Effects of Heat Treatment and Extrusion on Microstructure and Propertiesof A390 Alloy Hollow Billet Fabricated via DC Casting

Kesheng Zuo+1, Haitao Zhang+2, Ke Qin, Xing Han+1, Bo Shao+1 and Jianzhong Cui

Key Laboratory of Electromagnetic Processing of Materials, Ministry of Education, Northeastern University,Shenyang 110819, China

An A390 alloy hollow billet was fabricated via a DC casting process prior to heat treatment and extrusion. The microstructural evolutionand properties of the A390 alloy were studied by means of optical microscopy, scanning electron microscopy (SEM), electrical conductivity,tensile and hardness tests. The results show that primary Si particles experienced a slight change in that the sharp edges and corners becamesmooth. The eutectic Si particles were partially modified, the morphology changed from needle-like to rod-like and granular and the particle sizeincreased during homogenization and T6 heat treatment. Some coarse primary Si particles fractured with cracks inside or separated into smallerones, while fine eutectic Si particles being uniformly distributed in the matrix with a more rounded morphology after the extrusion process.Intermetallic phases were still distributed at the grain boundary after homogenization and T6 heat treatment, while the extrusion process wassuccessful in changing the distribution of intermetallic phases to be uniform in the matrix. The Al8FeMg3Si6 (³) phase disappeared and it wassubstituted by the Al5Cu2Mg8Si6 (Q) and ª phase during homogenization, and Fe was dissolved into the Q phase. The electrical conductivityincreased after the heat treatment and extrusion process. The ultimate tensile strength (UTS) and hardness significantly increased after the T6heat treatment as a result of precipitate hardening. The tensile strength and hardness value were reduced after the extrusion process and that canbe ascribed to a small portion of primary Si particles being fractured with crack inside during the extrusion process. However, the elongationincreased significantly after the extrusion process and it can be attributed to the microstructural evolution, especially the changes of eutectic Siimproved the continuity of the matrix. [doi:10.2320/matertrans.M2015132]

(Received April 1, 2015; Accepted June 9, 2015; Published August 25, 2015)

Keywords: hollow billet, microstructure, extrusion, heat treatment, primary silicon, eutectic silicon

1. Introduction

Lightweight materials such as cast aluminum alloysprovide cost effective alternatives for cast iron in theautomobile industry, since reducing the vehicle weight is anapproach to increasing automobile fuel economy.1­3) How-ever, cast iron liners are needed on aluminum engine blocksin common design due to the hypoeutectic Al-Si alloys donot have the required tribological characteristics for engineblock material. But the cast iron liners also have disadvan-tages in weight, thermal conductivity, recyclability anddifferential thermal expansion compared to the aluminumalloys.4) Hypereutectic Al-Si alloys have the required wearresistance for engine block without liners and the firstapplication was developed as early as the 1970s.4) However,the application did not have great success due to the primarySi particles are coarse and uneven distributed in hypereutecticAl-Si alloy. To overcome such problems, various methodssuch as addition of nucleation agents5,6) and rapid solid-ification7,8) have been employed to refine the size of primarySi and the extent of macrosegregation. Alusil alloy(AlSi17Cu4Mg) with P refined can be used as engine blockwithout liners.9) The primary Si particles have a size in therange of 20­70µm in this alloy. However, this alloy can onlybe cast using the low-pressure die cast (LPDC) process andthe throughput time is not acceptable for mass production.10)

Peak in Germany11) have developed spray formed hyper-eutectic Al-Si alloy used for cylinder liner which arecurrently being used in Daimler-Benz automotive engines.However, extrusion and rotary swaging process are necessaryto eliminating the porosity formed in the spray forming

process. In addition, the production cost of this technology isrelative high as compared to the widely used DC casting.

In the present study, an A390 alloy hollow billet wasfabricated by DC casting process with fine primary Siparticles uniformly distributed in the matrix, and thenextruded into tube. The application prospect for the extrudedtube is in the manufacture of cylinder liners with a low costat mass production as compared to other processes. It hasbeen reported that solution treatment is able to modify themorphology of eutectic Si from needle-like to spherical inaddition to enable maximum solubility and homogeneity ofalloying elements into the matrix.12) However, heat treatmentshould be taken according to the specific processing routesdue to microstructure varies with solidification rates. Theeffects of heat treatment and extrusion on the micro-structure and properties of A390 alloy hollow billet wereinvestigated by microstructural examination, electrical con-ductivity, tensile and hardness tests.

2. Experimental Procedure

2.1 DC casting of A390 alloy hollow billetA390 alloy (the nominal composition of Al-17%Si-

4.2%Cu-0.55%Mg and refined by 1% of Al-4.5%P masteralloy, in mass%) hollow billet with a size of ¯60mm/¯164mm © 2000mm was fabricated via the DC castingprocess. As shown in Fig. 1(a), as the melt is poured into thewater cooled mold and Cu mold (with appropriate taperangle), it starts to solidify and forms a solid shell, then thesolid shell is withdrawn from the mold, the outer surface ofhollow billet is directly cooled by water from the mold, whileno cooling water is injected directly onto the inner wall ofthe hollow billet. With this method, no risk of vaporizingexplosion would be caused even though a breakout of melt

+1Graduate Student, Northeastern University+2Corresponding author, E-mail: haitao_zhang@epm.neu.edu.cn

Materials Transactions, Vol. 56, No. 9 (2015) pp. 1591 to 1598©2015 The Japan Institute of Metals and Materials EXPRESS REGULAR ARTICLE

occurs at the inner wall, nor did any hanging would happensdue to relative slow solidification shrinkage at the inner wall.

The casting parameters for fabrication of the hollow billetwere: melt temperature into the molten pool of 770°C,casting speed of 110mm/min, volume of cooling rate formold of 90 l/min, volume of cooling rate for Cu core of30 l/min.

2.2 Extrusion of A390 alloy hollow billetPrior to the tube extrusion, for the purpose of releasing

of residual stress and elimination of dendrite segregation ofnon-equilibrium phases, the hollow billet was homogenizedat 480°C for 8 h and then furnace cooled to room temper-ature. Thereafter, the hollow billet was machined into a sizeof ¯70mm/¯156mm © 350mm. Indirect hot extrusion wascarried out in a 16MN press using a container diameter of7 inches and mandrel diameter of 66mm (illustrated inFig. 1(c)).

The hot extrusion procedure was determined as follows:the hollow billet was heated to a temperature of 410­420°Cand indirectly extruded to tube with a extrusion speed of 15­20mm/s. The extruded tube has an inner and outer diameterof 66mm and 84mm, respectively. Thus, the extrusion ratiowas about 10.1 : 1.

2.3 CharacterizationThe as-cast hollow billet and extruded tube were given

artificial age hardening treatment (T6) consisting a sequenceof solutionizing, quenching and age hardening. The solutionheat treatment was conducted at 490°C for 8 h followed bywater quenching (25°C) and artificial aging was carried outat 170°C for 8 h. The microstructures were observed byoptical microscope and scanning electron microscope (SEM).Image analysis method was used to calculate average size(equivalent diameter computed from the area of Si particle),aspect ratio (ratio between major axis and minor axis ofellipse equivalent to Si particle), roundness (permeter 2/(4³ *

area)) and area fraction of the primary and eutectic Siparticles. Each average value was taken from 20 measure-ments. For observation of three-dimensional (3D) morphol-ogies of Si particles, the samples were deep-etched with20mass% sodium hydroxide solution and then characterizedby using SEM. Electrical conductivity as a percentage of theinternational homogenized copper standard (%IACS) wasmeasured using a Sigmascope smp10 electrical conductivitymeter. Brinell hardness was measured using a 250 kg loadand 5mm diameter indenter (HB < 130), or a 750 kg loadand 5mm diameter indenter (HB ² 130), with a holding timeof 30 s. Tensile test samples were conducted with two typesof geometry: cylindrical bar for hollow billet and arc-shapedstrip for extruded tube. The gauge section of cylindricalsamples was ¯6mm © 30mm, while the arc-shaped sampleswas 12.5mm wide © 50mm long and the thickness wasmeasured according to the tube. The room temperature(25°C) and elevated temperature (250°C) tensile tests wereconducted according to GB/T 16865-1997 with a cross-headspeed of 1mm/min.

3. Results and Discussion

3.1 Microstructures3.1.1 Primary Si and eutectic Si

Figure 2(a) and (b) show the distribution of primary Siparticles at 25 times magnification in as-cast and extrudedconditions, respectively. It is clearly shown that the primarySi particles were uniformly distributed in the hollow billetand extruded tube. The DC casting process rather than theextrusion process plays the crucial role in determining thedistribution of primary Si particles, as no significantimprovement in the distribution of primary Si particles seenby the extrusion process. A 3­5mm thickness of Si depletedregion (shown in Fig. 1(b)) existed at the inner wall ofhollow billet, which results from the remelting of eutecticstructures with low melting point due to the high latent heatof the Al-Si eutectic reaction and relative low coolingcapacity at the inner wall. The Si depleted region can bemachined away before the extrusion process, thus which hasno negative effect on the down streaming processes. Basedon 20 micrographs at 25 times magnification, the imageanalysis on primary Si particles size distribution in as-castand extruded conditions were carried out and the results aregiven in Fig. 3(a) and (b), respectively. The frequency ofprimary Si particle size larger than 30 µm dropped from14.45% in as-cast condition to 9.53% in extruded condition.Meanwhile, the average size and area fraction of primary Siparticles in extruded condition are slightly lower than that in

(a) (b)

(c)

Fig. 1 (a) Schematic illustration showing the DC casting procedure ofhollow billet, (b) cross section of hollow billet, (c) schematic illustrationshowing the extrusion procedure of tube.

(a) (b)

Fig. 2 Optical micrographs showing the distribution of primary Siparticles: (a) as-cast, ©25, (b) extruded, ©25.

K. Zuo et al.1592

as-cast condition. This can be attributed to a small portion ofcoarse primary Si particles were fractured into smaller onesduring the extrusion process.

Figure 4 shows the microstructures of A390 under differ-ent conditions at 500 times magnification. Primary Si particlesize and morphology remained almost the same after thehomogenization and T6 heat treatment except for the sharpcorners of primary Si particles got smoothen. However, thedifferences in the eutectic Si particles size and morphologyare evident. Eutectic Si particles exhibited needle-like in as-cast condition (Fig. 4(a)) while the particles experiencedspheroidization and coarsening during the homogenizationand T6 heat treatment, and fractured into smaller size andmore rounded morphology during the extrusion process.Image analysis of eutectic Si particle size, aspect ratio androundness based on micrographs at 500 times magnificationwere carried out and the results are summarized in Table 1.The A390 alloy had needle-like eutectic Si particles with thehighest aspect ratio value of 3.63 « 0.25, highest roundnessvalue of 4.86 « 0.50, and smallest size of 3.20 « 0.38 µm inas-cast condition. The eutectic Si particle size increased to3.60 « 0.35 µm, aspect ratio decreased to 2.49 « 0.22 androundness decreased to 2.13 « 0.20 after homogenization,and the T6 heat treatment have similar modification effectson eutectic Si particles. Homogenization and T6 heattreatment were successful in changing the needle-like eutecticSi particles into granular or rod-like morphology due to thereduction of interfacial energy between the Si particles andmatrix provided the driving force for spheroidization of theneedle-like eutectic Si particles. And the eutectic Si particle

size increases as a result of the dissolution of small ones andthe growth of large ones according to Ostwald ripening.However, some coarse eutectic Si particles remained longrod-like with high aspect ratio and roundness value after thehomogenization and T6 heat treatment. Therefore, highertemperature and longer time are required to allow maximumbreaking up of coarse eutectic Si particles. During theextrusion process, long rod-like eutectic Si particles inhomogenized condition fractured into smaller ones as theeutectic Si particle size decreased from 3.60 « 0.35 µm to3.28 « 0.03 µm and the morphology became more roundedas aspect ratio decreased from 2.49 « 0.22 to 1.87 « 0.01and roundness decreased from 2.13 « 0.20 to 1.91 « 0.15.What is more, the morphology can be further rounded andsize increased after the T6 heat treatment. Fine eutectic Siparticles were uniformly distributed in the matrix instead ofmerely distributing at the grain boundary after the extrusionprocess. It is widely accepted that the eutectic Si particles canbe effectively modified by the addition of Sr.13,14) However,the Sr weakens the refinement effect of P on primary Si inhypereutectic Al-Si alloy for the formation of Sr-P com-

(a)

(b)

Fig. 3 Histograms showing primary Si particle size distribution: (a) as-cast,(b) extruded.

(a) (b)b

(c) (d)

(e)

Fig. 4 Optical micrographs showing the microstructure of A390 alloy:(a) as-cast, ©500, (b) homogenized, ©500, (c) as-cast-T6, ©500,(d) extruded, ©500, (f ) extruded-T6, ©500.

Table 1 Summary of the results of image analysis on eutectic Si particlesize, aspect ratio and roundness in the microstructures of A390 alloyunder different conditions.

Size, µm Aspect ratio Roundness

as-cast 3.20 « 0.38 3.63 « 0.25 4.86 « 0.50

homogenized 3.60 « 0.35 2.49 « 0.22 2.13 « 0.20

as-cast-T6 3.85 « 0.38 2.42 « 0.39 1.86 « 0.17

extruded 3.28 « 0.03 1.87 « 0.01 1.91 « 0.15

extruded-T6 3.63 « 0.08 1.79 « 0.02 1.70 « 0.10

Effects of Heat Treatment and Extrusion on Microstructure and Properties of A390 Alloy Hollow Billet Fabricated via DC Casting 1593

pounds.15,16) In this study, it can be seen that the combinationof extrusion and T6 heat treatment provides a high-efficiencysolution for modification of eutectic Si in hypereutectic Al-Sialloy.

In order to observe the primary and eutectic Si particlesmore intuitively, deep etch on A390 alloy were carried out.Figure 5 shows the 3D morphology of primary Si andeutectic Si particles. Primary Si particles in all conditionsmainly exhibit block morphology but the sharp edges gotsmoothen after the homogenization and T6 heat treatment.The primary Si particle with a crack inside (as marked bycircle in Fig. 5(e)) clearly demonstrated that primary Siparticle can be fractured during the extrusion process. It isclearly seen that the eutectic Si particles exhibit lamellarmorphology in as-cast condition, and can be partiallymodified as mixed morphologies of thick lamellar and nearspherical eutectic Si particles are observed in the homogen-ized and T6 treated conditions. The eutectic Si particle sizereduced and morphology became more rounded after theextrusion process.3.1.2 Intermetallic phases

Figure 6 shows the SEM micrographs of intermetallicphases in A390 alloy under different conditions. Figure 6(a)

shows the intermetallic phases were concentrated at the grainboundary in as-cast condition. As seen in Fig. 6(c) and (e),there is no change in the distribution of intermetallic phasesafter the homogenization and T6 heat treatment but theamount of intermetallic phases distinctly decreased after theT6 heat treatment. As shown in Fig. 6(g) and (i), in extrudedcondition, fine intermetallic phases were uniformly distribut-ed in the entire matrix instead of merely distributing along thegrain boundary. This can be deduced that intermetallic phaseswere fragmentized by the extrusion force. In addition, thesolubility of Cu increases with temperature up to approx-imately 1.75mass% at the extrusion temperature (420°C)by the calculation of JMatPro software. Therefore, thesedissolved elements preferentially precipitated at the disloca-tions inside the matrix rather than merely at the grainboundary. In as-cast condition, the main intermetallic phasesare Al2Cu (ª), Al5Cu2Mg8Si6 (Q) and Al8FeMg3Si6 (³). The³ phase contained a small amount of Cu, as the composition(mass%) of ³ phase (shown in Fig. 6(b)) obtained by EDSanalysis is 46.88 Al, 10.21 Fe, 13.29 Mg, 25.39 Si and 2.21Cu. However, during homogenization, the metastable ³ phasedisappeared almost completely, and Fe was found in Q phase(light gray) and another complex phase (bright white), as the

(a)

(d)

(e)

(c)

(b)

Fig. 5 The SEM micrographs showing the 3D morphology of primary Si and eutectic Si particles: (a) as-cast, (b) homogenized, (c) as-cast-T6, (d) extruded, (e) extruded-T6.

K. Zuo et al.1594

composition (mass%) of Q phase is 43.82 Al, 19.28 Si, 18.92Cu, 16.34 Mg and 1.65 Fe, while the complex phase is 45.9Al, 6.66 Si, 34.16 Cu, 6.31 Mg and 6.97 Fe. As shown inFig. 6(d), the complex phase is rich of Cu and associated

with Q phase. Lasa17) found that ª phase particles formedsurrounding the dissolving ³ particles in Al-Si-Cu-Mgcasting alloy during solution treatment. In this study, thecomplex phase is considered as a mixture structure of ª phase

(g) (h)

(i) (j)

(a) (b)

(c) (d)

(e) (f)

Fig. 6 SEM micrographs showing the intermetallic phases in A390 alloy: (a)-(b) as-cast, (c)-(d) homogenized, (e)-(f ) as-cast-T6,(g)-(h) extruded, (i)-( j) extruded-T6.

Effects of Heat Treatment and Extrusion on Microstructure and Properties of A390 Alloy Hollow Billet Fabricated via DC Casting 1595

and Q phase transformed from ³ phase, and the Q phase isnot possible to be distinguished from ª phase in this complexstructure. As seen in Fig. 6(f ), (h) and ( j), the Q phaseparticles always associated with the ª phase particles inextruded or T6 treated conditions.

3.2 Electrical conductivityThermal conductivity is the property of a material’s ability

to conduct heat and impacts the engines fuel efficiency andemissions. A higher thermal conductivity of cylinder blockand liner makes the block easier to keep cool, reduces thecooling load and reduces the engine cooling system.Moreover, this enables the designers to develop engine withhigher operating temperatures which increases engine powerand reduces emissions. The measurement of electricalconductivity indirectly reflects the evolution of thermalconductivity during the heat treatment and extrusion process.Since the thermal conductivity is proportional to thetemperature and electrical conductivity, according to theWiedemann-Franz law.

Histograms that compare the electrical conductivity ofA390 alloy under different conditions are shown in Fig. 7.The A390 alloy had the lowest electrical conductivity at21.4% IACS in as-cast condition. The electrical conductivityhas an obvious increase as it reached to the peak value of33.1% IACS after homogenization and to 26.2% IACS asa result of T6 heat treatment. The electrical conductivityremained at the peak value in extruded condition anddecreased to 29.1% IACS after the T6 heat treatment. Incomparing the electrical conductivity of as-cast-T6 andextruded-T6 treated A390 alloy, the value in extruded-T6condition is higher than that in as-cast-T6 condition. Whenthermal conductivity is only considered, A390 alloy inextruded-T6 condition is superior to alloy in as-cast-T6condition when used as service status for cylinder liner.

The evolution of electrical conductivity ¬ (inverse ofresistivity) can be explained by a general form ofMatthiessen’ law,18,19) i.e.

¬ ¼ μ0 þXi

μiCi þ μppt

!�1

ð1Þ

where μ0 is the temperature dependent resistivity term,Pi μiCi is the summation of the resistivity contribution from

the various solid solution contribution, i.e. μi is the specificresistivity of the ith solute and Ci is the concentration ofthis solute, and μppt is the contribution of precipitates toresistivity.

The eq. (1) shows that both the solid solubility of atomsinto the matrix and the type and size of the precipitates affectthe electrical conductivity, which indicates that the change inelectrical conductivity is related to microstructural changes.The lowest electrical conductivity in as-cast condition canbe explained as follows: A390 alloy experienced a non-equilibrium solidification condition due to relative highcooling rate during DC casting. The Cu and Mg element cannot completely precipitated into the ª, Q and ³ phases, thusthe matrix kept some degree of supersaturation in as-castcondition. Moreover, the lamellar (in 3D morphology)eutectic Si particles wrecked the continuity of matrix.However, equilibrium was nearly reached after homogeniza-

tion, the number of solute atoms reduced and the latticedistortion decreased evidently in the matrix. Furthermore, themorphology of eutectic Si particles changed from needle-liketo granular and rod-like to some extent. As a result, A390alloy had the highest value of electrical conductivity afterhomogenization. Rod-like eutectic Si particles got shortenand intermetallic phases dispersed uniformly in the matrixafter the extrusion process, resulting in the electricalconductivity value remained at the same level as that in thehomogenized condition. The A390 alloy underwent asequence of dissolution of Cu and Mg-containing phasesand precipitation of dissolved elements during the T6 heattreatment. As indicated by S. F. Fang,20) the growth andcoarsening of precipitates reduces the amount of solute atomsin the matrix and releases the strain fields around thoseprecipitates, which results in an increase in electricalconductivity during aging. However, strain fields still existbecause the ª A and Q A metastable phases are semi-coherentwith the matrix when precipitate from the matrix duringaging. Therefore, the value of electrical conductivity inas-cast-T6 condition fell in between that in as-cast andhomogenized conditions. A390 alloy had a higher electricalconductivity in extruded-T6 condition as compared to thatin as-cast-T6 condition and this can be attributed to theimproved continuity of matrix as eutectic Si particles weremore rounded during extrusion process.

3.3 Mechanical propertiesThe tensile properties and hardness of A390 alloy under

different conditions are listed in Table 2. It can be seen thatthe tensile strength (·b) and hardness of A390 alloy tested at25°C increased from 271.6MPa and 143 in as-cast conditionto 415.1MPa and 172.6 as a result of T6 heat treatment,respectively. It should be noted that the yield strength (·0.2) isalmost the same as tensile strength (·b) in non-extrudedcondition. The ·0.2, ·b and hardness of A390 alloy increasedfrom 203.1MPa, 246.4MPa and 90.8 in extruded conditionto 383.4MPa, 386.9MPa and 158 as a result of T6 heattreatment, respectively. T6 heat treatment did not havean effect on elongation of as-cast A390 alloy while theelongation of extruded alloy decreased from 5.2% to 1.47%after T6 heat treatment. It is obvious that the ·0.2, ·b andhardness decreased after extrusion process. Whereas, theelongation can be obviously improved as it increased from

0

5

10

15

20

25

30

35

extruded -T6

extrudedas-cast -T6

Elec

trica

l con

duct

ivity

, % IA

CS

as-cast homogenized

Fig. 7 Histograms showing electrical conductivity of A390 alloy underdifferent conditions.

K. Zuo et al.1596

less than 0.5% in as-cast condition to 5.2% in extrudedcondition. The ·0.2 and ·b of A390 alloy tested at 250°Cdecreased to 223.1MPa and 227MPa in as-cast-T6 con-dition, while decreased to 182.5MPa and 186.7MPa inextruded-T6 condition, respectively, as compared to thosetested at room temperature. The ductility of extruded-T6treated A390 alloy showed a bit better at higher temperature(250°C) but no visible improvement was found in as-cast-T6treated A390 alloy.

The technology requirements from FAW-VolkswagenAutomotive Company for the cylinder liner are as follows:average primary Si size smaller than 30 µm; tensile strengthno less than 250MPa; elevate temperature tensile strength noless than 150MPa; hardness (HB) no less than 120. It isclearly seen from Table 2 that the T6 treated A390 alloymeets to these requirements and showed some superiority tothe spray-formed, extruded and T6 treated Al25Si4Cu1Mgalloy. Furthermore, the results show that good comprehensivemechanical properties of A390 alloy can be achieved by thecombination of extrusion and T6 heat treatment, since theelongation was improved obviously by the extrusion processand the strength was increased significantly after the T6 heattreatment.

The significant improvements in UTS and hardness for T6treated A390 alloy can be mainly attributed to Cu and Mgelements dissolved into the matrix during solution treatment,and then ªA and QA phases precipitated dispersively from thesupersaturated matrix during the aging treatment, which aresemi-coherent with the matrix can enhance the strength andhardness of A390 alloy.

In tensile testing, the samples fractured suddenly withoutnecking and the fracture surfaces were relative flat suggesting

they were brittle fracture in all conditions. This is consistingwith the results shown in Table 2, which shows that theelongation was no more than 5.2%. Figure 8 shows themicrographs of fracture morphologies of samples in as-cast-T6 and extruded-T6 conditions tested at 25°C and 250°C. Asseen in Fig. 8(a) and (c), no visible dimple was found at thefracture surface of as-cast-T6 treated A390 alloy. The fracturesurfaces of extruded-T6 treated A390 alloy show mixture ofcleavages of primary Si particles and shallow dimples, asseen in Fig. 8(b) and (d). Cracks are clearly seen insideprimary Si particles at the fracture surfaces and consideredas crack initiator for the material. During deformation, thematrix in the vicinity of the Si particles deforms initially, andthe stress imposed on Si particles increases as the stressimposed on samples increases. When the stress increases

Table 2 Tensile properties and hardness of hypereutectic Al-Si alloys underdifferent conditions.

·0.2/

MPa·b/

MPaElongation,

%

Hardness,HB

as-cast, 25°C 271.6 271.6 <0.5 143

as-cast-T6, 25°C 415.1 415.1 <0.5 172.6

extruded, 25°C 203.1 246.4 5.2 90.8

extruded-T6, 25°C 383.4 386.9 1.47 158

as-cast-T6, 250°C 223.1 227 <0.5 ®

extruded-T6, 250°C 182.5 186.7 2.07 ®

Al25Si4Cu1Mg,extruded-T6, 25°C21) 264.5 313.9 ® ®

Al25Si4Cu1Mg,extruded-T6, 250°C21) 159.1 212.1 ® ®

(a) (b)

(d)(c)

Fig. 8 SEM micrographs showing the fracture morphologies: (a) as-cast-T6, 25°C, (b) extruded-T6, 25°C, (c) as-cast-T6, 250°C,(d) extruded-T6, 250°C.

Effects of Heat Treatment and Extrusion on Microstructure and Properties of A390 Alloy Hollow Billet Fabricated via DC Casting 1597

enough to debonding and fracturing of the Si particles, cracksgenerate by the fracture of primary Si particles, thus stressconcentration will be more severe at the neighboring Siparticles and promote the possibility of a succession of cracksdeveloped in the neighboring Si particles. The eventualfracture of material occurs as matrix fractured to connectingthe cracks developed in Si particles. As some primary Siparticles were fractured and cracks generated inside theprimary Si particles during the extrusion process, less stresswas required to promote the fracture of Si particles. As aresult, the ·0.2 and ·b of extruded-T6 treated A390 alloy islower than those in as-cast-T6 condition. Meanwhile, theincreases in elongation of extruded A390 alloy can beexplained as follows: In as-cast condition, the needle-likeeutectic Si particles were interconnected and formed highlybranched network,22) thus the deformation of matrix wasconstrained between Si particles which resulted in brittlenessof the as-cast material. What is more, due to the coarseeutectic Si particles still exhibited long rod-like and dis-tributed at the grain boundary, as a result, no visibleimprovement was achieved in elongation after the T6 heattreatment. However, long rod-like eutectic Si particles brokenup obviously and improved the continuity of the matrixduring the extrusion process. In addition, casting defects suchas microporosities were eliminated. Therefore, the matrixexperienced a relative high strain prior to the fracture ofmaterial during deformation.

4. Conclusions

The microstructural evolution and properties of DC-castA390 alloy hollow billet before and after heat treatment andextrusion were studied. The following conclusions can bedrawn from the results of the present investigation:(1) During the homogenization and T6 heat treatment, no

remarkable change occurred on primary Si particleexcept for its sharp edges and corners became smooth.In contrast, the eutectic Si particles experiencedspheroidization and coarsening as the morphologychanged from needle-like to rod-like and granular andthe particle size increased. During the extrusion process,a small portion of coarse primary Si particles werefractured with cracks inside or separated into smallerparticles, the eutectic Si particle size was refined intomore rounded morphology.

(2) Intermetallic phases were mainly distributed at the grainboundary in the as-cast condition, and the homoge-nization and T6 heat treatment had no effect on thedistribution. Intermetallic phases became more uniformin the matrix during the extrusion process. During thehomogenization and T6 heat treatment, the Al8FeMg3-Si6 (³) phase disappeared and it was substituted by theAl5Cu2Mg8Si6 (Q) and ª phase, and Fe was dissolvedinto Q phase.

(3) The electrical conductivity had the lowest value in as-cast condition and increased after the heat treatment andextrusion process. The extruded-T6 treated alloy had ahigher electrical conductivity than as-cast-T6 treatedalloy. The changes of number of solute atoms andlattice distortion in the matrix and the morphology

changes of eutectic Si particles are responsible for thevariation in electrical conductivity.

(4) The yield strength (·0.2) and tensile strength (·b) weresignificantly increased after the T6 heat treatment due toprecipitate hardening. After the T6 heat treatment, the·b of as-cast and extruded A390 alloy are 415.1MPaand 386.9MPa at 25°C, 227MPa and 186.7MPaat 250°C, respectively. The decline in ·0.2 and ·b ofextruded A390 alloy is ascribed to a small portion ofprimary Si particles were fractured with crack insideduring the extrusion process. However, the elongationshowed remarkable increases after the extrusion processwhich can be attributed to the microstructural evolution,especially the spheroidization and coarsening of theeutectic Si particles improved the continuity of thematrix.

Acknowledgments

This work is supported by the National Natural ScienceFoundation of China (No. 51204046) and FundamentalResearch Funds for the Central Universities (No.N130409003).

REFERENCES

1) M. M. Haque and M. A. Maleque: J. Mater. Process. Technol. 77(1998) 122­128.

2) Y. Birol and F. Birol: Wear 265 (2008) 590­597.3) H. Yamagata, W. Kasprzak, M. Aniolek, H. Kurita and J. H.

Sokolowski: J. Mater. Process. Technol. 203 (2008) 333­341.4) R. Donahue and P. A. Fabiyi: SAE 2000 World Congress, (SAE,

Warrendale, PA, 2000).5) M. Zuo, X. F. Liu and Q. Q. Sun: J. Mater. Sci. 44 (2009) 1952­1958.6) C. L. Xu, H. Y. Wang, Y. F. Yang, H. Y. Wang and Q. C. Jiang: J. Alloy.

Compd. 421 (2006) 128­132.7) J. Kusui, A. Tanaka, K. Kubo, T. Watsuji and T. Yokote: United States

Patent Grant 5405576.8) C. Cui, A. Schulz, K. Schimanski and H. W. Zoch: J. Mater. Process.

Technol. 209 (2009) 5220­5228.9) A. N. K. Jadoon and R. A. Mufti: Tribol. 4 (2010) 61­73.10) E. Koya, Y. Hagiwara, S. Miura, T. Hayashi, T. Fujiwara and M.

Onoda: 1994 SAE International Congress and Exposition, (SAE,Warrendale, PA, 1994).

11) A. Leatham: Met. Powder Rep. 54 (1999) 28­30, 32­34, 36­37.12) M. A. Moustafa, F. H. Samuel and H. W. Doty: J. Mater. Sci. 38 (2003)

4507­4522.13) F. J. Tavitas-Medrano, J. E. Gruzleski, F. H. Samuel, S. Valtierra and

H. W. Doty: Mater. Sci. Eng. A 480 (2008) 356­364.14) C. Y. Yang, S. L. Lee, C. K. Lee and J. C. Lin: Wear 261 (2006) 1348­

1358.15) Y. H. Cho, H. C. Lee, K. H. Oh and A. K. Dahle: Metall. Mater. Trans.

A 39 (2008) 2435­2448.16) M. Zuo, D. G. Zhao, X. Y. Teng, H. R. Geng and Z. Zhang: Mater. Des.

47 (2013) 857­864.17) L. Lasa and J. M. Rodriguez-Ibabe: J. Mater. Sci. 39 (2004) 1343­

1355.18) P. Guyot and L. Cottignies: Acta Mater. 44 (1996) 4161­4167.19) B. Raeisinia, W. J. Poole and D. J. Lloyd: Mater. Sci. Eng. A 420

(2006) 245­249.20) S. F. Fang, M. P. Wang and M. Song: Mater. Des. 30 (2009) 2460­

2467.21) P. Krug, K. Marcus and F. James: 2006 SAE World Congress, (SAE,

Warrendale, PA, 2006).22) M. Timpel, N. Wanderka, B. S. Murty and J. Banhart: Acta Mater. 58

(2010) 6600­6608.

K. Zuo et al.1598