A liquid aluminum alloy electromagnetic transport process for high pressure die casting

Accepted Manuscript Title: A liquid aluminum alloy electromagnetic transport process for high pressure die casting Author: Xixi Dong Xiusong Huang Lehua Liu Liangju He Peijie Li PII: DOI: Reference:

S0924-0136(16)30084-X http://dx.doi.org/doi:10.1016/j.jmatprotec.2016.03.028 PROTEC 14765

To appear in:

Journal of Materials Processing Technology

Received date: Revised date: Accepted date:

21-1-2016 27-3-2016 29-3-2016

Please cite this article as: Dong, Xixi, Huang, Xiusong, Liu, Lehua, He, Liangju, Li, Peijie, A liquid aluminum alloy electromagnetic transport process for high pressure die casting.Journal of Materials Processing Technology http://dx.doi.org/10.1016/j.jmatprotec.2016.03.028 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

A liquid aluminum alloy electromagnetic transport process for high pressure die casting Xixi Donga, b, c, *, Xiusong Huanga, b, Lehua Liua, b, Liangju Hed, Peijie Lia, b a

Department of Mechanical Engineering, Tsinghua University, Beijing 100084, China b

State Key Laboratory of Tribology, Tsinghua University, Beijing 100084, China c

BCAST, Brunel University London, Uxbridge, Middlesex, UB8 3PH, United Kingdom d *

School of Aerospace, Tsinghua University, Beijing 100084, China

Corresponding author: Tel: +44 1895 2677468; Fax: +44 1895 269758 E-mail address: [email protected]

1

Abstract For the electromagnetic transport (EMT) of liquid aluminum alloy during high pressure die casting (HPDC) by the plane induction electromagnetic pump (EMP), how to improve EMT efficiency and control EMT stationarity are key problems. Magnetic-flow coupling analysis was used to reveal effects of structural design and transport process parameters on EMT efficiency and stationarity. The output pump height of plane induction EMP was optimized by matching of iron core width W, coil width W' and pump ditch width b, i.e., b values of bopt corresponding to 90% of the maximum output pump height, bopt/W = 1.27 and bopt/W' ≈ 1. Both EMT efficiency and stationarity are achieved under the optimum transport current 32 A. With the increase of the transport height from 350 mm to 500 mm, the EMT flow rate decreases from 4.28 kg/s to 3.59 kg/s, and the fillings of shot sleeve are always stationary. The transport tubes suffer a maximum positive pressure of 1.8×104 Pa and a minimum negative pressure of -1.42×104 Pa during EMT. For the liquid aluminum alloy soup occasions of 4.5 kg, 6.5 kg and 12.0 kg, the transport time could be shortened significantly from manipulator’s 16 s, 22 s and 38 s to EMT’s at most 2.195 s, 2.75 s and 4.28 s, respectively. The developed EMT process with plane induction EMP for HPDC is a process with low cost, high transport efficiency and stationarity.

Keywords: Aluminum alloy; Electromagnetic transport; High pressure die casting; Magnetic-flow coupling; Efficiency; Stationarity

2

1. Introduction The electromagnetic fields have been widely used in materials processing fields such as electromagnetic casting, electromagnetic brake and electromagnetic treatment during casting, plastic forming and welding. Le et al. (2007) reported the electromagnetic casting of magnesium alloys. Singh et al. (2014) simulated the electromagnetic field effect on transient flow during continuous casting. Zhang et al. (2015) fabricated the carbon fibers reinforced Al-Mg matrix composites by electromagnetic casting. Yu et al. (2008) and Miao et al. (2012) revealed the influence of electromagnetic brake on flow field during continuous casting. Wang et al. (2010) studied microstructure evolution of AlSi7Mg alloy under superheat electromagnetic stirring. Haghayeghi et al. (2015) investigated microstructure evolution of AA5754 aluminum alloy under electromagnetic ultrasonic merged fields. Psyk et al. (2011) reviewed the electromagnetic forming process. Cui et al. (2014) reported a electromagnetic incremental forming technology of aluminum alloy sheet. Weddeling et al. (2015) introduced an analytical methodology for the process design of electromagnetic crimping. Kore et al. (2008) reported the electromagnetic impact welding of aluminum to stainless steel sheets. Bachmann et al. (2014) developed an electromagnetic weld pool support system for high power laser beam welding. During high pressure die casting (HPDC), manipulator is usually adopted for the transport of liquid aluminum alloy from the soup mouth of the holding furnace to the shot sleeve of the HPDC machine, as shown in Fig. 1(b). The liquid aluminum alloy soup mouth is exposed directly to the air, and it causes problems such as oxidation of liquid aluminum alloy and energy dissipation, as shown in Fig. 1(c). In addition, part of the liquid aluminum alloy adheres to the outer surface of the ladle when ladling, and falls off during the subsequent transport process, which not only decreases material utilization, but also harmful to safe production. The move of the ladle should be controlled steadily for the prevention of the spill of liquid aluminum alloy out of the ladle, so the transport efficiency of liquid aluminum alloy by the manipulator process is relatively low. However, if electromagnetic transport (EMT) process was 3

adopted for the transport of liquid aluminum alloy during HPDC, i.e., liquid aluminum alloy is driven by the electromagnetic pump (EMP) and transported rapidly to the shot sleeve of the HPDC machine through the sealed transport tube, problems such as oxidation, adhesion and falling of aluminum alloy, and energy dissipation would be eliminated, and the transport efficiency of liquid aluminum alloy would be improved. Since the invention of the Cosworth Process (Campbell and Wilkins, 1982) which employed the EMP to transport liquid aluminum alloy during low pressure die casting, the application of the EMT process has been extended to high pressure, squeeze and gravity casting fields. Saito et al. (1989) developed the EMT apparatus for HPDC with outside mounted annular induction EMP. Mercer, II et al. (1995) designed the EMT apparatus for HPDC with submerged EMP. Cheng et al. (2000) reported the EMT apparatus for low pressure and squeeze casting of aluminum alloys with outside mounted planar direct current (DC) EMP. Miura et al. (2009) proposed the EMT apparatus for die casting with outside mounted annular induction EMP, and the transport of molten metal could be stopped by inert gas. The CMI Novacast inc. is improving and promoting the EMT apparatus with submerged EMP for pressure, squeeze and gravity casting. However, there are problems for the existing EMT process. The EMT process with submerged EMP is with high efficiency but also high cost, and it also has high demand for cooling. The EMT process with outside mounted planar DC EMP has high conductive and anticorrosive demands for the electrodes that are inserted directly into the highly active liquid aluminum alloy. The EMT process with outside mounted annular induction EMP requires the insertion of the heat resisting iron core to liquid aluminum alloy in the pump ditch to decrease the magnetic gap, and it is also inconvenient for cooling and maintenance. Thus EMT process with outside mounted plane induction EMP is developed to solve the above mentioned problems. In addition to the control of EMT accuracy of liquid aluminum alloy, how to improve EMT efficiency and control EMT stationarity are key problems for the application of the EMT process with plane induction EMP during HPDC. In our 4

earlier work (Dong et al., 2013; Dong et al., 2014), four structural parameters, i.e., magnetic gap, pump ditch height, iron core width and pump ditch width were found to have significant effects on the output efficiency of pump height of the plane induction EMP, the quantitative effects of magnetic gap on pump height was revealed, and the optimum pump ditch height was determined. However, the comprehensive effects of iron core width, coil width and pump ditch width on the output efficiency of pump height of the plane induction EMP are still unknown. The effects of the EMT process on the improvement of hydrogen porosity defect in aluminum alloy during gravity casting were reported (Dong et al., 2015). The flow fields, and the effects of transport tube structural parameters on the outflow efficiency and the outflow stationarity in the transport tube, were uncovered by magnetic-flow coupling analysis (Dong et al., 2015). Nevertheless, the effects of transport process parameters on the filling efficiency and stationarity of liquid aluminum alloy in the shot sleeve of the HPDC machine are still unclear. In this paper, the comprehensive effects of iron core width, coil width and pump ditch width on the output efficiency of pump height of the plane induction EMP were investigated by electromagnetic analysis; the effects of transport process parameters on the filling efficiency and stationarity of liquid aluminum alloy in the shot sleeve of HPDC machine, and pressure distribution during EMT were studied by magnetic-flow coupling analysis; and the transport efficiency of liquid aluminum alloy during HPDC by the EMT process was compared with the common manipulator process. 2. EMT for HPDC Fig. 1(a) shows the EMT process with plane induction EMP for the transport of liquid aluminum alloy during HPDC. Liquid aluminum alloy in the holding furnace flows into the rectangular section pump ditch of the plane induction EMP by gravity, and the induced current is formed in the pump ditch by the excitation of the travelling magnetic field that is generated by the EMP. The interaction of the magnetic field and the induced current generates the electromagnetic force, which transports the liquid aluminum alloy along the sealed transport tube to the shot sleeve of the HPDC 5

machine. Under the EMT process, the open soup mouth in the holding furnace that is reserved for the common manipulator transport process can be sealed, so problems such as oxidation of liquid aluminum alloy and energy dissipation are avoided. Since liquid aluminum alloy is transported along the sealed transport tube under the EMT process, problems such as adhesion and falling of aluminum alloy can be eliminated during the transport process.

Fig. 1. (a) EMT process with plane induction EMP and (b,c) Common manipulator process for the transport of liquid aluminum alloys during HPDC.

For the EMT process with plane induction EMP, the plane induction EMP is the core component, and it consists of the magnet yoke, the iron core, the coil and the rectangular section pump ditch, as shown by the up middle inset in Fig. 2. Special heat radiation insulation was applied for the resistance of heat radiation from high temperature liquid aluminum alloy in the pump ditch to the iron core and coils in the EMP. In addition, the output efficiency of the EMP was optimized in Section 4 that a small transport current of 32A applied in the coils is enough for the EMT process. Based on the above mentioned two points and experimental results, it is definite that two air-cool (see Fig. 1(a)) can maintain the temperature of iron core and coils approximately at room temperature during the EMT process. Thus common enamelled coils can be used for the plane induction EMP without costly water-cooling 6

system, and low cost and easy cooling can be realized. Preheating is applied for the avoiding of thermal shock to the furnace lining and the transport tube during the EMT process. High aspect ratio rectangular section pump ditch is adopted to decrease magnetic gap, so there is no need for the insertion of the heat resisting iron core to liquid aluminum alloy in the pump ditch, thus the reliability can be improved for the EMT process with plane induction EMP. In addition, two half-retractable structure is designed for the plane induction EMP, so it is convenient for installation and maintenance. The excitation current and frequency of the coil in the EMP are represented by I and f respectively, and the excitation current of the coil is termed as current for short in the following text. 3. Numerical model 3.1. Electromagnetic model of EMP 3.1.1. Assumptions and governing equations For the simplifying of calculation, the following assumptions were done for the electromagnetic analysis of the plane induction EMP. (1) Flow of the liquid aluminum alloy in the pump ditch was neglected. (2) The temperature rise of iron core and coil was neglected. (3) The relative permeability of iron core and magnet yoke was constant. The governing equations of the electromagnetic analysis model are given by Maxwell's equations:  E  

B t

(1)

 B   J

(2)

 B  0

(3)

J  E

(4)

Where E is the electric intensity, B is the magnetic flux density, t is time, J is the induced current density, μ is the permeability, and σ is the conductivity. By solving Eqs. (1-4), B and J are obtained. Then the volume electromagnetic 7

force is calculated by B and J

f V (N / m3 )  J  B

(5)

Finally, the electromagnetic force is obtained by the integration of f V

FMAG (N)   f V dV

(6)

3.1.2. Model of EMP The electromagnetic analysis model of the plane induction EMP was built in the finite element analysis (FEA) software ANSYS by APDL command, as shown by the up middle inset in Fig. 2. The magnet yoke, the iron core, the coil and the liquid aluminum alloy in the pump ditch were meshed by hexahedral elements, while the peripheral air surrounding them was meshed by tetrahedral elements, and the element type was "solid97". The most commonly used A380 aluminum alloy in HPDC was chosen for analysis, and the material parameters for electromagnetic analysis are listed by the magnetic parameters in Table 1. Three-phase alternating current density was loaded into the elements of the coils, then the magnetic flux parallel boundary condition was applied to the outer surface of the air, and the details can be seen in our former published work (Dong et al., 2013). Transient analysis was adopted to obtain the transient spatial distribution of the magnetic flux, the induced current and the electromagnetic force in liquid aluminum alloy in the pump ditch, and harmonic analysis was chosen to calculate the time average electromagnetic force for the sake of saving calculation time and storage space. The correctness of the electromagnetic analysis model of the plane induction EMP has been verified by the electromagnetic force testing experiment of aluminum plate (Dong et al., 2013) and the pump height testing experiment of liquid aluminum alloy (Dong et al., 2014). 3.2. Magnetic-flow coupling model of EMT 3.2.1. Assumptions and governing equations For the simplifying of numerical calculation, the following assumptions were made for the magnetic-flow coupling analysis of the EMT of liquid aluminum alloy during HPDC. (1) The liquid aluminum alloy was taken as uncompressible Newtonian fluid. 8

(2) The transport tubes were not deformable under high transport temperature. (3) Physical parameters of liquid aluminum alloy were only function of temperature. (4) Temperature rise in liquid aluminum alloy under the EMT process was neglected. For the magnetic-flow coupling analysis, the flow of the liquid aluminum alloy was considered, and the induced current density is: J   ( E  v  B)

(7)

Where v is the velocity. According to Eqs. (1-3,5,7), the volume electromagnetic force fV is obtained. For the incompressible fluid, the continuity equation is satisfied:

 v  0

(8)

Thus the Navier-Stokes governing equation of fluid motion coupling the volume electromagnetic force is:

 v    v   v    g  P  f  2v  f v  t 



(9)

Where ρ and μf are the density and viscosity of fluid, respectively, P is the pressure, and g is the acceleration of gravity. By solving Eqs. (1-3,5,7-9) simultaneously, multi-physics fields such as B, J, fV, v and P during the EMT process are obtained, and the mass flow rate Q is calculated by the integration of v along the cross section:

Q    v dS

(10)

3.2.2. Model of EMT According to the actual production situation, the magnetic-flow coupling analysis model of the EMT of liquid aluminum alloy during HPDC was built in the fluid analysis software FLUENT, as shown in Fig. 2. The capacity of the holding furnace and the shot sleeve was 500 kg and 15.3 kg liquid aluminum alloy, respectively. All of the components in the model were meshed by high-quality hexahedral element in the ICEM CFD mesh division module in ANSYS. For the pump ditch, the transport tube and the shot sleeve, boundary layer meshes were divided to ensure calculation accuracy, as shown by the bottom right enlarged view of 9

the shot sleeve in Fig. 2. For the EMT process, the transport height is the altitude difference between the surface of the liquid aluminum alloy in the holding furnace and the peak of the lift tube, which is represented by H, see Fig. 2. The most commonly used A380 aluminum alloy in HPDC was chosen for magnetic-flow coupling analysis. The transport current I and the transport height H are the main transport process parameters during EMT. The maximum designed I is 48 A, and the designed H ranges from 350 mm to 500 mm.

Fig. 2. Magnetic-flow coupling model of the EMT of liquid aluminum alloy during HPDC.

The scheme of magnetic-flow coupling analysis should be referred to Dong et al., 2015. The whole harmonic magnetic flux generated by the EMP was first obtained by the harmonic electromagnetic analysis of the EMP in section 3.1. Then the coordinates and the real and imaginary part of the obtained whole harmonic magnetic flux were added into the interface software (USTB, 2013) to calculate the .mag file about the harmonic magnetic flux in liquid aluminum alloy in the pump ditch. Finally, the prepared .mag file was imported into the magnetohydrodynamics (MHD) module in FLUENT for magnetic-flow coupling analysis. The settings of magnetic-flow coupling analysis should be referred to Dong et al., 2015. Transient and pressure-based analysis was adopted, and gravity was considered. The multiphase model was chosen as volume of fluid (VOF), and the viscous model was set as the standard k-ε model. Two phases were defined, one was the air and the other was the liquid A380 aluminum alloy. For the boundary condition, a standard 10

atmospheric pressure was set to surfaces S1, S2 and S3, and the no slip stationary wall condition with a roughness height of 50 μm was set to all of the walls. The time step size and max iterations per time step were set as 0.005 s and 30 respectively during the solving process. The material parameters for magnetic-flow coupling analysis are listed in Table 1. The magnetic parameter resistivity and the fluid parameter viscosity of the liquid A380 aluminum alloy were given by Sklyarchuk et al. (2009), and the fluid parameter density of the liquid A380 aluminum alloy was given by the casting software PROCAST. The correctness of the scheme and the settings of magnetic-flow coupling analysis model about EMT has been verified by our former published EMT experiments (Dong et al., 2015). Finally, the transient transport mass flow rate was recorded during solving, and the liquid A380 aluminum alloy-air phase distribution and the pressure distribution were obtained by post-processing. Table 1 Material parameters of the magnetic-flow coupling model of EMT during HPDC. Magnetic parameters EMP

Fluid parameters

Liquid A380 alloy (700℃)

Relative permeability

Liquid A380 alloy (700℃)

Relative

Resistivity

Density

Viscosity

Core

Yoke

Coil

Air

permeability

(×10-7Ω·m)

(kg·m-3)

(mPa·s)

10000

10000

1

1

1

3.011

2596

0.725

4. Results and discussion 4.1. Effects of structural parameters on EMP The plane induction EMP is the core drive component for the EMT of liquid aluminum alloy during HPDC, and the output efficiency of pump height of the plane induction EMP directly influences the EMT efficiency of liquid aluminum alloy. There are seven structural design parameters of the plane induction EMP, i.e., the magnetic gap δ, the polar distance τ, the thick of the iron core c, the height of the pump ditch h, the width of the iron core W, the width of the coil W', and the width of the pump ditch b, as shown by the front view and the left view of the developed experimental plane induction EMP in Fig. 3. It was found that δ and h significantly 11

influenced the output efficiency of pump height of the plane induction EMP, while the influences of τ and c on the output pump height of the EMP were slight (Dong et al., 2013; Dong et al., 2014). However, the comprehensive effects of W, W' and b on the output pump height of the plane induction EMP are unclear. In this section, the influence mechanism of W, W' and b on the output pump height of the EMP, and the optimum matching between W, W' and b are investigated.

Fig. 3. (a) Front view and (b) left view of the developed experimental plane induction EMP.

4.1.1. Effects of pump ditch and iron core width According to the harmonic electromagnetic analysis of the plane induction EMP in section 3.1, the time average output electromagnetic force FMAGX of the plane induction EMP was calculated. When f is 50 Hz, I is 24 A, h is 12 mm, τ is 180 mm, c is 22 mm and δ is 70 mm, the output FMAGX of the EMP versus b and W is shown in Fig. 4(a). FMAGX increases nearly linear with the increase of W, and it increases gradually with the increase of b. Based on the calculated FMAGX and Eq. (11), the output pump height ΔH of the plane induction EMP was determined, see Fig. 4(b). H (mm) 

FMAGX 109 bh g

(11)

According to Fig. 4(a) and Eq. (11), ΔH also increases nearly linear with the increase of W. However, ΔH first increases rapidly, then increases slowly to the maximum, and finally decreases slowly with the increase of b. When ΔH reaches the maximum ΔHmax, the corresponding b is defined as bmax. The liquid aluminum alloy in the pump ditch of the EMP should be heated during the EMT process. The larger b is, the more power is required for the heat of liquid aluminum alloy, so bmax is not 12

considered as the optimum b since it is too large. Eclectically, when ΔH reaches 90% of the maximum ΔHmax, ΔH is high enough and the corresponding b is appropriate, which can be considered as the optimum condition, and the corresponding ΔH and b are defined as ΔHopt and bopt, respectively. Thus the increase of ΔH versus b can be divided into three steps: (i) ΔH first increases rapidly toward ΔHopt with the increase of b toward bopt in step I; (ii) ΔH then increases slowly from ΔHopt to ΔHmax with the increase of b from bopt to bmax in step II; (iii) ΔH finally decreases slowly from ΔHmax with the increase of b from bmax in step III; as shown in Fig. 4(b). From Fig. 4(b), it is definite that the iron core width (W) and the pump ditch width (b) influence the output pump height of the EMP significantly. When W is designed as 110 mm, comparing with the condition b=W, the output pump height of EMP can be increased by 22.5% when b is designed as the optimum value bopt, as indicated by the circles in Fig. 4(b). The Precimeter company reported a plane induction EMP with b=W, and it is obvious that the output efficiency of pump height of Precimeter’s plane induction EMP is not in the optimum state.

Fig. 4. Effects of iron core width and pump ditch width on the output (a) effective electromagnetic force and (b) pump height of the plane induction EMP.

4.1.2. Influence mechanism of pump ditch width According to the transient electromagnetic analysis of the plane induction EMP in section 3.1, the transient spatial distribution of the magnetic flux density B, the induced current density J and the electromagnetic force FMAG in liquid aluminum 13

alloy in the pump ditch of the EMP were obtained, as shown in Fig. 5, which can be used to explain the effects of the width of the pump ditch on the output pump height of the plane induction EMP in section 4.1.1. From Fig. 5(a), B is mainly along the height direction (Y) of the pump ditch, and the component of B along Y direction is defined as BY. From Fig. 5(b), J forms three groups of swirl rings, and it contains two components, i.e., Jx along the length direction (X) and Jz along the width direction (Z) of the pump ditch, see Eq. (12). According to Eqs. (5,6), FMAG is generated by B and J. From Fig. 5(c), FMAG contains two components, i.e., FMAGX and FMAGZ, see Eq. (13). FMAGX is along the transport direction and contributes to the transport of liquid aluminum alloy, so it is the effective component. FMAGZ is perpendicular to the transport direction and contributes nothing to the transport of liquid aluminum alloy, so it is the invalid component. According to Eqs. (14,15), the effective FMAGX is generated by Jz and BY, and the invalid FMAGZ is generated by Jx and BY, thus Jz and Jx are the effective and invalid component of J, respectively.

J  J X  JZ

(12)

FMAG  FMAGX  FMAGZ

(13)

FMAGX   J Z  BY dV

(14)

FMAGZ   J X  BY dV

(15)

According to Fig. 5(b), the effective component of the induced current density Jz is along the width direction of the pump ditch of the plane induction EMP, and it increases with the increase of the pump ditch width b, so the effective electromagnetic force FMAGX increases with the increase of b based on Eq. (14), as shown in Fig. 4(a). According to Eq. (11), the output pump height ΔH of the plane induction EMP increases with the increase of FMAGX, but it also decreases with the increase of b. With the increase of b, when b is small and in step I, the increase speed of FMAGX is significantly higher than that of b, which results in the rapid increase of ΔH in step I; when b increases to step II, the increase speed of FMAGX is slightly higher than that of b, so it leads to the slow increase of ΔH in step II; when b increases to step III, the increase speed of FMAGX is slightly lower than that of b, thus ΔH decreases slowly in 14

step III, as shown in Fig. 4(b).

Fig. 5. Transient spatial distribution of (a) magnetic flux B (T), (b) induced current density J (A/m2) and (c) electromagnetic force FMAG (N) in liquid aluminum alloy in the pump ditch of the plane induction EMP.

4.1.3. Matching of iron core, coil and pump ditch width According to Fig. 4(b), when the thick of the coil c' is fixed at 13 mm, the optimum pump ditch width bopt varies with the iron core width W, which indicates that there is a matching between W, the coil width W' and b. Table 2 shows the variation of bopt versus W. When W increases from 90 mm, 110 mm and 130 mm to 150 mm, i.e., W' increases from 116 mm, 136 mm and 156 mm to 176 mm, the corresponding bopt increases from 132 mm, 140 mm and 147.2 mm to 154.9 mm, and bmax increases from 180 mm, 190 mm and 200 mm to 210 mm, respectively. bopt is significantly smaller than bmax, while its corresponding output pump height ΔHopt has reached 90% of ΔHmax. With the increase of W from 90 mm, 110 mm and 130 mm to 150 mm, the optimum pump ditch-iron core width ratio bopt/W decreases from 1.47, 1.27 and 1.13 to 1.03, and the corresponding output pump height ratio ΔH(b=bopt)/ΔH(b=W) also decreases proportionally from 1.47, 1.27 and 1.13 to 1.03, while the optimum pump ditch-coil width ratio bopt/W' decreases from 1.14, 1.03 and 0.94 to 0.88, separately. Thus the optimum pump ditch-iron core width ratio bopt/W and the optimum pump ditch-coil width ratio bopt/W' depend on the iron core width W, and the optimum matching between W, W' and b is determined by further discussion. 15

Table 2 variation of the optimum pump ditch width bopt versus the iron core width W.



ΔH b  bopt



W

bmax (mm)

bopt (mm)

W ' =W+2c'

bopt

(mm)

(ΔH=ΔH max)

(ΔH=0.9ΔH max)

(mm)

W

ΔH  b  W 

90

180

132.0

116

1.47

1.44

1.14

110

190

140.0

136

1.27

1.22

1.03

130

200

147.2

156

1.13

1.12

0.94

150

210

154.9

176

1.03

1.02

0.88

bopt W'

From Table 2, the optimum pump ditch width bopt depends on the iron core width W, which results in three different relationships between bopt and the coil width W': (i) bopt > W', (ii) bopt ≈ W' and (iii) bopt < W', as shown in Fig. 6(a), Fig. 6(b) and Fig. 6(c), respectively. According to the electromagnetic theory, the magnetic flux B in the magnetic gap of the plane induction EMP distributes on a width of W' that is directly below the coil, as indicated by the one-way arrows in Fig. 6. For the condition bopt > W', the magnetic flux can’t cover the liquid aluminum alloy near two sides of the pump ditch; thus the liquid aluminum alloy near two sides of the pump ditch is not driven by the electromagnetic force, and it is invalid for transportation, as shown in Fig. 6(a). For the condition bopt < W', the cover area of the magnetic flux is wider than the pump ditch, and there is non-utilization of part of the magnetic flux, which decreases the output efficiency of pump height of the plane induction EMP, as shown in Fig. 6(c). For the condition bopt ≈ W', the magnetic flux covers just the pump ditch and there is no non-utilization of the magnetic flux, and all of the liquid aluminum alloy in the pump ditch is driven by the electromagnetic force, while the output efficiency of pump height of plane induction EMP has reached 90% of the maximum, so it is the most effective condition, as shown in Fig. 6(b). Thus the optimum matching between W, W' and b to ensure the output efficiency of pump height of plane induction EMP is that, b values of bopt corresponding to 90% of the maximum output pump height and bopt ≈ W', i.e., bopt/W = 1.27 and bopt/W' ≈ 1.

16

Fig. 6. Schematic left views of three matching relationships between the optimum pump ditch width and the coil width for the plane induction EMP. (a) bopt > W', (b) bopt ≈ W' and (c) bopt < W'.

4.2. Effects of transport parameters on EMT 4.2.1. Effects of transport current Fig. 7(a) shows the effects of transport current I on the transient EMT mass flow rate of liquid aluminum alloy during HPDC when the transport height H is the minimum 350 mm. When I is 24 A, the transient EMT mass flow rate curve undergoes two jumps before reaching the stable mass flow rate, the first jump corresponds to the start of the outflow of liquid aluminum alloy from the outlet tube, and the second jump corresponds to the full fill of liquid aluminum alloy in the outlet tube. The liquid aluminum alloy fills only part of the outlet tube after the first jump, as indicated by the bottom left inset. The liquid aluminum alloy fills full of the outlet tube after the second jump, as indicated by the bottom right inset. So the EMT capacity is insufficient under 24 A, and the two jumps in the EMT mass flow rate curve due to the insufficiency of EMT capacity result in a turning point in the EMT mass curve, which is unbeneficial for the quantitative control of the EMT mass of liquid aluminum alloy during HPDC. When I are 32 A and 40 A, the transient EMT mass flow rate curves reach the stable mass flow rate only after one jump, which indicates that the liquid aluminum alloy fills full of the outlet tube immediately after outflow. Thus the EMT capacities are sufficient under 32 A and 40 A, and there are no turning points in the corresponding EMT mass curves, which is beneficial for the quantitative control of the EMT mass of liquid aluminum alloy during HPDC. 17

Fig. 7(b) shows the liquid aluminum alloy-air phase fraction in the shot sleeve of HPDC machine at the transient time t during EMT when the transport height H is the minimum 350 mm, and it presents the effects of transport current I on the filling stationarity of liquid aluminum alloy in the shot sleeve. Spatter can act as the direct criteria for the judgement of non-stationarity, while gas entrapment number can be used as the indirect criteria for the judgement of stationarity. When I are 24 A and 32 A, the corresponding stable EMT mass flow rates are 2.91 kg/s and 4.28 kg/s, respectively, and there is no spatter in the shot sleeve, also the gas entrapment number is small, so the fillings of liquid aluminum alloy in the shot sleeve can be considered stationary. When I is 40 A, the stable EMT mass flow rate is as high as 5.53 kg/s, and there is spatter in the shot sleeve, also the gas entrapment number is large, thus the filling of liquid aluminum alloy in the shot sleeve under 40 A can be considered unsteady. According to the above mentioned discussion, in order to the ensure both the EMT efficiency and stationarity of liquid aluminum alloy during HPDC, the optimum transport current is 32 A.

Fig. 7. Effects of transport current on (a) transport mass flow rate and (b) filling stationarity of shot sleeve for the EMT of liquid aluminum alloy during HPDC.

4.2.2. Effects of transport height When the transport current I is the optimum 32 A, the effects of transport height 18

H on the transient and stable EMT mass flow rate of liquid aluminum alloy during HPDC are shown in Fig. 8(a) and Fig. 8(b), separately. With the increase of H from 350 mm to 500 mm, all of the transient EMT mass flow rate curves reach the stable mass flow rate only after one jump, i.e., the liquid aluminum alloy fills full of the outlet tube immediately after outflow and the EMT capacities are sufficient. It is convenient for the quantitative control of the EMT mass of liquid aluminum alloy under different H, since there are no turning points in the EMT mass curves. When H increases from 350 mm, 400 mm and 450 mm to 500 mm, the stable EMT mass flow rate decreases from 4.28 kg/s, 4.12 kg/s and 3.86 kg/s to 3.59 kg/s, respectively.

Fig. 8. Effects of transport height on (a) transient transport flow rate, (b) stable transport flow rate and (c) filling stationarity of shot sleeve for the EMT of liquid aluminum alloy during HPDC.

Fig. 8(c) shows the liquid aluminum alloy-air phase fraction in the shot sleeve of HPDC machine at the transient time t during EMT when the transport current I is 32 A, and it presents the effects of transport height H on the filling stationarity of liquid aluminum alloy in the shot sleeve. With the increase of H from 350 mm to 500 mm, there is no spatter in the shot sleeve, also the gas entrapment number is small and decreases gradually, so the fillings of liquid aluminum alloy in the shot sleeve can 19

always be considered stationary, and the filling stationarity increases with the increase of H. Thus the developed EMT process for HPDC is both stationary and highly efficient with a transport flow rate of above 3 kg/s, and the EMT flow rate decreases with the increase of transport height, while the EMT stationarity increases with the increase of transport height. 4.3. Pressure distribution during EMT The reveal of the pressure distribution during the EMT process is important since it is the basis for the determination of the anti-pressure requirement of the ceramic transport tube. When the transport current is the optimum 32 A and the transport height is 400 mm, Fig. 9(a) shows the transient pressure distribution at time t=2s during the EMT of liquid aluminum alloy for HPDC, and Fig. 9(b) exhibits the transient liquid aluminum alloy-air phase fraction distribution corresponding to Fig. 9(a). From Fig. 9(a), the maximum positive pressure zone A is in the transition tube, and the minimum negative pressure zone B is in the pump ditch.

Fig. 9. Transient (a) pressure (Pa) distribution and (b) liquid aluminum alloy-air phase fraction distribution during EMT of liquid aluminum alloy for HPDC.

Fig. 10(a) shows the variation of the transient maximum positive pressure in zone A and the transient minimum negative pressure in zone B versus the transport 20

time during the EMT of liquid aluminum alloy for HPDC. Both the transient maximum positive pressure in zone A and the transient minimum negative pressure in zone B decrease with the increase of the transport height H, and they fluctuate periodically with a very slight decrease tendency versus the increase of transport time. However, the transient maximum positive pressure in zone A fluctuates slightly with the increase of both transport time and transport height, while the transient minimum negative pressure in zone B fluctuates significantly with the increase of transport time when the transport height is higher than 400 mm, and the fluctuation amplitude of the transient minimum negative pressure in zone B increases with the increase of transport height.

Fig. 10. (a) Transient maximum positive pressure and minimum negative pressure versus transport time and (b) overall maximum positive pressure and minimum negative pressure versus transport height during EMT of liquid aluminum alloy for HPDC.

Fig. 10(b) shows the variation of the overall maximum positive pressure in zone A and the overall minimum negative pressure in zone B versus the transport height during the EMT of liquid aluminum alloy for HPDC. With the increase of the transport height from 350 mm to 500 mm, the overall maximum positive pressure in zone A decreases from 1.8×104 Pa to 1.53×104 Pa, and the overall minimum negative pressure in zone B decreases from -0.76×104 Pa to -1.42×104 Pa. Thus the transport tubes suffer an overall maximum positive pressure of 1.8×104 Pa and an overall minimum negative pressure of -1.42×104 Pa during the EMT of liquid aluminum alloy for HPDC. Based on the reveal of pressure in the transport tube, ceramic tube with an anti-pressure capacity of 0.2MPa that is highly above the maximum positive pressure 21

and minimum negative pressure was selected as the transport tube, and it can effectively resist the damage to the transport tube by pressure during the EMT process. 4.4. Transport efficiency under EMT Fig. 11(a) shows the real HPDC production cycle of a 3 kg aluminum alloy casting when the liquid aluminum alloy is transported by the manipulator process, the total production cycle is 86 s, and the time for clamping (A), soup of liquid aluminum alloy (B), injection and cooling (C), mold opening (D), pickup and spraying (E) is 8 s, 15 s, 21 s, 8 s and 34 s, respectively. Fig. 11(b) shows the transient transport flow rate and mass of liquid aluminum alloy versus transport time during HPDC under the EMT process. When the transport current I is the optimum 32 A, the transport height H is the maximum 500 mm, the transport time of 3 kg liquid aluminum alloy is only 1.78 s under the EMT process, which is significantly shorter than the transport time of 15 s under the manipulator process.

Fig. 11. Comparison of liquid aluminum alloy transport efficiency by EMT with manipulator during HPDC. (a) HPDC cycle for a 3 kg aluminum alloy casting by manipulator and (b) transport flow rate and mass of liquid aluminum alloy versus transport time by EMT.

For the commonly used three types of HPDC machines in industry with clamping forces of 6300 kN, 8000 kN and 10000 kN, the soup amounts of liquid aluminum alloy are in the range of 1.5 ~ 4.5 kg, 2.5 ~ 6.5 kg and 8.0 ~ 12.0 kg, and the suggested transport time of liquid aluminum alloy under the common manipulator process is 16 s, 22 s and 38 s, respectively. For the maximum liquid aluminum alloy 22

soup amounts of 4.5 kg, 6.5 kg and 12.0 kg under the three soup occasions, according to the calculated lowest EMT flow rate under the maximum transport height of 500 mm in Fig. 11(b), the transport time of liquid aluminum alloy is at most 2.195 s, 2.75 s and 4.28 s under the EMT process, and it is obvious that the EMT transport time could be largely shortened under the EMT process. Thus the developed EMT process with plane induction EMP is a process for the transport of liquid aluminum alloy during HPDC with both high transport efficiency and stationarity. 5. Conclusions (1) The output pump height of plane induction EMP was optimized by the matching of the iron core width W, the coil width W' and the pump ditch width b, i.e., b values of bopt corresponding to 90% of the maximum output pump height, bopt/W = 1.27 and bopt/W' ≈ 1. (2) Both the EMT efficiency and stationarity of liquid aluminum alloy during HPDC are achieved under the optimum transport current 32 A. With the increase of the transport height H from 350 mm to 500 mm, the EMT flow rate decreases from 4.28 kg/s to 3.59 kg/s, while the fillings of liquid aluminum alloy in the shot sleeve of HPDC machine are always stationary, and the filling stationarity increases with H. (3) The transient maximum positive pressure in the transition tube and the transient minimum negative pressure in the pump ditch decrease with the increase of H, and they fluctuate with a very slight decrease tendency versus the increase of transport time. The transport tubes suffer an overall maximum positive pressure of 1.8×104 Pa in the transition tube and an overall minimum negative pressure of -1.42×104 Pa in the pump ditch during the EMT of liquid aluminum alloy for HPDC. (4) For the liquid aluminum alloy soup occasions of 4.5 kg, 6.5 kg and 12.0 kg during HPDC, the transport time of liquid aluminum alloy can be shortened significantly from manipulator process’s 16 s, 22 s and 38 s to EMT process’s at most 2.195 s, 2.75 s and 4.28 s, respectively. The developed EMT process with plane induction EMP for HPDC is a process with low cost, high transport efficiency and stationarity. 23

Acknowledgements The work was supported by the National Basic Research Program of China (2013CB632203), the National Science and Technology Major Project of the Ministry of the Science and Technology of China (2011ZX04001-071) and the National Key Technology R&D Program of China (2011BAE21B00).

24

References Bachmann, M., Avilov, V., Gumenyuk, A., Rethmeier, M., 2014. Experimental and numerical investigation of an electromagnetic weld pool support system for high power laser beam welding of austenitic stainless steel. J. Mater. Process. Technol. 214, 578-591. Campbell, J., Wilkins, P.S.A., 1982. New developments in light alloy founding. In: Institute of British Foundrymen, 79th Annual Conference, Brighton, Britain, pp. 250-253. Cheng, J., Peng, Y., Hou, J., 2000. Planar DC electromagnetic pump for casting of Al-alloy. China Patent CN 1255768-A. Cminovacast, 2015. www.cminovacast.com. Cui, X.H., Mo, J.H., Li, J.J., Zhao, J., Zhu, Y., Huang, L., Li, Z.W., Zhong, K., 2014. Electromagnetic incremental forming (EMIF): A novel aluminum alloy sheet and tube forming technology. J. Mater. Process. Technol. 214, 409-427. Dong, X., He, L., Huang, X., Li, P., 2015. Coupling analysis of the electromagnetic transport of liquid aluminum alloy during casting. J. Mater. Process. Technol. 222, 197-205. Dong, X., He, L., Huang, X., Li, P., 2015. Effect of electromagnetic transport process on the improvement of hydrogen porosity defect in A380 aluminum alloy. Int J Hydrogen Energy 40, 9287-9297. Dong, X., He, L., Mi, G., Li, P., 2014. Simulative and experimental study on electromagnetic pump supplying liquid alloys during casting: modeling and effects of design parameters on pump height. Appl. Mech. Mater. 470, 441-446. Dong, X., Mi, G., He, L., Li, P., 2013. 3D simulation of plane induction electromagnetic pump for the supply of liquid Al-Si alloys during casting. J. Mater. Process. Technol. 213, 1426-1432. Haghayeghi, R., Ezzatneshan, E., Bahai, H., 2015. Experimental-numerical study of AA5754 microstructural evolution under electromagnetic ultrasonic merged fields. J. Mater. Process. Technol. 225, 103-109. 25

Kore, S.D., Date, P.P., Kulkarni, S.V., 2008. Electromagnetic impact welding of aluminum to stainless steel sheets. J. Mater. Process. Technol. 208, 486-493. Le, Q., Guo, S., Zhao, Z., Cui, J., Zhang, X., 2007. Numerical simulation of electromagnetic DC casting of magnesium alloys. J. Mater. Process. Technol. 183, 194-201. Mercer, II, W.E., King, H.L., Baker, C.F., 1995. Method and apparatus for handling molten metals. US Patent 5,407,000. Miao, X., Timmel, K., Lucas, D., Ren, Z., Eckert, S., Gerbeth, G., 2012. Effect of an Electromagnetic Brake on the Turbulent Melt Flow in a Continuous-Casting Mold. Metall. Mater. Trans. B 43B, 954-972. Miura, K., Teruyama, Y., Tomita, R., 2009. Induction type electromagnetic pump for molten metal used in e.g. die casting, has nozzle for ejecting gas towards top portion of weir which changes from up-gradient to down-gradient in supply direction of molten metal. Japan Patent JP 2009000689-A. Precimeter, 2015. www.carli-ea.com/eng/c3btb.htm. Psyk, V., Risch, D., Kinsey, B.L., Tekkaya, A.E., Kleiner, M., 2011. Electromagnetic forming-A review. J. Mater. Process. Technol. 211, 787-829. Saito, H., Sagamihara., Kubota, S., Yokohama., Motomura, N., Zama., 1989. Electromagnetic pump type automatic molten-metal supply apparatus. US Patent 4,828,460. Singh, R., Thomas B.G., Vanka, S.P., 2014. Large eddy simulations of double-ruler electromagnetic field effect on transient flow during continuous casting. Metall. Mater. Trans. B 45B, 1098-1115. Sklyarchuk, V., Plevachuk, Y., Yakymovych, A., Eckert, S., Gerbeth, G., Eigenfeld, K., 2009. Structure Sensitive Properties of Liquid Al–Si Alloys. Int. J. Thermophys. 30, 1400-1410. University of Science and Technology Beijing., 2013. Interface software between ANSYS and the MHD module (Magnetohydrodynamics) in FLUENT V1.0. China Software Copyright 2013SR008175. Wang, J., Li, P., Mi, G., Zhong, Y., 2010. Microstructural evolution caused by 26

electromagnetic stirring in superheated AlSi7Mg alloys. J. Mater. Process. Technol. 210, 1652-1659. Weddeling, C., Demir, O., Haupt, P., Tekkaya, A., 2015. Analytical methodology for the process design of electromagnetic crimping. J. Mater. Process. Technol. 222, 163-180. Yu, H., Wang, B., Li, H., Li, J., 2008. Influence of electromagnetic brake on flow field of liquid steel in the slab continuous casting mold. J. Mater. Process. Technol. 202, 179-187. Zhang, J., Liu, S., Zhang, Y., Dong, Y., Lu, Y., Li, T., 2015. Fabrication of woven carbon fibers reinforced Al-Mg (95-5 wt%) matrix composites by an electromagnetic casting process. J. Mater. Process. Technol. 226, 78-84.

27

Figure Captions Fig. 1. (a) EMT process with plane induction EMP and (b,c) Common manipulator process for the transport of liquid aluminum alloys during HPDC. Fig. 2. Magnetic-flow coupling model of the EMT of liquid aluminum alloy during HPDC. Fig. 3. (a) Front view and (b) left view of the developed experimental plane induction EMP. Fig. 4. Effects of iron core width and pump ditch width on the output (a) effective electromagnetic force and (b) pump height of the plane induction EMP. Fig. 5. Transient spatial distribution of (a) magnetic flux B (T), (b) induced current density J (A/m2) and (c) electromagnetic force FMAG (N) in liquid aluminum alloy in the pump ditch of the plane induction EMP. Fig. 6. Schematic left views of three matching relationships between the optimum pump ditch width and the coil width for the plane induction EMP. (a) bopt > W', (b) bopt ≈ W' and (c) bopt < W'. Fig. 7. Effects of transport current on (a) transport mass flow rate and (b) filling stationarity of shot sleeve for the EMT of liquid aluminum alloy during HPDC. Fig. 8. Effects of transport height on (a) transient transport flow rate, (b) stable transport flow rate and (c) filling stationarity of shot sleeve for the EMT of liquid aluminum alloy during HPDC. Fig. 9. Transient (a) pressure (Pa) distribution and (b) liquid aluminum alloy-air phase fraction distribution during EMT of liquid aluminum alloy for HPDC. Fig. 10. (a) Transient maximum positive pressure and minimum negative pressure versus transport time and (b) overall maximum positive pressure and minimum negative pressure versus transport height during EMT of liquid aluminum alloy for HPDC. Fig. 11. Comparison of liquid aluminum alloy transport efficiency by EMT with manipulator during HPDC. (a) HPDC cycle for a 3 kg aluminum alloy casting by manipulator and (b) transport flow rate and mass of liquid aluminum alloy versus transport time by EMT. 28

Tables Table 1 Material parameters of the magnetic-flow coupling model of EMT during HPDC. Table 2 variation of the optimum pump ditch width bopt versus the iron core width W.

29