Acta Metallurgica Sinica(English Letters), 2017, 30(5): 409-432
A Review on Grain Refinement of Aluminum Alloys: Progresses, Challenges and Prospects
Ren-Guo Guan, Di Tie
School of Materials Science and Engineering, Northeastern University Shenyang 110819 China
Ren-Guo Guan, Di Tie. A Review on Grain Refinement of Aluminum Alloys: Progresses, Challenges and Prospects. Acta Metallurgica Sinica(English Letters)[J], 2017, 30(5): 409-432 doi:10.1007/s40195-017-0565-8

Abstract:

Aluminum becomes the most popular nonferrous metal and is widely used in many fields such as packaging, building transportation and electrical materials due to its rich resource, light weight, good mechanical properties, suitable corrosion resistance and excellent electrical conductivity. Grain refinement, which is obtained by changing the size of grain structure by different techniques, is a preferred method to improve simultaneously the strength and plasticity of metallic materials. Therefore, grain refining of aluminum is regarded as a key technique in aluminum processing industry. Up to now, there have been a number of techniques for aluminum grain refining. All the techniques can be classified as four categories as follows: grain refining by vibration and stirring during solidification, rapid solidification, the addition of grain refiner and severe plastic deformation. Each of them has its own merits and demerits as well as applicable conditions, and there are still some arguments in the understanding of the mechanisms of these techniques. In this article, the research progresses and challenges encountered in the present techniques and the future research issues and directions are summarized.

Key words: Aluminum Grain refinement ; Vibration and stirring ; Rapid solidification ; Grain refiner ; Severe plastic deformation
1 Introduction

Aluminum becomes the most popular nonferrous metal and is widely used in many fields due to its rich resource, light weight, good mechanical properties, suitable corrosion resistance and excellent electrical conductivity. All the mentioned advantages are positive deciding factors for designers, manufacturers and industrial users who are always on the lookout for better-performing materials and innovative processes. Current industrial applications of aluminum alloys include but not limited to aerospace, automotive, marine, building, packaging, mechanical industry, 3C and energy distribution [1, 2]. In January 2015, the European Commission issued the European Metallurgical Roadmap of 2050. Its long-term goal was to develop metals and metal-based matrix composites with high strength, high ductility, corrosion and wear resistance and to give priority to improve the strength, formability and corrosion resistance of 2, 5, 6 and 7 series aluminum alloys as well as to expand the use of heat-resistant aluminum alloys [3]. In addition, high-strength aluminum alloy was listed as one of the key development plans in Advanced Manufacturing Partnership (AMP) of the USA [4]. Aluminum Industry Roadmap of 2020 was also put forward by the Aluminum Industry Association of the USA [5]. The goal was to reduce aluminum production costs by 25%, energy consumption by 25% and to develop a new generation of aluminum alloy materials with zero pollution emissions [5]. United States Council for automotive materials had made the aluminum alloy industry roadmap as well. According to this plan, in order to expand the application of aluminum alloy in automobile industry, the corrosion, design, preparation, processing, microstructure control and material jointing of aluminum alloys would be comprehensively studied [6, 7, 8]. Besides the mentioned traditional usages, Al has been used as matrix materials in metal matrix composites fabricated stir casting, powder metallurgy and mechanical milling as well [9, 10, 11, 12].

With increasing demands for better mechanical performance of aluminum alloys, various technologies were applied during the past decades for achieving higher strength as well as ductility of aluminum alloys [13], among which purifying and grain refinement were two key issues according to their outstanding effects on improving aluminum products’ mechanical performance. Similar as the importance of purifying in melting process, grain refinement is crucial for the products’ mechanical properties during and after solidification process. Grain refinement becomes a preferred method to improve simultaneously the strength and plasticity of metallic materials. The most evident improvement brought by grain refinement is elevated strength at room temperature, which can be theoretically explained by the Hall-Petch relation (Eq. 1) [14]:

$$\sigma_{\text{y}} = \sigma_{0} + k_{\text{y}} \cdot d^{{ - \frac{1}{2}}}, \ \ (1)$$

where σy is the yield stress, σ0 is a material constant of the starting stress for dislocation movement (or the resistance of the lattice to dislocation motion), ky is the strengthening coefficient (a constant specific to each material), and d is the average grain diameter. Hall-Petch strengthening (or grain boundary strengthening) is an important method of strengthening materials by grain refining; meanwhile, ductility is improved with increasing grain numbers. Compared with precipitation strengthening and deformation hardening, grain refinement can simultaneously improve strength, toughness and ductility and reduce casting defects, such as segregation and porosity. Moreover, it eliminates the columnar structure and therefore improves the quality of wrought alloys through improving formability. Grain refinement playing an important role in both cast and wrought aluminum alloys was gradually realized with aluminum’s application in industry. Reduced mechanical properties have been noted in plate products when as-cast grain is not fine and uniform [15]. It has been reported that twinned columnar grains reduced the formability, yield strength and tensile elongation to fracture of alloy [16], and a coarse microstructure may result in a variety of surface defects in alloys used in rolled or extruded form for applications [17]. Besides, hot cracking in the shell zone of a cast ingot is more severe if the grain structure is not equiaxed or fine [18]. An equiaxed structure allows a high casting rate to be achieved before hot crack occurring. There are also several other benefits from grain refinement, such as homogenous distribution of second phases, microporosity in a fine scale, better feeding to eliminate shrinkage porosity, improved ability to achieve a uniform anodized surface, better strength and fatigue life [19]. Therefore, most cast aluminum products with high mechanical performance are grain refined currently [20].

The development of grain refinement is shown in Table 1. Grain refining can be carried out during and after solidification process. These methods could be classified into four categories, namely grain refining by vibration and stirring (VS) during solidification, rapid solidification (RS), the addition of grain refiner (GR) and severe plastic deformation (SPD). It is taken for granted that grain refining by vibration and stirring is due to dendritic arm breakage under shear force induced by VS. Despite majority of researchers’ agreement to this viewpoint, it was still argued by some other scholars. Actually, the heterogeneous nucleation on the mold wall and convection could contribute to grain refinement. Moreover, mechanism of liquid infiltration in grain boundary induced by plastic bending of dendrite arm and re-melting of dendritic root induced by ripening and stirring was further put forward [21]. These disputes may be lasted for a long time. Further exploring, these issues may broaden the field of technology and bring about a revolution in condensed materials.

Table 1

Development of grain refinement methods

Subjected to solidification theory, a number of rapid solidification methods were developed for grain refining and metal glass (MG) preparation. Some of them, for instance, roll casting and spray deposition have been successfully used in industries for producing high-performance materials, but there were still issues in application, such as the purity, the efficiency and cost were not satisfactory [22].

It is now accepted that the presence of both potent nucleant particles and sufficient solutes are essential for effective grain refining during solidification process [23]. Despite the common recognition of these two essentials, the details of grain refinement mechanisms have still some ambiguities. One of the major problems is to determine the exact factors that control the efficiency of grain refinement. It appears that something important is still missing in the current understanding of grain refinement. It is worth noting that the spotlight of previous studies on grain refinement mechanism of aluminum was primarily focused on the Al-Ti-B and Al-Ti-C systems as they are the most common grain refiners used in the foundry [24]. However, it is believed that other solute elements may also produce effective grain refinement in Al, which may provide fresh insight into the factors that control the grain refining efficiency [25].

Grain refinement of aluminum in solid state generally relies on recrystallization which can be mostly achieved by classical thermo-mechanical treatment and severe plastic deformation technologies. It should be noted, for heat treatable aluminum alloys, deformation hardening and precipitation strengthening also provide important contribution to alloy strengthening, but these factors may decrease the material’s ductility. As far as thermo-mechanical treatment is concerned, a wide variety of conventional heat treatment processes have been developed and used in industry. In severe plastic deformation (SPD), high strain is imposed on metal samples using special tools, so that free flow is prevented and large deformation occurs under a hydrostatic pressure [47]. A critical requirement in SPD processing is that the material experiences a very high strain without introducing any significant changes in the overall dimensions of the workpiece. Actually, SPD method had successfully been applied to manufacture high-performance sword in ancient times, which is now called multi-direction forging. Its principle lies in the recrystallization that is induced by multi-direction forging and thermal action. Although the history of this type of processing may be traced back more than 2000 years ago to the metal-working practices of ancient China, the principle of SPD was revealed within the last few decades with the help of sophisticated analytical tools. According to this principle, a variety of SPD methods were developed and ultra-fine and nano-grained materials were successfully prepared. Processing by SPD is now a favorable procedure for the fabrication of bulk solids with exceptionally small grain sizes. However, most SPD methods face the challenges like long processing time, high cost and low efficiency. Therefore, dealing with these problems becomes an important issue, which is crucial for the development of aluminum processing industry.

Up to now, there have been a lot of techniques developed for aluminum grain refining, but a systematic summary of these methods is still absent. The main objective of the present review is to summarize the primary types of technologies currently applied for grain refining of aluminum alloys: grain refining by vibration and stirring (VS) during solidification, rapid solidification (RS), the addition of grain refiner (GR) and severe plastic deformation (SPD). Each of them has its own merits and demerits as well as applicable conditions, and there are still some arguments in the understanding of the mechanisms of these techniques. This article reviews the latest research progresses and challenges encountered in the present techniques, and the future research issues and directions are summarized.

2 Grain Refinement by Vibration and Stirring (VS)

Stirring and vibration during solidification are widely used for grain refining of aluminum. Various stirring methods were developed, including mechanical stirring, electromagnetic stirring and bubble mixing [48, 49, 50]. Wannasin et al. [51] investigated bubble mixing process and obtained fine-grained alloys. It was concluded that bubble stirring could increase the number of primary nucleation in the literature. As far as vibration is concerned, it mainly consists of mechanical vibration and ultrasonic vibration [52, 53]. Several theories toward different technologies were established to explain the mechanism of grain refining during stirring and vibration as follow.

(1) Shear fracture mechanism of dendritic arm. This theory was supported by Jackson and Flemings [54, 55]. Shear force caused by the stirring of viscous fluid was considered to have a shear effect on dendrite and led to dendritic arm breakage [56]. However, there was no obvious fracture defects observed inside the crystal of the stirred slurry. Some scholars pointed out that stirring could not induce dendrite fracture but could make elastic or plastic bending of the dendritic arm [57, 58].

(2) Mechanism of liquid infiltration in grain boundary induced by plastic bending of dendritic arm. Plastic bending of the dendritic arm and the infiltration of liquid in grain boundaries were observed in several studies [57, 59, 60]. Both high-angle and low-angle grain boundaries were obtained in the primary grain of the stirred slurry, so recrystallization behavior existed in the stirring process [61]. Thus, the mechanism of liquid infiltration in grain boundary induced by plastic bending of dendritic arm was proposed [62]. It was believed that when the orientation of high-angle grain boundary was more than 20°, the grain boundary energy could exceed that of the solid-liquid interface two times. In this situation, dendritic arms could detach from their mother grains [61]. But Pilling’s calculation results showed that only when the relative motion velocity of primary phase and the liquid phase was large enough, the dendritic arm bending could happen [62].

(3) Re-melting of dendritic root induced by ripening and stirring. Flemings and Martinez pointed out that solute segregation could easily occur at the dendritic root in the Ostwald ripening process, and solute was easy to be pushed into this region, which would decrease equilibrium melting point because of solute enrichment [63]. The equilibrium melting point at dendritic root was also decreased due to the existence of curvature, which can be expressed by Eq. (2) [64]. So the dendritic root might be re-melted and detach from its mother grain [55].

$$\Delta T_{\text{r}} = - r\frac{{2T_{\text{m}} V_{\text{s}} \sigma }}{{\Delta H_{\text{m}} }}, \ \ (2)$$

where ∆Tr is the change in melting points caused by solid phase curvature, σ is solid-liquid interfacial tension, Tm is the melting temperature, Vs is the molar volume of pure solid, ∆Hm is the molar enthalpy of fusion, and r is the average curvature of liquid-solid interface.

Hellawell believed that the stirring could produce strong fluctuations of temperature and composition, and stress may concentrated at the dendritic root, which caused re-melting of dendritic root [65].

(4) Crystal rain. Campbell and Southin believed that in the early stage of solidification, there was a large degree of undercooling on the surface of mold wall [66, 67]. Because the density of solid phase was bigger than that of liquid phase, nuclei formed on the wall fell into the melt and led to the formation of crystal rain. Some of these nuclei would be re-melted by heat flux that was caused by convection, and the rest would be the cores of grains growth [66].

(5) As far as ultrasonic vibration is concerned, the mechanism of grain refinement is mainly based on two phenomena: cavitation and acoustic streaming induced by ultrasonic vibration [68]. The cavitation is a formation, expansion and collapse process of bubbles in the melt [69]. The collapse of bubbles can produce high pressures on melt layers in a very short time and enhances nucleation rate and dendrite fragmentation. The pressures of collapse bubbles can cause the decrease in activation energy in nucleation, and the pressures are high enough to break up the clustered particles as well as dendritic arms leading to higher nucleation rate [70, 71]. Moreover, it was speculated that the bubble collapse could also increase the melting point of the melt near the broken bubbles, so the effective supercooling degree was improved. Partial undercooling might occur around the broken bubbles and therefore nucleation rate increased [70].

2.1 Electromagnetic Stirring (EMS)

The application of electromagnetic field in aluminum alloy melt can also refine grain size, and the schematic diagram of electromagnetic stirring is shown in Fig. 1. The magnetic field can induce the Lorenz force which causes pressure difference and intense convection in the melt. In conventional casting, the melt first solidifies in the low-temperature zone near the mold wall, followed by the heterogeneous nucleation. If there is no convection in the melt, the nucleus will grow into dendrites. However, due to the presence of electromagnetic field, the crystal nucleus on the mold wall could be swept away by strong convection and dispersed into the entire melt, which increases nuclei number and leads to grain refinement [72]. Accompanying grain refinement, structure spheroidizing also takes place in the melt in an electromagnetic field [73]. According to some previous studies, the convection induced by electromagnetic field can promote the uniformity of the distributions of solute and temperature in melt, which enables primary α-Al grains grow with nearly the same rate in all directions [74]. Besides this explanation, intense convection in the melt can also reduce the thickness of unstable layer around primary α-Al grains, which results in the decrease in the constitutional undercooling degree, so the unstable growth of nuclei is effectively avoided [75]. It should be pointed out that Joule heat generated by the induced current can also affect the solidification process. The electrical resistance of the liquid is higher than that of the solid phase, and the resistance at the interface of them is much larger. When the electric current flows through the melt, Joule heat induced by the solid-liquid interface at dendritic root reaches its maximum value. The heat accumulating at roots will promote the fusing of the dendritic arms [76] and further lead to grain refining and spheroidizing. Hellawell et al. [77] believed that the EMS could produce severe fluctuations of temperature and composition, which results in stress concentration at dendritic roots. All the mentioned factors cause re-melting of dendritic root in grain ripening.

Fig. 1

Schematic diagram of electromagnetic stirring

In 1922, Mcneill obtained the patent of the USA on controlling solidification process by EMS [78]. In 1947, Dreyfus in Switzerland manufactured the first device and applied it into arc furnace steelmaking [32]. In 1952, electromagnetic stirring at the secondary cooling zone of caster was realized in Germany [79]. In 1973, the industrial application of electromagnetic stirring technology in continuous casting was realized by the SAFE Company in France [80]. From then on, a variety of EMS devices were developed and applied toward casting industries including Nippon Steel Corporation and Kawasaki Steel Corporation in Japan, Forges & Acieries de Dillingen in German, Rotelec in France, Ispat Sidbec in Canada, POSCO in South Korea, etc. [81, 82, 83]. This technique was now also widely used in China [84]. According to electromagnetic stirring position, this method contains mold electromagnetic stirring (M-EMS), secondary cooling electromagnetic stirring (S-EMS) and final solidification electromagnetic stirring (F-EMS), which are summarized in Table 2. EMS can not only refine grain size but also homogenize the distribution of chemical composition. The application history of electromagnetic stirring in the ferrous metal is relatively earlier. At present, its application in aluminum alloy casting process is also popular, such as in ordinary aluminum alloy, special aluminum alloy and high-purity aluminum. In recent years, more researchers focused on microstructure refining by EMS. A variety of aluminum alloys, such as A356, 4045, 6061, 7075, Al-Si alloys, were treated by EMS during casting, and the primary α-Al was significantly refined [85]. For instance, Zuo et al. [86] found that the low-frequency electromagnetic field has a significant influence on the grain size of 7075 aluminum alloy. Ingot microstructure became very uniform and fine when EMS was applied on the melt during solidification process. The average grain size was refined from 230 to 42 μm and grain morphology transformed from coarse dendrites to equiaxed grains.

Table 2

Classification of electromagnetic stirring

Electromagnetic field has an obvious grain refining effectiveness. Compared with mechanical stirring method, EMS prevents the loss of stirring rods. However, the control of the parameters of this method is relatively strict. In addition, the capacity of the melt that can be stirred by EMS is fewer than mechanical stirring. Despite these shortcomings, it still has a very good prospection of industry applications. Several research directions should be addressed at present including solidification theory under EMS, EMS technology of large-volume melt, strong magnetic field stirring technique and multiple-field electromagnetic stirring method. With development of electromagnetic stirring technology, more extensive application of electromagnetic stirring technology is promising.

2.2 Vibration Method (VM)
2.2.1 Mechanical Vibration (MV)

Mold vibration experiment to modify the as-cast microstructure of components can be dated back to 1870s. Chernov found that the application of mechanical vibration during solidification of steel caused grain refinement [27]. Mechanical vibration is usually loaded on the bottom of the crucible. Exciting force can be controlled by adjusting vibration amplitude and frequency. For the forced sinusoidal vibration of the melt produced by mechanical vibration, the exciting force F is [87]:

$$F≈α_{max}m, \ \ (3)$$

where αmax is the vibration peak acceleration, which can be calculated by,

$$\left| {\alpha_{ \hbox{max} } } \right| = \left| A \right|\left( {2\pi f} \right)^{2}, \ \ (4)$$

where |A| is the vibration amplitude and f is the vibration frequency.

Generally, vibration intensity is evaluated by vibration acceleration which can be controlled by adjusting the vibration frequency and amplitude of vibration mold. Exciting force increases with increasing vibration intensity.

It is believed that vibration can refine microstructure by promoting nucleation by many studies. Limmaneevichitr et al. and Taghavi et al. [88, 89] have studied the influence of mechanical vibration on the microstructure of A356 alloy. The results showed that the average grain size of primary α-Al became finer and more globular as the degree of vibration increased. Comparing with as-cast sample, grain size of the alloy was refined from 1200 to 174 µm at a vibration frequency 50 Hz for 15 min [89]. The refinement grain effect was markedly enhanced when the vibration was introduced at larger solid fractions and lower pouring temperatures. With the increase in vibration frequency and vibration amplitude, the size of α-Al and secondary dendritic arm spacing decreased [90].

Mechanical vibration can not only refine grain but also improve the capacities of degassing and feeding during casting [91]. The vibration is usually applied to continuous casting mold for the purpose to improve the stability of casting process. This method fully proved effective on grain refining by laboratory investigation, but mass production in industrial still encountered some problems, one of which was that vibration might induce casting defects such as porosity and thermal fragment [92, 93]. Furthermore, a reliable vibration mold requires stable anti-shock performance. Therefore, efforts should be focused on the refining mechanism study of vibration casting, vibration mold design and the vibration casting system for large-volume melt.

2.2.2 Ultrasonic Vibration (UV)

The research of ultrasonic vibration for metallurgical applications can be traced back to 1930s. The introduction of high intensity ultrasonic vibration to metal melt can bring out a noticeable reduction in grain size. Ultrasonic vibration is usually applied at the bottom or the top of molten metal. A typical schematic view of ultrasonic treatment equipment is depicted in Fig. 2. Eskin had studied the microstructures of pure aluminum before and after ultrasonic vibration treatment [29]. The results showed that the morphology of solidification microstructure transformed from columnar dendritic structure to fine equiaxed grain when the ultrasonic vibration was applied. St. John et al. found that grain refinement was achieved only when ultrasonic treatment was applied in the temperature range from above liquidus 20 °C to a certain temperature below the liquidus [94]. The same phenomenon was also observed by Khalifa [69]. Haghayeghi et al. [95] investigated the effects of frequency on the refinement of microstructure. For 10 and 14 kHz of frequency, no microstructure refinement was found. When the frequency was increased to 17.5 and 20 kHz, the grain size was refined from 118 to 68 and 60 µm, respectively. Eskin suggested that the cavitation threshold that significantly affected the solidification microstructure of aluminum alloys was around 17 kHz [96].

Fig. 2

Schematic view of ultrasonic equipment

Cavitation and acoustic streaming induced by ultrasonic vibration can benefit degassing and homogenizing of the melt [97]. Ultrasonic vibration can refine grain size and homogenize the distribution of nanoparticles in the melt. It was reported that magnesium matrix composites with uniform nano-SiC reinforced particles could be prepared by this method, but the scale of the composite was very small [98]. The difficulty of manufacturing high energy ultrasonic equipment has become an important bottleneck in the industrial application. The limitations of high-power ultrasonic equipment and high-temperature resistance ultrasonic horn restrict the large-scale industrial applications [99], so further efforts should be made in these areas. The comparison of mechanical vibration and ultrasonic vibration is shown in Table 3.

Table 3

Comparison of mechanical vibration and ultrasonic vibration

3 Grain Refinement by Rapid Solidification (RS)

It is known that the nucleation work decreases with increasing cooling rate, and the nucleation rate increases with decreasing nucleation work. Therefore, the nucleation rate increases with the increase in cooling rate, which is the basic theory of rapid solidification process used as a grain refining method [100, 101, 102, 103]. Especially, rapid solidification can be used to prepare alloys that can easily segregate or coarsen. This method was actually developed on the base of the investigation by Duwez in 1960s [104]. It was found that non-equilibrium crystal could be obtained when liquid metal was cooled at an adequate cooling rate [105, 106, 107, 108, 109]. Since then, rapid solidification was applied to prepare different kinds of fine-grain metal materials. In 1990s, this method was used to manufacture amorphous alloy and became a hot research topic [110, 111, 112, 113, 114, 115, 116]. As far as cooling rate is concerned, the cooling rate of large-scale casting is less than 0.1 K/s and that of metal mold casting is usually in the range of 1-1000 K/s. Once the cooling rate reaches 103 K/s, it can be classified into rapid solidification. Generally, the cooling rate of rapid solidification can reach more than 106 K/s by using special devices. In order to obtain high cooling rate, at least one-dimensional size of the processed material should be small enough [117, 118, 119, 120]. Up to now, many techniques have been developed to achieve rapid solidification, which are used for preparing metal glass (MG) or fine-grain materials [121, 122, 123, 124, 125, 126]. The cooling rates of different solidification methods are listed in Table 4. Several typical rapid solidification techniques such as melt spinning method, gas atomization method, copper mold casting, liquid forging technique, water-quenching method and high-pressure die casting were widely studied in recent years for manufacturing metal glass [28, 30, 31, 34, 127]. In melt spinning method, the cooling rate can reach over 103 K/s by a high-speed rotating roll. In gas atomization method, liquid metal is usually dispersed into tiny droplets by high-speed gas flow, and the cooling rate can reach 102-104 K/s. In copper mold casting method, copper is used due to its high thermal conductivity to achieve rapid solidification of small pieces of metal materials. In liquid forging process, the punch penetrates into alloy melt and leads to rapid cooling of melt. In water-quenching method, a quartz tube containing molten alloy is quenched into cold water to achieve a cooling rate in the range of 102-103 K/s. In high-pressure die casting, alloy melt completely rushes into the mold within a short time under high pressure, and a high cooling rate can be achieved. The most popular rapid solidification techniques that were widely used in metal processing industries are twin-roll casting and spray deposition method. The application of twin-roll strip casting in aluminum alloy is successful, but the efficiency is low. Spray deposition can refine metal grain size to nanoscale. It can also refine and spheroidize hard phases in needle shape as well as avoid composition segregation. However, the materials prepared by spray deposition are usually porous.

Table 4

Cooling rates of different solidification methods

3.1 Twin-Roll Casting (TRC)

Twin-roll casting was first proposed by Bessemer. It is a short process technology which combines casting and hot rolling to produce metal strip directly from liquid metal [26]. Schematic diagram of twin-roll casting is shown in Fig. 3. Liquid metal is poured into the gap between two rotating rolls cooled by inside water, and thin solidified layer forms on the surface of each roll when liquid metal contacts the rolls; subsequently, the thickness of solidified layer increases. When both sides of the solidified layer meet at the final point of solidification, the casting process is over, and the casting strip is dragged out of the gap by the rolls simultaneously [133, 134]. During twin-roll casting, the cooling rate of liquid metal can be as high as 102-103 K/s which is much higher than that of conventional casting [129, 130].

Fig. 3

Schematic diagram of twin-roll casting

The development of twin-roll casting is listed in Table 5. Twin-roll casting was firstly patented in 1857, but its first application in industry was about a century later due to the limitations of control technology and measurement devices [135]. In 1951, twin-roll casting was used in industrial application for the first time by American Hunter Engineering [136]. In 1960, 3C (continuous caster between cylinders) twin-roll caster was developed by Scal Company in France [137]. Inclined twin-roll caster was developed by Hunter Engineering in 1962 [138]. In 1985, Nippon Metal Industry Co. developed unequal diameter twin-roll caster [139]. 3C twin-roll caster and inclined twin-roll caster are the most popular twin-roll casters. Twin-roll casting is widely used worldwide to produce metal strips at present. Hunter Engineering in the USA, Davy Ltd in the UK, and Pechiney Company in France produced aluminum strip in 1980s [140]. A twin-roll casting production line for producing austenitic stainless steel was built by Thyseen Krupp in 1999 [141]. In 2002, the first castrip was developed by American Nucor Corporation to produce ultra-thin cast strip of carbon steel and stainless steel [142]. Magnesium alloy plate of 1860 mm in width and 1 mm in thickness was produced by a twin-roll casting production line of Germany SZMT Company in 2002 [140], and the prepared plate was used in the automobile door frame of Volkswagen Automobile Company [140]. The thin strip continuous casting of aluminum alloy has been widely applied in industry at present, and it was gradually applied for manufacturing magnesium alloy strip.

Table 5

Comparison of different twin-roll casters

In the past 50 years, twin-roll casting developed rapidly and was successfully commercially applied. Hypereutectic Al-Si alloy strip, 5182 alloy strip, 1050 alloy strip, 7050 alloy sheet and 6111 alloy strip were successfully prepared [143, 144]. Wang and Zhou produced 7050 alloy sheet with the average grain size of 30-50 μm by twin-roll casting, and they found that the average grain size decreased and became more uniform with the increase in roll casting speed and roll gap wideness [143]. Haga et al. [144] from Osaka Institute of Technology prepared 3003, Al-6Si and Al-12Si alloy sheet with the thickness of 1-3 mm by a melted drag twin-roll caster (MDTRC) at the speed of 30 m/min in 2001. Haga et al. [145] developed a high-speed twin-roll caster, and Al-3Mg-0.6Si alloy sheet was prepared with the speed of 60 m/min. The technology was also applied for preparation of magnesium alloy strip in 1980s, and the first twin-roll strip casting of magnesium alloy was developed by Commonwealth Scientific & Industrial Research Organization in 2000 [146]. Its application is more and more extensive [147, 148]. Twin-roll semisolid casting was developed by combing vibrating slope and rolling mill, and the mechanical properties of the casting strip were significantly improved compared with conventional casting processed products [149, 150].

Though twin-roll casting has the outstanding advantages of low cost and energy saving, it also has limitations like low casting speed, restriction of alloys and poor mechanical properties [151]. The efficiency of twin-roll casting is determined by the diameter and cooling strength of rolls, and aluminum alloys with large freezing zone are difficult to be produced by twin-roll caster. When the freezing zone of aluminum alloy is wide, strip cannot completely solidify before the strip was rolled. So twin-roll casting is limited in aluminum alloys with narrow freezing zones [151]. In addition, the uneven flow behavior of the liquid metal in the molten pool usually results in poor surface quality of the strip, and the fluctuation of surface level of the molten pool may cause solidification defects [152].

The directions of future developments should include improving the rolling speed of twin-roll casting, modifying strip quality including surface precision and inner microstructure, as well as expanding the types of alloys and thickness of the product. For this purpose, efforts should be made in two issues: to improve the cooling capability of the roll and to develop semisolid rolling technology.

3.2 Spray Deposition Method (SD)

SD method is another important rapid solidification technique usually used for manufacturing fine-grain alloys as raw materials of powder metallurgy. It is also called Osprey method or spray casting that is shown in Fig. 4. In this method, liquid metal stream is atomized by inert gas at a certain pressure into the droplets which are then propelled away from the region of atomization by airflow. Heat exchanges tempestuously between the droplets and airflow, and the cooling rate of droplets can reach 103-105 K/s. The solidified droplets are collected by the substrate. By controlling the continuous movement of the substrate, large preforms can be produced in a variety of geometries including billets, tubes and strips [131].

Fig. 4

Schematic diagram of spray deposition

The comparison of different spray deposition methods is shown in Table 6. Spray forming was first proposed by Singer at Swansea University in 1968, and it was reported for the first time in 1970 [35, 153, 154]. In 1974, Brooks, a student of Singer, applied spray deposition in forging, developed a series of aluminum alloys suitable for spray deposition. He designed the first complete set of SD equipment and founded the Osprey Company [155, 156]. Subsequently, Grant et al. in MIT developed liquid dynamics compaction (LDC) process which was similar to spray deposition [156, 157, 158]. In 1988, Lavernia from California University developed spray co-deposition process to prepare metal matrix composites [159, 160]. At Drexel University, Lawley proposed reactive spray forming for preparing metal matrix composites in situ [161, 162]. Nowadays, spray deposition has been widely used in industry to produce various products, such as Al-25Si-4Cu-1Mg alloy engine cylinder liner used in Mercedes Benz car at PEAK company, Al-Si alloy cycle engine blade used in Mazda car at Japan Sumitomo light metal company, high-speed tool steel and alloy steel at UK Aurora company, and stainless steel tube at United States Naval Laboratory [163].

Table 6

Comparison of different spray deposition methods

Spray deposition can refine grain size to nanoscale. It can also refine and spheroidize hard phases in needle shape such as Al-Fe phases as well as avoid composition segregation, so it is an ideal solution for processing difficult-forming metal and easy segregation alloys. However, it is difficult to achieve complete metallurgical bonding between the solidified droplets, and it is hard to fully get rid of gas bubbles in material matrix. Thus, the materials prepared by spray deposition are usually porous. Future studies should focus on precise control of atomization process parameters, expanding the diversity of products and improving the density of product.

3.3 Vibrating Cooling Slope Process (VCS)

New rheo-casting with a cooling slope for casting was developed by UBE Company in Japan [40]. In this process, liquid alloy was poured onto a slope, fine primary α-Al grains could be obtained due to high cooling rate and melt flow. High cooling rate induced by the slope led to heterogeneous nuclei generating on the slope, and they were continuously sheared off the slope surface by metal flow. A large quantity of nuclei was then distributed in the slurry, which resulted in fine grains in the casting. Haga et al. has developed the rheo-rolling process by integrating cooling slope with rolling mill, and the mechanical properties of the strip prepared by the process were obviously improved as comparing with normal casting alloy [164, 165, 166]. However, the process stability was not ideal and the parameters needed to be carefully controlled, because slurry adhered to the slope surface occasionally. In order to deal with this problem and improve the slurry quality, Guan et al. [44, 167] applied vibration on the slope and vibrating cooling slope process was then developed. The schematic diagram of vibrating cooling slope is shown in Fig. 5. The average cooling rate model of the vibrating cooling slope process has also been established,

$$\overline{\Delta T}_{\text{l}} = \frac{{\overline{{h_{\text{l}} }} L(T_{{\text{l}}\infty } - T_{\text{bs}} )}}{{c_{\text{l}} \rho_{\text{l}} u_{{\text{l}}0} d_{{\text{l}}0} }} - \frac{{f_{\text{s}} L_{\text{m}} }}{{c_{\text{l}} }} \ \ (5)$$

where $$\overline{ΔT_l}$$ is the average cooling rate, $$\overline{h_l}$$ is the average convection coefficient between metal melt and cooling slope, L is the length of slope, Tl∞ is the temperature of melt out of temperature boundary layer, Tbs is the temperature of upper surface of the slope, cl is the specific heat of melt, ρl is the density of melt, ul0 is the initial flow speed of melt, dl0 is the initial thickness of melt, Lm is solidification latent heat, fs is solid fraction, C and n are empirical parameters, x is flow length, g is acceleration of gravity, H is casting height, θ is slope angle, v is the viscosity of melt, f is vibration frequency, A is vibration amplitude, Pr is Prandtl number, and λ is the heat conductivity coefficient of metal melt. As far as A356 alloy is concerned, the average cooling rate of the vibrating cooling slope process varies in the range of 10-360 K/s which is significantly higher than that of common casting process.

Fig. 5

Schematic diagram of vibrating cooling slope

Based on theoretical calculation, it is believed that the shear force induced by metal flow and vibration is not large enough to shear off dendritic arms, but the shear force can cause the heterogeneous nuclei detaches from the slope surface. According to this study, a nucleation model of vibrating cooling slope process was put forward [168],

$$I = \frac{n}{\eta } \times \frac{{N_{\text{A}}^{{\frac{1}{3}}} k_{\text{B}} T^{3} T_{\text{m}}^{2} \alpha_{\text{m}}^{2}\Delta S^{2} f^{2} (\theta )}}{{6r_{\text{a}}^{5} V^{{\frac{1}{3}}}\Delta H^{2}\Delta T^{2} }}\exp \left[ - m\frac{{16\pi \alpha_{\text{m}}^{ 3}\Delta S^{3} T^{2} T_{\text{m}}^{2} f^{3} (\theta )}}{{3\Delta H^{2}\Delta T^{2} N_{\text{A}} k_{\text{B}} }}\right], \ \ (6)$$

where NA is the Avogadro number, kB is the Boltzmann constant, T is the temperature of melt, Tm is the liquidus of the alloy, αm is the structural factor, ΔS is the molar entropy change, θ is the solid-liquid interface wetting angle, f(θ) is the wetting angle factor, ra is atomic radius, V is the molar volume of melt, ΔH is solidification latent heat, ΔT is the undercooling of the melt, η is dynamic viscosity, m is the influence factor of critical nucleation energy, and n is the influence factor of diffusion coefficient under coupling effects of shear and vibration.

A356 alloy part was prepared through combining vibrating slope and squeeze casting. Figure 6 shows the comparison of microstructures of A356 alloy that were prepared by different processes. The results showed that semisolid squeeze casting by vibrating cooling slope could refine grain size and eliminate eutectic phase segregation which usually occurred in traditional squeeze casting [169]. The ultimate tensile strength and elongation of A356 alloy were increased by 12% and 21%, respectively, as comparing with that of traditional squeeze casting [159]. Rheo-rolling process was also developed through combining vibrating slope and rolling mill, and the mechanical properties of the strip obtained by the processes were significantly improved as well [170, 171]. The tensile strength and elongation of Mg-3Sn-1Mn-1La (wt%) alloy strip prepared by the process were increased by 37% and 89% as comparing with that of Mg-3Sn-1Mn-0.87Ce (wt%) alloy processed by conventional casting [171].

Fig. 6

Comparison of microstructures of A356 alloy that were prepared by different processes

In cooling slope method, opened surface usually induced oxide of aluminum. So closed cooling channels with inert gas protection are adopted. As a novel convenient processing technology, vibrating cooling slope method effectively deals with the problem of process stability and it has the advantages of low cost and high efficiency, which promise its good application prospect.

4 Grain Refinement by Grain Refiner (GR)

Grain refinement is required in the casting process of aluminum alloy since it reduces the casting defects and improves casting properties. The addition of grain refiner to aluminum alloy has become a common industrial practice to achieve grain refinement of casting aluminum alloy. The grain size of aluminum alloy depends on the solidification process which consists of nucleation and growth of α-Al. In the nucleation stage, tremendous heterogeneous nucleation sites of α-Al are introduced into aluminum alloy melt by the addition of grain refiner. The heterogeneous nucleation can also be accelerated by solution elements which can induce constitutional supercooling. In the growth of α-Al grain, the solution elements (or insoluble particles) can aggregate between grains and then restrict the growth of grain. Hence, grain size is refined due to high nucleation rate and restraining of grain growth. The carbide-boride theory, the peritectic theory, the duplex nucleation mechanism and solute theory are put forwarded to illustrate the mechanism of grain refining by grain refiner.

Cibula proposed that heterogeneous nucleation of α-Al was promoted by carbide or boride particles in solidification process of aluminum alloy [33]. Those particles had the characters of high melting point and small size and acted as nucleation sites of α-Al, such as carbide particles, boride particles, Ca-contained compound and Nb-base compound. When the grain refiner was added into aluminum alloy melt, it was important whether the refiner supplied sufficient effective nucleation sites for inducing heterogeneous nucleation. Zhang et al. suggested the Ti, V, Zr and Nb elements induced constitutional supercooling and supplied heterogeneous nucleation sites [172, 173]. Ning et al. added Al-Ti-C refiner into aluminum alloy melt, and they observed that the TiC particles gathered and acted as nucleation sites of α-Al [174]. Mondal et al. [175] showed that the high melt point Ca-contain intermetallic phases, i.e., CaAl2, could act as nucleation sites. Nafisi and Birol et al. [176, 177] proposed that the nucleation sites in Al-B master alloy were AlB2 phase, and grain refinement of aluminum alloy was achieved by the addition of Al-B. Zhang et al. [178] showed that the Al3Fe phases in Al-Fe master alloy refined grain of pure Al, and Al-Fe master alloy promoted nucleation rate and restricted the growth of α-Al grain. Bolzoni et al. [179] proposed that tremendous nucleation sites (i.e., Al3Nb, NbB2 and Nb-base intermetallic particles) were introduced by the addition of Nb-B alloy. Farahani et al. and Patakham et al. [180, 181] showed that the additions of Zr (in the form of Al-Zr master alloy) and Sc (in form of Al-Sc master alloy) introduced massive heterogeneous nucleation sites. Pan et al. [182] observed that the addition of La and B induced the formation of LaB6 phase, and better grain refinement of aluminum alloy was achieved by the LaB6 phases. However, Fan et al., Gruzleski et al. and Zhou et al. [183, 184, 185] believed that the insoluble particles (i.e., TiB2) did not act as nucleation sites of α-Al in the absence of other soluble elements. Effective grain refinement of aluminum alloy was observed by the addition of Al-Ti-B refiner. Zhou et al. [185] also observed that there was no grain refinement when TiB2 particles were separately added into high pure Al, but TiB2 showed grain refinement in commercial pure Al. They argued that the impurity (i.e., Fe, Ti, Si) in commercial pure Al activated grain refinement of TiB2. Murty et al. [186] proposed that the refiner contained TiAl3, TiB2 and AlB2 particles showed a better grain refinement of aluminum alloy than that only contained single particles. Wang et al. [187] suggested that TiB2 and TiAl20Ce particles in Al-Ti-B-Re refiner acted as nucleation sites of α-Al. Nie et al. proposed that Er restrained the growth of α-Al grain and agglomeration of TiB2 particles [188, 189, 190].

The peritectic theory revealed that the grain refinement by the refiner derives from peritectic reaction. Crossley et al. [191] proposed binary peritectic reaction induced by TiAl3

L+Al3Ti→α-Al(solid)

Guzowski et al. [192] proposed ternary peritectic reaction by the addition Al-Ti-B master alloy,

L+Al3Ti+(TiB2)→α-Al+(TiB2)

Murty et al. [186] proposed binary eutectic reaction by the addition Al-B master alloy in aluminum alloy melt,

L→α-Al+AlB2

Antonio et al. [193] proposed ternary peritectic reaction for Al-Ti-B master alloy,

L→α-Al+TiAl3+(Al,Ti)B2

However, the reliability of this reaction remains to be clarified. Backerud et al. [194] proposed the peritectic hulk theory. It was suggested that TiAl3 was surrounded by a hulk formed from TiB2, so the dissolution rate of TiAl3 was decreased by the hulk. Once the composition of Ti reached a critical value, peritectic reaction took place. However, Fan et al. and Mohanty et al. [183, 184] suggested that peritectic reaction was not the main grain refinement of refiner. Because the TiAl3 particles are unstable in aluminum alloy melt, the amount of Ti introduced by refiner is less than the requirement for peritectic reaction [183].

Duplex nucleation mechanism considered that nucleation included two stages [184]. The first stage is that TiAl3 gathers on the surface of TiB2, and then α-Al nucleates on the site that is composed of TiAl3 and TiB2. The hypernucleation theory proposed by Jones is similar with the duplex nucleation mechanism [195]. However, Horsfield et al. doubted that the source of Ti on the surface of TiB2 was uncertain, and Schumacher et al. argued that the duplex nucleation mechanism was inaccuracy [196, 197].

The solute theory considered that the solute element, i.e., Ti, and nucleation particles were necessary for grain refinement of aluminum alloy. The constitutional supercooling caused by solute element promoted heterogeneous nucleation, and the segregation of solute element near the solid-liquid interface restricted the grain growth of aluminum alloy. Zhang et al. studied the effects of Ti, V, Zr, Nb, Cu, Mg and Si on grain refinement of aluminum alloy [173, 174]. It was revealed that the grain refinement potency from high to low was, Ti > V, Zr, Nb > Cu, Mg, Si. The Cu, Mg and Si elements induced grain refinement by constitutional supercooling. Ti elements exhibited the best grain refinement of aluminum alloy since it induced constitutional supercooling and supplied heterogeneous nucleation sites. Murty et al. [198] observed that the refinement effectiveness of aluminum alloy did not change with the increase in Ti (B) content when the content exceeded 0.03 wt%. Wang et al. [187] also observed that the refinement effectiveness of Al-Ti-B-Re did not change when the addition of refiner exceeded 0.2 wt%. Zhang et al. [199] showed that the grain refinement was promoted by the addition of rare earth element. They suggested that rare earth wrapped around the TiAl3 and reduced the surface energy of TiAl3, indicating the peritectic reaction was accelerated. With the development of testing method in recent years, some researchers tried to explain the mechanism of grain refinement with the lattice matching at the solid substrate. Nevertheless, no specific mechanism is found. The arguments in the present theories of grain refiner are summarized in Table 7. And Fig. 7 shows the schematic of nucleation in different theories.

Table 7

Arguments of different theories on mechanism of grain refinement by grain refiner

Fig. 7

Schematic of nucleation in different theories

In the grain refinement of aluminum alloy, the relative grain size (RGS) of alloy can be calculated by the following model [200],

$${\text{RSG}} = 1- \left( {\frac{{m_{ 1} c_{ 0} }}{{m_{ 1} c_{ 0} - \Delta T_{\text{n}} }}} \right)^{{ 1 / P}},\ \ (7)$$

where m1 is liquidus slope, c0 is concentration of solute element in alloy, ΔTn is supercooling for nucleation, and P is supercooling coefficient. This model was verified by pure Al and Al-Si alloy with addition of Al-Ti master alloy [200].

The different products require different grain refiner, so the composition of refiner becomes complex with the development of refiner. The grain refiner of aluminum alloys include Ca, Al-Ti, Al-B, Al-Sr, Al-Fe, Al-Zr, Al-Sc, Al-Ti-B, Al-Ti-C, Al-Ti-Be, Al-Nb-B, Al-Ti-B-Re, Al-Ti-B-C-Re. Nowadays, the most widely used grain refiner is Al-Ti-B master alloy. The different grain refiners are summarized in Table 8.

Table 8

Different grain refiners

Mondal et al. [175] observed that the yield strength of 7178 alloy was increased by 50 MPa by the addition of 1 wt% Ca. Birol et al. [177] proposed that the fine and spherical grain of Al-Si-10 Mg alloy was obtained by the addition AlB3 refiner. Bolzoni et al. [179] suggested that the average grain size of A356 alloy was decreased to ca. 500 μm by the addition Nb-B refiner. Farahani et al. [180] suggested that the best grain refining efficiency of 7178 alloy was achieved by addition of 0.03 wt% AlB3 refiner. Sofyan et al. [213] applied Al-Ti master alloy to A333 alloy, and they observed that addition of 0.078 wt% Ti increased the hardness and tensile strength by 7.17% and 33.4%, respectively. Kori et al. [198] observed that Al-B refiner showed a better grain refining efficiency to Al-Si alloy than to Al-Ti refiner.

Rodríguez et al. [214] suggested that Al-5Ti-1B refiner had obvious grain refining effectiveness to Al-Si-Fe alloy. Fakhraei et al. [215] modified the Al-20%Mg alloy by the addition 1 wt% Al-5Ti-1B refiner, and they observed that the average grain size was decreased to 80 μm and the original dendrite was modified to star-like grain. The tensile strength of Al-Mg alloy was increased from 168 to 253 MPa, and elongation was increased from 1.2% to 2.4% by the addition Al-5Ti-1B refiner. Alipour et al. showed that the mechanical properties of Al-12Zn-3Mg-2.5Cu alloy were increased by the addition of Al-5Ti-1B refiner, and the grain size of the alloy was obviously decreased from 480 to 40 μm [216]. Ebrahimi et al. [217] added Al-5Ti-1B refiner to Al-Zn-Mg-Cu alloy, and they observed improved mechanical properties. Liu et al. [210] suggested that the Al-5Ti-0.75C grain refiner showed a better grain refining efficiency to Al-5Cu alloy than Al-Ti-B refiner. The average grain size of Al-5Cu alloy was refined to 50 μm by the addition Al-5Ti-0.75C refiner, and the hardness value was increased to 64.6 HRB. Ning et al. [174] proposed that the addition of 0.5 wt% Al-5Ti-0.2C showed the best grain refinement to 7085 alloy. Murty et al. [218] suggested that Al-Ti-C refiner showed a better grain refinement to Al-Si alloy and pure Al than Al-Ti-B refiner. Liu et al. [207] proposed that the main phase in Al-3B-5Sr master alloy was SrB6. The average grain size of A356 alloy was decreased from 1000 to 300 μm after the addition of Al-3B-5Sr refiner, and the tensile strength and elongation of A356 alloy were increased by 26.2% and 4.5%, respectively. Bolzoni et al. [208] proposed that the Al-Nb-B alloy showed the same grain refinement to pure Al with commercial Al-5Ti-1B refiner. Guan et al. [219] found that Al-5Ti-1B refiner prepared by electromagnetic stirring and rheo-extrusion had a better grain refinement of Al than the refiner prepared by conventional process. In addition, Sun et al. proposed that Al-5Ti-1B refiner prepared under ultrasound showed a better grain refinement of Al alloy than of conventional Al-5Ti-1B refiner [205, 206].

The Al-5Ti-0.8B-0.2C refiner developed by Liu et al. showed a better grain refining efficiency than of Al-5Ti-1B refiner [202, 207]. The average grain size of pure Al added with 0.2 wt% Al-5Ti-0.8B-0.2C refiner was decreased from 3500 to 170 μm, and the grain refinement did not fade in 60 min. They suggested that the segregation trend of TiB2 in refiner alloy was weakened by the introduction of C element. Kang et al. [187] showed that the microstructure of Al-Zn-Mg-Cu alloy was refined by the addition Al-5Ti-1B-Re refiner, and the secondary dendrite arm space was also decreased. Nie et al. [188] observed that the average grain size of Al-10Zn-1.9 Mg-1.6Cu-0.12Zr alloy was decreased to 40 μm by the addition of Al-5Ti-1B-Er.

Some solute elements in master alloy have negative effects on the grain refining effectiveness of refiner, for example, Zr, Cr and Mn elements can weaken the grain refining effectiveness of Al-5Ti-1B alloy. Some solute elements V, Cr, etc. can attach on the surface of effective nucleation particles and reduce the nucleation potency of the particles, which induces the fading of grain refinement. The solute element can also replace the primary elements in effective nucleation particles and change the lattice constant, indicating these particles cannot act as nucleate sites anymore. Murty et al. [183] proposed that the effectiveness was faded by some elements, and those elements showed different fading potency of grain refinement. The elements with fading potency from low to high were B < Ti < Cr < Ni < Fe < Mg < Zr. They observed that the addition of both Al-1Ti-3B and Al-Sr showed a good effect on grain refinement and mechanical properties. However, Lozano et al. [220] observed the poisoning of grain refinement when Al-3Ti-1B and Al-10Sr were added in Al-12Si alloy together. Qiu et al. [221] proposed that Sr element in aluminum melt decreased the amount of TiAl3 by reacting with Ti and weakened the grain refining effectiveness of Al-5Ti-1B refiner. Bolzoni et al. [209, 222] proposed that Nb-based refiner showed effective grain refining efficiency to Al-Si alloy and had no fading behavior. Many studies proposed that rare earth element relived the poisoning of grain refinement, but its mechanism is not clear yet [192, 207].

Due to the limitation of in situ observation method, it is difficult to observe the nucleation process in refining. Moreover, the chemical reactions of elements are complex, so the refinement mechanisms of grain refiner are still not clear yet. It is important to clarify the refining mechanism since it is the theoretical basis for the exploitation of new refiner. The grain refining efficiency of refiner is affected by the morphology, size and amount of nucleation particle, so it is also important to clarify this effect. The researches of refiner for aluminum alloy are lack of pertinence, and the exploitation of specialized refiner for a given aluminum alloy is one of the developing directions. There are still many challenges in improving the grain refining efficiency of refiner. The grain refining efficiency is usually unstable in industrial practice. Therefore, the stability of grain refiner needs to be improved. To develop new generation of grain refiners with the features of low energy consumption, low cost and high efficiency is important for solving the present problems.

5 Grain Refinement by Severe Plastic Deformation (SPD)

Plastic deformation in thermo-mechanical processing is an effective way to control grain size. Many kinds of thermo-mechanical treatments have already been applied into the production of aluminum alloys, so the conventional methods will not be described in this review. SPD producing ultra-fine grains in metallic materials by imposing a high plastic strain has been widely studied for decades. Several techniques are now available for producing requisite high strains: equal-channel angular pressing (ECAP) [223], high-pressure torsion (HPT) [224], accumulative roll-bonding (ARB) [225], cyclic extrusion-compression (CEC) [226] and friction stir welding (FSW) [227, 228].

Accumulative continuous extrusion forming (ACEF) [46], multi-directional forging (MDF) [229], repetitive corrugation and straightening (RCS) [43], twist extrusion (TE) [42], tube cyclic extrusion-compression (TCEC) [45] and so on. In MDF method, the specimen is compressed and stretched continuously along with the change in axial load direction during deformation, which leads to severe plastic deformation in metal. In RCS method, a work piece is repetitively bent and straightened without significantly changing the cross-sectional geometry and large plastic strain is achieved through cyclic bending and straightening. In TE method, the specimen passes through a rectangular channel with a rotating section in the middle to achieve larger plastic deformation, and the size and shape of the work piece are not changed after twist extrusion. In TCEC method, the tubular specimen is placed between the chamber and mandrel, and deformation happens when the specimen passes the neck zone. Among these methods, ECAP, HPT, ARB, CEC and FSW are widely studied, and ACEF is relatively a new type of SPD method. ECAP process has the disadvantages of short service life of the die and complex process [230], so wide application of ECAP is limited. Most present SPD methods have shortcomings of low efficient and high cost. Hence, to develop SPD process with high efficiency, low cost and high yield would be of great significance for industrial applications. Different SPD methods are summarized in Table 9.

Table 9

Different SPD methods

5.1 Equal-Channel Angular Pressing (ECAP)

Equal-channel angular pressing (ECAP) was first developed by Segal and co-workers [36]. ECAP device contains two channels which are equal in cross section and intersect at an angle, as shown in Fig. 8. After about 40 years’ development, ECAP has become one of the most widely applied SPD techniques.

Fig. 8

Schematic of equal-channel angular pressing (ECAP) process

Valiev proposed a model for grain refinement [231]. In this model, a very high dislocation density was introduced in the early stage of ECAP, which led to the formation of intragranular structure consisting of cells with thick cell walls and low angles of misorientation. With increasing strain, the thickness of the cell walls decreased by recovery through dislocation slipping. Ultimately, an array of ultra-fine grains separated by high-angle grain boundaries was formed [232]. In Iwahashi’s study, the grain size of pure aluminum was decreased from 0.1 mm initially to 1 μm after one pass of ECAP, but the microstructure consisted of parallel bands of elongated subgrains [233]. Yamashita’s study demonstrated the increase in grain size of pure aluminum with increasing pressing temperature in ECAP, and further grain refinement was achieved through the introduction of Mg in solid solution [234]. The grain size of 1070 alloy after 8 passes of ECAP was measured as 0.3 μm compared with the initial grain size of 50 μm in Tolaminejad’s study [235], which showed that when 8 passes of ECAP were performed the microstructure mainly consisted of high-angle grain boundaries and became more homogeneous. In Xu’s study, a regular array of grains was achieved after 4 passes of ECAP on pure Al. An average size of 1.4 μm was obtained, and there was a significant increase in the numbers of high-angle grain boundaries [236]. Xu et al. [237] applied 8 and 12 passes of ECAP processing on Al-1Mg (wt%) alloy, and the ultra-fine-grained structure with the grain size of ~700 nm was achieved, while the fraction of high-angle grain boundaries also increased slightly.

Based on the traditional ECAP process, several new process methods were developed including but not limited to: rotary-die ECAP [238], equal-channel angular extrusion (ECAE) [239], multi-pass ECAP [240], ECAP-conform [241], ECAE with rotating tooling [242], parallel tubular channel angular pressing (PTCAP) [243] and high-temperature equal-channel angular pressing [244]. These processing methods widen the application areas of ECAP process. However, more homogeneous distribution of microstructure in the processed material is still the primary object for all techniques derived from ECAP process. On the other hand, ECAP process also has disadvantages like short service life of the die and complex processing procedures. How to figure out these problems is the key of development of this method.

5.2 High-Pressure Torsion (HPT)

In high-pressure torsion (HPT), a metal disk is subjected to a high pressure with concurrent torsional straining [38, 245]. Processing by HPT is a continuous operation in which the sample remains within the HPT facility throughout the processing operation, as shown in Fig. 9 [246].

Fig. 9

Schematic of high-pressure torsion (HPT) process

In processing by high-pressure torsion (HPT), it is possible to continuously generate strain in a forward direction in monotonic HPT (m-HPT) or to reverse the direction of straining in cyclic HPT (c-HPT). Kawasaki’s study suggested the tendency to produce an inhomogeneous microstructure near the centers of pure Al disks became more widespread in c-HPT [246]. Zhilyaev’s experimental results exhibited that the grain size of 1070 alloy after HPT process was 1.2-1.5 μm in the center of the disk, while a smaller grain size of 1.0 μm near the periphery was observed [247]. In Sakai’s study, Al-3Mg-0.2Sc (wt%) alloy was processed by HPT to produce a grain size of 0.15 μm. By comparison, processing of this alloy by ECAP gave a grain size of 0.20 μm. The processing of disks by HPT usually leads to a central core region where the hardness is less than the outer region and the microstructure is relatively coarse. The grain size of this core region decreases with increasing numbers of turns in HPT [248]. Ultra-fine-grained hypoeutectic Al-7Si (wt%) alloy with homogenous microstructure was achieved by HPT processing under an applied pressure of 6.0 GPa at room temperature by Wang et al. [249]. The micro-hardness of the HPT processed Al-7Si alloy was significantly improved with the increasing number of HPT turns [249].

Though high-pressure torsion has the outstanding advantages of uniform deformation, low deformation resistance and fewer contamination, it also has limitations, such as limited shape and size of products, impossibility of continuous industrial production and complexity of technical parameters.

5.3 Accumulative Roll-Bonding (ARB)

Accumulative roll-bonding (ARB) process was proposed to develop ultra-fine grains by introducing severe plastic strain in materials. Stacking of materials and conventional roll-bonding are repeated in the process. First, a strip is neatly placed on top of another strip. The interfaces of the two strips are surface treated in advance in order to enhance bonding strength. The two layers of material are jointed together by stacking and rolling. Then, the rolled material is sectioned into two equal halves. The sectioned strips are surface treated, stacked and roll bonded again. The whole process is repeated for several times to obtain ultra-fine microstructure, as shown in Fig. 10 [41]. In Saito’s study, ARB process was successfully applied to 1100 alloy and 5083 alloy, and submicron grains with very high strengths were obtained (304 MPa in 1100 alloy and 551 MPa in 5083 alloy) [41]. Lee’s study showed the ultra-fine grains with clear grain boundaries began to appear at the third cycle ARB process in 6061 alloy. The fraction of ultra-fine grains increased with the number of ARB cycles. The specimen processed by six-cycle ARB was almost composed of ultra-fine grains of 500 nm in average diameters. The further ARB process up to eight cycles made the size and aspect ratio of the ultra-fine grains smaller. The tensile strength of the ARB processed 6061 alloy after eight cycles increased to 363 MPa which was about three times of the initial strength [250]. The ARB process was also successful in grain refinement in 1050 alloy according to Gashti’s study [251]. TEM micrographs revealed that with increasing the number of ARB cycles, the average grain size of samples decreased [251]. Huang’s study showed that the microstructure of the deformed 1100 samples was subdivided by high-angle grain boundaries and low-angle dislocation boundaries [252]. ARB process can prepare ultra-fine-grain materials by using conventional rolling mill, so this process can be performed without special equipments. However, stacking, surface treating and roll-bonding have to be carried out in each cycle, which makes the method complicated and inefficient [252].

Fig. 10

Schematic of accumulative roll-bonding (ARB) process

5.4 Cyclic Extrusion-Compression (CEC)

The CEC method was invented to allow arbitrarily large strain deformation of a sample with the preservation of the original sample shape [37]. In CEC method, a sample is pushed from one cylindrical chamber with a diameter D into the second chamber with the same dimensions. Because a die with a narrower neck is installed between the two chambers, deformation happens when the sample passes through the die (Fig. 11). Severe plastic deformation can be accumulated through reciprocating metal flow from one chamber to the other [37].

Fig. 11

Schematic of cyclic extrusion-compression (CEC) process

In Yeh’s study, both the strength and ductility of Al-12Si (wt%) alloy were improved significantly after six passes of CEC [253]. CEC’s effect on grain refinement of 2024 alloy was revealed by Yuan et al. [254]. It was found that grain size decreased with the increase in CEC passes. The grain sizes of 2.9 and 1.9 μm were obtained after 5 passes and 20 passes CEC, respectively [254]. Richert’s study showed that during cyclic extrusion-compression of aluminum microstructure evolved from a cell block structure at lower strains to an equiaxed cell (subgrain) structure at large strain [255]. This structural transformation was assisted by localized glide and shear. It was believed that shear zone formed in the process of extrusion. The crossing and proliferating of the shear zone led to the fragmentation of microstructure, which gradually evolved into equiaxed cell (subgrain). The primary merit of CEC is that the arbitrarily strain deformation can be obtained by controlling number of CEC passes. Its demerits include limited extrusion ratio, restriction of die’s refining capacity and impossibility of continuous industrial production.

5.5 Friction Stir Welding (FSW)

Friction stir welding (FSW) is a solid-state jointing process [39, 256] It was developed by the Welding Research Institute of England in 1991. High-temperature-resistant material was made into a welding tool with a certain shape, and the welding tool was inserted into the welding seam. Friction heat generated in welding edge due to high-speed rotating of welding tool, and therefore, the treated metals are welded together (Fig. 12).

Fig. 12

Schematic of friction stir welding (FSW) process

An Al-Li alloy was treated in FSW method by Jata et al. to establish the mechanism of microstructural evolution in the dynamical recrystallization (DRX) region of FSW. Grain boundary misorientations generated in the DRX region were observed to be between 15° and 35°. Grains formed in the DRX region by a continuous dynamical recrystallization [257]. Su’s study showed that 7050 alloy exhibited a pancake microstructure with unrecrystallized regions containing subgrains of 1-2 μm after FSW process, and a large amount of dislocations were introduced [258]. Cabibbo’s study showed the microstructural evolution of 6056 plate in FSW process [259]. The microstructure was divided into three major regions: (1) fine equiaxed grains in the nugget at the weld center, (2) highly elongated grains with very small cells and grain size of 12 μm in the broad retreating side thermo-mechanically affected zones (TMAZ) and in the narrow advancing side and (3) slightly elongated coarse grains with grain size of 300 μm in the heat-affected zones (HAZ). In Ma’s study [260], the functions of FSW were: (1) to break up the coarse Si particles and disperse them into the matrix increase; (2) to break up the coarse aluminum dendrites and refine the grain structure; (3) to break up the coarse precipitates and dissolve part or most of them into the matrix; and (4) to eliminate the casting porosity. FSW was successfully applied to A356 alloys, and it brought out a uniform fine-grained structure with grain size of 5-8 μm. Uniform distribution of fine Si particles with an average size of 2-3 μm and an average aspect ratio of ~2.0 was observed [260].

Friction stir welding has the outstanding advantages of good dimensional stability and excellent metallurgical properties in the joint area, low pollution and low energy consumption. On the other side, the challenges still exist in low processing speed, narrow range of processed materials and limited shape of the joint.

5.6 Accumulative Continuous Extrusion Forming (ACEF)

Preparation of ultra-fine-grained material with short process, referred as continuous rheo-extrusion and accumulative continuous extrusion forming (ACEF), was developed on the basis of severe plastic deformation mechanism in continuous expanding rheo-extrusion, as shown in Fig. 13 [261]. In this method, the raw material rod prepared by continuous rheo-extrusion is fed into the entrance of the roll-shoe gap between the rotating extrusion wheel and shoe, and the rod is subjected to the driving force τd from side wall of extrusion wheel and resistance force τr from shoe in the roll-shoe gap. Therefore, the shear flows generate and lead to deformation of the processed rod. Subsequently, shear deformation occurs at the exit of roll-shoe gap like ECAP method. Finally, extrusion deformation happens in the extrusion mold. Three-stage continuous deformations accumulate in one single process, and therefore, efficient grain refining induced by severe plastic deformation is achieved. During ACEF, the grain size of the alloy was dramatically refined through continuous dynamical recrystallization (CDRX) attributed to severe plastic deformation and deformation heat [261]. Figure 14 shows electron backscatter diffraction (EBSD) maps of Al-0.2Sc-0.1Zr alloy before and after ACEF. Figure 14a shows that the grains in raw material rod were coarse and the grain size was significantly refined from 100 to 800 nm through CDRX after ACEF (Fig. 14b).

Fig. 13

Schematic of accumulative continuous extrusion forming (ACEF) process

Fig. 14

EBSD maps of Al-0.2Sc-0.1Zr alloy before and after ACEF: a before ACEF; b after ACEF

The ACEF process has the following characteristics:

1. Short process and large deformation.

The equivalent strain in one pass ACEF is,

$$\varepsilon = 2\left( {\frac{1}{\sqrt 3 }\cot \frac{\Phi }{2} + 2 1 {\text{n}}\frac{{D_{ 1} }}{{D_{ 0} }}} \right) + \alpha,\ \ (8)$$

where Φ is the turning corner angle of metal flow direction at the exit of roll-shoe gap, D0 is the diameter of sizing zone of extrusion, D1 is the diameter of maximum extending section, and α is the equivalent shear strain in the roll-shoe gap, α = 0.4-0.53. In ACEF process, Φ = 90o, D0 = 10 mm, D1 = 25 mm and α = 0.4, so one pass equivalent strain is 5.22. For ECAP, in the case of Φ = 90o and curvature angle φ = 30o, one pass equivalent strain is 1.02. Therefore, one pass equivalent strain in ACEF is equal to ca. 5.12 times of passes equivalent strain in ECAP. Therefore, ACEF is a short large-deformation process.

2. It can prepare ultra-fine-grained materials with various section sizes.

Ultra-fine-grained materials with various section sizes can be prepared by replacing the extrusion mold in the final pass.

3. It can prepare infinite-long products.

The infinite-long products can be fabricated by this method as long as the alloy is supplied continuously.

4. High productivity of ultra-fine grain materials.

6 Conclusions

The research of grain refinement has a long history, and studies on this technology run through the whole metal processing process from liquid to solid state. Until now, more advanced and efficient methods are still being explored. The arguments still exist in the understanding of grain refinement mechanism, and a deeper understanding of aluminum refinement from physical and chemical point of view will surely bring out a new leap in the relevant science.

The four grain refinement methods classified in this paper have respective advantage: VS method is easy to perform; RS is suitable for refining nano-sized material; the addition of GR is the most widely used; SPD method can produce ultra-fine-grained materials during deforming process. For future studies on aluminum grain refinement, the efforts could be made toward the following areas including but not limited to: EMS technology of large-volume melt, higher-magnetic-field stirring technique, multiple-field electromagnetic stirring method, vibration mold design, vibration casting for large-scale melt, high-power ultrasonic vibration equipment, high-temperature resistance ultrasonic horn, efficiency of twin-roll casting, broadening alloy types and product sizes of twin-roll casting, high-efficiency SPD process with low cost, and new generation of grain refiners with the features of low energy consumption, low cost and high efficiency. In sum, environment-friendly, high efficiency, low cost and high performance should be the development goal for all aluminum grain refinement technologies.

Acknowledgements

The authors thank for the supports of the National Natural Science Foundation of China under Grant Nos. 51474063, 51674077 and 51504065 and appreciate the help of the literature surveys by Yang Zhang, Ning Su, Xiang Wang, Diwen Hou, Lianze Ji, Wei Wang and Wei Song.

The authors have declared that no competing interests exist.

 [1] M. Covaciu, A.R. Kennedy, Acta Metall.Sin.(Engl. Lett.) 28(8), 1034(2015) URL Porous Al structures have been made using a porogen leaching method but with additional novel steps. In this way, Al-0.25Cu-0.6Si-1.0Mg alloys with interconnected, open cells have been made, with high levels of porosity(up to 89%). The use of 4 to 5 mm diameter sodium chloride droplets combined with interspersion of the metal powder and droplets in the die and dissolution of the salt before sintering was found to enable easier interspersion of the two components, resulting in good connectivity of the beads via the formation of ‘‘windows' ' between the pores, and rapid elimination of salt from the structure. Porous Al with densities in the range of 0.28–0.86 g/cm3 exhibited compressive yield strengths in the range of 0.14 to 7.5 MPa, with no difference being observed between the samples made by this route and by more conventional dissolution of the salt after sintering. These novel steps widen the possibility for porous metal manufacture to other materials, not limited by the need to have a sintering temperature below that for the melting point of salt. [本文引用: 1] [2] A. Lukaszek-Sołek, Acta Metall.Sin.(Engl. Lett.) 28(1), 22(2015) URL Results of an experimental and modelling study of forming processes in the AA2099 Al–Cu–Li alloy, for a wide range of temperatures, strains and strain rates, are presented. The analyses are based on tensile testing at 20 °C at a strain rate of 0.02 s-1and uniaxial compression testing in the temperature range 400–550 °C at strain rates ranging from0.001 to 100 s-1, for constant values of true strain of 0.5 and 0.9. The stability of plastic deformation and its relationship with a sensitivity of stress to strain rate are considered. The power dissipation efficiency coefficient, g(%), and the flow instability parameter, n B 0, were determined. The complex processing maps for hot working were determined and quantified, including process frames for basic forging processes: conventional forging and for near-superplastic and isothermal conditions. A significant aspect is the convergence of power dissipation when passing through the 500 °C peak.Deformation, temperature and strain-rate-dependent microstructures at 500 °C for strain rates of 0.1, 1, 10 and 100 s-1are described and analysed for the conventional die forging process frame, corresponding to 465–523 °C and strain rates of50–100 s-1. [本文引用: 1] [3] R. Verma, A. Ghosh, S. Kimand, C. Kim, Mater. Sci. Eng. A 191(1), 143 (1995) URL A preliminary investigation of thermomechanical processing of 5083 aluminum plate (Al-4.7%Mg-0.7%Mn) was undertaken to develop a fine-grain sheet for superplastic forming applications. Significant differences in grain size and the extent of superplasticity are seen in hot-rolled vs. cold-rolled sheets, with tensile elongations exceeding 600% for the cold-rolled alloy. Additionally, a separate fine-grain sheet of the same alloy, produced by Alusuisse Co., was studied in greater detail. Superplastic deformation behavior of this sheet was investigated under uniaxial tension over the temperature range of 500-565 [本文引用: 1] [4] The U.S. DRIVE Partnership,Materials Technical Team Roadmap (U.S. Driving Research and Innovation for Vehicle efficiency and Energy sustainability, American, 2013) p 5 [本文引用: 1] [5] The Aluminum Association, Aluminum Industry Technology Roadmap (The Aluminum Association, Washington D.C., 2003) p. 7 [本文引用: 3] [6] The United States Automotive Materials Partnership, Magnesium vision2020: A North American Automotive Strategic Vision For Magnesium (Department of Energy National Energy Technology Laboratory,American, 2004) p. 3 [本文引用: 1] [7] J.X. Lü, W.X. Yang, S.K. Wu, X.D. Zhao, R.S. Xiao, Acta Metall.Sin.(Engl. Lett.) 27(4), 670(2014) [本文引用: 1] [8] R. Haase, R. Müller, D. Landgrebe, P. Scholz, M. Riemer, Acta Metall.Sin.(Engl. Lett.) 28(12), 1518(2015) URL The global trends towards improving fuel efficiency and reducing CO2 emissions are the key drivers for lightweight solutions. In sheet metal processing, this can be achieved by the use of materials wi [本文引用: 1] [9] A. Canakci, T. Varol, Powder Technol. 268, 72(2014) URL Solid state recycling of AA7075 chips by hot pressing is a novel processing technique that uses remarkably lower energies compared to conventional recycling by casting. In this study, AA7075 chips were recycled to fabricate SiC particle (2.5, 5 and 1002wt.%) and Al powder (10, 20, 30 and 5002wt.%) reinforced AA7075 alloy matrix composites by the combination of powder metallurgy and hot pressing techniques. The effects of the amounts of SiC particles and Al powder on the microstructure of the recycling AA7075/Al–SiC composites were investigated. The density, porosity and hardness of the composites were also examined. It was observed that the Al powders in the AA7075/Al–SiC composites were distributed non-homogeneously and randomly between the AA7075 chip regions. The results showed that the green density of the AA7075/Al–SiC composites increased with increasing amounts of Al powder, while SiC showed the opposite effect. The hot pressed density (final density) of the composites decreased with increasing amounts of Al powders and SiC content. An increase in the weight percentage of the SiC particles (from 0 to 1002wt.%) in the AA7075/Al–SiC composites improved the hardness of the recycled composites due to the high hardness of the SiC particles. [本文引用: 1] [10] M. Mukherjee, U. Ramamurty, F. Garcia-Moreno, J. Banhart, Acta Mater. 58(15), 5031(2010) URL The application of different cooling rates as a strategy to enhance the structure of aluminium foams is studied. The potential to influence the level of morphological defects and cell size non-uniformities is investigated. AlSi6Cu4 alloy was foamed through the powder compact route and then solidified, applying three different cooling rates. Foam development was monitored in situ by means of X-ray radioscopy while foaming inside a closed mould. The macro-structure of the foams was analysed in terms of cell size distribution as determined by X-ray tomography. Compression tests were conducted to assess the mechanical performance of the foams and measured properties were correlated with structural features of the foams. Moreover, possible changes in the ductile鈥揵rittle nature of deformation with cooling rate were analysed by studying the initial stages of deformation. We observed improvements in the cell size distributions, reduction in micro-porosity and grain size at higher cooling rates, which in turn led to a notable enhancement in compressive strength. [本文引用: 1] [11] T.P.D. Rajan, B.C. Pai, Acta Metall.Sin.(Engl. Lett.) 27(5), 825(2014) [本文引用: 1] [12] T. Varol, A. Canakci, S. Ozsahin, Acta Metall.Sin.(Engl. Lett.) 28(2), 182(2015) [本文引用: 1] [13] D. Lloyd, Int. Mater. Rev. 39(1), 1(1994) URL AbstractParticle reinforced metal matrix composites are now being produced commerically, and in this paper the current status of these materials is reviewed. The different types of reinforcement being used, together with the alternative processing methods, are discussed. Depending on the initial processing method, different factors have to be taken into consideration to produce a high quality billet. With powder metallurgy processing, the composition of the matrix and the type of reinforcement are independent of one another. However, in molten metal processing they are intimately linked in terms of the different reactivities which occur between reinforcement and matrix in the molten state. The factors controlling the distribution of reinforcement are also dependent on the initial processing method. Secondary fabrication methods, such as extrusion and rolling, are essential in processing composites produced by powder metallurgy, since they are required to consolidate the composite fully. Other methods, suc... [本文引用: 1] [14] A. Chokshi, A. Rosen, J. Karch, H. Gleiter, Scr. Mater. 23(10), 1679 (1989) [本文引用: 1] [15] K.T. Kashyap, T. Chandrashekar, Mater. Sci. 24(4), 345 (2001) [本文引用: 1] [16] D.A. Granger, J. Liu, JOM 35(6), 54 (1983) [本文引用: 1] [17] C.W. Granger, Oxford B Econ. Stat. 48(3), 213(1986) [本文引用: 1] [18] S. Kou, Acta Mater. 88, 366(2015) URL Cracking during solidification is a serious defect in castings and welds. When solidification shrinkage and thermal contraction of the semisolid and its surrounding solid are obstructed, tensile deformation can be induced in the semisolid to cause cracking along grain boundaries that are not fed with sufficient liquid. A criterion for cracking was derived, focusing on events occurring at the grain boundary including separation of grains from each other, lateral growth of grains toward each other, and liquid feeding between grains. An index for the susceptibility of an alloy to cracking during solidification was also proposed, that is, | dT / d ( f S 1/2 )| near ( f S ) 1/2 =1, where T is temperature and f S the fraction solid in the semisolid. The index affects: (a) the lateral growth rate of two neighboring grains toward each other to bond together to resist cracking, and (b) the length of the grain-boundary liquid channel through which feeding has to occur to resist cracking. The index was verified with experimental data in casting and welding of Al alloys. [本文引用: 1] [19] N.S. Reddy, A.P. Rao, M. Chakraborty, B.S. Murty, Mater. Sci. Eng., A 391(1), 131 (2005) URL Neural networks, which are known for mapping non-linear and complex systems, have been used in the present study to model the grain-refinement behavior of Al–7Si alloy. The development of a feed forward neural network (FFNN) model with back-propagation (BP) learning algorithm has been presented for the prediction of the grain size, as a function of Ti and B addition level and holding time during grain refinement of Al–7Si alloy. Comparison of the predicted and experimental results shows that the FFNN model can predict the grain size of Al–7Si alloy with good learning precision and generalization. [本文引用: 1] [20] X.F. Liu, Z.G. Zhang, Z. Gao, X.F. Bian, JOM 52(2), 47 (2000) [本文引用: 1] [21] Z. Yang, C.G. Kang, Z.Q. Hu, Metall. Mater. Trans. A 36(10), 2785 (2005) [本文引用: 1] [22] G. Manogharan, B. Yelamanchi, R. Aman, Z. Mahbooba, JOM 68(3), 842 (2016) URL Over the years, many studies have been conducted to study and analyze the grain structures of metal alloys during additive manufacturing to improve mechanical properties. In particular, columnar grains are observed predominantly during rapid solidification of molten metal. This leads to lower mechanical properties and requires expensive secondary heat-treatment processes. This study is aimed at disrupting the formation of columnar grain growth during rapid solidification using ultrasonic vibration and analyzes the effects on grain structure and mechanical properties. A gas-metal arc welder mounted on a Rep-Rap-based low-cost metal 3 Dimension printer was used to deposit ER70S-6 mild steel layers on a plate. A contact-type ultrasonic transducer with a control system to vary the frequency and power of the vibration was used. The effects of ultrasonic vibration were determined from the statistical analysis of microstructure and micro-indentation techniques on the deposited layer and heat-affected zone. It was found that both frequency and interaction between frequency and power had significant impact on the refinement of average grain size up to 10.64% and increased the number of grains by approximately 41.78%. Analysis of micro-indentation tests showed that there was an increase of approximately 14.30% in micro-hardness due to the applied frequency during rapid solidification. A pole diagram shows that application of vibration causes randomization of grain orientation. Along with the results from this study, further efforts in modeling and experimentation of multi-directional vibrations would lead to a better understanding of disrupting columnar grains in applications that use mechanical vibrations, such as welding, directed energy deposition, brazing, etc. [本文引用: 1] [23] Z.N. Chen, H.J. Kang, G.H. Fan, J.H. Li, Y.P. Lu, J.C. Jie, Y.B. Zhang, T.J. Li, X.G. Jian, T.M. Wang, Acta Mater. 120, 168(2016) URL The enhanced grain refining efficiency of B in treating hypoeutectic Al-Si alloys has been identified for more than 30 years. However, its exact underlying mechanism remains to be explored. Here, we report that AlB 2 could be a highly efficient nucleating substrate for 伪-Al in Al-Si alloys with high Si concentrations. The addition of AlB 2 alone without the presence of Si does not result in a significant grain refinement. The enhanced grain refining efficiency can be mainly attributed to the enhanced heterogeneous nucleation of AlB 2 caused by Si. It is proposed that the creation of a layer of SiB 6 at the interface between AlB 2 and Al may reduce the crystallographic mismatch, which can significantly improve the nucleating potency of AlB 2 . Furthermore, the formation of SiB 6 layer may occur for only a limited span of time in Al-Si melt, and it may further transform to other stable phase (e.g. Si). This investigation reveals the nucleating behavior of AlB 2 influenced by solute Si and thereby sheds some novel insights on the grain refinement mechanism of Al alloy. [本文引用: 1] [24] I. Naglič, A. Smolej, M. Doberšek, P. Mrvar, Mater. Charact. 59(10), 1458(2008) URL This work deals with the influence of TiB particles already present, and added in the form of Al–5Ti–1B grain refiner, on the effectiveness of Al–3Ti–0.15C grain refiner in an Al–Fe alloy. We found that Al–3Ti–0.15C grain refiner is ineffective in the case of the addition of both Al–5Ti–1B and Al–3Ti–0.15C grain refiners to an Al–Fe alloy. Despite the presence of TiB particles in the Al–Fe alloy, Al–3Ti–0.15C grain refiner is, however, effective when added on its own to an Al–Fe alloy. The effectiveness and ineffectiveness of Al–3Ti–0.15C grain refiner with Al–Fe alloys were also confirmed by the presence of TiC particles and TiB particles in the centres of the α-Al grains. Recalescence on solidification measured by in-situ thermal analysis can explain the effectiveness and ineffectiveness of Al–3Ti–0.15C grain refiner in Al–Fe alloys. [本文引用: 1] [25] X.J. Wang, X. Cong, A. Muhammad, S. Hanada, H. Yamagata, W.H. Wang, T. Nonferr, Metal. Soc. 24(7), 2244 (2014) [本文引用: 1] [26] H. Bessemer, UK Patent, No. 11317(1857) [本文引用: 2] [27] K. Kocatepe, C.F. Burdett, J. Mater. Sci. 35(13), 3327(2000) [本文引用: 2] [28] X. Zhang, A. Atrens, Mater. Sci. Eng. A 159(2), 243 (1992) URL ABSTRACT The recalescence characteristics in rapid solidification of copper was studied using a thermokinetic model. The effects of the heat transfer coefficient, the melt thickness and the nucleation temperature were investigated. Results showed that lower nucleation temperature and thinner melt lead to a longer recalescence effect while larger heat transfer coefficient results in a weaker recalescence effect. A dimensionless number was derived to measure the extent of recalescence. [本文引用: 3] [29] G.I. Eskin, Metallurgiya (Moscow, 1965), p. 234 [本文引用: 2] [30] C. Yang, Q.R. Cheng, L.H. Liu, Y.H. Li, Y.Y. Li, Intermetallics 56, 37 (2015) [本文引用: 3] [31] G. Wang, J. Shen, J.F. Sun, Y.J. Huang, J. Zou, Z.P. Lu, Z.H. Stachurski, B.D. Zhou, J. Non-Cryst. Solids 351(3), 209 (2005) URL The superplastic deformation behavior and superplastic forming ability of the Zr 41.25 Ti 13.75 Ni 10 Cu 12.5 Be 22.5 (at.%) bulk metallic glass (BMG) in the supercooled liquid region were investigated. The isothermal tensile results indicate that the BMG exhibits a Newtonian behavior at low strain rates but a non-Newtonian behavior at high-strain rates in the initial deformation stage. The maximum elongation reaches as high as 1624% at 656聽K, and nanocrystallization was found to occur during the deformation process. Based on the analysis on tensile deformation, a gear-like micropart is successfully die-forged via a superplastic forging process, demonstrating that the BMG has excellent workability in the supercooled liquid region. [本文引用: 3] [32] M.H.Z. Ling, N.E. Shao, Light Metals 2, 59 (1989) [本文引用: 2] [33] A. Cibula, J. Inst. Metals 80, 1 (1951) [本文引用: 3] [34] D.J. Wang, Y.J. Huang, J. Shen, Y.Q. Wu, H. Huang, J. Zou, Mater. Sci. Eng.,A 527(21), 5750 (2010) [本文引用: 3] [35] A.R.E. Singer, Metals Mat. 4(6), 246 (1970) [本文引用: 3] [36] V.M. Segal, Mater. Sci. Eng. A 197(2), 157 (1995) URL Abstract It is shown that simple shear can be considered a “near ideal” deformation method for structure and texture formation in metal-working. Equal channel angular extrusion (ECAE) is a special industrial process employed to realize this method. In comparison with traditional metal-working operations, this process has a number of advantages; the most important is that extra-large, strictly uniform and unidirectional deformations can be produced under relatively low pressure and load for massive products. The unusual effects that are observed in both the structure and physical-mechanical properties of various metals and alloys suggest many new applications of the simple shear method in materials synthesis and processing. [本文引用: 2] [37] M. Richert, H.J. McQueen, J. Richert, Can Metall. Quart. 37(5), 449 (1998) [本文引用: 3] [38] A.P. Zhilyaev, T.G. Langdon, Prog. Mater Sci. 53(6), 893(2008) URL High-pressure torsion (HPT) refers to the processing of metals whereby samples are subjected to a compressive force and concurrent torsional straining. Although the fundamental principles of this procedure were first proposed more than 60 years ago, processing by HPT became of major importance only within the last 20 years when it was recognized that this metal forming process provides an opportunity for achieving exceptional grain refinement, often to the nanometer level, and exceptionally high strength. This review summarizes the background and basic principles of processing by HPT and then outlines the most significant recent developments reported for materials processed by HPT. It is demonstrated that HPT processing leads to an excellent value for the strength of the material, reasonable microstructural homogeneity if the processing is continued through a sufficient number of torsional revolutions and there is a potential for achieving a capability for various attractive features including superplastic forming and hydrogen storage. The review also describes very recent developments including the application of HPT processing to bulk and ring samples and the use of HPT for the consolidation of powders. [本文引用: 2] [39] W.M. Thomas, J.C. Needham, M.G. Church, P. Templesmith, C.J. Dawes, GB Patent Application 9125978.8 (December 1991) and US Patent US5460317 (October 1995) [本文引用: 2] [40] C.P. Hong, J.M. Kim, Solid State Phenomena Trans. Tech. Publications 116, 44 (2006) [本文引用: 2] [41] Y. Saito, H. Utsunomiya, N. Tsuji, T. Sakai, Acta Mater. 47(2), 579(1999) URL A novel intense plastic straining process named accumulative roll-bonding (ARB) is proposed. First, a strip is neatly placed on top of another strip. The two layers of material are joined together by rolling like a roll-bonding process. Then, the length of rolled material is sectioned into two halves. The sectioned strips are again stacked and roll-bonded. The whole process is repeated again and again. The process can introduce ultra-high plastic strain without any geometrical change if the reduction in thickness is maintained to 50% every rolling pass. The process has been applied to commercial aluminum (1100), Al鈥揗g alloy (5083) and interstitial free (IF) steel. Well-bonded bulk materials were successfully obtained. After several cycles of ARB, ultra-fine (sub-micron) grain structure with large misorientations, i.e. polycrystal, was formed and the materials were strengthened dramatically. [本文引用: 3] [42] Y. Beygelzimer, V. Varyukhin, S. Synkov, D. Orlov,Mater. Sci.Eng., A 503(1), 14 (2009) URL We present an experimental study of the kinematics of twist extrusion (TE) and show that TE has the following properties: (i) as in equal-channel angular pressing (ECAP), the mode of deformation in twist extrusion is simple shear. Unlike in ECAP, there are two shear planes; one of them is perpendicular and the other is parallel to the specimen axis. (ii) The following processes are present during twist extrusion: vortex-like flow with large strain gradient, stretching and mixing of metal particles. We argue that, due to these properties, TE opens possibilities for investigating and forming new microstructures. It has already been successfully used to obtain ultrafine-grained microstructures with good properties in Al, Cu and Ti alloys. [本文引用: 2] [43] J.Y. Huang, Y.T. Zhu, H. Jiang, T.C. Lowe, Acta Mater. 49(9), 1497(2001) URL The microstructures and dislocation configurations in nanostructured Cu processed by a new technique, repetitive corrugation and straightening (RCS), were studied using transmission electron microcopy (TEM) and high resolution TEM. Most dislocations belong to 60掳 type and tend to pile up along the {111} slip planes. Microstructural features including low-angle grain boundaries (GBs), high-angle GBs, and equilibrium and non-equilibrium GBs and subgrain boundaries were observed. Dislocation structures at an intermediate deformation strain were studied to investigate the microstructural evolutions, which revealed some unique microstructural features such as isolated dislocation cell (IDC), dislocation tangle zones (DTZs), and uncondensed dislocation walls (UDWs). [本文引用: 2] [44] R.G. Guan, F.R. Cao, L.Q. Chen, J.P. Li, C. Wang, J. Mater. Process. Tech. 209(5), 2592(2009) URL A vibrating wavelike sloping plate process (VWSP) was proposed. Heat and solute transformations, nucleation mechanism and grain growth as well as microstructure evolution were investigated. It is shown that the sloping plate can provide strong undercooling, and a large quantity of heterogonous nuclei appear on the sloping plate surface, wavelike flow and vibration can enable heterogonous nucleus to escape off the plate, which lead to nucleus multiplication. Moreover, wavelike flow and vibration improve solute diffusion coefficient and cause uniform solute and temperature field, which lead to eruptive nucleation. Grain growth has two typical ways, direct globular growth and dendritic growth. Under relative uniform temperature and solute fields, some grains can keep stable growth surface, go on growing with the round surface and finally maintain their globular structure. However, there are always some grains that grow along a certain preferred direction, but under wavelike flow and vibration their dendritic arms break and transform into near spherical structure. During the casting process, microstructural evolution from globular/dendritic structure to globular/equiaxed structure and to globular structure was observed. [本文引用: 2] [45] A. Babaei, M.M. Mashhadi, Mater. Charact. 95(3), 118(2014) URL 61Tubes of AA1050 for the first time were successfully SPD processed by TCEC.61The grain size was refined to 550nm after two cycles of TCEC.61Notable increase in the strength and decrease in the elongation were documented.61The microhardness increased to 49Hv from the initial value of 29Hv.61Good homogeneity in the microhardness distribution was recorded. [本文引用: 2] [46] N. Su, R. Guan, X. Wang, Y.X. Wang, W.S. Jiang, H.N. Liu, J. Alloys Compd. 680, 283(2016) URL The grain of an Al 0.2Er alloy (wt%) was refined by accumulative continuous extrusion forming (ACEF). Microstructure of the Al Er alloy subjected to ACEF was investigated through electron backscatter diffraction (EBSD) and transmission electron microscopy (TEM). The results revealed that the formation of ultra-fine grains results from a type of continuous dynamic recrystallization (CDRX) that is induced by large strains. With the accumulation of strain, LAGBs gradually introduced and transformed into HAGBs, which leads to an increase in the average orientation angle. The micro-grain alloy is refined through repeated passes of the continuous extrusion forming process and the average grain size of the Al 0.2Er (wt%) alloy decreased from 26.01渭m to 1.3渭m in four passes. The Al 0.2Er (wt%) alloy had a smaller average grain size than that of pure aluminum. The formation of precipitates in Al Er alloy can stimulate the nucleation of CDRX by increasing the local strain energy and slow the growth of recrystallization grains by limiting grain boundary migration that causing grain refinement. [本文引用: 2] [47] C. Cai, Y. Beygelzimer, L.S. Toth, J.J. Fundenberger, Mater. Lett. 159, 253(2015) URL 61Protrusions occurring during severe plastic deformation of aluminum are characterized.61Strain is estimated in the protrusions analytically and a value of about 5.0 is found.61The same plastic strain is obtained using the disorientation distribution functions.61The protrusions present shear textures in twist extrusion. [本文引用: 1] [48] F. Simbarashe, Q.D. Hu, J.G. Li, Mater. Lett. 65(14), 2248(2011) URL With the purpose of upgrading the purity of a dilute industrial aluminum alloy during Ohno Continuous Casting (OCC) process through the use of forced melt stirring, the critical stirring intensity required under a casting speed of 2mm/min was numerically determined. The determined critical stirring intensity induces convective velocity magnitudes into the melt which are effective in transporting solutes from the solute boundary layer into the bulk melt. It was found that, with adequate melt stirring, production of substantially long and high purity ingots during OCC is possible before their purity diminishes to lower than the initial alloy purity. [本文引用: 1] [49] C.G. Kang, J.W. Bae, B.M. Kim, J. Mater. Process. Tech. 187, 344(2007) [本文引用: 1] [50] L. Zhang, X.P. Dong, J.Q. Li, K. Li, Z.K. Zhang, W.J. Wang, Z.T. Fan, J. Cent. South Univ. Technol. 18(6), 1789(2011) [本文引用: 1] [51] J. Wannasin, R.A. Martinez, M.C. Flemings, Scr. Mater. 55(2), 115(2006) URL A new technique to achieve grain refinement using gas bubbles to agitate a molten metal during solidification is reported. The effects of the processing conditions on the degree and homogeneity of the grain refinement are discussed. Results show that a fully grain-refined structure can be obtained using suitable processing conditions. [本文引用: 1] [52] J.W. Zhao, S.S. Wu, L.Z. Xie, P. An, Y.W. Mao, Trans. Nonferr. Metal. Soc. 18(4), 842(2008) [本文引用: 1] [53] A. Ramirez, M. Qian, B. Davis, T. Wilks, D.H. Scr. Mater. 59(1), 19(2008) URL High-intensity ultrasonic treatment (UT) for grain refinement of magnesium alloys has been investigated using a novel theoretical approach in order to better understand its grain-refining potential and the mechanism of nucleation. The process demonstrated significantly superior grain-refining potency to carbon inoculation for Al-containing magnesium alloys but inferior potency to zirconium for Al-free alloys. Details revealed by applying the theoretical approach to ultrasonic grain refinement provide new clues to understanding the mechanism of grain nucleation by UT. [本文引用: 1] [54] K.A. Jackson, J.D. Hunt, Trans. TMS AIME 236, 1129 (1966) [本文引用: 1] [55] M.C. Flemings, Metall. Trans. A 22(5), 957 (1991) URL No Abstract available for this article. [本文引用: 2] [56] D.H. Kirkwood, 5th International Conference on Semi-Solid Processing of Alloys and Composites, Proceedings (Golden, Colorado, 1998) p. 33 [本文引用: 1] [57] Z. Fan, Int. Mater. Rev. 47(2), 49(2002) [本文引用: 2] [58] J.L. Wang, Y.H. Su, C.Y.A. Tsao, Scr. Mater. 37(12), 2003 (1997) URL Hot compression tests were performed on supersaturated Cu-3.46 wt.% Ti alloy at the temperatures of 873 K and 973 K. Constant true strain rates of 1.4 x 10(-4) s(-1) and 6.9 x 10(-3) s(-1) were used. The flow curves were characterized by a single peak followed by continuous flow softening until the sample fracture. The strain to the flow stress peak was found to depend on both the deformation temperature and the applied strain rate. The flow softening (flow stress decrease) was more pronounced during deformation at 873 K for the used strain rates. Strain hardening and precipitation process were together responsible for the initial hardening of the material until the flow stress maximum has been reached. Moreover, the strain localization in a form of coarse slip and shear bands was intensified with increased strain value. It resulted in additional sites for discontinuous precipitation beside the high angle grain boundaries and might be responsible for the flow stress decrease at larger strains. The microstructure of hot deformed samples did not reveal any evidence for dynamic recrystallization during hot deformation in the presence of precipitates. However, dynamic particles coarsening within shear bands was observed and has been assumed to be responsible for further strain softening of the hot deformed sample. It was also suggested that the flow stress was partly reduced due to dynamic recovery which intensified in the course of discontinuous growth and particles coarsening within shear bands. As expected, the flow stress value was affected by the discontinuous precipitation process more effectively in those samples which were deformed at higher temperatures and low strain rates. Flow localization was significantly reduced during hot deformation of the material containing the structure transformed by discontinuous precipitation. (C) Acta Metallurgica Inc. [本文引用: 1] [59] M.N. Mohammed, M.Z. Omar, M.S. Salleh, K.S. Alhawari, P. Kapranos, Sci World J2013, 16 (2013) [本文引用: 1] [60] A. Vogel, R. Doherty, B. Cantor, Proceedings of International Conference on Solidification (University of Sheffield, England, 1979) p. 518 [本文引用: 1] [61] R. Doherty, H.I. Lee, E. Feest, Mater. Sci. Eng. 65(1), 181(1984) URL The non-dendritic microstructure of stircast metals is discussed in terms of the shape stability produced by a high density of growing solid particles. This high density appears to be due to the fragmentation of the initial dendrites. The model of fragmentation due to the formation and subsequent melting of high energy grain boundaries is discussed and is shown to be compatible with a wide range of microstructural observations. New observations are reported on the formation of large-angle grain boundaries in bent aluminium single crystals, on the build-up (and collapse) of a solid-like shear strength in metallic slurries as the stirring conditions are varied and on the stir casting of eutectic alloys. The behaviour of the metallic slurries is briefly compared with recent observations on the dramatic loss of strength of porous sintered powders when exposed to liquid solutions that wet the grain boundaries. [本文引用: 2] [62] J. Pilling, A. Hellawell, Metall. Mater. Trans. A 27(1), 229 (1996) URL It is generally accepted that liquid agitation during alloy solidification assists in crystal multiplication, as in dendrite fragmentation and the detachment of side arms in the mushy region of a casting. Even without deliberate stirring by electromagnetic or mechanical means, there is often vigorous interdendritic fluid flow promoted by natural thermosolutal convection. Interdendritic fluid flow rates in metals might be as high as 10 mm s{sup {minus}1}. It is the purpose of this article to examine whether such fluid flow can cause mechanical deformation of dendrites, sufficient to cause side arms to bend or break. Metals are so ductile at their melting points that applied forces could only be expected to cause bending, as opposed to fracture, although there are no reports of which the authors are aware of dendritic arms being mechanically bent in this way. The following estimates demonstrate why even bending is not to be expected. In this analysis, the authors shall estimate the stress at the root of a secondary dendrite arm of aluminum arising from the action of a flow of molten metal past the dendrite arm. [本文引用: 2] [63] M.C. Flemings, R.A. Martinez, Sol. St. Phen. 116-117, 1(2006) [本文引用: 1] [64] C.M. Burba, J. Janzen, Thermochim. Acta 615, 81 (2015) [本文引用: 1] [65] A. Hellawell, Proceedings of the 4th International Conference on Semi-Solid Processing of Alloys and Composites (University of Sheffield, England, 1996) p. 60 [本文引用: 1] [66] J. Campbell, Int. Met. Rev. 26(1), 71(1981) [本文引用: 2] [67] R.T. Southin, J. Ins. Met. 94, 401 (1966) [本文引用: 1] [68] W. Khalifa, Y. Tsunekawa, Trans. Nonferr. Metal. Soc. 26(4), 930(2016) URL The objective of the current study was to investigate the use of ultrasonic melt treatment technology in the production of grain-refined billets of the AC7A alloy, which was intended for subsequent use as a feedstock in forming operations. The experiments included the application of ultrasonic vibrations to the molten alloys via direct and indirect techniques. Several process parameters such as pouring temperatures (several temperatures between 740 and 660 °C), and treatment time (from 12 min down to 2 min) were investigated. The experiment included continuous ultrasonic treatment from the liquid to the semisolid states. The results showed that both treatment techniques were viable for producing billets with the desirable microstructural characteristics. The optimum treatment conditions were the short treatment time (2 to 3 min), from about 660 °C down to 615 °C for the indirect treatment technique, and from 660 °C to 635 °C for the direct treatment technique. The resulting microstructures, at three positions along the height of the ingot, were characterized by fine, non-dendritic α(Al) grains in the order of a hundred microns, as compared to few thousands of microns for the conventional cast ingots. The intermetallic particles were also refined in size and modified in morphology by the ultrasonic treatment. The operating mechanisms by which the ultrasonic vibrations altered the ingot microstructures were discussed and analyzed. [本文引用: 1] [69] S. Chakma, V.S. Moholkar, Chin. J. Eng. 2013, 14(2013) [本文引用: 2] [70] D. Aquilano, E. Costa, A. Genovese, F.R. Massaro, M. Rubbo, Prog. Cryst. Growth Charact. 46(1-2), 59(2003) URL Perfect single three dimensional crystals with the shape of bubbles can nucleate and grow encompassing gas filled cavities previously formed on solid substrates in aqueous solutions. Bubbles are double walled, both internal and external faces belonging to the cleavage calcite rhombohedron which separate the growth solution from the enclosed gas cavity. The bubbles thickness amounting to a few nanometers, while the size of the bubble is strictly related to that of the cavity ( of the order of a few micrometers), the surface-to-volume ratio is the highest obtained till now for 3D single crystals. A tentative model of nucleation and growth mechanisms of these peculiar objects is proposed. Growth morphology modifications due to the effect of impurities ( Li + and acetate ions) or to foreign gas bubbling in the growth solution (methane) are also described. A peculiar attention is paid to both mechanical and thermodynamic conditions to be fulfilled in order to nucleate crystals around gas cavities from solution and melt. By analysing growth and patterning of Cacarbonate hollow shells in reverse microemulsions along with the growth from melt of sodium disilicate around gas microcavities, we found the criterion ruling the nucleation and growth of poly- or single crystals around a gas cavity and outlined the major role played by the nature of the substrate. Finally, is proposed a new interpretation of hollow shaped crystals, recently found in methane-derived carbonate rocks (Western Alps-Italy) and in late Cenozoic dolomites of the Bahamas. [本文引用: 2] [71] W. Khalifa, Y. Tsunekawa, M. Okumiya, Transactions of the American Foundry Society (AFS, American, 2010), p. 91 URL Transactions of the American Foundry Society American Foundry Society, 2000- Vol. 108 (2000)- [本文引用: 1] [72] X.P. Ma, Y.J. Li, Y.S. Yang, J. Mater. Res. 24(8), 2670(2009) URL [Ma, Xiaoping; Li, Yingju; Yang, Yuansheng] Chinese Acad Sci, Inst Met Res, Shenyang 110016, Peoples R China.;Yang, YS (reprint author), Chinese Acad Sci, Inst Met Res, Shenyang 110016, Peoples R China;ysyang@imr.ac.cn [本文引用: 1] [73] L.Z. He, X.H. Li, P. Zhu, Y.H. Cao, Y.P. Guo, J.Z. Cui, Mater. Charact. 71, 19(2012) URL The evolutions of coarse constituent phases in homogenized 7085 aluminum alloy at different conditions with or without the application of 1202T high magnetic field were examined by using differential scanning calorimetry, scanning electronic microscope, energy dispersive spectroscopy and X-ray diffraction. It is found that the main constituent phases including quaternary phase T(AlZnMgCu), Al 7 Cu 2 Fe, and AlTiCuFeSi are present in as-cast 7085 alloy. During homogenization, α02+02T eutectics become discontinuous and spheroidized, and Al 2 CuMg phase nucleates and grows along α02+02T eutectics. High magnetic field significantly accelerates the melting of quaternary phase T and Al 2 CuMg phase. When the alloy homogenized at 46002°C/1002h02+0248002°C/802h with 1202T magnetic field, the least amount of constituent phases is obtained. [本文引用: 1] [74] K. Sugiura, K. Iwai, ISIJ. Int. 45(7), 962(2005) URL Refining mechanism of solidified structure in which a static magnetic field and an alternating electrical current are simultaneously imposed on the local area of a metal or an alloy, has been experimentally examined using a Sn-10mass%Pb alloy. The refining period of the structure was examined by changing the imposing period of the electromagnetic vibration on the sample and it was confirmed that the refining period was the initial stage of solidification. The refining region was specified to be around the electrodes by inserting a stainless steel wire net in the sample while it was independent of the inserting position of the electrodes. Convection was induced by the electromagnetic vibration because temperature difference in the sample drastically decreased as soon as the vibration was excited in the sample. From these experimental results, the estimated mechanism in this process is that dendrite tips around the electrodes are cut off by the electromagnetic vibration in the initial stage of the solidification and it spread to the whole area of the sample by the convection induced by the electromagnetic vibration. Furthermore, nucleation is induced by an intense electromagnetic vibration. [本文引用: 1] [75] S. Ji, Z. Fan, M.J. Bevis, Mater. Sci. Eng., A 299(1), 210 (2001) URL Based on the extensive experience in injection moulding of polymeric materials, a twin-screw rheomoulding process has been developed in our laboratory for near net-shape production of engineering components. The rheomoulding equipment consists of a liquid metal feeder, a twin-screw extruder with closely intermeshing, self-wiping and co-rotating screws, a shot assembly and a central control unit. The fluid flow in the twin-screw rheomoulding process is characterised by high shear rate and high intensity of turbulence. The experimental results of rheomoulded Sn–15wt.% Pb and Mg–30wt.% Zn alloys have demonstrated that the twin-screw rheomoulding process is capable of producing small and near mono-sized solid particles distributed uniformly in a fine-grained eutectic matrix. Compared with other existing semi-solid metal processing techniques, the twin-screw rheomoulding process has the following advantages: small and spherical solid particles of near mono-size, chemical and microstructural uniformity throughout the component, accurate control over a large range of solid volume fractions, lower overall component cost due to low cost of feedstock materials, and shorter cycle time. [本文引用: 1] [76] X.L. Liao, Q.J. Zhai, C.J. Song, W.J. Chen, Y.Y. Gong, Mater. Sci.Eng., 466(1),56 (2007) [本文引用: 1] [77] A. Hellawell, S. Liu, S.Z. Lu, JOM 49(3), 18 (1997) [本文引用: 1] [78] J.D. Mcneill, US patent US1419280 (13 June 1922) [本文引用: 1] [79] S. Kunstreich, Rev. Metall. 100(4), 395(2003) URL ABSTRACT Electromagnetic stirring (EMS), also called induction stirring, uses rotating or travelling magnetic fields to generate a motion of the liquid steel or to control the liquid steel flow with the purpose of obtaining metallurgical improvements. Part 1 of this article has been published in No. 4, April 2003, of the Revue de Metallurgie-CIT and presents a historical review of EMS in billet, bloom and slab casters as well as the basics of electromagnetism and fluid flow that are relevant for the understanding EMS. Part 1 contains the following chapters with 17 figures : history (billets and blooms, slabs), electromagnetism (principle, Faraday and Lorentz, choice of frequency and performance, efficiency, electrical characteristics), fluid flow (flow velocity, linear or rotational flow, rotational EMS for billets and blooms, linear EMS for slabs, magnetohydrodynamics). The present part 2 presents the metallurgical principles and the industrial applications and results. [本文引用: 1] [80] K. Fujisaki, IEEE Trans. Ind. Appl. 37(4), 1098(2001) URL To get the optimal operation of in-mold electromagnetic stirring through the fundamental characteristics, a three-dimensional magnetohydrodynamic calculation model was used, taking into consideration heat transfer and solidification as well as free surface. Comparison between measured velocity and calculated velocity is in agreement. A shadow method that calculates electromagnetic force in consideration of free surface transition is proposed to reduce computing time and maintain sufficient accuracy. A calculated free surface of mercury under a 200-Hz magnetic field shows an adequate agreement with the experimental one. Calculation results applied to the continuous casting process taking into consideration heat transfer and solidification are in good agreement with operation data. This model shows that electromagnetic stirring makes the solidified shell uniform and dynamic deviation of temperature stable [本文引用: 1] [81] A.A. Tzavaras, H.D. Brody, J. Met. 36(3), 31 (2013) [本文引用: 1] [82] B. Mao, H. Xiao, B. Yi, Cont. Cast 1(1), 1 (2015) [本文引用: 1] [83] L. Zheng, Q. Chen, H. Luo, J. Liu, Z. Zhou, J. Chin. Soc. Rare Earths 32, 61 (2014) [本文引用: 1] [84] B.Y. Ban, Y.L. Li, Q.X. Zuo, T.T. Zhang, J. Chen, S.Y. Dai, J. Mater. Process. Tech. 222, 142(2015) URL Aiming at developing a Si solidification purification process in the Al–Si alloy system, series of experiments on the effect of electromagnetic stirring on refining of metallurgical grade Si(MG–Si) were carried out. By the use of electromagnetic stirring, an increase of size and average mass of primary Si flakes with the increasing voltage of the electromagnet were obtained. Impurity elements formed intermetallic compounds with Al and Si, such as Al 7 Fe 2 Si(α phase) and Al 5 FeSi(β phase). An intermetallic compound type transformation with electromagnetic stirring was observed. The refined Si under the alternative electromagnetic field contains much lower B and P than those without the electromagnetic stirring. An apparent segregation coefficient was introduced to describe the segregation between solid Si and Al–Si melt. The apparent segregation coefficients of B and P are determined to be 0.37 and 0.17, respectively, when the voltage of the electromagnet is 21002V and cooling rate is 0.502K02min 611 . [本文引用: 1] [85] C.K. Jin, A. Bolouri, C.G. Kang, Metall. Mater. Trans. B 45(3), 1068 (2013) [本文引用: 1] [86] Y.B. Zuo, Z.H. Zhao, J.Z. Cui, Adv. Mater. Res. 402, 850(2011) URL Low frequency electromagnetic casting (LFEC) process was used to make 7050 aluminum alloy 飦162mm ingots and study its effect on the as-cast microstructure. Effects of electromagnetic field parameters such as frequency and current intensity on microstructures were systemically investigated. The results showed that LFEC has a significant grain refining effect on 7050 alloy. The microstructures of LFEC ingot from the border to the center of the cross section are all equiaxed or nearly equiaxed grains which are much finer and more uniform than those of DC cast ingot. It was also found that electromagnetic field frequency and current intensity play important roles on the microstructure refinement. The discussion was mainly focused on the mechanism of grain refinement by LFEC process. [本文引用: 1] [87] Z. Zhao, Z.T. Fan, X.P. Dong, B. Tang, D. Pan, J.Q. Li, China Foundry 7(1), 24 (2010) [本文引用: 1] [88] C. Limmaneevichitr, S. Pongananpanya, J. Kajornchaiyakul, Mater. Des. 30(9), 3925(2009) URL This study investigated the effects of mechanical vibration during solidification on the metallurgical structure of hypoeutectic aluminum鈥搒ilicon A356. A series of casting trials were conducted. Emphasis was placed on the morphological changes of the primary aluminum phase of the as-cast alloy, which was subjected to different levels of mechanical vibration at various values of pouring temperature and solid fraction. It was found that the average grain size of the primary phase became relatively finer and more globular as the degree of vibration increased. This suggested that during the solidification process, dendrites that formed normally in the liquid alloy were subsequently disturbed and fragmented by the mechanical vibration introduced into the melt. This effect was enhanced when the vibration was introduced into an alloy with a larger solid fraction, as was observed with solidification at lower pouring temperatures. In addition to the macrostructure examination, semi-solid properties were also assessed and reported using the Rheocasting Quality Index. It was shown that the introduction of mechanical vibration into the A356 melt with adequate solid fraction prior to complete solidification successfully resulted in an as-cast structure featuring semi-solid morphology. [本文引用: 1] [89] F. Taghavi, H. Saghafian, Y.H.K. Mater. Des. 30(5), 1604(2009) URL In the present paper, the effect of vibration frequency and time on the size of α-Al primary solid phase and density on an A356 aluminum alloy was studied. The vibration frequencies and times used were 10, 20, 30, 40 and 50 Hz and 5, 10 and 15 min, respectively. It was observed that the increase in vibration time up to 15 min and vibration frequency up to 50 Hz led to decrease in the size of α-Al phase. The results show that maximum grain refinement was occurred at 50 Hz and 15 min vibration conditions. At these conditions, the size of α-Al phase decreased down to 173 μm and 53% grain refinement was achieved. However, it was observed that the increase in frequency up to 50 Hz and time up to 15 min led to increase in the density of A356 aluminum alloy. The best density was obtained at 50 Hz and 15 min resulting in density of 2.68 g/cm 3. [本文引用: 2] [90] H.M. Guo, A.S. Zhang, X.J. Yang, M.M. Yan, Trans. Nonferr. Metal. Soc. 24(8), 2489(2014) [本文引用: 1] [91] M.G. Mousavi, The2012 International Conference on Advanced Material and Manufacturing Science (ICAMMS, Beijing, China, 2012) p.1306 [本文引用: 1] [92] W.M. Jiang, Z.T. Fan, D.F. Liao, D.J. Liu, Z. Zhao, X.P. Dong, Mater. Des. 32(2), 926(2011) URL In this study, microstructures and mechanical properties of A356 aluminum alloy obtained by expendable pattern shell casting process with vacuum and low pressure (EPSC-VL) were investigated. Meanwhile, microstructures and mechanical properties of A356 alloy among EPSC-VL, expendable pattern shell casting process under gravity casting (EPSC-G), lost foam casting with vacuum and low pressure (LFC-VL) and lost foam casting under gravity casting (LFC-G) were compared. The results showed that the microstructure of A356 alloy fabricated by EPSC-VL was finer and denser than the products from EPSC-G, LFC-G and LFC-VL processes, and its grain size was only 147.202μm compared to LFC-G (327.102μm). Moreover, its porosity defects were greatly reduced (0.16%) compared to LFC-G (1.97%). The tensile strength, elongation and hardness of castings produced by EPSC-VL were all considerably higher than the products from EPSC-G, LFC-G and LFC-VL. The A356 alloy made by EPSC-VL exhibited morphology of dimple fracture that was very deep and well-distributed. In addition, the castings produced by EPSC-VL process had better surface quality compared to LFC castings. [本文引用: 1] [93] F. Taghavi, H. Saghafian, Y.H.K. Mater. Des. 30(1), 115(2009) URL In the present paper, efficiency of mechanical vibration for producing semisolid slurry with thixotropic microstructure in an A356 aluminum alloy was studied. The effects of vibration frequency and time were investigated on the size and morphology of α-Al phase and grain refinement degree. It was observed that thixotropic microstructure could be obtained in the Al alloy by applying mechanical vibration. It was also found that size and morphology of α-Al phase were strongly affected by vibration frequency and time. [本文引用: 1] [94] G. Wang, M.S. Dargusch, M. Qian, D.G. Eskin, D.H. St. J. Cryst. Growth 408, 119 (2014) [本文引用: 1] [95] R. Haghayeghi, A. Heydari, R. Kapranos, Mater. Lett. 153, 175(2015) URL The effects of introducing ultrasonic vibrations prior to high pressure die casting have been experimentally investigated. Results suggest reductions of up to 73% in grain size and up to 5% on porosity levels due to cavitation, leading to enhanced structural performance. The typical values of ultimate tensile stress and yield stress increased respectively to 59002MPa and 50202MPa and elongation improved to 18%. Wetting of the oxides by the collapse of cavitation bubbles along with enhanced degassing appears to be the responsible mechanisms for these property improvements. The effect of relaxation time after ultrasonic vibrations examined and results indicate slight grain size increase from 68 to 800208m. Proving ultrasonic treatment above liquidus appears to be a stable phenomenon where its effects remaining active for a considerable time, making the process feasible for industrial scale applications. [本文引用: 1] [96] G.I. Eskin, Ultrason. Sonochem. 2(2), S137(1995) URL Abstract The real melt always contains non-wettable fine inclusions which are potential nuclei for cavitation and degassing. This paper deals with the nature of ultrasonic degassing and the industrial application of a relevant technology. [本文引用: 1] [97] H. Puga, S. Costa, J. Barbosa, S. Ribeiro, M. Prokic, J. Mater. Process. Tech. 211(11), 1729(2011) URL A novel MMM (Multi-frequency, Multimode, Modulated) ultrasonic (US) technology was used to refine the as cast microstructure and improve the mechanical properties of a AlSi9Cu3 alloy. Ultrasonic vibration was isothermally applied to the melt for 120 s at different temperatures slightly above the liquidus temperature of the alloy, using different electric power values, before pouring into a metallic mould. The microstructure of the cast samples was characterized by optical and scanning electron microscopy and energy dispersive spectrometry. Ultrasonic vibration promoted the formation of small 伪-Al globular grains, changed the size and morphology of intermetallic compounds and distributed them uniformly throughout the castings. Ultimate tensile strength and strain were increased to 332 MPa and 2.9%, respectively, which are 50% and 480% higher than the values obtained for castings produced without vibration. The microstructure morphology and the alloy mechanical properties were found to depend on the electric power and the melt temperature, and by using a suitable combination of these parameters it is possible to achieve high refinement efficiency by treating the melts in the liquid state. [本文引用: 1] [98] L.Y. Chen, J.Q. Xu, H. Choi, M. Pozuelo, X.L. Ma, S. Bhowmick, J.M. Yang, S. Mathaudhu, X.C. Li, Nature 528(7583), 539 (2015) [本文引用: 1] [99] G. Harvey, A. Gachagan, T. Mutasa, IEEE Trans. Ultrason. Ferroelectr. 61(3), 481(2014) PMID:24569252 URL Applications involving high-power ultrasound are expanding rapidly as ultrasonic intensification opportunities are identified in new fields. This is facilitated through new technological developments and an evolution of current systems to tackle challenging problems. It is therefore important to continually update both the scientific and commercial communities on current system performance and limitations. To achieve this objective, this paper addresses two key aspects of high-power ultrasonic systems. In the first part, the review of high-power applications focuses on industrial applications and documents the developing technology from its early cleaning applications through to the advanced sonochemistry, cutting, and water treatment applications used today. The second part provides a comprehensive overview of measurement techniques used in conjunction with high-power ultrasonic systems. This is an important and evolving field which enables design and process engineers to optimize the behavior and/or operation of key metrics of system performance, such as field distribution or cavitation intensity. [本文引用: 1] [100] W.J. Yao, N. Wang, B. Wei, Mater. Sci.Eng., 344(1),10 (2003) URL Droplets of Co–62.1%Si and Co–43.5%Si eutectic alloys with different sizes are rapidly solidified during containerless processing in drop tube. The microstructures of Co–62.1%Si eutectic alloy are mainly characterized by lamellar eutectic plus anomalous eutectic of CoSi 2 and Si phases, whereas those of Co–43.5%Si eutectic alloy are mainly characterized by lamellar eutectic plus anomalous eutectic of CoSi and CoSi 2 phases. For both alloys, the experimental results show that the microstructural evolution depends mainly on undercooling. In the case of Co–62.1%Si eutectic alloy, the smaller the droplet diameter, the larger the volume fraction of anomalous eutectic. When its diameter is very small, the droplet exhibits anomalous eutectic morphology entirely. As for Co–43.5%Si eutectic alloy, with the decrease of droplet size, the microstructural transition proceeds from lamellar eutectic to anomalous eutectic, even to dendrites. The nucleation rates of each phase have been calculated. The TMK eutectic growth and LKT/BCT dendritic growth theories are applied to analyze the rapid solidification process and investigate the microstructural transition mechanisms. The coupled zone around Co–43.5%Si eutectic alloy has also been calculated on the basis of TMK and LKT/BCT models, which covers a composition ranging from 40.8 to 43.8%Si. And calculated coupled zone of Co–62.1%Si covers a composition ranging from 60.2 to 81.6%Si. [本文引用: 1] [101] W.J. Yao, X.J. Han, B. Wei, J. Alloys Compd. 348(1), 88(2003) URL Ni–45%Mo hypoeutectic, Ni–47.7%Mo eutectic and Ni–50%Mo hypereutectic alloys are rapidly solidified during containerless processing in drop tube. The microstructures of Ni–47.7%Mo eutectic alloy are composed of lamellar eutectic plus anomalous eutectic of Ni and NiMo phases. When the droplet size decreases, the volume fraction of anomalous eutectic becomes larger. The structural morphology transforms into Ni dendrite plus lamellar eutectic in very small droplets which are highly undercooled. The microstructures of Ni–45%Mo hypoeutectic alloy are characterized by primary Ni dendrite plus lamellar eutectic, whereas those of Ni–50%Mo hypereutectic alloy consist of NiMo dendrite plus lamellar eutectic. For both off-eutectic alloys, the experimental results show that the microstructure evolution depends mainly on droplet size. In the case of Ni–45%Mo hypoeutectic alloy, with the decrease of droplet size, the primary Ni phase transforms from dendrites to equiaxed grains. As for Ni–50%Mo hypereutectic alloy, when droplets become smaller and smaller the microstructural transition proceeds from primary NiMo dendrite plus lamellar eutectic to anomalous eutectic. The calculated highest undercoolings of the three alloys are 226, 182 and 135 K, respectively. By classical nucleation theory, Ni phase is the primary phase to nucleate for Ni–47.7%Mo eutectic alloy. The TMK eutectic growth and LKT/BCT dendritic growth theories are applied to analyze the rapid solidification process and investigate the microstructural transition mechanisms. The coupled zone of Ni–Mo eutectic alloy has also been calculated on the basis of TMK and LKT/BCT models, which covers a composition range from 45.7% to 57.1% Mo. [本文引用: 1] [102] J. Cai, G.C. Ma, Z. Liu, H.F. Zhang, Z.Q. Hu, J. Alloys Compd. 422(1), 92(2006) URL This paper presents differences of microstructure between conventionally casting and rapidly solidified AZ91HP alloy. The experimental results showed that rapid solidification changed the morphology of microstructure. The grain size was refined and homogeneously distributed β-Mg 17Al 12 phase was obtained. High cooling rate increased nucleation sites of α-Mg, refined microstructure and decreased the proportion of brittle β-Mg 17Al 12 in the eutectic (α-Mg and β-Mg 17Al 12). The decrease of microporosity should be attributed to the grain refinement and application of copper mould. [本文引用: 1] [103] J. He, H.Q. Li, B.J. Yang, J.Z. Zhao, H.F. Zhang, Z.Q. Hu, J. Alloys Compd. 489(2), 535(2010) URL The solidification process of the immiscible alloys exhibit a unique opportunity in designing the composites with the spherical crystalline particles dispersed in the amorphous metal matrix. The typical Al–Pb immiscible alloy and the additional elements Ni, Y and Co were selected, and the Al 82.87 Pb 2.5 Ni 4.88 Y 7.8 Co 1.95 multicomponent immiscible alloy has been designed. The ribbon samples of the multicomponent alloy were prepared by using the melt spinning technique. The ribbons were characterized by the scanning electron microscopy (SEM). The phase constitution and transformation were studied by the X-ray diffraction (XRD) and the differential scanning calorimeter (DSC). It was revealed in the as-quenched ribbons the Al-based metallic glass matrix is embedded by the spherical crystalline Pb-rich particles. The microstructure evolution, the glass formation and the thermal stability of the as-prepared composite have been discussed in detail. A method has been developed based on the mechanism of the liquid–liquid phase transformation in the miscibility gap of the multicomponent immiscible alloy to produce the spherical crystalline particles in the amorphous matrix. [本文引用: 1] [104] P. Duwez, R.H. Willens, J.W. Klement, J. Appl. Phys. 31(6), 1136(1960) [本文引用: 1] [105] S.L. Sobolev, Acta Mater. 60(6), 2711(2012) URL A local non-equilibrium diffusion model (LNDM) for rapid solidification of binary alloys has been briefly reviewed and used to modify a number of solute trapping models with different solid–liquid interface kinetics. The LNDM takes into account deviation from local equilibrium of a solute diffusion field in bulk liquid on the basis that the exact solutions to hyperbolic diffusion equations govern the solute concentration and solute flux in bulk liquid under local non-equilibrium conditions. The LNDM leads to a velocity-dependent effective diffusion coefficient in bulk liquid ahead of the solid–liquid interface D b LNDM ( V ) mathContainer Loading Mathjax , which goes to zero when the interface velocity V 02→02 V Db , where V Db 02 is the bulk liquid diffusion speed. The results show an abrupt transition from diffusion-limited to purely thermally controlled solidification, with the diffusion coefficient in bulk liquid D b LNDM ( V ) mathContainer Loading Mathjax 02=020 and complete solute trapping K LNDM ( V )02=021 at a finite interface velocity V 02=02 V Db for any type of solid–liquid interface kinetics. The bulk liquid diffusion speed V Db is a critical parameter for the transition. The velocity dependence of partition coefficients K LNDM ( V ) has been calculated for different types of solid–liquid interface kinetics, with allowance for local non-equilibrium diffusion effects. The calculation shows that the local non-equilibrium partition coefficients K LNDM ( V ) reduce to the standard K ( V ) at low interface velocity ( V 026102 V Db ) and differ substantially at high interface velocity ( V 026502 V Db ). [本文引用: 1] [106] J.W. Fu, Y.S. Yang, J. Cryst. Growth 322(1), 84 (2011) [本文引用: 1] [107] T.K. Han, S.J. Kim, Y.S. Yang, A. Inoue, Y.H. Kim, I.B. Kim, Met. Mater. Int. 7, 91(2001) URL The influence of the addition of Ag to the crystallization and mechanical properties of the amorphous Zr 65 Al 7.5 Ni 10 Cu 17.5-x Ag x (x=5, 10 at.%) alloys was investigated. It was found that crystallization takes place through the two-stage process of Am→Am'+icosahedral→phase Zr 2 Cu+Zr 2 Ni+Zr 2 Al 3 . The icosahedral particles have a spherical morphology and their sizes range from 10 to 100nm. With the increase in the amount of Ag, thermal instability increases through the simultaneous decrease of T x and Δ T x . The σ f increases significantly from 1560 MPa at 0% V f to 2120 MPa at 62% V f for the 10% Ag alloy. [本文引用: 1] [108] D. Ma, W.Q. Jie, J. Cryst. Growth 210(4), 777 (2000) [本文引用: 1] [109] J.F. Li, W.Q. Jie, G.C. Yang, Y.H. Zhou, Acta Mater. 50(7), 1797(2002) URL The Fe alloy melts containing 7.5, 15, 22.5 and 30 at% Ni were bulk undercooled to investigate the structure evolution. When the undercooling of the four melts is lower than the critical value 110, 125, 175 and 325 K, respectively, only the stable face-centered cubic phase crystallizes. In this case a grain refinement caused by solid superheating is observed in all the alloys, but another grain refinement induced by recrystallization can merely occur in the Fe–30 at%Ni alloy undercooled by 190–325 K. Alternate crystallization of the metastable body-centered cubic phase occurs above the critical undercooling. It is indicated that the subsequent heterogeneous nucleation of the stable phase in the metastable solid and remaining liquid coexisting system is influenced not only by the morphology and surface area of the metastable solid, but also by the effective undercooling of the remaining liquid. On the basis of the experimental results and the theoretical analyses, a structure evolution map for bulk Fe–Ni system is constructed. [本文引用: 1] [110] A. Inoue, Meter. Trans. JIM 36(7), 866 (1995) [本文引用: 1] [111] Q.G. Meng, J.K. Zhou, H.X. Zheng, J.G. Li, Scr. Mater. 54(5), 777(2006) URL The kinetic viscosity ( η ) of superheated melts, thermal properties ( T x ,02 T m ,02 T L ) and X-ray diffraction analysis on the Pr-based bulk metallic glasses (BMG) are reported and discussed. A new refined concept, the superheated fragility defined as M ′ 02=02 E S δ x / k B , has been developed based on common solidification theory and the Arrhenius equation. The interrelationship between this kind of fragility and the glass-forming ability (GFA) is elaborated on and evaluated in Pr-based BMG and Al-based amorphous ribbon alloys. Using viscosity data of superheated melts, it is shown, theoretically and experimentally, that the fragility parameter M ′ may be used as a GFA indicator for metallic alloys. [本文引用: 1] [112] Q.G. Meng, S.G. Zhang, M.X. Xia, J.G. Li, X.F. Bian, J. Alloys Compd. 438(1), 77(2007) URL A series of Pr–Ni–Al bulk metallic glasses (BMGs) was prepared by a copper mold casting method. The effect of the alloy composition on the glass transition temperatures ( T g) and supercooled liquid regions (Δ T x) was studied. It showed that the increase in Pr content to a certain degree, though at which the alloy remained fully amorphous, caused the disappearance of T g and Δ T x. Continuous and isothermal annealing methods were used to investigate the disappearance mechanism of T g and Δ T x on the representative Pr 55Ni 25Al 20 and Pr 65Ni 25Al 10 BMGs. X-ray diffraction (XRD) analysis on crystallization products showed that the disappearance of T g and Δ T x for Pr 65Ni 25Al 10 was not due to the precipitation of different kinds of crystalline products but due to the decrease in glass stability (GS) resulted from the larger quantities of the first precipitating Pr 7Ni 3 phase at relative lower temperatures. As a whole, the alloy composition dependence of T g and Δ T x was also discussed in the framework of the three empirical rules for achievement of high glass-forming ability glasses. [本文引用: 1] [113] W. Yang, F. Liu, H.F. Wang, Z. Chen, G.C. Yang, Y.H. Zhou, J. Alloys Compd. 470(1), L13(2009) URL Applying melt-fluxing, how to predict the maximal recalescence temperature T R upon rapid solidification of bulk undercooled Cu 70 Ni 30 alloy has been investigated. Reproducible experimental results show that T R decreases with the increase of the initial melting undercooling 螖 T , and moreover, its value is much lower than that predicted by Level rule and Scheil equation, let alone the equilibrium liquidus temperature. Quantitative thermodynamic calculation, in combination with an analysis of solute distribution, demonstrates that, T R is affected by the difference of Gibbs free energy between the residual liquid and solid after recalescence. Subjected to sufficiently high 螖 T , T R should be nearly equal to the thermodynamic T 0 point corresponding to the concentration of the residual melt after recalescence. [本文引用: 1] [114] Y.Z. Chen, F. Liu, G.C. Yang, X.Q. Xu, Y.H. Zhou, J. Alloys Compd. 427(1), L1(2007) URL Adopting cyclic superheating and DSC analysis, solidification behavior of undercooled Cu 70Fe 30 alloy was investigated. The metastable phase separation is observed if the initial melt undercooling (Δ T) is above 56 K. The primary solidification of Fe-rich phase in the phase-separated melt is due to its high nucleation driving force i.e. a high Δ T. The coarsening of Fe-rich droplets is controlled by Δ T and can be ascribed to the Ostwald ripening and coagulation effect in the melt. [本文引用: 1] [115] J. Bai, J.S. Li, J. Wang, J. Cui, L.Y. Li, H.C. Kou, P.K. Liaw, J. Alloys Compd. 639, 131(2015) URL The deformation behavior of in-situ Ti 48 Zr 20 Nb 12 Cu 5 Be 15 bulk-metallic-glass (BMG) composites was investigated upon dynamic deformation. The present BMG composite exhibits good dynamic mechanical properties, combining high fracture strength (185002MPa) with remarkable plasticity (>10%) at the strain rate of 1.302×0210 3 02s 611 . Ductile to brittle transition occurs with the increase of strain rates, which can be ascribed to the deteriorated ability of dendrites to impede the propagation of shear bands at higher strain rates. An obvious positive to negative transition on strain rate dependence of flow stresses can be observed with the increasing strain rates. Detailed analysis reveals that the variation from the dendrite-dominated mechanism associated with dislocation movement to the matrix-dominated fracture related to thermal softening is responsible for the present transition. The constitutive equations based on their deformation mechanisms are established for describing the present transition. [本文引用: 1] [116] J.N. Mei, J.L. Soubeyroux, J.J. Blandin, J.S. Li, H.C. Kou, H.Z. Fu, L. Zhou, Intermetallics 19(1), 48 (2011) [本文引用: 1] [117] Q.J. Zhai, Y.L. Gao, W.B. Guan, K.D. Xu, Mater. Sci. Eng. A 441(1), 278 (2006) [本文引用: 1] [118] B. Yang, Y.L. Gao, C.D. Zou, Q.J. Zhai, E. Zhuravlev, C. Schick, Mater. Lett. 63(28), 2476(2009) URL The heterogeneous nucleation kinetics of a single undercooled tin droplet was measured at a cooling rate of 102×0210 4 02K/s by fast scanning non-adiabatic calorimetry. The nucleation rates were obtained by a statistical analysis of nearly 1000 identical events. Applying classical heterogeneous theory in conjunction with the available thermodynamic data, it was found that a single mechanism spherical cap model was capable of describing the experimental data. And the present study shows that fast scanning calorimetry allows studying nucleation in micron sized droplets and may open a way for detailed investigations of the influence of droplet size to discriminate surface and bulk nucleation. [本文引用: 1] [119] Q. Wang, R.P. Liu, Y.Q. Qian, D.C. Lou, Z.B. Su, M.Z. Ma, D.M. Herlach, Scr. Mater. 54(1), 37(2006) URL Droplets of pure silicon were dropped onto a chill plate of copper within an electromagnetic levitation facility. Through observation of morphologies grown on the quenched surfaces, direct evidence of growth mode transition from a continuous form to a lateral form was provided. [本文引用: 1] [120] G. Li, Y.P. Gao, Y.N. Sun, Z.H. Chi, R.P. Liu, Chin. Phys. B 17(9), 3412 (2008) [本文引用: 1] [121] Q.Y. Pan, W.D. Huang, X. Lin, Y.H. Zhou, J. Cryst. Growth 181(1), 109 (1997) [本文引用: 1] [122] Q.Y. Pan, X. Lin, W.D. Huang, Y.H. Zhou, G.L. Zhang, Mater. Res. Bull. 33(11), 1621(1998) URL Laser rapid solidification experiments were performed on a Cu-Mn alloy to investigate its microstructure evolution as a function of growth rate. By taking transverse and longitudinal sections of the laser traces, the resulting morphology and the corresponding growth rate were quantitatively measured by scanning electron microscopy (SEM) and image analysis. As the growth rate increased, the morphological transition from dendrite --> fine cellular crystal --> banded structure --> fully cell-free microstructure, involving planar front growth, was clearly observed in the laser molten pool. By determining these transition conditions and comparing them with the predictions of theoretical models, we found that the origin of the banding can be well elucidated by the current Carrard-Gremaud-Zimmermann-Kurz (CGZK) phenomenological model, while the critical condition of planar front growth was in reasonable agreement with Mullins-Sekerka (M-S) stability theory. (C) 1998 Elsevier Science Ltd. [本文引用: 1] [123] Y.L. Ren, R.L. Zhu, J. Sun, J.H. You, K.Q. Qiu, J. Alloys Compd. 493(1), L42(2010) URL The glass formation, microstructure and thermal stability of Mg 60 Cu 10 Ni 16 Nd 14 bulk metallic glass (BMG) were investigated by using copper mould casting, X-ray diffraction, scanning electron microscopy, transmission electron microscopy and differential scanning calorimetry. The results show that at least 302mm in diameter of BMG with a larger supercooled liquid region up to 6102K was obtained for the designed composition. Phase-separated regions with diameter size <202μm are homogeneously distributed in the as-cast sample. Under uniaxial compressive loading, the designed Mg-based BMG demonstrates a remarkable macroscopic plastic strain of 2.5% and no work hardening behavior for the sample of 302mm in diameter at room temperature. The improvement of mechanical properties was explained by the extensive shear band formation, interaction, and multiplication due to the heterogeneous microstructure caused by the phase separation during solidification. The low yielding behavior for the BMG was suggested to come from the random formation of shear bands before reached the plateau stress. [本文引用: 1] [124] J.T. Fan, F.F. Wu, Z.F. Zhang, F. Jiang, J. Sun, S.X. Mao, J. Non-Cryst. Solids 353(52), 4707 (2007) [本文引用: 1] [125] C. Fan, D. Chen, P.K. Liaw, H. Choo, C. Benmore, J. Siewenie, G.L. Chen, J.X. Xie, C.T. Liu, Appl. Phys. Lett. 93(26), 1905 (2008) [本文引用: 1] [126] J.S. Yang, C. Zhou, J.X. Xie, Chin. J. Nonferr. Metals 14, 1804 (2004)Google Scholar [本文引用: 1] [127] J. Ding, Y.Y. Li, X.Z. Wang, J. Metastable Nanocrystal. Mater. 2-6, 539(1999) [本文引用: 1] [128] Q. Chao, MA Dissertation (Shanghai University, 2008) [本文引用: 2] [129] S.S. Park, G.T. Bae, D.H. Kang, I.H. Jung, K.S. Shin, N.J. Kim, Scr. Mater. 57(9), 793(2007) URL The microstructure and tensile properties of twin-roll cast (TRC) Mg–6% Zn–1% Mn (ZM61) alloys with varying Al contents were investigated. The microstructure of the alloys becomes finer with the addition of Al. The type of dispersoid particles also changes with the Al addition: from MgZn 2 in ZM61 alloy to Al 8 Mn 5 in the ZM61–1% Al (ZMA611) and ZM61–3% Al (ZMA613) alloys. ZMA611 alloy has a better combination of tensile properties than ZM61 alloy. However, ZMA613 alloy shows poor ductility due to the formation of coarse Mg 21 (Al,Zn) 17 phase during TRC. [本文引用: 2] [130] S.S. Park, Y.S. Park, N.J. Kim, Met. Mater. Int. 8(6), 551 (2002) [本文引用: 2] [131] P.S. Grant, Prog. Mater Sci. 39(4), 497(1995) [本文引用: 2] [132] R.G. Guan, Z.Y. Zhao, H.Q. Huang, C. Lian, R.Z. Chao, C.M. Liu, Acta Phys. Sin. 61(20), 206602 (2012) [本文引用: 1] [133] A. Miehe, U. Gross (2012) In IOP Conference Series: Materials Science and Engineering, IOP Publishing 33(1), 012017 [本文引用: 1] [134] J.T. Li, G.M. Xu, H.L. Yu, G. Chen, H.J. Li, C. Lu, J.Y. Guo, Int. J. Adv. Manuf. Technol. 85, 1007(2016) URL Electromagnetic fields were used in twin-roll casting (TRC) of aluminum alloy 5052 (AA5052) for improvement of the microstructure and mechanical properties. A static magnetic field induces an inhibiti [本文引用: 1] [135] W.Y. Zhang, D.Y. Ju, H.Y. Zhao, X.D. Hu, Y. Yao, Y.J. Zhang, J. Mater. Sci. Technol. 31(5), 517(2015) [本文引用: 2] [136] R. Cook, P.G. Grocock, P.M. Thomas, D.V. Edmonds, J.D. Hunt, J. Mater. Process. Tech. 55(2), 76(1995) URL Casting thin (typically 1–2 mm) has the advantage of reducing the downstream processing requirement and in addition it has been demonstrated that significant increases in productivity are possible for a wide range of alloys. However, the increased exit temperatures and increased separating forces associated with casting at thin gauges result in an increase in the tendency for the cast strip to stick to the caster rolls. The phenomenon of sticking is described and a number of methods of overcoming the problems are discussed. [本文引用: 1] [137] B.Q. Li, JOM 47(5), 29 (1995) [本文引用: 2] [138] J.D. Hwang, H.J. Lin, J.S.C. ISIJ Int. 36(6), 690(1996) URL ABSTRACT Good surface quality is essential for the twin roll cast stainless steel strip to be directly cold-rolled. Thus, surface quality of AISI 304 stainless steel is one of the main problems for the twin roll strip casting process to be put into commercial production. The surface defects commonly appeared in the as-cast strip of twin roll process may include longitudinal and transverse cracks or wrinkles as well as unevenness which is mainly caused by gas trapping and surface vibration of meniscus between the two rolls. In this study, a modeling software was applied to simulate the flow behavior of molten pool at quasi-steady state in inclined twin roll casting and to investigate the relationship between flow characteristics and surface quality of the produced thin strip. The flow characteristics was drawn from the calculated flow patterns and then correlated to the surface quality of thin strips produced in an experimental easter under the same casting condition as the calculations. Through this procedures, several correlations were then found. First, the surface wrinkles appeared on the thin strip when the pool levels were low. The problem was improved with a higher pool level due to less flow violence nor meniscus fluctuation. Second, the recirculation with a large vortex occurring around the triple intersection point of the roll/melt/air three phases was the main cause for the slag and/or dress trapping. Third, by using a higher roll angle along with an inclined feeding position nearer to the lower roll, a much stable flow on the meniscus surface could be obtained. [本文引用: 2] [139] K.I. Miyazawa, Sci. Technol. Adv. Mat. 2(1), 59(2001) URL Not Available [本文引用: 2] [140] K. Feng, MA Dissertation (University of Science and Technology Liaoning, 2007) [本文引用: 3] [141] N. Zapuskalov, ISIJ Int. 43(8), 1115(2003) URL Increased competition and the need to decrease steel product prices has combined with other factors such as growing salaries, increasing cost for electricity and raw materials to apply pressure to both mini and integrated mills. To compete in the future market and to maintain market share, both mills will need to use new efficient technologies capable of supplying steel strip products of high quality at low cost. Large investments in the development of strip casting process by the main world steel produces and plantmakers have already occurred due to its huge potential to substantially reduce the investment cost of steel strip production. Open literature on the continuous casting has been reviewed, with emphasis on strip casting of steel as new generation of continuous casting technology and has covered the following topics: current trends in production of steel strip; advantages of strip produced direct from the melt; strip casting process; strip quality; commercial strip casters and expenditure on research and development of the strip casting process. [本文引用: 1] [142] R. Wechsler, Scand. J. Metall. 32(1), 58(2003) URL Within the coming days the world's first commercial, direct twin-roll caster for carbon steel strip less than 2.0 mm, without further hot or cold rolling, will be started by Nucor in the United States. Although the process has been previously proven on a technical basis during full-scale development in Australia, the Nucor caster is unique in that it will have to demonstrate both its technical and economic advantages in a rigorous commercial environment. The advent of any new technology naturally raises the question of whether there is a need for that technology, and where the new technology will fit into the spectrum of existing processes. Experience to date indicates that twin-roll casters will quickly fill a latent demand for ultra-thin cast strip (UCS), i.e., a new product category for sheet steel. In the future, thick slab casters, thin slab casters and twin-roll casters will coexist and become increasingly specialized. The result will be both new and better products for steel consumers and satisfactory financial returns for steelmakers. The CASTRIP 庐 1 process produces steel strip between 0.7 and 2.0 mm in thickness, directly from liquid steel. For many applications the UCS product will be directly usable by end users, without further hot rolling, pickling or cold rolling. It can be galvanized, coated with aluminum/zinc, painted and formed without further processing, and will be available in strength levels from about 250 Mpa to more than 600 Mpa鈥攁ll using a single, simple chemistry. End-users will be able to buy a new product with price and properties intermediate between today's hot-rolled and cold-rolled steels. Among the unique features of the process is its much lower consumption of energy and reduced environmental emissions. [本文引用: 1] [143] H.B. Wang, L. Zhou, Y.W. Zhang, Y.H. Cai, J.H. Zhang, J. Mater. Process. Tech. 233, 186(2016) URL In the present study, strips of 7050 aluminum alloys were cast using twin roll casting technology (TRC) where roll casting speed, roll gap, cooling water flow rate and initial cooling water temperature were varied. The optimized process parameters and their effects on TRC of 7050 aluminum alloys strips were investigated. The optimized process parameters and properties were obtained to be roll gap of 1.802mm, roll casting speed of 11.402m/min (1202r/min), cooling water flow rate of 11.102m 3 /h and initial cooling water temperature of 20.802°C. The macro morphology, microstructure, hardness and yield stress of the strips were investigated using optical microscopy, scanning electron microscopy (SEM), energy dispersion spectroscopy (EDS) and Vickers hardness. The homogeneous microstructures and improved mechanical properties were obtained by increasing the roll casting speed and roll gap thickness. [本文引用: 2] [144] T. Haga, S. Suzuki, J. Mater. Process. Tech. 113(1), 291(2001) URL A melt drag twin roll caster (MDTRC) was devised in order to cast aluminum alloy strips at a speed higher than 3002m/min. A3003 alloy, A5182 alloy, Al–6mass%Si alloy and Al–12mass%Si alloy were cast into thin strips using the MDTRC. These alloys could be cast into strips at a speed of up to 6002m/min. The thickness of the strips was about 1–302mm. The MDTRC could cast thinner strips than the conventional twin roll caster for aluminum alloys. Lubricant was not used to prevent the sticking of the strip, as sticking did not occur due to the effect of the copper rolls and the low separating force. The mechanical properties and the microstructures of the strips that were cast using the MDTRC, were investigated in the present study. It was shown that MDTRC was useful for the high speed roll casting of aluminum alloy strips. [本文引用: 2] [145] T. Haga, M. Ikawa, H. Watari, S. Kumai, J. Achiev. Mater. Manuf. Eng. 21(1), 7 (2007) [本文引用: 1] [146] D. Liang, C.B. Cowley, JOM 56(5), 26 (2004) [本文引用: 1] [147] H. Watari, T. Haga, N. Koga, K. Davey, J. Mater. Process. Tech. 192, 300(2007) [本文引用: 1] [148] H. Zhao, P.J. Li, L.J. He, J. Mater. Process. Tech. 212(8), 1670(2012) URL The microstructure along the thickness of a strip produced by asymmetric twin-roll casting was found to be significantly inhomogeneous. There were many banded structures with flow form near the upper surface of the strip, whereas a fine dendritic structure dominated near the lower surface of the strip. Precipitates of Mg17Al12 and Al8Mn5 were dispersively distributed throughout the as-cast strip. Recrystallization of the strip during homogenization first occurred in the banded structures. After rolling and annealing, the strip consisted of fine grains measuring approximately 5渭m in size. It was concluded that the shear strain caused by the difference in linear velocity between the surfaces of the upper and lower rolls resulted in the banded structures of the strip. [本文引用: 1] [149] R.G. Guan, Z.Y. Zhao, X. Wang, C.G. Dai, C.M. Liu, Acta Metall.Sin.(Engl. Lett.) 26, 293(2013) URL A novel continuous semisolid rolling process for producing AZ31 alloy strip was developed. The process parameters were optimized, and microstructure and properties of AZ31 alloy prepared by the process were studied. The results reveal that primary grains of the strip become coarse, and the grain structure transforms from round shape to dendrite with the increment of casting temperature gradually. Eutectic phase fraction and primary grain size increase with the increment of roll speed. The primary grain size decreases firstly and then increases with the increment of the vibration frequency correspondingly. When the casting temperature is from 650℃ to 690℃, the roll speed is 0.069 m·s-1 , and the vibration frequency is about 80 Hz, AZ31 alloy strip with a cross section size of 4 mm×160 mm was prepared by the proposed process. The ultimate tensile strength and elongation are improved 1% and 57%, respectively. [本文引用: 1] [150] R.G. Guan, Z.Y. Zhao, Q.S. Zhang, C.S. Lee, Int. J. Adv. Manuf. Tech. 67, 917(2013) URL In this paper, the effect of temperature field on microstructure of A2017 alloy during an innovative continuous semisolid rolling process with a vibrating sloping plate device was studied. The results show that the alloy temperature decreases gradually from the entrance to the exit of the roll gap. In the backward slip zone, the isothermal lines have twice buckling. In the forward slip zone, the isothermal lines have once buckling. Semisolid region moves forward from the filling mouth to the exit of the roll gap with the increment of casting temperature, and the solid fraction increases from the entrance to the exit of the roll gap. The average grain size of the product increases with the increment of casting temperature, and the plastic deformation along the rolling direction happened obviously. According to the simulation and experiment, the proper casting temperature between 650 and 680聽掳C is suggested. A2017 alloy strip with good surface quality was obtained. The microstructure of the product is mainly composed of fine spherical or rosette grains which were elongated along rolling direction. [本文引用: 2] [151] T. Haga, K. Tkahashi, M. Ikawaand, H. Watari, J. Mater. Process. Tech. 153, 42(2004) [本文引用: 2] [152] S. Sahoo, S. Ghosh, Steel Res. Int. 85(2), 207(2014) URL Numerical simulation of transient fluid flow, heat transfer, and solidification during the starting of high speed twin roll strip casting was carried out to estimate the time required for steady state condition to set in and solidification front speed during the transient stage. The solidification front speed during the transient stage was expected to be higher, leading to different microstructure and thus it is necessary to know it in comparison to that during the steady state. The free surface of the pool was tracked using volume of fluids (VOFs) approach and an enthalpy-porosity method was employed to incorporate the phase change during solidification. Surface tension was incorporated into the momentum transfer equation and all the physical properties were treated as functions of temperature. The filling sequence and temperature profiles in the molten pool, along with the solidification front profile were numerically simulated. From the evolution of temperature profile the transient length of the cast strip was estimated. The prediction of temperature profile in the transient stage was experimentally validated using Jackson–Hunt theory. From the results of simulation the transient length of strip, which needs to be rejected for different casting speeds was estimated. It was found that when the speed is increased beyond 0.7979 up to 3.9865m65s611 the transient length did not increase appreciably. [本文引用: 1] [153] A.R.E. Singer, Int. J. Powder Metall. Powder Tech. 21(3), 219 (1985) [本文引用: 1] [154] M. Zhan, Z. Chen, H. Yan, J. Mater. Process. Tech. 202(1), 269(2008) URL For the spray deposited preforms with the porosity up to 10–15%, densification processing is necessary to produce fully dense products. In this paper, two novel processes named as “Frame-Confined Rolling” and “Ceramic Rolling” are presented. The experimental results show that these two methods can improve the workability of the porous preforms without forming cracks on surface and edge during rolling. [本文引用: 1] [155] R.G. Brooks, A.G. Leatham, J.S. Coombs, C. Moore, Metall. Met. Form 44(4), 157 (1977) [本文引用: 1] [156] E.G. Miravete, E.J. Lavernia, G.M. Trapaga, J. Szekely, N.J. Grant, Metall. Trans. A 20(1), 71 (1989) [本文引用: 2] [157] E.J. Lavernia, E.M. Gutierrez, J. Szekely, N.J. Grant, Int. J. Rapid Solidif. 4(1-2), 89 (1988) [本文引用: 1] [158] E.J. Lavernia, E. Gomez, N.J. Grant, Mater. Sci. Eng. 95, 225(1987) URL The structures and room temperature tensile properties of two magnesium alloys processed by the liquid dynamic compaction (LDC) rapid solidification technique were investigated. The two compositions selected for this study were Mg-8.4wt.%Al-0.2wt.%Zr and Mg-5.6wt.%Zn-0.3wt.%Zr. The magnesium alloys were spray-deposited as discs about 15 cm in diameter by 30 mm in thickness against a water-cooled copper substrate. The as-deposited grain sizes ranged from 30 to 60 μm in diameter, associated with a solidification rate of about 10 3 Ks 611, and consisted of equiaxed grains. Densification of the as-deposited LDC compacts was successfully achieved by both hot extrusion and hot isostatic pressing (HIP). The mechanical properties of the extruded or hot pressed, heat treated LDC products were determined at room temperature. The improvements in strength and elongation values of the extruded and heat treated LDC MgAlZr alloy over those of the equivalent ingot product were attributed to the microstructural refinements achieved by LDC and extrusion processing. The as-extruded LDC MgZnZr alloy exhibited high elongation values at reasonable strength levels. [本文引用: 2] [159] E.J. Lavernia, N.J. Grant, Mater. Sci. Eng. 98, 381(1988) URL The development of spray atomization and collection processes, i.e. Osprey, consolidated spray deposition (CSD) and liquid dynamic compaction (LDC) was motivated by the severe problems of oxide contamination associated with quenched reactive metal powders as well as by the need to minimize the overall number of processing steps from the liquid metal to fully dense product, perhaps in the near-net shape state. So far, these techniques have been successfully applied to numerous aluminum-, magnesium-, nickel- and iron-based alloys, yielding attractive combinations of properties, which include improvements in tensile strength, axial fatigue strength, fatigue crack growth resistance, superplasticity and magnetic properties. Furthermore, experimental and mathematical modeling studies have shown that the aforementioned property improvements depend on the production of a rapidly solidified, microstructurally homogeneous material during spray deposition. In turn, the achievement of such structures depends on the characteristics of the atomized droplets in flight and on impact, i.e. droplet size and distribution, relative proportions of solid and liquid, temperatures, droplet velocities, dendrite arm spacings, degree of undercooling etc. The purpose of this paper is to review the more important findings over the past decade of spray atomization and collection studies. [本文引用: 3] [160] B. Yang, F. Wang, J.S. Zhang, Acta Mater. 51(17), 4977(2003) URL An innovative spray-deposition technique has been applied to produce in situ TiC/Al and TiC/Al–20Si–5Fe–3Cu–1Mg composites. This technique provides a new route to solve the problems of losses and agglomeration of the reinforcement particles when they are injected into the spray cone of molten droplets during spray forming process. Experimental results have shown that the presence of needle-like Al 3 Ti and Al–Si–Fe compounds, which are detrimental not only to the fracture toughness, but also to the stability of the microstructure, can be eliminated completely from the final product by using a proper Ti:C molar ratio of 1:1.3 in the Ti–C–Al preforms and adding 5 wt% TiC particles to Al–20Si–5Fe–3Cu–1Mg alloy. Moreover, another major problem of coarsening of silicon particles usually encountered in the hypereutectic Al–Si alloys has also been solved by the technique. The silicon particles in the spray-deposited 5 wt% TiC/Al–20Si–5Fe–3Cu–1Mg composite were much refined (652 μm) compared to those (655 μm) obtained in the matrix alloy without TiC addition. The formation and elimination mechanisms of Al 3 Ti phase in TiC/Al composites can be explained based on thermodynamic theory. The modification of the microstructures in the spray-deposited Al–20Si–5Fe–3Cu–1Mg alloy can be interpreted in the light of the knowledge of atomic diffusion. The experimental results also showed that the ultimate tensile strength of the TiC/Al composites was improved over that of the unreinforced Al matrix. [本文引用: 1] [161] K. Peter, V. Domala, N.J. Grant, Meter. Trans. 14(8), 1599(1983) URL Aluminum alloy 7075 was modified by additions of 1.1 wt pct nickel and 0.8 wt pct zirconium, rapidly solidified by ultrasonic gas atomization, canned, cold compacted, hot extruded, and evaluated in terms of structure and properties. Significant improvements in tensile strength (627 MPa YS and 680 MPa UTS) and crack growth rates were realized, along with a decrease in fracture toughness (23.7 MPa√m) while maintaining ductility (10 pct elong.) as compared to nominal I/M 7075 behavior. The stress for 10 7 cycles fatigue life was greater than 275 MPa, which represents a 73 pct increase over that of I/M 7075. A variety of experiments was performed to evaluate effects on strength, ductility, and on structure. The variables were: powder size distribution, extrusion ratio, extrusion profile, different size fractions from the same lot of powder, and different locations of test bars in the several extrusions. Tensile properties, toughness, and fatigue properties were not importantly influenced by the location of test bars in the cross section or length of rectangular extruded bars. A comparison of mechanical properties from extruded bars prepared from 6153 μm powders vs 53 to 250 μm powders showed a small loss of ductility and fatigue stress for 10 7 cycles for the fine powder product. Higher extrusion ratios were beneficial for mechanical properties. [本文引用: 2] [162] B. Li, E.J. Lavernia, Acta Mater. 45(12), 5015(1997) URL An intermetallic matrix composite TiAl/TiB{sub 2} was spray formed with co-injection using an induction skull melting/spray forming technique. The volume percentage of reinforcement in the spray formed materials is approximately 35%. The microstructure of the spray formed composite is characterized by fine grained, equiaxed fully lamellar structures, with an average linear grain size of approximately 40 {micro}m. The distribution of TiB{sub 2} in the composite exhibits a layered morphology. On the basis of experimental observations of TiB{sub 2} distribution in oversprayed powders and related numerical analysis, this phenomenon was attributed to the segregation of TiB{sub 2} particles to the exterior region of the droplets and the subsequent deformation of the droplets along the direction normal to the substrate during impingement. Numerical analysis suggests that the most likely approach to improve the homogeneity of the distribution of TiB{sub 2} in the deposit is to decrease the co-injector/atomizer distance. [本文引用: 1] [163] D. Li, Z.T. Wang, D.Z. Ma, Light Alloy Fabr. Tech. 42(11), 1 (2014) [本文引用: 1] [164] T. Haga, J. Mater. Process. Tech. 130, 558(2002) [本文引用: 1] [165] T. Haga, S. Suzuki, J. Mater. Process. Tech. 137(1), 92(2003) URL A melt ejection twin roll caster (METRC) was devised for the high-speed roll casting of aluminum alloy to secure sound surfaces. This caster can cast A5182 alloy as strip at a speed of up to 120聽m/min. The strip surface can be improved by ejecting the melt onto the roll. The strip does not stick to the roll because the separating force is very small. Therefore, lubricant is not required in order to prevent the sticking of the strip at the roll. The microstructure of the strip cast using the METRC is very fine. [本文引用: 1] [166] T. Haga, J. Mater. Process. Tech. 111(1), 64(2001) URL The present paper describes the melt drag twin roll caster, a recently devised type of high-speed and semi-solid twin roll caster which requires no lubricant to prevent the strip-sticking to the rolls. The copper rolls and extremely small separating force (0.01鈥0.1 MPa) of the melt drag twin roll caster eliminates the tendency for the strip to adhere to the rolls, thus eliminating the need for a lubricant. Use of copper rolls and no use of the lubricant are significant factors in high-speed roll casting, and A5182 can be cast at the speeds up to 60 m/min. The microstructure of the A5182 strips cast by the melt drag twin roll caster is not columnar structure but equiaxed structure. The higher the roll speed, the smaller the microstructure. The strip thickness is at the range about from 2 to 3 mm. [本文引用: 1] [167] R.G. Guan, Z.Y. Zhao, H. Zhang, C. Lian, C.S. Lee, C.M. Liu, J. Mater. Process. Tech. 212(6), 1430(2012) [本文引用: 1] [168] X. Wang, R.G. Guan, N. Guo, Z.Y. Zhao, Y. Zhang, N. Su, Mater. Sci. Technol. 32(2), 154(2016) URL This paper proposes the influence factors of coupling effects of shearing flow and vibration on diffusion coefficient and critical nucleation energy during metal solidification. Based on this proposal, a metal solidification–nucleation–rate model under coupling effects of shearing flow and vibration is established. Verification experiment using Al–7Si alloy is carried out. When vibration frequency and melt flow velocity are zero, the results calculated by the above model agree with that calculated by Turnbull’s theory. The results calculated by the above model under coupling effects of shearing flow and vibration agree with the experimental results, with the error within 0·2–14·3%. So the established model can calculate and explain the nucleation rate of melt under coupling effects of shearing flow and vibration. [本文引用: 1] [169] R.G. Guan, Z.Y. Zhao, Y.D. Li, T.J. Chen, S.X. Xu, P.X. Qi, J. Mater. Process. Tech. 229, 514(2016) URL The microstructures of A356 alloy parts contained near-round or rosette α-Al grains and eutectics at grain boundaries. Melt processing by vibrating slope could refine grain size and eliminate eutectic phase segregation as well as regional segregation which usually occurred in traditional squeeze casting. Average grain diameter of α-Al grain increased from 4402μm to 6402μm, and grain average roundness decreased firstly from 2.9 to 1.7 then increased to 3.3 with the increase of casting temperature. At casting temperature from 67002°C to 68002°C, crack and cooling shut usually occurred in the products due to poor flow ability of the alloy. When the casting temperature was 69002°C, the alloy had a good filling ability and smooth product with fine near-round or rosette α-Al grains and homogenous eutectics has been obtained. The ultimate tensile strength and elongation of semisolid squeeze casting A356 alloy reached 23202MPa and 7%, which were improved by 12% and 21% respectively as comparing with that of traditional squeeze casting. [本文引用: 1] [170] Z.Y. Zhao, R.G. Guan, X. Wang, C.M. Liu, Acta Metall.Sin.(Engl. Lett.) 26, 447(2013) URL A novel semisolid rheo-rolling process of AZ91 alloy was proposed. The microstructure formation mechanism of AZ91 magnesium alloy during the process was studied. The results reveal that the eruptive nucleation and the heterogeneous nucleation exist. During the grain growth process, the grain breakage took place and transformed into fine spherical or rosette grains on the sloping plate gradually, the other grain growth style is direct globular growth. Due to the secondary crystallization of the remnant liquids in the roll gap, the microstructure of the strip becomes finer with the increment of the casting temperature from 650 °C to 690 °C. But when the casting temperature reached 710 °C, a part of the liquid alloy transformed into the eutectic phases, and the primary grains ripened to form coarse dendrites. In the casting temperature range from 650 °C to 690 °C, AZ91 alloy strip with fine spherical or rosette grains was prepared by the proposed process. [本文引用: 1] [171] R.G. Guan, Y.F. Shen, Z.Y. Zhao, R.D.K. Misra, Sci. Rep. 6 (2016). doi:10.1038/srep23154 [本文引用: 2] [172] F. Wang, Z. Liu, D. Qiu, J.A. Taylor, M.A. Easton, M.X. Zhang, Acta Mater. 61(1), 360(2013) URL The grain refining effect of four peritectic-forming solutes (Ti, V, Zr and Nb) as well as three eutectic-forming solutes (Cu, Mg and Si) on pure Al was investigated. Significant grain refinement is observed by the addition of peritectic-forming solutes, whereas the addition of eutectic-forming solutes only slightly decreases the grain size. The mechanisms underlying the grain refinement of these alloys were then studied by a new analytical methodology for assessing grain refinement that incorporates the effects of both alloy chemistry and nucleant potency. It is found that the low degree of grain refinement by the addition of eutectic-forming solutes is mainly attributed to the segregating power of solutes, i.e. the constitutional undercooling contribution. However, peritectic-forming solutes do not only cause grain refinement by their segregation power but, more importantly, they introduce copious potent nuclei into the melt and promote significant grain refinement via heterogeneous nucleation. [本文引用: 5] [173] F. Wang, D. Qiu, Z.L. Liu, J.A. Taylor, M.A. Easton, M.X. Zhang, Acta Mater. 61(15), 5636(2013) URL The mechanism underlying the grain refinement of cast aluminium by zirconium has been studied through examination of a range of Al alloys with increasing Zr contents. Pro-peritectic Al 3 Zr particles are reproducibly identified at or near the grain centres in grain-refined alloy samples based on the observations of optical microscopy, scanning electron microscopy and X-ray diffraction. From the crystallographic study using the edge-to-edge matching model, electron backscatter diffraction and transmission electron microscopy, it is substantiated that the Al 3 Zr particles are highly potent nucleants for Al. In addition, the effects of Al 3 Zr particle size and distribution on grain refinement has also been investigated. It has been found that the active Al 3 Zr particles are bigger than previously reported other types of active particles, such as TiB 2 for heterogeneous nucleation in Al alloys. Considering the low growth restriction effect of Zr in Al (the maximum Q -value of Zr in Al is 1.0聽K), it is suggested that the significant grain refinement of Al resulting from the addition of Zr can be mainly attributed to the heterogeneous nucleation facilitated by the in situ formed Al 3 Zr particles. [本文引用: 2] [174] J.J. Ning, X.Q. Li, R.P. Jiang, M.Z. Huang, J. Cent, South Univ. 46(8), 2837 (2015)Google Scholar [本文引用: 3] [175] D.P. Mondal, N. Jha, A. Badkul, S. Das, M.S. Yadav, P. Jain, Mater. Des. 32(5), 2803(2011) URL Aluminium 7178 alloys containing 1% calcium are used to study the effect of calcium addition on their microstructure and compressive deformation behaviour. The compressive deformation behaviour of aluminium alloy containing 1% calcium is studied at varying strain rates (10 612–10/s). The material is prepared using stir casting technique. The yield stress, flow stress and elastic limit are measured from the true stress–strain graph. The strain rate sensitivity and strain-hardening exponent was also determined for each material at different strain rates. Its microstructural characterization reveals that Ca particles act as grain refiners for primary base alloy and helps in improving the strength of the virgin alloy. An empirical relationship has been proposed to predict the flow curve of the alloys as a function of strain and strain rate. [本文引用: 2] [176] S. Nafisi, R. Ghomashchi, Mater. Sci. Eng. A 452, 437 (2007) [本文引用: 1] [177] Y. Birol, J. Alloys Compd. 513, 150(2012) URL The potential of AlB3 master alloy in the grain refinement of AlSi10Mg and AlSi12Cu foundry alloys was investigated and compared with that of the AlTi5B1 master alloy, the standard grain refiner for most aluminium foundries. The latter refines the grain structures of both alloys. However, this performance is not nearly as good as that obtained in wrought aluminium alloys with the same grain refiner. The Ti-free AlSi10Mg and AlSi12Cu alloys, on the other hand, exhibit very small grains for the entire range of holding times when inoculated with AlB3. This implies a remarkable grain refining efficiency, typical of grain refined wrought aluminium alloys, as well as a strong resistance to fading of the grain refinement effect. Lack of Ti in the melt allows the entire B to form AlB 2 particles, the perfect substrates, shortly before α-Al starts to crystallize. Aluminium castings can enjoy grains as small as those of the wrought alloys, well below 20002μm, with an addition of 0.0202wt% B provided that their Ti content is controlled. [本文引用: 2] [178] Y.J. Zhang, N.H. Ma, H. Yi, S. Li, H. Wang, Mater. Des. 27(9), 794(2006) URL The influence of various Fe contents on the grain refinement of commercial purity aluminum was investigated. After commercial purity aluminum melted, given amounts of Al–Fe master alloy were added into the molten in this study. Macroscopic observation indicates that the degree of grain refinement of commercial purity aluminum depends on the additions of Al–Fe master alloy. The case of 0.5 wt%Fe content was founded to be more effective in refining grain of commercial purity aluminum compared to the other Fe contents within the range studied. Basing on scanning electron microscopy analyses grain refining mechanism of Fe was discussed in this research. [本文引用: 1] [179] L. Bolzoni, M. Nowak, N.H. Babu, Mater. Des. 66, 376(2015) URL The potency of Nb–B inoculation for the refinement of Al–Si cast alloy has been demonstrated in Part I of this work by the systematic analysis of binary Al– x Si alloys (where x 02=021–1002wt.%). In Part II of this work the effect of Nb–B inoculation on commercial Al–Si alloys is assessed. Specifically, hypo-eutectic alloys such as LM24 (A380) and LM25 (A356) as well as near-eutectic LM6 (A413) Al–Si alloys are considered. The aim is to quantify the grain refinement and detect possible interaction with alloying elements commonly present in Al cast alloys, such as Mg, Fe, Cu, Mn and Zn. The in-depth analysis of the alloys solidified under wide range of cooling rates indicates that Nb–B inoculation does not only lead to a much finer microstructural features but also makes the final grain size far less sensitive to the cooling rate employed to solidify the material. Finally, the mechanism essential for the grain refinement of commercial Al cast alloys by Nb–B inoculation is determined on the base of SEM and thermal analysis results. It is found that in-situ formed Al 3 Nb and NbB 2 intermetallic particles (forming from the interaction of Al alloy/Nb powder/KBF 4 flux) are the heterogeneous nuclei responsible for the grain refining of Al cast alloys. [本文引用: 2] [180] M.V. Farahani, E. Emadoddin, M. Emamy, A.H. Raouf, Mater. Des. 54, 361(2014) URL The present study deals with an investigation on dry sliding wear behavior of grain refined Sc-free 7042 aluminum alloy by using a pin-on-disc wear test machine. Al–5Ti–1B and Al–15Zr master alloys were used as grain refining agents. The optimum amounts of added Ti and Zr in the alloy were found to be 0.0302wt.% and 0.302wt.%, respectively. Extrusion was carried out and T6 heat treatment ware applied for all rod specimens before testing. Significant improvement in mechanical properties was obtained with the addition of grain refiners. The worn surfaces were characterized by energy dispersive X-ray spectrometry microanalysis. Results showed that the wear resistance of unrefined alloy increased with the addition of both grain refiners. Furthermore, the worn surface studies showed a mixed type of wear mechanisms; delaminating, adhesive and abrasive which took place at higher applied load. [本文引用: 2] [181] U. Patakham, J. Kajornchaiyakul, C. Limmaneevichitr, J. Alloys Compd. 542, 177(2012) URL Grain refinement of the primary aluminum (α-Al) phase in a hypoeutectic Al–Si alloy using scandium (Sc) was studied to identify the grain refinement mechanism. Optical microscopy (OM), Scanning electron microscopy (SEM) with energy dispersive spectroscopy (EDS) and electron backscatter diffraction (EBSD) techniques were extensively used in this study. We found that Sc refined grains of primary aluminum. However, the grain refinement efficiency of Sc was considerably lower than that of titanium (Ti) in the Al–Si–Mg foundry alloy. It was evident that the precipitated Sc-containing phases acted as heterogeneous nucleation sites for the primary aluminum phase. The Sc-containing heterogeneous sites are irregular in shape with sizes between 3 and 5μm. At least three groups of nuclei based on their chemical composition were found, i.e., (i) Al and Sc, (ii) Al, Si, Mg, and Sc, and (iii) Al, Si, Mg, Sc, and Fe. Crystal orientation mapping showed primary aluminum dendrites with one orientation in each grain near Al3Sc particles. The grain refinement mechanism of Sc for aluminum relies on heterogeneous nucleation of Al3Sc particles, with less responsibly for grain growth restriction. Many intermetallic phases with Al, Si, Fe, Mg and Sc as their major components were found, and these phases could not effectively act as heterogeneous nuclei. [本文引用: 1] [182] Y. Chen, Y. Pan, T. Lu, S.W. Tao, J.L. Wu, Mater. Des. 64, 423(2014) URL In this paper, with the combinative addition of La and B elements, the grain refinement of Al–Si alloys with different contents of Si was achieved. Compared to individual addition of B element, the combinative addition of La and B elements can effectively refine the grains of Al–Si alloys. The addition of La element suppresses the mutual poisoning between Sr and B elements, benefiting the formation of a fully modified eutectic silicon structure in the Al–Si alloys. This work also indicates that the tensile properties, especially the elongation, of Al–Si alloys are enhanced with the addition of La element. [本文引用: 1] [183] Z. Fan, Y. Wang, Y. Zhang, T. Qin, X.R. Zhou, G.E. Thompson, T. Pennycookc, T. Hashimoto, Acta Mater. 84, 292(2015) URL Al–Ti–B is the most widely used grain refiner for many Al alloys. However, the precise mechanism of grain refinement is still not clear after 6002years of intensive research. This work aims to further our understanding on the grain refining mechanism involving Al–Ti–B-based grain refiners. Extensive high-resolution electron microscopy investigation has confirmed the existence of a Ti-rich monolayer on the (0020020021) TiB 2 surface, which is most likely to be a (1021022) Al 3 Ti two-dimensional compound (2DC). Further experimental investigation was carried out to understand the potency of TiB 2 particles and the stability of the Al 3 Ti 2DC. Our results showed that the potency of TiB 2 particles is significantly increased by the formation of a monolayer of Al 3 Ti 2DC on their surface. The Al 3 Ti 2DC forms at the liquid–Al/TiB 2 interface in concentrated Al–Ti solutions, but dissolves in dilute Al–Ti solutions, although the kinetics of both the formation and dissolution of Al 3 Ti 2DC are relatively sluggish. Effective grain refinement by the Al–5Ti–1B grain refiner is directly attributed to the enhanced potency of TiB 2 particles with the Al 3 Ti 2DC and sufficient free Ti solute in the melt after grain refiner addition to achieve the columnar-to-equiaxed transition. [本文引用: 4] [184] P.S. Mohanty, J.E. Gruzleski, Acta Metall. Mater. 43(5), 2001 (1995) [本文引用: 4] [185] L. Zhou, F. Gao, G.S. Peng, N.A. Baena, J. Alloys Compd. 689, 401(2016) URL In the present work, the effect of potent TiB 2 particle addition levels and impurities on the grain refinement of two Al compositions was investigated. Results showed that no effective grain refinement was achieved in high purity Al (HP-Al) up to particle density levels as high as 2.5脳10 14 /m 3 . In commercial purity Al (CP-Al) fine grain refined structures were observed with merely a particle density of 0.35脳10 13 /m 3 . As more particles were added into the CP-Al melt, the number of grains formed showed an overall increase while the calculated particle efficiency showed a general decrease, consistent with the prediction of interdependence theory. Such results show that the impurities in CP-Al play a significant role in refining the grain structures. [本文引用: 3] [186] B.S. Murty, S.A. Kori, M. Chakraborty, Int. Mater. Rev. 47(1), 3(2002) URL Grain refinement of aluminium and its alloys is common industrial practice. The field has been extensively investigated by many workers over the past 50 years, not only to develop efficient grain refiners for different aluminium alloys, but also to achieve an understanding of the mechanism of grain refinement. The present review confines itself to the literature on grain refinement by heterogeneous nucleation and alloying. Initially, the fundamentals of grain refinement by inoculants are outlined. The types of grain refiner, Al-Ti-B master alloys in particular, and their methods of manufacture are next discussed. The grain refining tests to assess the efficiency of the grain refiners and the grain refining behaviour of aluminium alloys are also discussed in brief. The performance of a grain refiner, as well as the response of an aluminium alloy to grain refinement, is influenced by the microstructure of the grain refiner as controlled by the process parameters involved in its preparation and the alloying elements present in the aluminium alloy. The roles of these factors, and particularly the roles of poisoning elements such as Si, Cr, Zr, Li, are reviewed. The paper also reviews the mechanisms of grain refinement, the fading and poisoning phenomena, and the trends in the development of new grain refiners for aluminium alloys containing poisoning elements. [本文引用: 2] [187] H.J. Wang, J. Xu, Y.L. Kang, M.O. Tang, Z.F. Zhang, J. Mater. Eng. Perform. 23(4), 1165(2014) URL High crack tendency is easy to occur during preparation of the castings and ingots of Al-Zn-Mg-Cu alloy, and it is most difficult to be solved due to the characteristics for the series alloy. As-cast microstructures and hot cracking tendency of Al-Zn-Mg-Cu alloy with different addition levels of Al-5Ti-1B-1Re were investigated in the paper. Moreover, solidification characteristics of the experimental alloys with different mass fractions of Al-5Ti-1B-1Re were analyzed, and the addition content of Al-5Ti-1B-1Re was optimized. These results indicate that the microstructure of the experimental alloys with Al-5Ti-1B-1Re refiner is fine obviously, the dendrite shape refined by Al-5Ti-1B-1Re becomes more globular, the grain boundary is smoother, and the SDAS is smaller compared with the alloy without Al-5Ti-1B-1Re. When the addition level of the grain refiner is less than 0.2%, the hot cracking tendency for the experimental alloy is reduced. The addition of 0.2% Al-5Ti-1B-1Re is best effective to improve the as-cast microstructures and hot cracking of Al-Zn-Mg-Cu alloy, and the best addition level of Al-5Ti-1B-1Re refiner is 0.2%. [本文引用: 3] [188] T.F. Ma, Z.Y. Chen, Z.R. Nie, H. Huang, J. Rare Earth 31(6), 622 (2013) [本文引用: 3] [189] Z.R. Nie, S.P. Wen, H. Huang, B.L. Li, T.Y. Zuo, Chin. J. Nonferr. Metals 21(10), 2361 (2011)Google Scholar [本文引用: 1] [190] Z.B. Xing, S.P. Lin, S.P. Wen, Z.R. Nie, J. Chin. Soc. Rare Earths 29(2), 158 (2011) [本文引用: 1] [191] F.A. Crossley, L.F. Mondolfo, J. Metals (New York) 191, 1143 (1951)Google Scholar [本文引用: 2] [192] M.M. Guzowski, G.K. Sigworth, D.A. Sentner, Metall. Mater. Trans. A 18(5), 603 (1987) [本文引用: 3] [193] J.M. Antonio, L.M. Lfo, Metal. Mater. Trans. B 2(2), 465 (1971) [本文引用: 1] [194] L. Backerud, S. Yidong, Aluminium 67(7-8), 780 (1991) [本文引用: 2] [195] G.P. Jones, Refining and Alloying of Liquid Aluminium and Ferro-Alloys (Aluminium-Verlag, Trondheim, 1985), p. 211Google Scholar [本文引用: 2] [196] D. Wearing, A.P. Horsfield, W. Xu, P.D. Lee, J. Alloys Compd. 664, 460(2016) URL TiB 2 particles are proven effective nucleants of commercial purity aluminium, resulting in smaller grains and hence greater desired mechanical properties; however, there is uncertainty as to the mechanism by which it operates. Here we clarify what happens in the initial stages by computing the total Gibbs energy change associated with four possible nucleation mechanisms, each characterised by the termination of the TiB 2 (0001) substrate (Ti or B) and the solid that forms on it (Al or Al 3 Ti). The appropriate solid//solid interfacial energies are derived from Density Functional Theory (DFT) calculations, while the bulk energies are derived from thermodynamic data, supplemented with strain energies calculated from DFT. Solid//liquid interfacial energies are estimated using simple models with parameters based on the literature and DFT calculations. The results suggest that the Ti termination of TiB 2 is more stable than the B termination in the melt, and that the direct formation of Al off a Ti-terminated TiB 2 substrate is the most favourable mechanism for the nucleation of Al rather than the previously proposed formation of a Al 3 Ti interlayer. On the B termination of TiB 2 , Al formation is more stable for thick solid layers, but this is much more uncertain for thin solid layers where it is possible that Al 3 Ti formation is more stable. [本文引用: 1] [197] P. Schumacher, A.L. Greer, J. Worth, P.V. Evans, M.A. Kearns, P. Fisher, A.H. Green, Mater. Sci. Technol. 14(5), 394(1998) URL AbstractGrain refinement in commercial aluminium alloys can be achieved by addition of Al–Ti–B master alloys, containing α aluminium, TiB2, and Al3 Ti. The final grain size depends on the kinetics of both nucleation and growth of solid in the liquid. Conventional solidification studies permit measurement of growth, but only indirect investigation of nucleation. By embedding grain refining particles in a glassy matrix of Al85Y8Ni5 Co2, novel microstructural studies of nucleation become possible. Such studies show that nucleation of α aluminium occurs on the basal faces of TiB2 particles coated with Al3Ti. Well defined crystallographic orientation relationships are observed, together with evidence for strong chemical interactions. The nucleation in the amorphous alloy can be related to phenomena common in conventional casting using grain refiners, such as contact time, fading, influence of titanium content, and poisoning. This increased understanding of the fundamental nucleation mechanism permits further i... [本文引用: 1] [198] S.A. Kori, B.S. Murty, M. Chakraborty, Mater. Sci.Eng., 283(1),94 (2000) URL The grain refining response of Al and Al–7Si alloy has been studied with various Al–Ti, Al–B and Al–Ti–B master alloys at different addition levels. The results show that Al–B and B rich Al–Ti–B master alloys cannot grain refine Al, while they are efficient grain refiners to Al–7Si alloy. The level of grain refinement saturates after 0.03% of Ti or B for most of the master alloys studied both at short and long holding times. The grain refining efficiency of some elements other than Ti and B on Al–7Si alloy has also been studied. Interestingly, all the elements studied (B, Cr, Fe, Mg, Ni, Ti and Zr) have resulted in some grain refinement of Al–7Si alloy at short holding time and have shown fading/poisoning on long holding, which increased in the order of B (no poisoning), Ti, Cr, Ni, Fe, Mg, Zr. Sr (0.02%) has been found to provide complete modification of the eutectic in Al–7Si alloy within 2 min, which is not lost even after long holding up to 120 min. Significant improvements in the mechanical properties have been obtained by a combination of grain refinement and modification to an extent that was not possible by either of them alone. [本文引用: 2] [199] H.R. Zhang, Z.B. Liu, Z.Z. Li, G.W. Li, H. Zhang, Acta Metall.Sin.(Engl. Lett.) 29(5), 414(2016) URL The cooling rate sensitivities of Al Ti B, RE and Al Ti B–RE refiners were investigated using laboratory experiments and the actual industrial applications of A356 automotive wheel via low pressure die casting technology. Their impact mechanisms on the microstructure and mechanical properties of the A356 alloy were discussed. The results demonstrated that the Al Ti B–RE refiner possessed most effective and synergetic refinement effects compared to the individual Al Ti B or RE refiners. The Al Ti B–RE refiner exhibited the least sensitivity to the cooling rate changes than the other refiners. The comprehensive properties of alloy wheel refined by the Al Ti B–RE refiner were improved significantly.The tensile strength, yield strength, and elongation of wheel spoke improved by approximately 11.3%, 10.8% and 44.1%,respectively. The property difference values of the tensile strength, yield strength, and elongation in different positions of the wheel decreased from 14.8%, 31.2% and 47.7% to 8.6%, 27.1% and 30.9%, respectively. [本文引用: 1] [200] M.A. Easton, D.H. Acta Mater. 49(10), 1867(2001) URL A model for the determination of relative grain size is developed based on the assumption that nucleant substrates are activated by constitutional undercooling generated by growth of an adjacent grain. The initial rate of development of constitutional undercooling is shown to be equivalent to the growth restriction factor and is a useful approximation for the effect of solute on grain size when potent nucleant substrates are present. However, when the nucleants are of a poor potency then a calculation of the fraction solid necessary to develop the constitutional undercooling required for nucleation needs to be performed. Using the model, trends in grain size observed experimentally by the addition of titanium to pure aluminium and to a typical casting alloy, AlSi7Mg0.3, can be predicted. It was also found that summing the individual growth restriction factors of each solute element in a multi-component alloy can grossly overestimate the actual value of the growth restriction factor for the alloy. [本文引用: 2] [201] J.F. Nie, X.G. Ma, H.M. Ding, X.F. Liu, J. Alloys Compd. 486(1), 185(2009) URL A kind of Al–Ti–C–B master alloy with a uniform microstructure is prepared using a melt reaction method. It is found that the average grain size of α-Al can be reduced from 3500 to 17002μm by the addition of 0.202wt.% of the prepared Al–5Ti–0.3C–0.2B and the refining efficiency does not fade obviously within 6002min. It is considered that the TiC x B y and TiB 261 m C n particles found at the grain center are the effective and stable nucleating substrates for α-Al during solidification, which accounts for the good grain refining performance. [本文引用: 1] [202] P.T. Li, X.G. Ma, Y.G. Li, J.F. Nie, X.F. Liu, J. Alloys Compd. 503(2), 286(2010) URL The effects of a trace amount of C addition on the microstructure and refining performance of Al–Ti–B master alloy were studied and the Al–5Ti–0.8B–0.2C master alloy was prepared. Particles were extracted from the Al–5Ti–1B and Al–5Ti–0.8B–0.2C master alloys. In the Al–5Ti–1B master alloy, TiB 2 particles exhibit hexagonal platelet and layered stacking morphologies, and agglomerate seriously in grain boundaries. After a trace amount of C is added, many paragenetic particles form in the Al–5Ti–0.8B–0.2C master alloy. They are composed of TiB 2 particles and small TiC particles. The small TiC particles grow on the surface of TiB 2 particles or are enchased in TiB 2 particles. The existence of TiC particles weakens the TiB 2 particle agglomerations obviously. Due to the microstructure improvement, the Al–5Ti–0.8B–0.2C master alloy shows much better refining performance than the Al–Ti–B master alloy. The average grain size of α-Al refined by 0.202wt.% the prepared Al–5Ti–0.8B–0.2C master alloy is about 19002μm. In addition, the refining efficiency of the Al–5Ti–0.8B–0.2C does not fade within 6002min. [本文引用: 2] [203] T.M. Wang, Z.N. Chen, H.W. Fu, J. Xu, Y. Fu, T.J. Li, Scr. Mater. 64(12), 1121(2011) URL Al–B master alloys were ascertained to be efficient grain refiners on pure aluminum. Without introducing any stirring during reaction, three experimental Al–3B master alloys were prepared via the halide salt route by varying the salt addition temperature, i.e. 750, 850 and 95002°C. All of them showed grain refining potency on pure aluminum and that with the 95002°C addition temperature was found to be the most efficient. Resistance to fading was also noted during holding after inoculation with this master alloy. [本文引用: 1] [204] P. Zhang, J.F. Nie, T. Gao, T. Wang, X.F. Liu, J. Alloys Compd. 601, 267(2014) URL In this study, a group of Al–Ti–C master alloy has been prepared in an atmosphere with different concentration of N 2 . The master alloys are analyzed by field emission scanning electron microscope (FE-SEM) and transmission electron microscope (TEM). It is discovered that N atoms can be doped into TiC x at a certain concentration of N 2 while it will form AlN if the concentration is higher. Adding Al–Ti–C master alloy with N doped TiC x particles in molten aluminum, the average grain size of α-Al can be reduced from 332002μm to 18902μm and the efficiency does not fade within 6002min. It is supposed that the grain refinement efficiency and stability of TiC x is improved obviously after N doping. This master alloy exhibits potential to promote the application of grain refiner in industries. [本文引用: 1] [205] Y.F. Han, D. Shu, J. Wang, B.D. Sun, Mater. Sci.Eng., 430(1),326 (2006) URL The microstructure and grain refining performance of an Al–5Ti–1B master alloy prepared under high-intensity ultrasound were investigated. With applying continuous high-intensity ultrasound vibrations in the reaction, the Al–5Ti–1B master alloy is successfully manufactured in 402min. Compared with conventional Al–5Ti–1B master alloys, the mean size and the size spread of TiB 2 particles in the prepared master alloy are evidently decreased. The narrower particle size spread significantly improves the grain refining performance of the master alloy, which proves the calculation predictions by Greer. Consequently, the limiting grain size of commercial purity aluminium refined by the new master alloy can reach 4502μm. [本文引用: 2] [206] Y.F. Han, K. Li, J. Wang, D. Shu, B.D. Sun, Mater. Sci.Eng., 405(1),306 (2005) URL The microstructure and grain refinement of an Al–Ti–B master alloy treated with high-intensity ultrasound in fabrication and remelting process have been investigated. The high-intensity ultrasound applied in remelting of commercial Al–5Ti–1B improves the microstructure and grain refinement performance of the master alloy. The reaction among halide salts and aluminium is accelerated when the melt is treated with ultrasound. With united ultrasonic treatments in fabrication and solidification of the master alloy, not only the morphology of TiAl 3 phase is improved, but also the particles in the agglomeration are in spawn-like form instead, which further improves the performance of the master alloy. The effects of acoustic cavitation and streaming on the Al–5Ti–1B master alloy are also discussed. [本文引用: 2] [207] X.L. Cui, Y.Y. Wu, T. Gao, X.F. Liu, J. Alloys Compd. 615, 906(2014) URL A novel Al–3B–5Sr master alloy has been prepared through in situ synthesis method in Al melt. The as-prepared master alloy was then used to modify A356 alloy. The microstructures of the Al–3B–5Sr alloy and modified A356 alloy were investigated. The results showed that the Al–3B–5Sr alloy mainly consisted of phases of α-Al and granular SrB 6 , while the Al–3B–5Sr–7Si alloy contained irregular blocky Al 4 Sr phase, AlB 2 and Si besides the above-mentioned phases. Satisfactory grain refining and modifying effects were obtained by the addition of Al–3B–5Sr alloy (0.502wt.%) to the A356 alloy. Meanwhile, the morphology of eutectic silicon was changed from needle/platelike form to fibrous/globular form. The sizes of the α-Al dendrites decreased from 65100002μm to 6530002μm. The mechanical properties of A356 alloy has been improved significantly with 0.502wt.% Al–3B–5Sr alloy addition. [本文引用: 4] [208] M. Nowak, W.K. Yeoh, L. Bolzoni, N.H. Babu, Mater. Des. 75, 40(2015) URL We recently reported that the combined employment of niobium and boron (i.e. Nb-based intermetallics formed in the melt by the addition of powders), instead of niobium or boron individually, is a highly effective way to refine the grain size of Al–Si alloys without the inconvenience of the poisoning effect typical of commercial Al–Ti–B master alloys. In this work the progress concerning the development of Al–xNb–yB master alloys, which are much more suitable for its use in aluminium foundries, is reported and discussed. Precisely, a first approach to produce Al–xNb–yB master alloys as well as its characterisation by means of EDS mapping and TEM is presented. The study is completed by testing the effectiveness of the produced Al–xNb–yB master alloys on pure aluminium and binary Al–10Si alloy as well as commercial hypoeutectic and near-eutectic Al–Si alloys. It is found that the approach employed to produce the Al–xNb–yB master alloys is suitable because the size of the primary α-Al dendrites is significantly reduced in each of the case investigated. [本文引用: 3] [209] L. Bolzoni, M. Nowak, N.H. Babu, A 628, 230 (2015) [本文引用: 2] [210] T. Wang, T. Gao, P. Zhang, J.F. Nie, X.F. Liu, J. Alloys Compd. 589, 19(2014) URL The microstructure and mechanical properties of Al–5Cu alloy refined by a new kind of Al–5Ti–0.75C master alloy were investigated. The results show that the master alloy has an excellent grain refining performance on Al–5Cu alloy, and the average grain size of α-Al is reduced from about 100002μm to 5002μm when 0.2 wt.% of the master alloy was added. It also indicates that the addition of Al–5Ti–0.75C considerably increases the quantity of θ ′ precipitates and decreases their sizes during heat treatment process. In addition, the tensile property, hardness and wear resistance of the Al–5Cu alloy are improved markedly after grain refinement. The improved mechanical properties are attributed to a fine equiaxed grain structure of the alloy, a dispersed distribution of second phases and a large number of fine θ ′ precipitates. [本文引用: 2] [211] E.Z. Wang, T. Gao, J.F. Nie, X.F. Liu, J. Alloys Compd. 594, 7(2014) URL The grain refinement limit of 6063 alloy inoculated by Al–5Ti–0.25C and Al–5Ti–1B master alloys and the impact of grain size on the mechanical properties, including the tensile strength and hardness, are investigated in the present article. It is revealed that the smallest average grain size of 6063 alloys refined by these two master alloys is the same under the present experimental condition, thus the grain refinement limit of 6063 alloy is 4002μm. However, the limit addition of the two master alloys corresponding to the grain refinement varies at 2.002wt.% and 3.002wt.% for Al–5Ti–0.25C and Al–5Ti–1B, respectively, and the grain refining performance of the former is much better than that of the later one before reaching the refinement limit. Furthermore, the mechanical properties of 6063 alloy are also improved first with the grain refinement, but turn to decrease after the same 1.502wt.% addition both for Al–5Ti–0.25C and Al–5Ti–1B. The highest ultimate tensile strength (UTS) and hardness (HB) of 6063 alloy obtained by the refinement of Al–5Ti–0.25C and Al–5Ti–1B are 27102MPa, 81.7 and 25002MPa, 80, respectively. Thus the UTS and hardness of 6063 alloy refined by the former master alloy are also much higher than those refined by the later one. It is worthy to note that the 6063 alloys with the smallest average grain size did not behave the highest mechanical properties and the possible mechanisms are also discussed. [本文引用: 2] [212] S.S. Ebrahimi, J. Aghazadeh, K. Dehghani, M. Emamy, S. Zangeneh, Mater. Sci. Eng. A 636, 421 (2015) [本文引用: 1] [213] B.T. Sofyan, D.J. Kharistal, L. Trijati, K. Purba, R.E. Susanto, Mater. Des. 31, S36(2010) URL Addition of grain refiner is an option to obtain higher mechanical properties of aluminium cast alloy. Grain refiner will react with molten aluminium and will form nucleant particles that initiate solidification. Therefore, the grain refiner will also be useful to control the solidification processes to reduce shrinkage formation. This study evaluated grain refinement in AA333 aluminium alloys by using Al–Ti granulated flux. Experiments were conducted by adding 0.056 and 0.07802wt.% Ti granulated flux to the molten aluminium during rotary gas bubble floatation and it was subsequently transferred into the Low Pressure Die Casting (LPDC) holding furnace. The pressure in the LPDC holding furnace was initially 25602kPa and no stirring was applied. The melt was injected to cylinder head moulds with cycle time of 18002s. Samples were cut at the thin and thick sections of the cylinder head to analyze the effects of cooling rate on grain refinement. Fading of grain refiner was studied for the period of 402h. Hardness testing and microstructural observation were conducted. The results showed that addition of Al–Ti granulated flux for 0.056 and 0.07802wt.% Ti, increased the hardness and lowered the dendrite arm spacing (DAS). The refinement is more significant in thin samples, because of the assistance by higher cooling rate. The mechanism of grain refinement of Al–Ti granulated flux in AA333 alloy is a combination of high growth restriction factor by the large amount of solute and nucleation by Al 3 Ti particles. Fading of grain refiner was detected by the increase in DAS, the lowering of Ti concentration in the melt and the reduction in hardness and strength. The pressure utilized in LPDC was not adequate to give stirring effect to prevent fading. [本文引用: 1] [214] S.H. Rodríguez, R.E.G. Mater. Manuf. Process. 27(6), 599(2012) URL In present article, influence of grain refinement (Al5Ti1B) on the morphology of iron-rich 尾-phase intermetallic compounds in Al-12Si-Cu-Mg alloy having two addition levels of iron as alloying element has been reported. The mechanical properties such as tensile strength and ductility of base alloy and the grain refined alloys were measured. The effect of grain refiner on quality index of the alloys on the basis of mechanical properties was also assessed. It was observed that the higher iron content in an alloy results in coarser iron-rich 尾-phase intermetallic compounds in more quantity than low iron content alloy. The addition of grain refiner in an alloy modifies the distribution of iron-rich intermetallic compounds and changes their morphology from needle to blocky and Chinese script form besides reducing their average length and aspect ratio, which in turn improves the mechanical properties significantly. Further, it was observed that alloy with high iron (2.68聽wt%) content required more quantity of grain refiner for better refinement and mechanical performance than the alloy with low iron content. Addition of grain refiner also increased the quality index of the low level iron alloy by about 50%, while that in case of high level iron alloy increased up to 136%. [本文引用: 1] [215] O. Fakhraei, M. Emamy, H. Farhangi, A 560, 148 (2013) [本文引用: 1] [216] M. Alipour, M. Emamy, Mater. Des. 32(8), 4485(2011) URL In this study the effect of Al–5Ti–1B grain refiner on the structural characteristics and hardness of Al–12Zn–3Mg–2.5Cu aluminum alloy has been investigated. The alloy was produced by modified strain-induced melt activation (SIMA) process. Reheating condition to obtain a fine globular microstructure was optimized. The specimens subjected to deformation ratio of 40% (at 30002°C) and various heat treatment times (5–4002min) and temperature (550–62002°C) regimes were characterized in this study. Microstructural study was carried out on the alloy by the use of optical and scanning electron microscopy (SEM) in both unrefined and Ti-refined conditions. The results showed that for the desired microstructures of the alloy during SIMA process, the optimum temperature and time are 57502°C and 2002min respectively. The hardness test results of the alloy also revealed that T6 heat treatment is more effective in hardness enhancement of all specimens in comparison with SIMA processing. [本文引用: 1] [217] S.S. Ebrahimi, J. Aghazadeh, K. Dehghani, M. Emamy, S. Zangeneh, Mater. Sci. Eng. A 636, 421 (2015) [本文引用: 1] [218] G.V. Kumar, B.S. Murty, M. Chakraborty, J. Alloys Compd. 396(1), 143(2005) URL Al–5Ti–0.8C and Al–5Ti–1.2C master alloys have been successfully prepared by reaction of K 2 TiF 6 salt and graphite powder with molten Al. While the Al–5Ti–0.8C consists of some TiAl 3 particles in addition to TiC particles in the Al matrix, the Al–5Ti–1.2C master alloy revealed the presence of only TiC particles. The grain refining efficiency of these two master alloys has been compared with that of the conventional Al–5Ti–1B master alloy on Al and Al–7Si alloy at different addition levels. Al–5Ti–1.2C master alloy was found to be the most efficient grain refiner for Al amongst the grain refiners studied. Even in case of Al–7Si alloy, the Al–5Ti–0.8C and Al–5Ti–1.2C master alloys performed better than conventional Al–5Ti–1B master alloy. However, the Al–5Ti–1.2C master alloy shows poor response to grain refinement in Al–7Si alloy at higher addition levels than the Al–5Ti–0.8C master alloy, indicating poisoning. [本文引用: 1] [219] Z.Y. Zhao, R.G. Guan, X.H. Guan, J. Zhang, X.P. Sun, H.N. Liu, Mater. Manuf. Process. 30(10), 1223(2015) [本文引用: 1] [220] J.A. Lozano, B.S. Peña, Scr. Mater. 54(5), 943(2006) URL The modification of eutectic die castings of Al–Si alloys using strontium and/or titanium refinement was studied. The results show that the most adequate microstructure is obtained from the alloy modified with the highest percentage of Ti. [本文引用: 1] [221] K. Qiu, R.C. Wang, C.Q. Peng, N.G. Wang, Z.Y. Cai, C. Zhang, Trans. Nonferr. Met. Soc. China 25(11), 3546 (2015) URL The effects of Mn and Sn on the microstructure of Al–7Si–Mg alloy modified by Sr and Al–5Ti–B were studied. The results show that the columnar dendrites structure is observed with high content of Sr, indicating a poisoning effect of the Al–5Ti–B grain refinement. In addition, Sr intermetallic compounds distribute on the TiB2particles, which agglomerate inside the eutectic Si. The mechanism responsible for such poisoning was discussed. The addition of Mn changes the morphology of iron intermetallic compounds fromβ-Al5FeSi toα-Al(Mn,Fe)Si. Increasing the amount of Mn changes the morphology ofα-Al(Mn,Fe)Si from branched shape to rod-like shape with branched distribution, and finally convertsα-Al(Mn,Fe)Si to Chinese script shape. The microstructure observed by transmission electron microscopy (TEM) shows that Mg is more likely to interact with Sn in contrast with Si under the effect of Sn. Mg2Sn compound preferentially precipitates between the Si/Si interfaces and Al/Si interfaces. [本文引用: 1] [222] L. Bolzoni, M. Nowak, N.H. Babu, J. Alloys Compd. 623, 79(2015) URL In a recent work we reported that the combined addition of Nb and B has higher grain refinement effectiveness than the addition of these individual chemical elements. In this work a commercial Al–12Si–0.6Fe–0.5Mn alloy is employed to fully characterise the refining potency of Nb–B inoculation. It is found that the addition of Nb powder and KBF4 flux to Al–Si melt introduces potent nuclei with low lattice mismatch with the primary α-Al dendrites (i.e. low undercooling) and, thus, promote noteworthy grain refinement through heterogeneous nucleation. Nb–B inoculation induces a significant increment of the volumetric number of grains, decreases the undercooling needed for solidification and makes the grain size of the Nb–B inoculated material less sensitive to the cooling rate employed to solidify the material. [本文引用: 1] [223] G. Faraji, P. Yavari, S. Aghdamifar, M.M. Mashhadi, J. Mater. Sci. Technol. 30(2), 134(2014) [本文引用: 1] [224] K. Rodak, J. Pawlicki, J. Mater. Sci. Technol. 27(11), 1083(2011) [本文引用: 1] [225] H.S. Liu, B. Zhang, G.P. Zhang, J. Mater. Sci. Technol. 27(1), 15(2011) [本文引用: 1] [226] J.B. Lin, X.Y. Wang, W.J. Ren, X.X. Yang, Q.D. Wang, J. Mater. Sci. Technol. 32, 783(2016) [本文引用: 1] [227] H. Barekatain, M. Kazeminezhad, A.H. Kokabi, J. Mater. Sci. Technol. 30(8), 826(2014) [本文引用: 1] [228] R. Adalarasan, M. Santhanakumar, A.S. Sundaram, J. Mech. Sci. Technol. 28(1), 301(2014) URL Friction welding is a solid state joining process in which the quality of welded joint is influenced by the input parameter setting. The objective of the present study is to conduct experimental investigation of the bond strength and hardness of the friction welded joints involving AA 6061 and AA 6351 alloys by conducting experiments designed by Taguchi鈥檚 L 9 orthogonal matrix array. A systematic approach becomes essential to find the optimal setting of friction welding parameters. Hence a new approach named grey-principal component analysis (G-PCA) is presented in which the principal component analysis (PCA) is used to generate weights for the grey relational coefficients obtained in the grey relational analysis (GRA). The results of the confirmation experiment conducted with the optimal setting predicted by the G-PCA have shown improvements in the performance characteristics. Hence G-PCA can be used for experimental welding optimization. [本文引用: 1] [229] A. Takayama, X. Yang, H. Miura, T. Sakai, Mater. Sci. Eng. A 478(1), 221 (2008) PMID:1988972 URL Annealing processes in copper, processed by multi-directional forging to strains of 07 02=020.4–6.0 at 30002K, were studied at 503–57302K. Strain-induced ultrafine-grained copper shows mainly grain coarsening behaviour, which is categorized in three stages, i.e. (1) an incubation period, (2) a rapid and limited grain growth and finally (3) a classical (normal) grain growth. That is, in situ or continuous static recrystallization (cSRX) takes place in stages 1 and 2. The annealing characteristics and the mechanisms of cSRX are discussed with reference to those of conventional discontinuous SRX (dSRX). [本文引用: 1] [230] W. Blum, Y.J. Li, Y. Zhang, J.T. Wang, Mater. Sci. Eng. A 528(29), 8621 (2011) URL Pure Cu was subjected to severe plastic predeformation by p = 1, 2, 4, 8, 16 and 24 passes of equal channel angular pressing (ECAP) on route B C at ambient temperature and subsequently tested in uniaxial compression parallel to the extrusion direction at constant rate or constant stress and temperatures from ambient temperature up to 418 K. The maximum compressive strength of the ECAPed Cu varies in a systematic fashion with p, until a steady state is finally reached between p = 8 and 16 where the rate sensitivity of flow stress is maximal. The results are quantitatively interpreted in terms of the boundary structure, considering the superposition of hardening due to refinement of low-angle boundaries and softening due to enhanced thermal recovery at high-angle boundaries. Beyond the maximum the compressive strength declines with strain for relatively low rate and/or elevated temperature of compression. This is explained by dynamic grain coarsening towards the new steady state developing in compression. [本文引用: 1] [231] R.Z. Valiev, Y.V. Ivanisenko, E.F. Rauch, B. Baudelet, Acta Mater. 44(12), 4705(1996) URL Structural evolutions in an Armco iron subjected to severe plastic deformation by torsion under high pressure are anlysed with conventional and high resolution electron microscopes. The substructure observed at low strains appears to shrink with increasing deformation and transforms at very high strains into grain boundaries. The resulting grain size decreases down to a constant submicrometric value. Meanwhile, the material strength, as revealed by micro hardness measurements, levels out. Dislocation densities and internal stress levels are used to discuss the structural transformations. Hydrostatic pressure and deformation temperature are believed to modify the steady-state stress level and structural size by impeding the recovery processes involving diffusion. [本文引用: 1] [232] T.G. Langdon, Mater. Sci. Eng. A 462(1), 3 (2007) URL Equal-channel angular pressing (ECAP) is a convenient processing tool for introducing very significant grain refinement, typically to the submicrometer level, in a wide range of metals. It is shown by experiment that processing by ECAP produces very similar microstructures in single crystals and in polycrystalline materials. Thus, after a single ECAP pass, aluminum single crystals and polycrystalline high-purity aluminum both exhibit microstructures consisting of bands of elongated subgrains and the experiments on single crystals have established unambiguously that these bands lie with their longer axes oriented parallel to the primary slip system. A model for grain refinement is developed incorporating the major experimental observations. Calculations of the shearing patterns for different processing routes lead to the conclusion that an equiaxed microstructure is achieved most rapidly in ECAP when slip occurs on three orthogonal planes over a wide range of angles: an example is route B C where the sample is rotated by 90掳 in the same sense about the longitudinal axis after every pass through the ECAP die. [本文引用: 1] [233] Y. Iwahashi, Z. Horita, M. Nemoto, T.G. Langdon, Acta Mater. 45(11), 4733(1997) URL Experiments were conducted to investigate the development of an ultra-fine grain size during equal-channel angular (ECA) pressing of high purity aluminum with an initial grain size of 651.0 mm. The results show that, under ECA pressing conditions giving a strain of 651.05 on each passage through the die, the microstructure is reasonably homogeneous after a single pressing and consists of parallel bands of elongated subgrains, having an average length of 654 μm, and these subgrains are further divided by boundaries with very low angles of misorientation. Repetitive pressings were conducted on the same samples, up to a total of 10 passages through the die, with the samples pressed either without rotation (route A) or after rotating through 180° between each pressing (route C). It is demonstrated that the misorientations of the subgrain boundaries increase with repetitive pressings until ultimately both routes lead to a similar equiaxed ultra-fine grain size of 651 μm after 10 pressings, but the microstructural evolution is enhanced using route C where there is a more rapid transition into an array of high angle grain boundaries. The results suggest that, at least for high purity aluminum, an ultra-fine microstructure close to optimum may be obtained after only 4 pressings provided the sample is rotated through 180° between each pressing. [本文引用: 1] [234] A. Yamashita, D. Yamaguchi, Z. Horita, T.G. Langdon, Mater. Sci. Eng. A 287(1), 100 (2000) [本文引用: 1] [235] B. Tolaminejad, K. Dehghani, Mater. Des. 34, 285(2012) URL In the present research, equal channel angular extrusion (ECAE) of commercial purity aluminum (1070) was conducted using route B C . For ECAE processing a proper die set was designed and constructed. Electron backscatter diffraction (EBSD) and X-ray diffraction (XRD) analyses were used to evaluate the microstructure and texture of the extruded materials. The results reveal two distinct processing regimes: from 1 to 4 passes the microstructure evolves from elongated subgrains to a rather equiaxed array of ultrafine grains and from 4 to 8 passes there is no strict change in the average grain size. The boundary misorientation angle and the fraction of high-angle boundaries increase rapidly up to 4 passes and at a slower rate from 4 to 8 passes. Also, the variation of hardness and yield stress with number of extrusion was documented up to 8 passes. The present results showed that first ECAE pass has resulted in enhancement of mechanical properties more than four times over the annealed condition. Further ECAE processing has resulted in slight improvement. Based on two strengthening mechanisms, variations of the strength as a function of the pass numbers were related to the calculated dislocation densities and the average boundary spacing. [本文引用: 1] [236] J. Xu, X.C. Zhu, L. Shi, D.B. Shan, B. Guo, T.G. Langdon, Adv. Eng. Mater. 17(7), 1022(2015) URL A very high-purity (99.999%) aluminum was processed by equal-channel angular pressing (ECAP) at room temperature through 1 to 8 passes using a die with a channel angle of 90°. Analysis shows that processing by ECAP produces an ultrafine-grained (UFG) structure with a grain size of ≈1.365μm and with microhardness and microstructural homogeneity. The mechanical properties and the fracture behavior were evaluated using micro-tensile testing after ECAP processing. A micro-forming process was used to fabricate a micro-turbine at ambient temperature and subsequent examination demonstrated that UFG pure aluminum gives higher strength and more uniform mechanical properties by comparison with conventional coarse-grained pure aluminum. The results confirm the very significant potential for using UFG pure aluminum for micro-forming at ambient temperature. [本文引用: 1] [237] C. Xu, Z. Horita, T.G. Langdon, Mater. Sci. Eng. A 528(18), 6059 (2011) URL Samples of an Al–1% Mg solid solution alloy were processed by equal-channel angular pressing (ECAP) at room temperature for totals of 1–12 passes and the microstructures of the processed samples were examined using orientation imaging microscopy. The results demonstrate the alloy achieves a reasonably stable microstructure after 6 passes through the ECAP die with an ultimate equilibrium grain size of 65700 nm. Measurements show both the fraction of high-angle boundaries and the average boundary misorientation increase with increasing numbers of passes up to 6 passes but thereafter there is only a minor additional increase up to 12 passes. [本文引用: 1] [238] Y. Nishida, H. Arima, J.C. Kim, T. Ando, Scr. Mater. 45(3), 261(2001) [本文引用: 1] [239] A. Azushima, K. Aoki, Mater. Sci. Eng. A 337(1), 45 (2002) URL The development of an ultrafine grain size during repeated shear deformation of side extrusion of ultra-low-carbon steel with an initial grain size of approximately 150 μm and a tensile strength of 200 MPa was investigated. Experiments were carried out at room temperature and at a low extrusion speed of 2 mm min 611 . Repeated side extrusions with lateral pressure were carried out for up to ten passes for the specimen extruded without rotation (route A). Uniform shear deformation was obtained by the side extrusion with the lateral pressure and an equivalent strain after a single side extrusion was 1.15. After ten passes of the side extrusions, an ultrafine-grained steel with a grain size of 0.5 μm×0.2 μm was realized and the tensile strength was over 1000 MPa. [本文引用: 1] [240] K. Nakashima, Z. Horita, M. Nemoto, T.G. Langdon, Mater. Sci. Eng. A 281(1), 82 (2000) URL A multi-pass facility was fabricated for equal-channel angular (ECA) pressing which gave a total strain of approximately 5 on a single passage through the die. Experiments on high purity aluminum show that, when comparisons are made at the same total strains, both the hardness and the evolution of the microstructure are identical when using the multi-pass facility or when repetitively pressing samples through a standard die containing a single shearing plane. [本文引用: 1] [241] G.J. Raab, R.Z. Valiev, T.C. Lowe, Y.T. Zhu, Mater. Sci. Eng. A 382(1), 30 (2004) URL In this paper, we report a new severe plastic deformation (SPD) technique, which combines equal channel angular pressing (ECAP) with Conform, to process ultrafine grained (UFG) materials in a continuous manner. ECAP in its original form can only process short metal bars and is labor intensive. Conform is a technique that has been used to continuously form metals into various shapes. By combining these two techniques, we were able to produce UFG structures in an Al wire and to significantly increase its strength. [本文引用: 1] [242] D.J. Alexander, J. Mater. Eng. Perform. 16(3), 360(2007) URL Several new concepts for possible methods of severe plastic deformation (SPD) of bulk quantities of materials are presented. The first of these are variations of equal channel angular extrusion (ECAE) in which the conventional fixed die is replaced by rotating tools, for the inner die corner, the outer die corner, or both corners. Other methods share some characteristics of ECAE in that they use shearing strains to deform the material; these are reversed shear spinning and transverse rolling. Deformation sequences for a cylindrical or annular workpiece that deform the workpiece while eventually restoring the initial workpiece geometry can be performed by numerous processes. These techniques can be used to accumulate high strains by repeated deformation cycles. These methods offer possible alternatives to ECAE and high-pressure torsion, with potential benefits that include different and larger workpiece geometries, simplified tooling design, lower tooling loads, ease of lubrication, automated or reduced part handling, and, in some cases, potentially continuous operation. It is hoped that these suggestions will prompt new examination of alternative methods for SPD. [本文引用: 1] [243] G. Faraji, A. Babaei, M.M. Mashhadi, K. Abrinia, Mater. Lett. 77(19), 82(2012) URL A new severe plastic deformation (SPD) technique entitled PTCAP is proposed to be suitable for deforming cylindrical tubes to extremely large strains. The tube constrained between mandrel and die is pressed by a first punch into a tubular angular channel with two shear zones and the diameter of the tube is increased and then pressed back to the initial dimension by the second punch. This process was applied to a commercially pure copper and a significant grain refinement resulted to a mean grain size of 150–30002nm was achieved even after a single pass PTCAP. Finite element (FE) results showed that the equivalent plastic strain of about 302±020.05 could be achieved at the end of the second cycle of PTCAP while it is about 302±020.4 at the end of the recently developed TCAP process. So, there is excellent strain homogeneity through the length and thickness of the PTCAP processed tube. Also, from the numerical results it was obtained that 57% lower loads are required for the PTCAP compared to TCAP. So, this new SPD process has two important advantages of excellent strain homogeneity and requiring a lower load. [本文引用: 1] [244] G. Sha, J.H. Li, W. Xu, K. Xiad, W.Q. Jie, S.P. Ringer, Mater. Sci. Eng. A 527(20), 5092 (2010) URL The use of high-temperature (33002°C) equal-channel angular pressing (ECAP) is demonstrated to promote precipitation of a fine and uniform dispersion of the FCC β 1 phase in an Mg–Nd–Gd–Zn–Zr alloy. Significantly, this process induces a hardening reaction in the alloy, where isothermal ageing at this temperature leads only to softening. The evolution of microstructure is characterized using transmission electron microscopy and scanning electron microscopy. The nucleation and growth of precipitates during the high-temperature ECAP are discussed. This research highlights a new approach to engineer precipitate microstructures via the application of severe plastic deformation so as to extend the property space of high-temperature Mg alloys. [本文引用: 1] [245] G.F. Zhang, X. Sauvage, J.T. Wang, N. Gao, T.G. Langdon, J. Mater. Sci. 48(13), 4613(2013) URL A Cu-11.802wt% Al alloy was quenched in iced water from a high temperature (85002°C) to introduce a martensitic phase and then the alloy was processed using quasi-constrained high-pressure torsion (HPT). The micro-hardness and the microstructures of the unprocessed and severely deformed materials were investigated using a wide range of experimental techniques (X-ray diffraction, optical microscopy, scanning electron microscopy, transmission electron microscopy, and high- resolution TEM). During HPT, a stress-induced martensite–martensite transformation occurs and an $$\alpha^{\prime}_{1}$$ martensite phase is formed. In the deformed material, there are nanoscale deformation bands having high densities of defects and twins in the $$\alpha^{\prime}_{1}$$ martensite. It was observed that a high density of dislocations became pinned and accumulated in the vicinity of twin boundaries, thereby demonstrating a strong interaction between twin boundaries and dislocations during the HPT process. [本文引用: 1] [246] M. Kawasaki, T.G. Langdon, Mater. Sci. Eng. A 498(1), 341 (2008) URL In processing by high-pressure torsion (HPT), it is possible to continuously strain in a forward direction in monotonic HPT (m-HPT) or to reverse the direction of straining in cyclic HPT (c-HPT). Experiments were conducted to compare the effects of torsionally straining high-purity Al using m-HPT and c-HPT through up to 4 turns at room temperature under a pressure of 6.0聽GPa. The appearance of the microstructural damage was examined on the surface of each disk and values of the Vickers microhardness were recorded both along disk diameters and over the total surfaces to permit the construction of color-coded contour maps. Although the inhomogeneities in the microstructures decreased with increasing numbers of turns in m-HPT, the experiments show that microstructural evolution is slower when using c-HPT. It is concluded that reversals in the direction of straining during HPT processing provide an opportunity for manipulating the hardness values attained in HPT. [本文引用: 2] [247] A.P. Zhilyaev, K. Ohishi, T.G. Langdon, T.R. McNelley, Mater. Sci. Eng. A 410(12), 277 (2005) [本文引用: 1] [248] G. Sakai, Z. Horita, T.G. Langdon, Mater. Sci. Eng. A 393(1), 344 (2005) URL An aluminium-2024 alloy was prepared in the annealed and the solution treated conditions. Specimens in both conditions were processed by high-pressure torsion (HPT) at temperatures of either 293 or 673 K giving grain sizes for both conditions of 650.15 and 650.30 μm for these two processing temperatures, respectively. Tensile testing was conducted at 673 K and superplastic elongations were achieved in both the annealed and the solution treated samples, with a maximum elongation of 570% for an annealed specimen processed by HPT at 293 K. To check on the significance of microstructural inhomogeneities across the HPT disks, samples were machined either from the center of the disks or from an off-center position. The experimental results show that higher elongations are achieved at the off-center position. The similarity between the annealed and the solution treated conditions after HPT is attributed to a dissolution of precipitates during HPT processing. [本文引用: 1] [249] X. Wang, M. Nie, C.T. Wang, S.C. Wang, N. Gao, Mater. Des. 83, 193(2015) URL High-pressure torsion (HPT) was used to produce hypoeutectic Al–7Si alloy samples having a range of microstructures to investigate the effect of the grain refinement on its corrosion behavior in 3.502wt.% NaCl solution for the first time. Optical microscopy measurements reveal that with the HPT processing increased from 1/4 to 10 revolutions under an applied pressure of 6.002GPa, brittle coarse silicon particles and intermetallic phases were effectively broken into ultrafine-grained particles and redistributed homogeneously into the Al-rich matrix. Open-circuit potential and polarization curves results exhibit that corrosion resistance of the Al–7Si alloy in NaCl solution was significantly enhanced upon high torsion strains, with corrosion rate reduced from 7.4102μm02y 611 for the as-received sample to 1.6802μm02y 611 for the 10-turn processed sample. Electrochemical impedance spectroscopy analysis combined with characterization of the corroded samples using scanning electron microscopy and energy dispersive X-ray spectroscopy indicates that the enhancement in corrosion performance of the Al–7Si alloy is due to the breakage of coarse silicon particles and intermetallic phases, the microstructure homogeneity and the increased HPT-induced active sites. It is demonstrated that microstructure refinement through HPT processing can significantly improve both microhardness and corrosion properties of the Al–7Si alloy. [本文引用: 2] [250] S.H. Lee, Y. Saito, T. Sakai, H. Utsunomiya, Mater. Sci. Eng. A 325(1), 228 (2002) URL Accumulative roll-bonding (ARB) process is an intense plastic deformation process that has been performed for a 6061 aluminum alloy to develop ultra-fine grains below 1 渭m in diameter and to improve mechanical properties. The ARB process up to eight cycles is performed at ambient temperature under unlubricated conditions. The ultra-fine grains surrounded by clear boundaries begin to appear at the third cycle, and the specimen after eight cycles shows a microstructure covered with ultra-fine grains with an average diameter of 310 nm. The tensile strength of the ARB processed 6061 alloy increases with the number of ARB cycles (equivalent total strain), and after eight cycles it reaches the maximum of 363 MPa, which is about three times of the initial. On the other hand, the elongation drops abruptly at the first cycle, above which it decreases progressively with the number of ARB cycles. The hardness of the specimens ARBed by one, three and five cycles varies inhomogeneously in the thickness direction; having peak values near the surface and the center. This is due to the redundant shear strain and wire brushing. The results show that the ARB process is effective for grain refinement and strengthening of 6061 alloy. [本文引用: 1] [251] S.O. Gashti, J. Alloys Compd. 658, 854(2016) URL In this work, the effects of grain size and dislocation density on strain hardening behavior of ultrafine grained (UFG) AA1050 produced by 9-cycle accumulative roll bonding (ARB) process were studied. Transmission electron microscopy (TEM) micrographs indicated that UFG structure with the average grain size of 270±11nm achieved after nine cycles of ARB. The dislocation density which was measured by microhardness value increased with an increase in the number of ARB cycles. The dislocation density increased from about 7×10 12 m 612 for annealed sample to above 2×10 13 m 612 after nine cycles ARB. Tensile tests results revealed that with increasing the number of ARB cycles, mechanical properties of the samples improved. The Hollomon analysis was used to investigate the strain hardening behavior of AA1050 samples. The results showed that the strain hardening exponent decreased with increasing the number of ARB cycles. All ARBed samples showed high strain hardening rate in initial stage of deformation, which increased with increasing dislocation density and grain refinement. [本文引用: 2] [252] X. Huang, N. Tsuji, N. Hansen, Y. Minamino, Mater. Sci. Eng. A 340(1), 265 (2003) URL The microstructure in commercial purity aluminum deformed from medium to high strain ( ε vM =1.6–6.4) by accumulative roll-bonding (ARB) at 473 K was quantitatively examined by transmission electron microscopy. It was found that a sub-micrometer lamellar structure characterizes the microstructure at high strains ( ε vM >1.6), and that the lamellar boundary spacing decreases and the misorientation across the lamellar boundaries increases with increasing rolling strain. This characteristic evolution has also been observed during conventional cold-rolling of commercial purity aluminum. However, a comparison between the two processes shows a significant difference in the evolution of the microstructural parameters. These differences are discussed based on the different processing conditions characterizing ARB and conventional rolling, respectively. [本文引用: 2] [253] J.W. Yeh, S.Y. Yuan, C.H. Peng, Metall. Mater. Trans. A 30(9), 2503 (1999) URL A reciprocating extrusion (RE) process has been developed for producing A1-12 wt pct Si bulk alloys with fine and uniform microstructures and superior properties. Two starting forms were used: disks produced by the hammer-and-anvil method and cast billets produced by casting. Variations of micro-structure and mechanical properties with the number of extrusion passes are investigated for these two starting forms. The results show that the porosity along the interfaces between the rapidly solidified layers could be completely eliminated to give a sound matrix. The Si-phase particles in both cases could be refined and distributed uniformly. The strength and ductility of all specimens are also enhanced, until the microstructure reaches an optimum state, as the number of extrusion passes increases. The tensile properties of the rapidly solidified Al-Si alloys are found to be superior to those of ingot-processed alloys, due to the inherent finer particles produced by rapid solidification. The mechanism for the improvement of the microstructures and properties is also discussed. [本文引用: 1] [254] S.Y. Yuan, J.W. Yeh, C.H. Tsau,Mater. Trans. JIM 40(3), 233 (1999) URL A reciprocating extrusion process was used to work the 2024 aluminum alloy ingots at 723 K so that refined microstructures and improved mechanical properties could be obtained. The number of extrusion passes was monitored to see their effect upon microstructure and properties. As the number of extrusion passes was increased, the grains became finer but after 5 passes a limiting grain size of about 2 渭m was approached. On the other hand, the dispersoids and inclusions became smaller and more equiaxed, and more uniform in the matrix up to 20 passes. Grain refinement results from the repeated partial recrystallization and the increased number of inclusions and dispersoids to enhance recrystallization during reciprocating extrusion. The refinement of dispersoids and inclusions was attributable to three possible cracking mechanisms: bending mechanism, short-fiber loading mechanism and shear mechanism depending on the shape of a particle. Redistribution of particles is driven by the cyclic plastic flows of extrusion and compression during reciprocating extrusion.In the first 5 passes, yield strength and ultimate tensile strength showed a small decrease but elongation increased much. After more passes, their variations became small. The overall loss of strength is about 10% and the increase of elongation is about 54%. The combination of strength and elongation is still significantly improved. The decrease in strength is mainly due to the coarseness of S鈥 precipitates which is related to the loss of quenched vacancies. The large improvement of elongation is mainly attributable to the refinement of inclusions. [本文引用: 2] [255] M. Richert, Q. Liu, N. Hansen, Mater. Sci. Eng. A 260(1), 275 (1999) URL ABSTRACT Polycrystalline pure aluminium (99.99%) has been deformed at room temperature by the Cyclic-Extrusion–Compression (CEC)-method to strains in the range 0.9–60 (1–67 cycles). At different strains, the microstructure and local crystallography have been characterised in particular by transmission electron microscopy. It has been found that the microstructure develops from a cell block structure into an almost equiaxed structure of cells and subgrains, that the spacing between the boundaries subdividing the structure is almost unaffected by the strain and that the misorientation across these boundaries increases with the strain over the whole strain range. At the largest strain, the average misorientation across the deformation induced boundaries is 6525°. The flow stress in compression is measured after the cyclic deformation and it is found that the flow stress increases with strain towards a saturation level which is reached at a relatively low strain. The discussion comprises the effect of deformation mode and plastic strain over a large strain range on the microstructural evolution and mechanical behaviour of aluminium. [本文引用: 1] [256] G. Liu, L.E. Murr, C.S. Niou, J.C. Scr. Mater. 37(3), 355(1997) URL In the as-produced condition the room temperature strength (similar to 6 Gpa) of Textron Specialty Materials' 50 mu m CVD SiC fiber represents the highest value thus far obtained for commercially produced polycrystalline SiC fibers. To understand whether this strength can be maintained after composite processing conditions, high temperature studies were performed on the effects of time, stress, and environment on 1400 degrees C tensile creep strain and stress rupture on as-produced, chemically vapor deposited SiC fibers. Creep strain results were consistent, allowing an evaluation of time and stress effects. Test environment had no influence on creep strain but 1 hour annealing at 1600 degrees C in argon gas significantly reduced the total creep strain and increased the stress dependence. This is attributed to changes in the free carbon morphology and its distribution within the CVD SiC fiber. For the as-produced and annealed fibers, strength at 1400 degrees C was found to decrease from a fast fracture value of 2 GPa to a 100-hour rupture strength value of 0.8 GPa. In addition a loss of fast fracture strength from 6 GPa is attributed to thermally induced changes in the outer carbon coating and microstructure. Scatter in rupture times made a definitive analysis of environmental and annealing effects on creep strength difficult. Published by Elsevier Science Ltd. [本文引用: 1] [257] K.V. Jata, S.L. Semiatin, Scr. Mater. 43(8), 743(2000) URL Friction stir welding (FSW) is a solid state joining process 1,2,3 that uses a rapidly-rotating, non-consumable high strength tool-steel pin that extends from a cylindrical shoulder (Figure 1). The workpieces to be joined are firmly clamped to a worktable; the rotating pin is forced with a pre-determined load into them and moved along the desired bond line. Frictional heating is produced from the rubbing of the rotating shoulder with the workpieces, while the rotating pin deforms (i.e. 'stirs') the locally-heated material. To produce a high integrity defect-free weld, process variables (RPM of the shoulder-pin assembly, traverse speed, the downward forging force) and tool pin design must be chosen carefully. FSW can be considered as a hot-working process in which a large amount of deformation is imparted to the workpiece through the rotating pin and the shoulder. Such deformation gives rise to a weld nugget (whose extent is comparable to the diameter of the pin), a thermomechanically-affected region (TMAZ) and a heat-affected zone (HAZ). Frequently, the weld nugget appears to comprise equiaxed, fine, dynamically recrystallized grains whose size is substantially less than that in the parent material. The objective of the present research was to develop a basic understanding of the evolution of microstructure in the dynamically recrystallized region and to relate it to the deformation process variables of strain, strain rate, and temperature. Such a correlation has not been attempted before perhaps because of the difficulty in quantifying the process variables. To overcome such difficulties, recent work 4 to measure and model the local temperature transients during FSW was utilized, and an approximate method was employed to estimate the strain and strain rate in the weld nugget. [本文引用: 1] [258] J.Q. Su, T.W. Nelson, R. Mishra, M. Mahoney, Acta Mater. 51(3), 713(2003) URL The grain structure, dislocation density and second phase particles in various regions including the dynamically recrystallized zone (DXZ), thermo-mechanically affected zone (TMAZ), and heat affected zone (HAZ) of a friction stir weld aluminum alloy 7050-T651 were investigated and compared with the unaffected base metal. The various regions were studied in detail to better understand the microstructural evolution during friction stir welding (FSW). The microstructural development in each region was a strong function of the local thermo-mechanical cycle experienced during welding. Using the combination of structural characteristics observed in each weld region, a new dynamic recrystallization model has been proposed. The precipitation phenomena in different weld regions are also discussed. [本文引用: 1] [259] M. Cabibbo, H.J. Mater. Sci. Eng. A 460, 86 (2007) [本文引用: 1] [260] Z.Y. Ma, A.L. Pilchak, M.C. Juhas, J.C. Williams, Scr. Mater. 58(5), 361(2008) URL Friction stir processing (FSP) is a novel metal-working technique that provides microstructural modification and control in the near-surface layer of metal components. FSP of cast Al and Mg alloys resulted in the break-up of coarse dendrites and secondary phases, refinement of matrix grains, dissolution of precipitates and elimination of porosity, thereby improving the mechanical properties of the castings significantly. Cast Ti alloys do not contain harmful secondary phases and the dendritic structure is masked by the 尾 to 伪 allotropic transformation, but the coarse lamellar structure is undesirable. In this article, microstructure and particle refinement, accelerated dissolution of precipitates and alloy systems suitable for FSP were addressed. [本文引用: 2] [261] Y.F. Shen, R.G. Guan, Z.Y. Zhao, R.D.K. Acta Mater. 100, 247(2015) URL We elucidate the potential significance of nanosized, Al 3 (Sc, Zr), precipitates in obtaining ultrafine-grained structure in an Al–0.2Sc–0.1Zr alloy during the novel process, referred as accumulative continuous extrusion forming (ACEF). The grain size of the alloy was dramatically refined from 10002μm to 80002nm through continuous dynamic recrystallization (CDRX). The effectiveness of nanosized precipitates on CDRX was pronounced with increase in the ratio of the volume fraction ( F v ) to the diameter ( d ) of the Al 3 (Sc, Zr) precipitates. Nanosized Al 3 (Sc, Zr), precipitates promoted grain refinement through three mechanisms: (i) the precipitates facilitated retention of high dislocation density in the alloy by promoting the generation of dislocation and pinning dislocation slip, which increased the driving force for CDRX, (ii) promoted the formation of deformation bands, providing sites for activation of CDRX, and (iii) activated CDRX near the grain boundary. [本文引用: 2]