Precipitation hardened steels are widely used in various engineering fields due to their high strength, high fracture toughness, good ductility and ease of machinability. As two kinds of common precipitates, Cu-rich and NiAl phases play an important role on the mechanical properties of steels. The obvious effects of Mn on the precipitate evolution of Cu-rich phase and NiAl phase in steel have been observed respectively. However, the effect of Mn is complex, when Cu-rich phase and NiAl phase exist at the same time. This work aims to reveal the effects of Mn on the co-precipitation of Cu-rich phase and NiAl phase in steel. Fe-Cu-Ni-Al and Fe-Cu-Ni-Al-Mn steels were aged at 500 ℃ for different times after solution treatment at 900 ℃ for 2 h. Hardness testing indicates that by adding 2.17%Mn, Fe-Cu-Ni-Al-Mn steel shows a peak hardness of 420 HV which is 80 HV higher than Fe-Cu-Ni-Al steel (about 340 HV). And Fe-Cu-Ni-Al-Mn steel reaches the peak hardness at 1 h which is 1 h earlier as compared with Fe-Cu-Ni-Al steel at 2 h. Moreover, the peak hardness plateau of Fe-Cu-Ni-Al-Mn steel only lasts for 7 h which is far less than that of Fe-Cu-Ni-Al steel. All in all, the addition of Mn enhances the effect of precipitation hardening at early aging, and accelerates the whole process of precipitation hardening. Atom probe tomography (APT) results reveal that Mn increases the nucleation rate of precipitates at early ageing, accelerates the growing and coarsening of precipitates and then accelerates the separation of the Cu-rich phase and NiAl phase. This is due to Mn can reduce the energy for nucleation and accelerate the diffusion rate of elements in the matrix, while the partial substitution of Mn for Al in the NiAl phase can form point defects which can accelerate the diffusion rate of Cu in NiAl phase.
To investigate the formation mechanism of sink vortex during ladle teeming, the effects of some factors such as Coriolis force, the position of the ladle shroud and initial tangential velocity of the fluid on the vortex formation process have been studied using numerical simulation combined with experiments. In addition, the evolution tendencies of tangential and radial velocities of the fluid over radial position were studied at certain initial tangential velocity. The results show that as for fully settled fluid, Coriolis force is the major reason for sink vortex formation and the spinor near the shroud is the initial driving force. There is no obvious effect of the ladle shroud position on the critical height of vortex for fully settled fluid, while the critical height of vortex significantly decreases with increasing shroud eccentricity for the fluid with a certain initial velocity, and the tangential motion is the main driving force for vortex formation in this case. The initial tangential velocity affects the critical height significantly. The larger the initial angular velocity is, the earlier vortex occurs and the bigger the critical height of vortex is. As a result, keeping the fluid settled for some time is an effective measure to delay vortex during ladle teeming. The relationship between the start height of vortex (HSS) and initial angular velocity (ω) can be expressed as HSS=0.11+2.85ω-4.04ω2+1.95ω3, and that of the height of air column extending to shroud (HCS) and ω expressed as HCS=0.09+1.49ω-0.79ω2, both of the fitting degrees are higher than 0.99.
Among various hardening factors of steels, precipitation hardening has the least embrittlement vector value except grain refinement hardening. Giving full play to the precipitation hardening of microalloyed carbonitrides is an important aspect in the development of microalloyed high strength steels. Recently, the research on behaviors of precipitation and development of microalloyed high strength steels is mainly focused on these relatively simple microalloyed steels including single V, single Ti, Ti-V and Ti-Mo microalloyed steels, while paid less attention on complex microalloyed steels such as Ti-V-Mo steels. Therefore, it is expected to provide a theoretical basis and a practical significance for the development of Ti-V-Mo microalloyed high strength steel. Various hardening increments at different coiling temperatures were calculated. Meanwhile, the effect of coiling temperatures on yield strength and the influence of MC particles on uniform elongation were discussed by means of OM, EBSD, TEM, XRD and physical-chemical phase analysis. The results show that Ti-V-Mo steel has the best mechanical properties with ultimate tensile strength of 1134 MPa, yield strength of 1080 MPa, elongation of 13.2% and uniform elongation of 6.8% at coiling temperature of 600 ℃. The precipitation hardening increment was high to about 444~487 MPa due to the mass fraction of about 72.6% of total precipitates with a size of ?10 nm. In addition, precipitation hardening and grain refinement hardening are the main mechanisms to improve the strength of Ti-V-Mo steel, while the variation in precipitation hardening increment causes a significent difference in yield strength. With the coiling temperature increases from 500 ℃ to 600 ℃, the ultimate tensile strength and yield strength increase continuously, but the uniform elongation increases slowly instead of decreasing, which is mainly attributed to an increase of precipitation hardening increment.
Ni-based single crystal (SX) superalloys have been used as blades in aero-space industry and land-based applications due to their excellent high-temperature properties. However, residual strain is introduced into as-cast SX superalloy blades during the manufacturing process, such as casting, grinding or shot peening, and so on. Recrystallization (RX) occurs easily during subsequent high temperature heat treatment. In previous work, it is believed that RX has detrimental effect on the mechanical properties of SX superalloy. Furthermore, in order to improve the mechanical properties, more and more refractory elements, such as W, Re, Mo, Ta, are added into SX superalloys. However, so far, few reports about the effect of refractory elements on the RX in as-cast SX superalloys have been available. In the present work, the effect of Re and W on the RX behavior of as-cast Ni-based SX superalloy was studied. Single crystal superalloys with different Re and W contents were annealed at 1230~1330 ℃ after indened using Brinell hardnesstester. It is found that RX grains form at the surface under indentation and grow preferentially along the dendritic cores. Subsequent growth of RX is impeded by the residual coarse γ' and γ +γ' eutectics in the interdendritic regions. Both the volume fraction of γ +γ' eutectics and γ' solvus temperature are increased with the addition of Re and W, which are attributed to the increase of RX threshold temperature. For all SX superalloys studied in this work, RX area increases with the increase of annealing temperature due to the dissolution of γ' and γ+γ' eutectics. At the same annealing temperature, in comparison to Re, W shows more effect to inhibit RX growth. Additionally, SX superalloy containing both Re and W has the smallest RX area in the present experiments.
A Hf-containing Ni-based alloy was used as the interlayer alloy of TLP bonding for the 2nd (CMSX-4, as-cast condition) and 3rd (SXG3, standard heat treatment condition) generation Ni-based single crystal superalloys containing Re in this work, and the microstructure, composition and micro-hardness of bonding zone were characterized. The results show that the TLP bonding of CMSX-4 and SXG3 alloy were completed after bonded at 1290 ℃ in vacuum for 24 h. These TLP bonding process of CMSX-4 and SXG3 alloys can be explained well using classical TLP model. The diffusion affected zone was not observed during the TLP bonding process. In addition, the heat treatment process of CMSX-4 is shortened by 24 h resulted from the solid solution heat treatment of CMSX-4 alloy has been completed after the process of TLP bonding. The isothermal solidification stage of SXG3 alloy was also accelerated due to the precipitation of HfC at the bonding temperature, resulting in the reduced Hf concentration of Hf in the melting zone. This work also indicates that the interfacial stability of low angle grain boundaries can be investigated by the TLP bonding. The critical misorientation value for discontinuous precipitation of SXG3 alloy along TLP bonding grain boundaries by Hf-containing interlayer alloy was in between 10° and 17° after heat treatment at 1150 ℃.
A well known feature of ferromagnetic materials is the time dependent behavior of the magnetic polarization, i.e. magnetic viscosity, which arises from thermal activation over energy barriers. It is found that magnetic parameters, such as the fluctuation field (Hf) and the exchange interaction length (lex), have a close relationship with the microstructure of the materials. Therefore, investigation on magnetic viscosity is helpful to understand the coercivity mechanism of ferromagnetic materials. In this work, ingots with nominal composition Nd8.5Fe76Co5Zr3B6.5Dy1, Nd9.5Fe75Co5Zr3B6.5Nb1 and Nd9.5Fe75.4Co5Zr3B6.5Ga0.6 were prepared by arc-melting pure metals Nd, Fe, Co, Zr, Dy, Nb, Ga and Fe-B alloy in an argon atmosphere. A small portion of an ingot weighing about 5 g was re-melted in a quartz nozzle and ejected onto a rotating copper wheel in a range of 10~30 m/s. The annealing treatment was carried out at 690~710 ℃ for 4~5 min. Vibrating sample magnetometer (VSM), XRD and TEM were used to study magnetic viscosity behavior and exchange interaction for Nd2Fe14B/α-Fe nanocomposite permanent alloys. Furthermore, the relationship among exchange interaction, microstructure and magnetic property was discussed. For the nanocomposite Nd8.5Fe76Co5Zr3B6.5Dy1, Nd9.5Fe75Co5Zr3B6.5Nb1 and Nd9.5Fe75.4Co5Zr3B6.5Ga0.6 alloys, Hf and lex were obtaind by sweep rate measurement. The Hf were 4.80, 4.87 and 5.09 kA/m, and lex were 4.53, 4.41 and 4.20 nm for permanent Nd8.5Fe76Co5Zr3B6.5Dy1, Nd9.5Fe75Co5Zr3B6.5Nb1 and Nd9.5Fe75.4Co5Zr3B6.5Ga0.6 alloys, respectively. It suggested that the lex had a minor change. The Nd9.5Fe75Co5Zr3B6.5Nb1 alloy had the strongest exchange interaction among three alloys in this work. It is due to a refined microstructure and uniform distribution of grains. Furthermore, the behavior of the irreversible susceptibility (χirr) as a function of applied magnetic field (H) was investigated. A single sharp peak could be seen near coercive field in the χirr-H curve in three alloys, suggesting that the magnetization reversal was a uniform reversal process. The Nd9.5Fe75.4Co5Zr3B6.5Ga0.6 alloy exhibited a sharper and narrower peak, indicating a more rapid change in magnetization and a strong interaction between adjacent magnetic phases. Since exchange interaction of neighboring grains favors the nucleation of reversed domains, remanence enhancement is generally achieved at the expense of coercivity. Among three alloys, Nd9.5Fe75.4Co5Zr3B6.5Ga0.6 alloy showed the optimum magnetic properties, that is, the coercivity Hc=687.56 kA/m, the remanence Br=0.92 T, the maximum magnetic energy product (BH)max=120.88 kJ/m3. It was mainly due to consisting of well-coupled grains with near perfect alignment of the easy magnetization direction, which improved the remanence and maximum energy product.
Among the traditional cast magnesium alloys system, Mg-3Nd-1Zn alloy with high strength and heat resistance, has been widely applied in aeronautics such as in engine box and wing rib of airplane. In present research, the as-cast Mg-2.7Nd-0.6Zn-0.5Zr (NZ31) alloy was solution treated and then aged at temperatures ranging from 200 ℃ to 300 ℃. The microstructures and mechanical properties of the aged specimens especially at relatively high temperature (225~300 ℃) were systematically characterized by OM, SEM and TEM. A new kind of precipitates distributed in line was clearly found in the specimens aged at high temperature (225~275 ℃) for a short time (15~30 min), which corresponded to a significant enhancement of hardness and tensile strength at room temperature. The TEM results showed that the precipitates distributed in line had a composition of Mg12Nd and a granular shape, and mainly formed along the (0001)Mg basal plane. The special distribution of the granular Mg12Nd precipitates was effective barriers to the basal slips and may probably restrain the intergranular coordination during tensile deformation, leading to a large strengthening effect, that is, the yield strength and ultimate tensile strength, of the NZ31 alloy aged at 250~275 ℃ for 20~30 min, increased by about 70% and 29% respectively, in comparison with only the solutionized one.
Microsegregation is the unbalanced distribution of alloying element between solid and liquid phases in dendritic scale during solidification. The solute redistribution usually leads to the formation of brittle secondary phase, which is harmful to the workability and final mechanical properties of casting component. It has been accepted that fluid flow plays a critical role in mass transfer during solidification and thus altering the microsegregation pattern. High static magnetic field has been considered as an effective way to control the convection in solidification. In this work, the impact of the high static magnetic field on the microsegregation was investigated. Al-4.5Cu (mass fraction, %) alloy was directionally solidified from <001> seed crystal under various magnetic fields with a constant pulling rate of 50 μm/s and temperature gradient of 101 K/cm. OM and BSE were applied to characterize the microstructure of the solidified samples. The fraction of Al2Cu second phase was obtained by software analysis from the transverse and longitudinal sections. The results show that the Al-4.5Cu alloy solidifies in dendritic morphology. The formation of second phase is significantly affected by the magnetic field. Without magnetic field, the continuous network of second phase is observed at grain boundaries. In the presence of the magnetic field, the second phase is disconnected in the grain boundaries and dispersed in grains. The fraction of the second phase is reduced with the increase of the magnetic field. EDS area scan was carried out to measure the concentration of Cu solute in dendritic scale. Isoconcentration contour maps of Cu in the plane perpendicular to the primary dendrite trunk were drawn. The concentration profiles of Cu were plotted from the measured data and the effective partition coefficient ke was calculated. It is found that the redistribution of Cu solute in interdendritic region is greatly altered by the magnetic field. When the intensity of the magnetic field increases, the concentration profile and the ke decrease. The disturbance of the Cu solute in the plane perpendicular to the primary trunk suggests the existence of fluid flow in the interdendritic region. The above phenomena could be attributed to the dendritic scale thermoelectric magnetic convection (TEMC) as well as the second flow driven by the TEMC. The azimuthal TEMC and meridional second flow will bring about stirring in mushy zone and lead to the modification of solute transport during solidification process.
Ti-6Al-4V alloys are widely used in aero engine blades for their unique properties, such as high specific strength, high specific stiffness and high fatigue strength. Aero engine blades usually suffer a variety of cyclic loading during the period of services, which finally results in fatigue failure. Fatigue life of materials is known to highly depend on the surface quality. Consequently, more and more researches about the influence of machined surface roughness on the fatigue behavior of materials have been carried out in the last decades. However, there are less relevant results about the relationship between surface roughness and very high cycle fatigue (VHCF) properties of Ti-6Al-4V alloy. To investigate the effects of surface roughness on fatigue properties of Ti-6Al-4V alloy under very high cycle fatigue regimes, ultrasonic fatigue tests were conducted at the conditions of 20 kHz and stress ratio R1=-1 at room temperature in air. During ultrasonic fatigue testing, each specimen was water-cooled. The specimen surfaces were cut and grinded which gave different surface roughnesses. The surface roughness was characterized using profilometry. In order to explain the high dependence of stress-fatigue life curves on the surface roughness, an approach based on the finite element analysis of measured surface topography was proposed. The results show that the VHCF property of Ti-6Al-4V alloy was significantly affected by surface roughness. The critical flaw size was 0.49~1.10 μm when the ratio between spacing and height of circumferential grooves was between 2~10. When surface roughness was smaller than the critical flaw size, surface roughness exerted no influence on fatigue life. While surface roughness was greater than critical flaw size, fatigue life decreased with increasing surface roughness. Surface roughness played a more important role in long life regime than that in VHCF regime in which with the growth of surface roughness, the crack initiation site changed from single one to two or more ones, as well as changed from inside to subsurface. When the surface roughness was large enough, all cracks initiated from surface even in super long life regime.
The service environment faced by marine equipment and its key friction pair parts are much more severe than that on land surface. The life cycle and safety of the hydraulic and power transmission system, which directly get in touch with the seawater, depends largely on the tribological behavior of the components in the seawater. Titanium alloy is an ideal material used for ocean engineering, however due to its poor friction performance its life cycle may be shortened when working in the environment with friction and wear. In order to improve the tribological performance of titanium alloy in seawater, laser processing was used to build super hydrophobic with grid and dot micro-structure on Ti6Al4V alloy surface. Tribological performance was evaluated by HSR-2M high speed reciprocating friction test machine in artificial seawater, and compared with in water (distilled water). The results show that the friction coefficients and wear losses (volume) of super hydrophobic Ti6Al4V alloy surface are significantly smaller than that of the Ti6Al4V alloy substrate. The friction coefficients of surface with dot and grid reduced by 17.8% and 11.7%, and wear losses (volume) reduced by 36.8% and 57.5% respectively in artificial seawater. The friction coefficient of super hydrophobic Ti6Al4V alloy surface in artificial seawater is smaller than that in water while the wear loss has the opposite phenomena. The tribological performances of titanium alloy in artificial seawater are significantly improved by the build of super hydrophobic Ti6Al4V alloy surface.
Many components in secondary side of pressurized water reactors (PWRs) are made of carbon steels and low alloy steels. The corrosion products produced by the flow accelerated corrosion (FAC) of these components can deposite on the surface of steam generator (SG) tubes and decrease the heat transfer efficiency of SG tubes. Moreover, the enrichment of foreign ions (e.g. Cl- and Pb2+) occurs with the sedimentation of corrosion products and causes the local environment degradation, and thus accelerates the failure of SG tubes. In order to decrease the FAC of carbon steels and low alloy steels, pH controllers are often added to adjust the pH value of secondary water. The water chemistry environment of secondary side in PWRs has experienced various treatment techniques, such as phosphate treatment, all volatile treatment (AVT), morpholine (MPH) treatment, ethanolamine (ETA) treatment, and boric acid treatment. In comparision with AVT, ETA can significantly reduce the concentration of Fe in the steam-water phase region and water supply system because of its higher alkalinity and lower molar concentration in feedwater. Due to the high resistance to corrosion and stress corrosion cracking in high temperature and high pressure water, alloy 800 is often used as steam generator tubes in nuclear power plants and thus becomes increasingly attractive among researchers. However, few studies focus on the effect of ETA on the corrosion behavior of alloy 800 in high temperature and high pressure water. This work mainly aims to investigate the corrosion behavior of alloy 800 in NaOH and ETA solutions at 300 ℃ by potentiodynamic polarization curve, electrochemical impedance spectra (EIS), SEM and XPS. The electrochemical results demonstrate that the addition of ETA decreases the current density of anodic and cathodic reactions, and increases the corrosion potential of alloy 800. Besides, ETA addition significantly increases the resistance of inner oxide layer and makes the oxide film more compact, which increases the film resistance of alloy 800 in high temperature water. Through the morphology observation and composition analysis, it is found that ETA addition can promote the formation of Cr-rich layer and increase the ratio of chromium in the oxide films although the deposition of magnetite is enhanced on the surface of alloy 800. For stainless steels and nickel-based alloys in high temperature water, the Cr-rich oxide layer can inhibit the diffusion process of O and metal ions, and reduces the corrosion rates of alloys. Therefore, the corrosion resistance of alloy 800 is enhanced after ETA is added in high temperature water.
CeO2 is an important rare earth oxide and can be used in automotive exhaust three-way catalysts on the basis of its oxygen storage capability. Ion doping is an effective method to enhance the oxygen storage capability of CeO2. And when doping a cation whose size is smaller than Ce4+ and valence is lower than +4, it tends to evolve more defects. It is known that defects play important roles in enhancing the oxygen storage capability of CeO2. Therefore, In ion was selected as a dopant cation which matches above two factors of size and valence. In this work, a series of CeO2 with different content of In3+ were synthesized via a two-step process. The precursor was synthesized by a solvothermal method at 200 ℃ using a mixture solvent of (CH2OH)2 and H2O, as well as Ce(NO3)36H2O and In(NO3)3?4.5H2O as Ce and In sources, respectively. CeO2 was obtained after the precursor was calcined at 500 ℃ for 2 h in air. It was found that the solid solubility of In3+ in CeO2 was 1% (molar fraction). The doping of 1%In3+ in CeO2 almost had no impact on the morphology of multilayered structure. However, a second phase of small particles appeared and there were some changes of the morphology of multilayered structure when the concentration of In3+ increased further. The specific surface area of the 1%In3+ solid solution was 100 m2/g, which was th highest among all the samples, and undoped CeO2 (92 m2/g) ranked second. When the content of In3+ was above the solid solubility, i.e., 1%In3+, the specific surface area decreased. The low temperature oxygen storage capability could be improved from 3.6×10-4 mol/g for undoped CeO2 to 4.4×10-4 mol/g for 1%In3+-doped CeO2. When the In3+ content was greater than or equal to 3%, the low temperature oxygen storage capability decreased at the beginning, and then almost no change. Lattice parameter decreased and the concentration of Ce3+ and oxygen vacancy increased by the doping of In3+. Moreover, lattice parameter, the specific surface area, concentration of oxygen vacancy and low temperature oxygen storage capacity could mark a turning point for 1%In3+. It could be found that the low temperature oxygen storage capability was in relation to both the specific surface area and the concentration of oxygen vacancy of CeO2. In addition, the low temperature reduction peaks shifted towards lower temperatures with the addition of In3+.
The formation of a uniform Fe-Al inhibition layer with a proper thickness at steel interface during continuous hot-dip galvanizing process is a crucial issue for industrial production. The inhibition layer prohibits the nucleation and growth of brittle Fe-Zn intermetallic compounds which deteriorate the adhesion of the galvanizing coating and result in an inhomogeneous distribution of the coating. The inhibition layer was identified to be Fe2Al5 with some Zn dissolved in it. But Fe2Al5 inhibition layer was damaged with galvanized time increasing, will lose the inhibition to the Fe-Zn reaction. Nevertheless, there is no systematic and comprehensive investigation the causes of the inhibition layer is damaged. The aim of this work is to clarify the destabilization mechanism of Fe2Al5 inhibition layer. In the present study, the mass fraction of 0.2%Al was added into the zinc bath at 450 ℃ for hot-dip galvanizing. SEM was used to observe the structure characteristics of the hot-dip galvanized coating. EDS was used to quantitatively analyze the micro area components of phases and also used its line scan and mapping scan to qualitatively analyze the element change of the coating cross section. By means of the Miedema model and the Toop model, the thermodynamic values of the binary Fe-Al, Fe-Zn and ternary Fe2Al5Znx (η) intermetallic compounds (IMC) in the coatings were calculated. The fundamental reason for the Fe-Zn reaction caused by Fe2Al5 destabilization with galvanized time increasing was analyzed. The results show that because Fe-Al IMC which is generate preferentially had more stable thermodynamic property than Fe-Zn IMC, the continuous Fe2Al5 intermetallic compound inhibition layer was produced preferentially at steel and zinc bath interface which inhibit the Fe-Zn reaction. However, with the galvanized time increasing, Fe2Al5 destabilization which led to the loss of inhibitory effect of Fe-Zn reaction and produced FeZn10 (δ) . There are two kinds of destabilization mechanism of Fe2Al5 inhibition layer, one is that the local depletion of Al at Fe2Al5 and zinc bath interface result in erosion of Fe2Al5 by Zn and the formed Fe2Al5Znx caused the decrease of the systematic thermodynamic stability which led to erosion and decomposition of Fe2Al5 by Zn. At the same time, FeZn10 (δ) phase was produced between the Fe2Al5 and zinc bath interface. The phase transformation process can be described as: Fe2Al5→η→L+η→L+η+δ→L+δ. The other kind of destabilization mechanism is Zn diffused to the steel substrate by Fe2Al5 grain boundaries and directly produced δ phase between Fe2Al5 and steel substrate interface, which caused outburst of Fe2Al5. The two kinds of Fe2Al5 destabilization mechanism are mutual coexistence and mutual competition, in particular conditions may be a mechanism to occupy absolute advantage.
MCrAlY (M=Ni and/or Co) coatings are widely used as overlays or bond coats for thermal barrier coatings due to their good performance against high temperature oxidation and hot corrosion. Usually, high Al content in the MCrAlY coatings can benefit the performance and lifetimes of the coatings. However, MCrAlY coatings usually contain only restricted Al content because high Al content might lead to brittleness and potential crack. Design of gradient coating can be used to solve the problem, since it can provide a balance between high Al content and high stress bearing ability. Therefore, much attention has been paid to coatings with gradient structures, and these coatings show good oxidation and corrosion resistance. In this work, a gradient and a conventional NiCoCrAlYSi coating were prepared by arc ion plating technique and subsequent annealing treatment. Cyclic oxidation tests of the two coatings were carried out between room temperature and 1000 ℃. The hot corrosion tests of the coatings were performed in two different mixed salts of 75%Na2SO4+25%K2SO4 and 75%Na2SO4+25%NaCl (mass fraction) at 900 ℃. The results indicated that the gradient coating possessed a graded distribution of Al-rich outer layer and Cr-rich inner layer after annealing treatment, and it showed better performance of re-healing alumina scale due to its possession of more β phase as Al reservoir during the cyclic oxidation. The degradation process of the gradient coating was favorably retarded by the formation of Cr(W, Re)-rich precipitates in the interdiffusion zone. In sulphates, the two coatings showed good corrosion resistance. The presence of NaCl aggravated the corrosion extent of the two coatings. Compared with the conventional coating, the gradient coating postponed the formation of internal oxidation and sulfidation, resulting from the gradient distribution of Al-enriched outer layer and Cr-enriched inner layer.
The solidification path of alloy reveals the detailed relationship between the solute concentration in liquid and the temperature during the solidification process. The best and most accurate method to predict the solidification path of multicomponent/multiphase alloys is to establish proper microsegregation modeling coupled with phase diagram calculations according to the CALPHAD method. Recently, several alloy systems such as Al-Cu, Al-Mg and Cu-Mg have been developed, which have aroused the interest of many researchers. Up to now, the research about Al-Cu-Mg ternary alloy, especially containing higher Mg content, is relatively rare. The purpose of the present work is to investigate the solidification path of Al-6.32Cu-25.13Mg (mass fraction, %) ternary eutectic alloy at different cooling rates and solid back diffusion coefficients by an extended unified microsegregation model coupled with Thermo-Calc. Solidification experiments and subsequent microstructural characterization are combined with numerical calculation of solidification paths. It was shown that the cooling rates Rf had no obvious effect on the solidification path which was (L+α)→(L+α+T)→(L+α+β+T); but the solid back diffusion coefficient Φ had a great effect on the solidification path, which evolved gradually from (L+α)→(L+α+T)→(L+α+β+T) into (L+α)→(L+α+T) when Φ increased from 0 to 1. The volume fractions of primary α phase Vα, binary eutectic V2E and ternary eutectic V3E at each solidification path were calculated. It was shown that V2E decreased with the increase of Rf whereas V3E increased and Vα was almost invariant. The dependence of V2E, V3E and Rf were determined by linear regression analysis given as: V2E=-2.5lgRf+64.9, V3E=2.5lgRf+22.12. The increase in Φ led to increases in Vα and V2E and decrease in V3E. The predicted solidification paths and volume fractions of Al-6.32Cu-25.13Mg ternary eutectic alloy at different cooling rates were in good agreement with experimental results.