were not found for T6 material under the plastic strain amplitude of 4 x 10(-3).

It is well known that creep at room temperature remarkably occurs in alpha-Ti alloys. Cyclic creep also appears. Moreover, Ti alloys have different crystal structures and various mechanical properties. Although the cyclic creep behavior and the cyclic softening behavior have been studied, integrated effect of them on fatigue life has not been clarified yet. In this study, fatigue behavior under stress ratio R=0, on Ti alloys of three crystal structures, is investigated. Influence of the cyclic creep and fatigue damage, and influence of the crystal structure to fatigue life are examined. As the results, it is found that as to the alpha-Ti, creep is the dominant damage factor. On the other hand, as to the beta-Ti alloy, fatigue is the dominant damage factor. As to the alpha + beta-Ti alloy, both creep and fatigue are the dominant damage factors (intermediate type). Furthermore, it is found that the area ratio and grain diameter of alpha phase, the spacing between different phases, should influence on cyclic creep strain rate.

In this work, the JIS AD12.1 (almost the same as A383.1) aluminum alloy was cast in a green sand mold. The restraint force from the sand mold and the contraction of the casting were measured dynamically from the solidifying temperature to the shake-out temperature using a dedicated device. Then, FEM (Finite Element Method) thermal stress analyses of the experiment were performed. The analyses adopted four types of representative constitutive equations and the mechanical properties of the green sand mold, which were quoted from previous research articles. As verification, this study dynamically compared the simulated restraint force and the contraction of casting with measured results and examined which mechanical properties are important for expressing the restraint force of the sand mold. This verification is the first attempt in the world. As a result, the simulated restraint force was estimated to be over ten times as large as the measured result in each type of equation because the yield stress of the sand mold used in our experiment was lower than those quoted from previous studies. The yield stress measured by a uniaxial compression test was 1/20 of the quoted values. When the measured yield stress was adopted in the simulation, the simulated restraint force and contraction approached the measured results. The yield stress of the sand mold was a dominant factor in the restraint force simulated by the thermal stress analyses. The yield stress of the green sand mold used in the casting process should be measured to predict the residual stress using FEM thermal stress analyses. (C) 2013 Elsevier B.V. All rights reserved,

In this work, the JIS AD12.1 (almost the same as A383.1) aluminum alloy was cast in a green sand mold. The restraint force from the sand mold and the contraction of the casting were measured dynamically from the solidifying temperature to the shake-out temperature using a dedicated device. Then, FEM (Finite Element Method) thermal stress analyses of the experiment were performed. The analyses adopted four types of representative constitutive equations and the mechanical properties of the green sand mold, which were quoted from previous research articles. As verification, this study dynamically compared the simulated restraint force and the contraction of casting with measured results and examined which mechanical properties are important for expressing the restraint force of the sand mold. This verification is the first attempt in the world. As a result, the simulated restraint force was estimated to be over ten times as large as the measured result in each type of equation because the yield stress of the sand mold used in our experiment was lower than those quoted from previous studies. The yield stress measured by a uniaxial compression test was 1/20 of the quoted values. When the measured yield stress was adopted in the simulation, the simulated restraint force and contraction approached the measured results. The yield stress of the sand mold was a dominant factor in the restraint force simulated by the thermal stress analyses. The yield stress of the green sand mold used in the casting process should be measured to predict the residual stress using FEM thermal stress analyses. (C) 2013 Elsevier B.V. All rights reserved,

Nickel-based superalloys have been applied to gas turbine and aircraft jet engine parts due to superior high temperature strength and corrosion resistance. However, casting defect such as solidification cracking often occurs. In order to increase productivity of precision casting and various shape casting processes, predicting the occurrence of solidification cracking by using CAE (Computer Aided Engineering) should be essential. Therefore, it is necessary to obtain mechanical properties in the state of solid-liquid coexistence.
In the previous reports, high temperature mechanical properties were examined but tensile rate dependency wasn't examined. In this study, high temperature (around solidus temperature) mechanical properties and tensile rate dependency of their alloy were examined by using originally developed tensile test.

Nickel-based superalloys have been applied to gas turbine and aircraft jet engine parts due to superior high temperature strength and corrosion resistance. However, casting defect such as solidification cracking often occurs. In order to increase productivity of precision casting and various shape casting processes, predicting the occurrence of solidification cracking by using CAE (Computer Aided Engineering) should be essential. Therefore, it is necessary to obtain mechanical properties in the state of solid-liquid coexistence.
In the previous reports, high temperature mechanical properties were examined but tensile rate dependency wasn't examined. In this study, high temperature (around solidus temperature) mechanical properties and tensile rate dependency of their alloy were examined by using originally developed tensile test.

Nickel-based superalloys have been applied to gas turbine and aircraft jet engine parts due to superior high temperature strength and corrosion resistance. However, casting defect such as solidification cracking often occurs. In order to increase productivity of precision casting and various shape casting processes, predicting the occurrence of solidification cracking by using CAE (Computer Aided Engineering) should be essential. Therefore, it is necessary to obtain mechanical properties in the state of solid-liquid coexistence.
In the previous reports, high temperature mechanical properties were examined but tensile rate dependency wasn't examined. In this study, high temperature (around solidus temperature) mechanical properties and tensile rate dependency of their alloy were examined by using originally developed tensile test.

In order to predict the hot tearing during DC casting by using thermal stress analysis, constitutive equations in both solid–liquid coexisting and below solidus of alloy are inevitable. However, previous constitutive equations used in hot tearing predictions have no less than one problem as follows. (1) Testing methods of obtaining material constants weren't appropriate. Firstly, elastic strain reversibility was unconfirmed. Secondary, flat distribution of temperature in gauge length of specimen wasn't guaranteed. Thirdly, strain was measured from not local strain but cross head displacement. Fourth, the melt-back phenomenon was unavoidable in partial melting method which was caused by homogenization of segregation structure. (2) Temperature dependence of strain rate sensitivity of stress wasn't considered. (3) Some material constants were not obtained experimentally but guessed. In this study, elasto-viscoplastic constitutive equations (Hooke's and Norton's laws) for both partial solidification and below solidus were developed. In order to obtain material constants experimentally, two tensile tests which the problem (1) was solved were conducted using Al–5mass%Mg alloy. They were partial solidification tensile tests (above solidus) and high temperature tensile test with high frequency induction (below solidus). Then, temperature dependence of elastic, viscoplastic properties were investigated and material constants were obtained. Furthermore, they were compared with other previous testing methods.

New tentative semi-solid injection molding machine for magnesium alloy which was able to control solid fraction from 0 to 30% was developed. The solidification cracking susceptibility of the heat-resistant magnesium alloy containing 4% Al and 3% Ca fabricated by semi-solid injection molding and die-casting were investigated. The critical strain for solidification cracking of the magnesium alloy was obtained by U-type hot cracking test using in-situ observation method. High temperature ductility curve between liquidus and solidus temperatures were compared with semi-solid injection molding and die-casting. As result, solidification cracking susceptibility of semi-solid injection molding process with solid fraction of 25% was better than that of die-casting.

This paper describes the results of shaking table tests and effective stress analyses to investigate effect of seismic reinforcement for gravity type quay walls using ground anchors. As a result, the following conclusions are obtained. 1) Caisson displacement is constrained with the ground anchors due to the tensile force. 2) Dynamic behavior of quay walls with ground anchors is not affected by the difference of rubble mound layer thickness. 3) The relationship of the ground anchors between the tensile force and the elongation shows a hysteresis loop. 4) Effect of the ground anchors is confirmed by the analyses considering of both non-linear characteristic of the sand and the ground anchors.

True stress–true strain curve of JIS AD12.1 (A383.0) alloy was obtained by tensile test in the partial solidification state for predicting residual stress and distortion of aluminum castings. It was found that tensile strength in true stress–true strain curves decreased with increasing temperature and decreasing strain rate. In addition, most of true stress–true strain curves showed constant stress with increasing strain. Thus, the constitutive equation taking account of strain-rate-dependence and stress saturation in the partial solidification state should be used in the thermal-stress analysis during solidification. Material constants of Norton creep law of JIS AD12.1 alloy in the partial solidification state were successfully obtained for the first time in the world.

Most studies on CSSC (cyclic stress-strain curve) have been conducted in the condition of the stress ratio R=-1. There are few reports on the effect of mean stress on CSSC. In this study, the stress-controlled fatigue tests under R=0 and -1 were performed on JIS S25C (AISI 1020), IF steel, A2024-T6, A6061-T6 and Ti-6Al-4V alloys. CSSCs of R=0 and -1 were compared. Creep strain (ratcheting strain) was also measured to investigate dominating mechanism of fatigue life of these alloys. It was found that the effect of mean stress on CSSC was different with each alloy. The dominating mechanism of JIS S25C and IF steel was creep strain. However, the dominating mechanism of A2024-T6 and A6061-T6 was plastic strain amplitude. The dominating mechanism of Ti-6Al-4V alloy was both creep strain and plastic strain amplitude. Thus, fatigue life reduction caused by mean stress in Ti-6Al-4V alloy was remarkably occurred than the other alloys.

The prediction of residual stress in a stress lattice shape casting (stress lattice) has been conducted and discussed by some researchers via the Finite Element Method (FEM). However, most of the previous studies used the first-order tetrahedral element, which has poor analysis accuracy in problems including bending. The use of the first-order tetrahedral element makes the verification of these studies uncertain because the bending deformation essentially occurs in the stress lattice casting. This study first shows that the thermal stress analysis for the stress lattice should use the element that can represent the bending deformation in principle for bending of the thin parts. Second, the simulated residual stress was compared with the measured value. The thermal stress analysis successfully predicted the residual stress of the stress lattice casting with and 11 pct difference. In addition to the prediction of the residual stress, it is important from the viewpoint of the productivity of castings to reveal the effect of the shake-out temperature on the residual stress. However, in the previous studies, conclusions concerning the effect of the shake-out temperature on the residual stress were not consistent (i.e., the one study said the higher shake-out temperature decreased the residual stress, and another study said a higher shake-out temperature increased the residual stress). Therefore, the current study first discusses the reason for the inconsistent conclusions in the previous studies. Second, stress lattice castings were cast and shaken out at various shake-out temperatures. Then, the current study validated the effect of the shake-out temperature on the residual stress. Consequently, the experimental results supported the conclusion of Kasch and Mikelonis that the shake-out at higher temperature contributed to the increase of the residual stress in the casting. © The Minerals, Metals &amp
Materials Society and ASM International 2013.

The load on flange castings in sand molds was gradually increased beginning from the end of the solidification process until the final cooling stage. The maximum tensile load on the flange castings in furan sand molds was larger than that of the flange castings in green sand molds. With the furan sand mold, permanent deformation in the flange castings occurred beginning from the end of the solidification process until reaching a temperature of approximately 250°C. The mechanical interaction between the casting and the sand mold should be considered for more accurate stress calculations, particularly in furan sand molds. © 2012 Elsevier B.V.

True stress-true strain curve of JIS AD12.1 (A383.0) alloy was obtained by tensile test in the partial solidification state for predicting residual stress and distortion of aluminum castings. It was found that tensile strength in true stress-true strain curves decreased with increasing temperature and decreasing strain rate. In addition, most of true stress-true strain curves showed constant stress with increasing strain. Thus, the constitutive equation taking account of strain-rate-dependence and stress saturation in the partial solidification state should be used in the thermal-stress analysis during solidification. Material constants of Norton creep law of JIS AD12.1 alloy in the partial solidification state were successfully obtained for the first time in the world. © 2013 The Japan Institute of Light Metals.

The prediction of residual stress in a stress lattice shape casting (stress lattice) has been conducted and discussed by some researchers via the Finite Element Method (FEM). However, most of the previous studies used the first-order tetrahedral element, which has poor analysis accuracy in problems including bending. The use of the first-order tetrahedral element makes the verification of these studies uncertain because the bending deformation essentially occurs in the stress lattice casting. This study first shows that the thermal stress analysis for the stress lattice should use the element that can represent the bending deformation in principle for bending of the thin parts. Second, the simulated residual stress was compared with the measured value. The thermal stress analysis successfully predicted the residual stress of the stress lattice casting with and 11 pct difference. In addition to the prediction of the residual stress, it is important from the viewpoint of the productivity of castings to reveal the effect of the shake-out temperature on the residual stress. However, in the previous studies, conclusions concerning the effect of the shake-out temperature on the residual stress were not consistent (i.e., the one study said the higher shake-out temperature decreased the residual stress, and another study said a higher shake-out temperature increased the residual stress). Therefore, the current study first discusses the reason for the inconsistent conclusions in the previous studies. Second, stress lattice castings were cast and shaken out at various shake-out temperatures. Then, the current study validated the effect of the shake-out temperature on the residual stress. Consequently, the experimental results supported the conclusion of Kasch and Mikelonis that the shake-out at higher temperature contributed to the increase of the residual stress in the casting. © The Minerals, Metals &amp
Materials Society and ASM International 2013.

New tentative semi-solid injection molding machine for magnesium alloy which was able to control solid fraction from 0 to 30% was developed. The solidification cracking susceptibility of the heat-resistant magnesium alloy containing 4% Al and 3% Ca fabricated by semi-solid injection molding and die-casting were investigated. The critical strain for solidification cracking of the magnesium alloy was obtained by U-type hot cracking test using in-situ observation method. High temperature ductility curve between liquidus and solidus temperatures were compared with semi-solid injection molding and die-casting. As result, solidification cracking susceptibility of semi-solid injection molding process with solid fraction of 25% was better than that of die-casting. © 2013 The Japan Institute of Light Metals.

In order to predict the hot tearing during DC casting by using thermal stress analysis, constitutive equations in both solid-liquid coexisting and below solidus of alloy are inevitable. However, previous constitutive equations used in hot tearing predictions have no less than one problem as follows. (1) Testing methods of obtaining material constants weren't appropriate. Firstly, elastic strain reversibility was unconfirmed. Secondary, flat distribution of temperature in gauge length of specimen wasn't guaranteed. Thirdly, strain was measured from not local strain but cross head displacement. Fourth, the melt-back phenomenon was unavoidable in partial melting method which was caused by homogenization of segregation structure. (2) Temperature dependence of strain rate sensitivity of stress wasn't considered. (3) Some material constants were not obtained experimentally but guessed. In this study, elastoviscoplastic constitutive equations(Hooke's and Norton's laws)for both partial solidification and below solidus were developed. In order to obtain material constants experimentally, two tensile tests which the problem (1) was solved were conducted using Al-5mass%Mg alloy. They were partial solidification tensile tests(above solidus) and high temperature tensile test with high frequency induction (below solidus). Then, temperature dependence of elastic, viscoplastic properties were investigated and material constants were obtained. Furthermore, they were compared with other previous testing methods. © 2013 The Japan Institute of Light Metals.

The load on flange castings in sand molds was gradually increased beginning from the end of the solidification process until the final cooling stage. The maximum tensile load on the flange castings in furan sand molds was larger than that of the flange castings in green sand molds. With the furan sand mold, permanent deformation in the flange castings occurred beginning from the end of the solidification process until reaching a temperature of approximately 250°C. The mechanical interaction between the casting and the sand mold should be considered for more accurate stress calculations, particularly in furan sand molds. © 2012 Elsevier B.V.

Most studies on CSSC (cyclic stress-strain curve) have been conducted in the condition of the stress ratio R=-1. There are few reports on the effect of mean stress on CSSC. In this study, the stress-controlled fatigue tests under R=0 and -1 were performed on JIS S25C (AISI 1020), IF steel, A2024-T6, A6061-T6 and Ti-6Al-4V alloys. CSSCs of R=0 and -1 were compared. Creep strain (ratcheting strain) was also measured to investigate dominating mechanism of fatigue life of these alloys. It was found that the effect of mean stress on CSSC was different with each alloy. The dominating mechanism of JIS S25C and IF steel was creep strain. However, the dominating mechanism of A2024-T6 and A6061-T6 was plastic strain amplitude. The dominating mechanism of Ti-6Al-4V alloy was both creep strain and plastic strain amplitude. Thus, fatigue life reduction caused by mean stress in Ti-6Al-4V alloy was remarkably occurred than the other alloys.

Most studies on CSSC (cyclic stress-strain curve) have been conducted in the condition of the stress ratio R=-1. There are few reports on the effect of mean stress on CSSC. In this study, the stress-controlled fatigue tests under R=0 and -1 were performed on JIS S25C (AISI 1020), IF steel, A2024-T6, A6061-T6 and Ti-6Al-4V alloys. CSSCs of R=0 and -1 were compared. Creep strain (ratcheting strain) was also measured to investigate dominating mechanism of fatigue life of these alloys. It was found that the effect of mean stress on CSSC was different with each alloy. The dominating mechanism of JIS S25C and IF steel was creep strain. However, the dominating mechanism of A2024-T6 and A6061-T6 was plastic strain amplitude. The dominating mechanism of Ti-6Al-4V alloy was both creep strain and plastic strain amplitude. Thus, fatigue life reduction caused by mean stress in Ti-6Al-4V alloy was remarkably occurred than the other alloys.

To predict and control the residual stress present in sand castings manufactured via CAE (Computer Aided Engineering), the mechanical interaction between the casting and the sand mold during cooling must be determined experimentally. A device was developed in this study to determine the load on the casting caused by the resistance of the mold and the contraction of the casting during cooling. Our device consists of two modules that work simultaneously: a module containing a load cell, for measuring the load on the casting during cooling and a module containing an LVDT (Linear Variable Differential Transformer) for measuring the contraction of the casting during cooling. In performance verification testing, the device enabled the simultaneous measurement of the load on the sand casting and the contraction of the casting. This measurement was performed dynamically during the cooling process. Additionally, for the case where the contraction of the casting was hindered by the sand mold, the permanent deformation of the casting after shake out (which leads to residual stress in the casting) was successfully measured using our device. (C) 2012 Elsevier B.V. All rights reserved.

To predict and control the residual stress present in sand castings manufactured via CAE (Computer Aided Engineering), the mechanical interaction between the casting and the sand mold during cooling must be determined experimentally. A device was developed in this study to determine the load on the casting caused by the resistance of the mold and the contraction of the casting during cooling. Our device consists of two modules that work simultaneously: a module containing a load cell, for measuring the load on the casting during cooling and a module containing an LVDT (Linear Variable Differential Transformer) for measuring the contraction of the casting during cooling. In performance verification testing, the device enabled the simultaneous measurement of the load on the sand casting and the contraction of the casting. This measurement was performed dynamically during the cooling process. Additionally, for the case where the contraction of the casting was hindered by the sand mold, the permanent deformation of the casting after shake out (which leads to residual stress in the casting) was successfully measured using our device. (C) 2012 Elsevier B.V. All rights reserved.

Detection of solidification cracking of JIS AC7A aluminum alloy and Al–7%Si alloy was attempted by using acoustic emission method. In order to detect AE of solidification cracking, we investigated about AE signals of gas porosity, shrinkage cavity, and fracture in the semi-solid tensile test. The frequency bands of these casting defects were 150 kHz or less, so the source characterization of AE signals were difficult by that. Therefore we discussed and suggested value of integrated AE peak volt as an indicator of crack initiation. As a result, we proved value of time integral (IAP) is beneficial for eutectic alloy and value of temperature integral (TIAP) is beneficial for noneutectic alloy. We also suggested value of time temperature integral (DIAP) as the criterion of crack initiation which is beneficial for eutectic and non-eutectic alloy.

Nickel-based superalloys have been applied to gas turbine and aircraft jet engine parts due to superior high temperature strength and corrosion resistance. However, casting defect such as solidification cracking often occurs. In order to increase productivity of precision casting and various shape casting processes, predicting the occurrence of solidification cracking by using CAE (Computer Aided Engineering) should be essential. Therefore, it is necessary to obtain mechanical properties in the state of solid-liquid coexistence. In this study, we try to get high temperature (around solidus temperature) mechanical properties of the nickel-based superalloy such as Inconel 718 by using originally developed tensile test. In the previous reports, flat distribution of temperature in the gage length, and crack initiation strain were not ensured. On the other, the developed device ensured within 3 degrees C of the temperature distribution in the gage length as to the Inconel 718. Tensile strain was measured by using in-situ observation of marker on the surface of the specimen. As a result, Stress-Strain curve, crack initiation stress, and crack initiation strain of Inconel 718 were obtained.

Nickel-based superalloys have been applied to gas turbine and aircraft jet engine parts due to superior high temperature strength and corrosion resistance. However, casting defect such as solidification cracking often occurs. In order to increase productivity of precision casting and various shape casting processes, predicting the occurrence of solidification cracking by using CAE (Computer Aided Engineering) should be essential. Therefore, it is necessary to obtain mechanical properties in the state of solid-liquid coexistence. In this study, we try to get high temperature (around solidus temperature) mechanical properties of the nickel-based superalloy such as Inconel 718 by using originally developed tensile test. In the previous reports, flat distribution of temperature in the gage length, and crack initiation strain were not ensured. On the other, the developed device ensured within 3 degrees C of the temperature distribution in the gage length as to the Inconel 718. Tensile strain was measured by using in-situ observation of marker on the surface of the specimen. As a result, Stress-Strain curve, crack initiation stress, and crack initiation strain of Inconel 718 were obtained.

Nickel-based superalloys have been applied to gas turbine and aircraft jet engine parts due to superior high temperature strength and corrosion resistance. However, casting defect such as solidification cracking often occurs. In order to increase productivity of precision casting and various shape casting processes, predicting the occurrence of solidification cracking by using CAE (Computer Aided Engineering) should be essential. Therefore, it is necessary to obtain mechanical properties in the state of solid-liquid coexistence. In this study, we try to get high temperature (around solidus temperature) mechanical properties of the nickel-based superalloy such as Inconel 718 by using originally developed tensile test. In the previous reports, flat distribution of temperature in the gage length, and crack initiation strain were not ensured. On the other, the developed device ensured within 3 degrees C of the temperature distribution in the gage length as to the Inconel 718. Tensile strain was measured by using in-situ observation of marker on the surface of the specimen. As a result, Stress-Strain curve, crack initiation stress, and crack initiation strain of Inconel 718 were obtained.

Detection of solidification cracking of JIS AC7A aluminum alloy and Al-7%Si alloy was attempted by using acoustic emission method. In order to detect AE of solidification cracking, we investigated about AE signals of gas porosity, shrinkage cavity, and fracture in the semi-solid tensile test. The frequency bands of these casting defects were 150 kHz or less, so the source characterization of AE signals were difficult by that. Therefore we discussed and suggested value of integrated AE peak volt as an indicator of crack initiation. As a result, we proved value of time integral (IAP) is beneficial for eutectic alloy and value of temperature integral (TIAP) is beneficial for noneutectic alloy. We also suggested value of time temperature integral (DIAP) as the criterion of crack initiation which is beneficial for eutectic and non-eutectic alloy.

Detection of solidification cracking of JIS AC7A aluminum alloy and Al-7%Si alloy was attempted by using acoustic emission method. In order to detect AE of solidification cracking, we investigated about AE signals of gas porosity, shrinkage cavity, and fracture in the semi-solid tensile test. The frequency bands of these casting defects were 150 kHz or less, so the source characterization of AE signals were difficult by that. Therefore we discussed and suggested value of integrated AE peak volt as an indicator of crack initiation. As a result, we proved value of time integral (IAP) is beneficial for eutectic alloy and value of temperature integral (TIAP) is beneficial for noneutectic alloy. We also suggested value of time temperature integral (DIAP) as the criterion of crack initiation which is beneficial for eutectic and non-eutectic alloy.

Effect of pre-aging conditions on T5 heat treatment behavior of Al–9.0%Si–0.3%Mg die-casting alloy was investigated by using hardness measurement, differential scanning calorimetric analysis (DSC) and transmission electron microscopy (TEM). T5 treated alloy showed the positive effect on the two-step aging at pre-aging temperatures between 273 and 343 K, unlike T6 treated alloy which showed the negative effect at pre-aging temperatures below 343 K. The positive effect on two-step aging seems to be the characteristic behavior at T5 heat treatment. Difference of two-step aging behavior between T5 and T6 treatments seems to result from the difference of cluster (1) formed in the pre-aging process below 343 K. Higher water-quenching temperature after casting tended to increase the hardness after artificial aging.

Because it has been difficult to predict crack sensitivity depending on alloy composition during aluminum direct chill casting (DC casting), a new prediction method based on the relationship between the calculated solid fraction and temperature was developed for Al-Mn and Al-Mg series aluminum alloys. In this work, two crack indexes are suggested. The first index is brittle temperature range (BTR). The second index is based on the strain rate difference in the mushy region. These indexes are quantitatively calculated by using thermodynamic software such as Thermo-calc. This method is verified by DC casting of A3000 and A5000 series aluminum alloys. It can be utilized in the alloy design stages to control the crack sensitivity before casting. [doi:10.2320/matertrans.L-M2010823]

Because it has been difficult to predict crack sensitivity depending on alloy composition during aluminum direct chill casting (DC casting), a new prediction method based on the relationship between the calculated solid fraction and temperature was developed for Al-Mn and Al-Mg series aluminum alloys. In this work, two crack indexes are suggested. The first index is brittle temperature range (BTR). The second index is based on the strain rate difference in the mushy region. These indexes are quantitatively calculated by using thermodynamic software such as Thermo-calc. This method is verified by DC casting of A3000 and A5000 series aluminum alloys. It can be utilized in the alloy design stages to control the crack sensitivity before casting. [doi:10.2320/matertrans.L-M2010823]

Solidification cracking is one of the defects of casting. In order to increase productivity of DC casting and various shape casting processes, predicting the occurrence of solidification cracking by using CAE (Computer Aided Engineering) should be essential. Therefore, it is necessary to obtain mechanical properties of solid-liquid coexistence alloys. However, in the previous reports, flat distribution of temperature in the gage length was not ensured. In this study, tensile test device for a semi-liquid Al–Mg alloy with 2°C of the temperature distribution in the gage length (Max. 10 mm) was developed. Tensile strain was measured by using in-situ observation of marker on the surface of the specimen. As the result, stress-strain curves and fracture strain in various temperatures on the semi-liquid Al–4.3mass%Mg alloy were measured. It is also found that tensile strain tends to be higher at neighborhood of the crack. Thus, strain depends on the gage length. From the observation of the fracture surface and estimation of Scheil-Gulliver micro-segregation, microstructure of the specimen were supposed to change due to the heating process.

Seismic behavior of sheet pile quay walls with ground anchor for seismic reinforcement is not clarified enough, because few case study of seismic behavior has been conducted. In this study, shaking table tests with a scale of 1 to 17 model in 1G field were conducted to investigate seismic behavior of sheet pile quay walls with ground anchor. And effective stress analyses were conducted to simulate seismic behavior of sheet pile quay walls with ground anchor in this tests. By arrangement of the ground anchors, displacement of the quay wall was fairly constrained and effectiveness of them was confirmed by the test and the analysis.

In our previous works, tensile test devices for both semi-liquid and semi-solid aluminum alloys were developed. In this report, comparison of mechanical properties such as ultimate tensile strength (UTS) and fracture strain between semi-liquid and semi-solid Al–Mg alloy was examined. Difference of the mechanical properties will be caused by the microstructural change during heating process in the tensile test of semi-liquid alloy. By constant load creep test to the semi-liquid alloy, about 90% deformation of permanent and 10% of elastic deformation were found. Thus, in the thermal-stress analysis, solid–liquid co-existence aluminum alloy should be dealt as visco–elastic or viscoelasto–plastic material rather than elasto-plastic material.

It is necessary to obtain the mechanical properties of semi-solid alloy in order to predict the occurrence of solidification cracking, one of the serious casting defects of aluminum alloy die-casting and DC casting, by computer simulation. New in-situ measuring method for mechanical properties on semi-solid aluminum alloy by horizontal tensile test, based on that of Oya and Kitaoka et al., was developed. The strain was determined by measuring the displacement between dendrites defined as the makers. Influence of gage length on the fracture strain was clarified. Fracture strain and tensile strength of semi-solid Al–7mass%Si alloy were obtained.

Effect of pre-aging conditions on T5 heat treatment behavior of Al-9.0%Si-0.3%Mg die-casting alloy was investigated by using hardness measurement, differential scanning calorimetric analysisDSCand transmission electron microscopy TEM. T5 treated alloy showed the positive effect on the two-step aging at pre-aging temperatures between 273 and 343 K, unlike T6 treated alloy which showed the negative effect at pre-aging temperatures below 343 K. The positive effect on two-step aging seems to be the characteristic behavior at T5 heat treatment. Difference of two-step aging behavior between T5 and T6 treatments seems to result from the difference of cluster1formed in the pre-aging process below 343 K. Higher water-quenching temperature after casting tended to increase the hardness after artificial aging.

Solidification cracking is one of the defects of casting. In order to increase productivity of DC casting and various shape casting processes, predicting the occurrence of solidification cracking by using CAE (Computer Aided Engineering) should be essential. Therefore, it is necessary to obtain mechanical properties of solid-liquid coexistence alloys. However, in the previous reports, flat distribution of temperature in the gage length was not ensured. In this study, tensile test device for a semi-liquid Al-Mg alloy with 2°C of the temperature distribution in the gage length (Max. 10mm) was developed. Tensile strain was measured by using in-situ observation of marker on the surface of the specimen. As the result, stress-strain curves and fracture strain in various temperatures on the semi-liquid Al-4.3mass%Mg alloy were measured. It is also found that tensile strain tends to be higher at neighborhood of the crack. Thus, strain depends on the gage length. From the observation of the fracture surface and estimation of Scheil-Gulliver micro-segregation, microstructure of the specimen were supposed to change due to the heating process.

It is necessary to obtain the mechanical properties of semi-solid alloy in order to predict the occurrence of solidification cracking, one of the serious casting defects of aluminum alloy die-casting and DC casting, by computer simulation. New in-situ measuring method for mechanical properties on semi-solid aluminum alloy by horizontal tensile test, based on that of Oya and Kitaoka et al., was developed. The strain was determined by measuring the displacement between dendrites defined as the makers. Influence of gage length on the fracture strain was clarified. Fracture strain and tensile strength of semi-solid Al-7mass%Si alloy were obtained.

In our previous works, tensile test devices for both semi-liquid and semi-solid aluminum alloys were developed. In this report, comparison of mechanical properties such as ultimate tensile strength(UTS)and fracture strain between semiliquid and semi-solid Al-Mg alloy was examined. Difference of the mechanical properties will be caused by the microstructural change during heating process in the tensile test of semi-liquid alloy. By constant load creep test to the semi-liquid alloy, about 90% deformation of permanent and 10% of elastic deformation were found. Thus, in the thermal- stress analysis, solid-liquid co-existence aluminum alloy should be dealt as visco-elastic or viscoelasto-plastic material rather than elasto-plastic material.

In our previous works, tensile test devices for both semi-liquid and semi-solid aluminum alloys were developed. In this report, comparison of mechanical properties such as ultimate tensile strength(UTS)and fracture strain between semiliquid and semi-solid Al-Mg alloy was examined. Difference of the mechanical properties will be caused by the microstructural change during heating process in the tensile test of semi-liquid alloy. By constant load creep test to the semi-liquid alloy, about 90% deformation of permanent and 10% of elastic deformation were found. Thus, in the thermal- stress analysis, solid-liquid co-existence aluminum alloy should be dealt as visco-elastic or viscoelasto-plastic material rather than elasto-plastic material.

In order to develop high quality aluminum alloy products, modeling to predict a behavior of solute distribution in dendrite and precipitation within the inter-cell region quantitatively is important. However Scheil's and Brody-Flemings's equations can't find the precipitation behavior during solidification because a diffusion in liquid phase is not considered. Numerical analysis methods typified by "Phase Field" are more effective but a calculation time has not been enough short to apply industrial use, yet. Consequentially a mathematical model is developed based on the Matsumiya's calculus of finite differences for analyzing inter-cell microsegregation and precipition. Diffusion of solutes in both the solid and liquid is taken into consideration, and realistic mesh shape to calculate microsegregation is conducted for finite difference calculation at this Matsumiya's model. But the calculus hardly predict precipitation behavior, since liquid concentration shift by precipitation during solidification is not considered. Then we improved the model to estimate precipitation of intermetallic compounds by thermodynamics database in this work. Additionally evaluation method of microsegregation used by EPMA was developed Consequently, we built a new segregation model, which could predict both the microsegregation and the kinds and amount of intermetallic compounds.

Seismic behavior of gravity type quay walls with ground anchor for seismic reinforcement is not clarified enough, because few case study of seismic behavior of gravity type quay walls with ground anchor for seismic reinforcement has been conducted. In this study, we conducted shaking table test about caisson type quay wall with ground anchor for seismic reinforcement and evaluated relation between the deformation of caisson type quay wall and the tension of ground anchors. And we calculated safety factor of sliding, rotation, bearing capacity, based on result of shaking table test, because we researched relation between safety factor and deformation mode of caisson type quay walls.

Carbon fiber reinforced aluminum alloy composites (CF/Al composites) are expected to be applied to electric power cable due to superior specific strength and specific modulus. It is reported that CF/Al composites form aluminum carbide (Al4C3) at the interface between carbon fiber and aluminum alloy. However, in operating condition (300°C, 36 years), the growth of Al4C3 and tensile strength of CF/Al composites have not been clarified. In this study, at first, pitch-based CF (XN-60)/Al composites are fabricated with ultrasonic infiltration method and held at 300, 450, 500°C for a given length of time. Secondary, influence of holding temperature on the quantity of Al4C3 was investigated. Thirdly, relationship between the quantity of Al4C3 and tensile strength of CF/Al composites was examined. As holding temperature increased, the quantity of Al4C3 increased and tensile strength decreased. Reaction kinetics calculation indicated that remaining strength versus theoretical strength (ROM) of the CF/Al composites, held at 300°C for 36 years, was 0.81. That is, it should be better for applying the pitch-based CF/Al composites to electric power cable than the PAN-based CF/Al composites.

In order to develop high quality aluminum alloy products, modeling to predict a behavior of solute distribution in dendrite and precipitation within the inter-cell region quantitatively is important. However Scheil's and Brody-Flemings's equations can't find the precipitation behavior during solidification because a diffusion in liquid phase is not considered. Numerical analysis methods typified by "Phase Field" are more effective but a calculation time has not been enough short to apply industrial use, yet. Consequentially a mathematical model is developed based on the Matsumiya's calculus of finite differences for analyzing inter-cell microsegregation and precipition. Diffusion of solutes in both the solid and liquid is taken into consideration, and realistic mesh shape to calculate microsegregation is conducted for finite difference calculation at this Matsumiya's model. But the calculus hardly predict precipitation behavior, since liquid concentration shift by precipitation during solidification is not considered. Then we improved the model to estimate precipitation of intermetallic compounds by thermodynamics database in this work. Additionally evaluation method of microsegregation used by EPMA was developed. Consequently, we built a new segregation model, which could predict both the microsegregation and the kinds and amount of intermetallic compounds.

Detection of AE signals of solidification cracking of Al–7mass%Si alloy was attempted with in situ microscopic observation on the surface of casting using high speed camera. During solidification, it was found that AE signal from the casting with crack was higher than AE signal from the casting without crack. As a result of observation of casting using high speed camera, it was found that a rise of AE signal corresponded to initiation of solidification cracking. This stage was the stage of eutectic solidification. Thus, AE signals of solidification cracking should be generated during eutectic solidification. However, AE signals from the casting without cracking were also detected during solidification. This means that AE signals other than solidification cracking should be generated during solidification. Therefore, AE signals detected during eutectic solidification is not necessarily AE signals of solidification cracking. In order to detect AE signals of solidification cracking, it is necessary to separate AE signals of solidification cracking and other than solidification cracking.

Detection of solidification cracking of Al-7 mass%Si alloy was attempted by using acoustic emission method. According to our previous report, it was found that it is necessary to separate AE of solidification cracking from other estimated sources such as gas porosity, shrinkage cavity, or eutectic solidification to detect AE of solidification cracking. Then, we investigated about AE signals of gas porosity, shrinkage cavity, and eutectic solidification. As a result, AE of eutectic solidification was not generated or was too small to detect. AE of shrinkage cavity and gas porosity formation were detected at the start of eutectic solidification, and the frequency band was 150 kHz or less. The frequency band of AE accompanied with fracture of solidifying castings was also 150 kHz or less. Therefore, it was difficult to separate AE of solidification cracking from gas porosity, shrinkage cavity, or eutectic solidification by the frequency band. However, integrated AE peak volt during solidification correlated with crack ratio. When Solidification cracking occurred, Integrated AE peak volt was about more than 1000 V during solidification. In situ detection of solidification cracking will be possible by investigating integrated AE peak volt.

Carbon fiber reinforced aluminum alloy composites.CF/Al composites.are expected to be applied to electric power cable due to superior specific strength and specific modulus. It is reported that CF/Al composites form aluminum carbide.Al4C3.at the interface between carbon fiber and aluminum alloy. However, in operating condition.300°C, 36 years., the growth of Al4C3 and tensile strength of CF/Al composites have not been clarified. In this study, at first, pitch-based CF.XN-60./Al composites are fabricated with ultrasonic infiltration method and held at 300, 450, 500°C for a given length of time. Secondary, influence of holding temperature on the quantity of Al4C3 was investigated. Thirdly, relationship between the quantity of Al4C3 and tensile strength of CF/Al composites was examined. As holding temperature increased, the quantity of Al4C3 increased and tensile strength decreased. Reaction kinetics calculation indicated that remaining strength versus theoretical strength.ROM.of the CF/Al composites, held at 300°C for 36 years, was 0.81. That is, it should be better for applying the pitch-based CF/Al composites to electric power cable than the PAN-based CF/Al composites.

The cooling water used in vertical aluminum DC casting is generally induced to just under the mold to prevent molten aluminum break out. However, the cooling water leaks into the air gap between the mold and the surface of ingot, so water splashes over the top of the molten aluminum, resulting in uneven cooling. But the quantitative survey on the splash break behavior and the change in the cooling condition has not been conducted. Hence, to quantify the splash break condition and the cooling capability, we have made the "Cooling water simulator" capable of freely controlling the potential splash causing factors such as temperature of cooling water, water flow rate, air gap amount, and angle of the water flow, reproducing the splash, and measuring the heat transfer value. The experimental result reveals that the temperature of cooling water is not a factor highly influencing on heat flux and splash, but the splash behavior highly depends on three factors which are the water flow rate, the angle of water flow, and the distance between the cooling water hit point level and the cooling water exit level. And that the heat flux also increases when splash occurs. In addition, the splash break parameter was drawn based on the above three factors, and the correlation between splash break parameter and heat flux was clarified. Use of splash break parameter enables setting of the casting condition where the maximum heat flux is obtained while preventing splash.

Detection of AE signals of solidification cracking of Al-7mass%Si alloy was attempted with in situ microscopic observation on the surface of casting using high speed camera. During solidification, it was found that AE signal from the casting with crack was higher than AE signal from the casting without crack. As a result of observation of casting using high speed camera, it was found that a rise of AE signal corresponded to initiation of solidification cracking. This stage was the stage of eutectic solidification. Thus, AE signals of solidification cracking should be generated during eutectic solidification. However, AE signals from the casting without cracking were also detected during solidification. This means that AE signals other than solidification cracking should be generated during solidification. Therefore, AE signals detected during eutectic solidification is not necessarily AE signals of solidification cracking. In order to detect AE signals of solidification cracking, it is necessary to separate AE signals of solidification cracking and other than solidification cracking.

For improving the electrical efficiency of a hybrid car, relationship between microstructure and thermal properties of carbon nanotube, carbon fiber reinforced aluminum composites were examined as heat sink plate. In this study, influence of interfacial thermal resistance between fiber and binder, fiber diameter and binder volume fraction on thermal conductivity of the composites were examined. Firstly, in order to examine the effectivity for binding the fibers by the binder, the interfacial thermal resistance between fiber and binder was compared with that between fiber and matrix. In order to examine the influence of fiber diameter, the thermal conductivity of multi walled carbon nanotube preforms was compared with that of carbon fiber preforms. As the result, interfacial thermal resistance between fiber and binder was lower than that between fiber and matrix. Secondly, it was revealed that the larger fiber diameter leads to the higher thermal conductivity of the composite. Finally, the binder volume fraction was influenced by the microstructure so that the simply increasing and decreasing tendency was not found for thermal conductivity of the composites. By means of simulation for thermal conductivity, it was suggested that highly constricted fiber+binder part would deteriorate the thermal conductivity of the composites.

Since it has been difficult to predict a crack sensitivity depending on alloy composition during aluminum DC casting, a new prediction method based on a relation between calculated solid fraction and temperature was developed concerning Al-Mn and Al-Mg series aluminum alloys. In order to understand that the surface crack is located at ingot surface, two cracking modes were considered. They were due to "difference in strain between water chilling surface and mushy surface (crack mode Z)" and "difference in strain rate between surface and just inner surface (crack mode Y)". It's common knowledge that the material is cracked at high solid fraction region since melt cannot go into a crack gap and ductility is very low. Therefore we assumed the high solid fraction region was from 0.75 to 0.95 (Region II) and set up two indexes which are calculated from temperature difference (ΔT Π) and difference in a temperature drop per unit solid fraction ( ΔRΠ/ΔTΠ) within the Region II. By using the both indexes, which were called "Brittle temperature range" and "Parameter of the strain rate difference" respectively, the crack sensitivities through IX casting experiments were suitably represented.

When stainless steels are sensitized, intergranular corrosion resistance decreases and sensitization is assumed to be one of the causes of stress corrosion cracking under severe environment in nuclear reactors. For various stainless steels, Laser surface melting (LSM) with CO2 laser have been reported. In this study YAG laser that is able to transmit energy with the optical fiber is used. Influence of LSM condition on intergranular corrosion resistance was examined. Experimental results indicated that by LSM sensitized Type 316 stainless steel restored intergranular corrosion resistance as the material after solution treatment. Moreover, The material that sensitized in low temperature (LTS) 500 degrees C for 24 hours after LSM exhibited the same degree of intergranular corrosion resistance as the material after solution treatment. This would be caused by lower 6 ferrite volume fraction in the solidified microstructure after LSM than that in arc welding metal.

Carbon fiber reinforced aluminum alloy composites (CF/AL composites) are expected to be applied to electric power cable due to superior specific strength and specific modulus. But it is reported that CF/AL composites form aluminum carbide (Al 4C 3) at the interface between carbon fiber and aluminum alloy. However, influence of operating temperature, in usual 300°C, on the growth of Al 4C 3 and tensile strength of CF/AL composites have not been clarified. In this study, at first, PAN-based CF (M40J)/AL composites are fabricated with ultrasonic infiltration method and hold at 300, 450, 500°C for 64 h. Secondary, influence of holding temperature on the quantity of Al 4C 3 was investigated. Thirdly, relationship between the quantity of Al 4C 3 and tensile strength of CF/AL composites was examined. As holding temperature increased, the quantity of Al 4C 3 increased and tensile strength decreased. By examining extracted fiber with SEM, it was found that tensile strength should be decreased because Al 4C 3 corroded carbon fiber. Kinetics calculation indicated that remaining strength of the CF/AL composites, held at 300°C for 36 years, was nought. As a result, it will be difficult to apply the CF/A1 composites to electric power cable.

The die casting process and its alloys have been developed in recent years for automobile body parts such as B-pillars. However, it is known that die casting alloys with high ductility and fracture elongation often show a higher susceptibility to cracking during solidification than conventional AI-Si alloys. Thus, it is important to estimate and control the susceptibility to cracking during solidification before trial casting or mass-production. in this study, as a representative non-heat treatment type alloy, Al-4.Swt%Mg (JIS AC7A, AA 514) aluminum alloy was used. The effect of the addition of silicon and grain refiner on the reduction of the susceptibility to cracking was examined. in order to evaluate the susceptibility to cracking, both the "I-beam casting cracking test" and the "TIG spot welding cracking test" were carried out. As a result, the addition of Ti + B worked as a grain refiner on both testing methods. The susceptibility to cracking was significantly reduced by the addition of Ti + B in both the I-beam casting and the weld crater. It was found that the finer grain size led to lower susceptibility to cracking. Furthermore, the susceptibility to cracking of the die casting product decreased with the addition of Ti + B. (C) 2008 Elsevier B.V. All rights reserved.

Multi-walled carbon nanotubes (MWCNTs) should be attractive for the reinforcement of metal-matrix composites, because of their high strength, high modulus and high thermal conductivity. However, the fiber diameter of MWCNTs is hundreds of times smaller than that of carbon fiber. This causes difficulty in infiltration into the MWCNT preform. Moreover, the threshold pressure which was applied to the preform will cause preform deformation. Therefore, knowledge of preform compressive properties which are the buckling strength and elastic modulus are necessary to fabricate the composites. In this study, at first, wettability of the basal plane of graphite by molten aluminum or magnesium was measured using the sessile drop method. Moreover, trial fabrication of MWCNT-reinforced aluminum or magnesium alloy composites was carried out by squeeze casting. As a result, these composites were fully infiltrated. An order-of-magnitude agreement was found between the estimated threshold pressure and the applied infiltration pressure to the MWCNT preform. (C) 2008 Elsevier B.V. All rights reserved.

Carbon fiber reinforced aluminum alloy composites (CF/Al composites) are expected in aerospace and electric power cable industries due to superior specific strength and specific modulus. But, it is known that CF/Al composites form aluminum carbide (Al4C3) at the interface between carbon fiber and aluminum alloy when CF/Al composites are fabricated. However, effects of type of carbon fiber (PAN, pitch) on growth mechanism of Al 4C3 and tensile strength of CF/Al composites have not been clarified. In this study, at first, CF/Al composites are fabricated with ultrasonic infiltration method. Secondary, effects of type of carbon fiber and fabricating time on quantity and size of Al4C3 were investigated. Thirdly, effects of quantity and size of Al4C 3 on tensile strength of CF/Al composites were examined. The length of Al4C3 increased with increase in fabricating time for PAN-based composites. It was suggested that the numbers of nucleating sites of Al4C3 increased with an increase in fabricating time for pitch-based composites. As the result, as to the PAN-based composites, it should be controlled less than 100 nm of the length of Al4C3 to inhibit degradation of tensile strength. As to the pitch-based composites, fabricating time should be shorter.

As a potential material for high temperature services, unidirectional sintered SiC fiber reinforced SiO2-30 mol%Al2O3 composite with BN interphase was synthesized by liquid state hot pressing. The obtained composite had dense structure, and both the mullite and glassy SiO2 phases were found in the matrix. No chemical reaction among the constituents was observed. The composites were exposed to air at 1573 K up to 3600 ks and then oxidation behavior was investigated. Three-point flexural test was performed at room temperature after each exposure. Flexural strength decreased with an increase in the exposure time and then leveled off in 360 ks. This tendency corresponds to the crystallization Of SiO2 glass in the matrix. Change in the stress field which is caused by the crystallization and phase transformation Of SiO2 would lead to the reduction of the strength. It was confirmed that oxidation progressed along the fibers and the oxidized depth was limited within approximately 200 mu m independent of exposure time (3.6 similar to 3600 ks). The oxide layer with approximately 10 pm thickness including silicon, aluminum and oxygen was found on the surface of the composites. This layer will prevent the inner oxidation within 200 mu m.

Continuous M40J carbon fiber reinforced aluminum magnesium alloy composite wires have been fabricated using ultrasonic infiltration. The infiltration phenomenon is examined from the viewpoint of acoustic cavitation. The ease of infiltration of the molten alloys was found to be proportional to the maximum intensity of the acoustic cavitation. The ease of infiltration and the intensity were enhanced by the addition of surfactant elements into the molten aluminum. Thus, a decrease in surface tension caused an increase in the generation of acoustic cavitation thereby resulting in infiltration. Therefore, the generation of the acoustic cavitation is an infiltration controlling factor during the use of ultrasonic vibration. (c) 2006 Elsevier Ltd. All rights reserved.

In order to fabricate continuous carbon fiber-reinforced aluminum alloy matrix composites, various infiltration methods such as gas pressure infiltration, CVD-infiltration, and ultrasonic infiltration methods have been developed. Among these methods, the ultrasonic infiltration method is the simplest. In this study, the effects of ultrasonic power, the diameter of the hole of the horn, fabricating speed, and magnesium content on the ease of infiltration are investigated. As the results, both an ultrasonic power of 200 W and the addition of more than 2.4 mass% Mg are indispensable to infiltrate molten aluminum alloy into a PAN-based M40J carbon fiber bundle, which has 6000 filaments. Contrariwise, the tensile strength and relative strength (ROM ratio) of the obtained composites decreased from 1100 MPa (0.7) at both 2.4 and 4.7 mass% Mg contents to 800 MPa (0.5) at 10 mass% Mg content. This was probably caused by an increase in the content of the Al3Mg2 intermetallic compound. Consequently, the addition of magnesium is effective in improving the infiltration; however, it causes the strength of the composites to decrease. It is found that in this process, the optimum magnesium content in aluminum from the viewpoints of ease of infiltration and strength was 4.7 rnass%. (c) 2007 Elsevier Ltd. All rights reserved.

As the fabricating process of continuous M40J carbon fiber reinforced aluminum alloys composite wires by using ultrasonic infiltration method, the mechanism of the infiltration of molten alloys into the bundle of the carbon fibers was examined especially from the viewpoints of wettability, acoustic cavitation and threshold pressure for infiltration. It was found that the infiltratability of the alloys was proportional to the maximum intensity of the acoustic cavitation. Both the infiltratability and the intensity were enhanced by the addition of surfactant elements for molten aluminum. Thus, decrease in surface tension will cause the generation of acoustic cavitation to increase. When the ultrasonic vibration is applied to molten aluminum alloys, acoustic cavitation would be formed on/in the bundle of carbon fibers. Then the shock wave, which was caused by the cavitation collapsed, would lead to leave the distance between fibers in the bundle. From the results of the direct observation of the bundle in the aqueous solution using high speed camera during applied ultrasonic vibration, the diameter of the bundle was increased by generating the cavitation. Thus, this phenomenon will also cause the decrease in the threshold pressure for infiltration. Therefore, both the generation of the acoustic cavitation and increase in the interval between fibers will be the controlling factors of the infiltration in the fabricating process of this kind of composites by using ultrasonic vibration.

In order to investigate behavior of acoustic cavitation in molten aluminum alloys when ultrasonic vibration was applied, we measured sound pressure in molten aluminum alloys by means of acoustic emission (AE). The frequency analysis was performed to the obtained sound pressure. As a result, the characteristic frequency of acoustic cavitation would be 5, 10 and 260 kHz. The maximum intensity at 260 kHz was increased by addition of magnesium in molten aluminum at the ultrasonic power of more than 200 W drastically. The reason why the addition of magnesium led to promote the generation of the cavitation would be caused by its relatively low surface tension and high vapor pressure. From the viewpoint of the cavitation threshold, both the decrease in the surface tension of the molten aluminum alloy and the increase in the vapor pressure will lead to generate the acoustic cavitation. Therefore, addition of zinc of 8.7 mass% and bismuth of 0.5 mass % also caused to generate acoustic cavitation.

M40J carbon fiber reinforced aluminum-2.4, 4.7 and 10mass% magnesium alloy composite wires were continuously fabricated using ultrasonic infiltration method. The tensile strength of the composites varied from 800 to 1200 MPa caused by change in cross sectional area, volume fraction of fibers and so on. Moreover, their tensile strength decreased with the increase in Mg content in aluminum. In the case of the Al-10mass%Mg matrix composites, Al 3Mg 2 intermetallic compound was found on the surface of fibers and continuously formed between fibers in the matrix like bridging. The decrease in the strength would be caused by premature fracture and bridging of the intermetallic compound between fibers. As the result, the tensile strength of the CF/Al composites was controlled by cross sectional area and amount and shape of Al 3Mg 2 intermetallic compound in aluminum matrix.

In this study, we have developed a novel manufacturing process of "preform wires" which are semi-finished materials used for the fabrication of CF/Al composites. The preform wires were continuously fabricated using an ultrasonic infiltration method. The manufacturing apparatus fundamentally consisted of preheating furnace, ultrasonic horn and transducer and taking up reel for winding. Influence of each processing parameter on the infiltration ratio was investigated. As the optimum condition, after carbon fiber was heated at 973 K for desizing, the ultrasonic was applied at 200 W to molten Al-Mg alloy for the infiltration of the molten alloy into the bundle. The addition of magnesium into molten aluminum improved the infiltratability. CF/Al-4.7%Mg preform wires with a tensile strength of 1100MPa was obtained at the fabricating speed of 0.22 m/s. It is clear that the ultrasonic infiltration method is effective to fabricate CF/Al preform wires.