OPTIMIZATION OF PROCESS PARAMETERS TO IMPROVE COMBINATION IN DUPLEX ROLLER SLEEVE Zhi-hong Guo Key Lab of Metastable Materials Science and Technology and College of Material Science and Engineering, Yanshan University, Qinhuangdao, China Hebei Key Laboratory of Material Near-Net Forming Technology and College of Material Science and Engineering of Hebei University of Science and Technology, Shijiazhuang, China Fu-ren Xiao and Bo Liao Key Lab of Metastable Materials Science and Technology and College of Material Science and Engineering, Yanshan University, Qinhuangdao, China Su-ling Lu and Rui-ling Liu Hebei Key Laboratory of Material Near-Net Forming Technology and College of Material Science and Engineering of Hebei University of Science and Technology, Shijiazhuang, China

Copyright Ó 2016 American Foundry Society DOI 10.1007/s40962-016-0078-7

Abstract In this study, to improve the interfacial bonding between the inner sleeve and outer layer of a duplex roller sleeve, a model of the uphill casting of a roll sleeve was constructed with a commercially available process modeling software and simulated the temperature fields in the casting process. Furthermore, the effects of the pouring temperature, the preheating temperature of the steel sleeve and the preheating temperature of the sand on the boundary temperature between the casting and 1045 steel sleeve were simulated. The results show that the lower boundary temperature (1268 °C) between the casting and 1045 steel sleeve leads to poor bonding strength because the metallurgical bonding layer does not form. Increasing the pouring temperature (1450–1600 °C), the preheating temperature of the steel sleeve (600–1200 °C) and the preheating temperature of the sand (60–260 °C) can

increase the boundary temperature, which favors the formation of a metallurgical bonded layer and enhances the bonding strength between the casting layer and 1045 steel sleeve. Finally, 1550 °C pouring temperature, and the preheating temperature of the inner sheath and casting sand of 1000 and 160 °C, respectively, are adopted in casting a roll sleeve, when other conditions remain unchanged. The results of the samples collected in the middle of the casting suggest that the maximum surface temperature of the inner sheath should be above the solidus temperature and a metallurgical bonded layer will form.

Introduction

a roller core of low alloy or carbon steel. These have been extensively used because of lower cost.1,2 The bonding properties of two or more material workpieces are of major concern.3,4 The duplex roller sleeve is normally prefabricated by casting a layer of wear-resistant material on the outside surface of the steel sleeve. For the prefabricated duplex roller sleeve,

As a principal part in rolling mill, the mill roll must have excellent wear resistance and sufficient fracture toughness; therefore, the duplex rolls have been widely used. The duplex rolls consist of a wear resistant high-alloy cast iron or steel with

International Journal of Metalcasting

Keywords: roller sleeve, uphill casting, temperature field, metallurgical bonding, 1045 steel sleeve, high chromium cast iron

Materials and Procedures

the interface bonding strength between the cast wear-resistant material and the steel sleeve become a very important factor. The duplex roller sleeve, which needs the formation of a complete metallurgical bonding layer between the cast wearresistant materials and the steel sleeve during the casting process.

Problem Descriptions and Model Construction The duplex roller sleeve consisted of high-chromium cast iron on the outside and 1045 steel sleeve on the inside. The thicknesses of the high-chromium cast iron and sleeve of ASTM 1045 steel are 75 and 20 mm, respectively. The chemical composition of HCCI is given in Table 1. The duplex roller sleeve was manufactured by means of uphill casting. However, a lot of problems were found in the duplex roller sleeves. The bonding strength of the interface zone was very poor, even peeling of the HCCI layer occurred in some duplex roller sleeves, as shown in Figure 1a. Figure 1b, c shows the microstructure of the interface zone. The interface zone with about 20 lm thickness could be observed, while the interface zone having smooth layer on 1045 steel surface and weak combination only appeared at the grain boundary of 1045 steel (Figure 1b). The microstructure of 1045 steel show ferrite and pearlite. Further observation found that the interface zone was a carbide-free layer composed of laminar structures (Figure 1c), which is different from the microstructure of HCCI consisted of large amounts of (Fe,Cr)7C3 distributed in the matrix of martensite and retained austenite.11,12 The results indicate that the interface boundary may have a lower temperature and shorter holding time during the pouring and cooling processes; as a result, the metallurgical bonding layer does not form. From this point of view the lack of metallurgical bonding formation, the interface temperature and holding time at high temperature are very important factors affecting the formation of the metallurgical bonding layer. Therefore, some possible factors that may influence the temperature field at the interface were simulated.

However, during the actual production process, there are many parameters affecting the quality of the metallurgical bonding layer, such as the casting parameters, the state of steel sleeve and the mold parameters.5–7 So various parameters could affect the microstructure and properties of the bonding layer; therefore, it is difficult to optimize these parameters and improve the microstructure and properties of the metallurgical bonding layer in the actual production. The commercially available process modeling software (ProCast) has been extensively employed to simulate the complex casting process.8–10 The simulation software has allowed us to conduct many virtual experiments to arrive at a of optimal process parameters for the duplex roller sleeve. In this work, the causes of cracks or peeling of the duplex roller sleeve made by uphill casting of high-chromium casting iron (HCCI) on the outer-surface of ASME 1045 steel were analyzed. The modeling of temperature field during the casting and cooling process of the duplex roller was established by the computer-aided engineering software. Furthermore, the effects of some key process parameters, such as preheating temperature of the steel sleeve, the pouring temperature of HCCI and the preheating temperature of the casting sand mold, and the interface temperature of outside surface of the sleeve of ASME 1045 steel were simulated. The simulation results were then used to optimize the process parameters of uphill casting and improve the adhesion strength of the two materials.

Table 1. Chemical Composition of the Casting (wt%) Materials

C

Si

Mn

P

S

Cr

Ni

Mo

V

W

Cr20

2.1/2.3

0.5/0.6

0.3/0.5

\0.03

\0.02

21/23

1.5/1.6

1.2/1.3

0.15/0.25

0.2

Figure 1. Flaking photos (a); microstructure (b) and SEM micrographs (c) of the roll sleeve.

International Journal of Metalcasting

Figure 2. Schematic representation (a) and the grid data (b) of the roll sleeve by uphill casting process. A, B—actual measurement point; C—casting outer-surface temperature point; D—1045 steel sleeve outer-surface temperature point.

A geometric model of the uphill casting process of the duplex roller sleeve is shown in Figure 2. Point A and B represent the locations of temperature measurement, and point C and D represent the outer and inner surface of the cast layer, respectively. In order to construct the mathematical model, the following assumptions were established: (1) The incoming liquid metal is evenly distributed only at the inlet of the gravity flow part to simulate the smooth filling process of the channel; (2) The liquid metal is incompressible Newton fluid; (3) The inlet is on the cross section of runner, and the filling velocity is parallel to the direction of gravity. The natural convection of the liquid metal is described by the mass conservation equation and the Navier–Stokes equations.9,13 Boundary Conditions and Loading

2000 W/m2 K was assigned between metal and metal; 500 W/m2 K was assigned to the metal–sand interface; and 200 W/m2 K was assigned to the sand–sand interface.14 The external cylindrical surface of the mold was divided into two parts: the cast iron mold and the others. A convection coefficient varying over time was applied to the interface between the two parts and air. The ambient temperature of the pit surrounding air was approximately 40 °C. The convection coefficient of 3 W/m2 °C was applied to the inner cylindrical surface of the sheath. The ambient temperature of the surrounding air was approximately 600 °C. An initial temperature of 75 °C was applied to the cold mold and coating. The inner sleeve was preheated to 600 °C. An initial temperature of 1530 °C was applied to the casting, representing the pouring temperature of the material used in casting the roll sleeve. The other parts’ temperature was 60 °C. A casting speed of 8.5 kg/s was applied to the channel flow.

In this work, the cold mold was F12101; the sand boxes were low carbon steel; the runner channel was silica-sand; and the coating was a type of sand made by mixing quartz powder, bentonite, sodium carbonate and water at a proper ratio. However, the material performance parameters of the casting were not selected from the material data. According to the chemical composition (Table 1), the material performances, such as the enthalpy curve and the solidification path, were computed using Thermo-Calc software 4.0 based on TCFE7 database. The liquidus temperature of 1320 °C and the solidus temperature of 1275 °C were measured by DSC (differential scanning calorimetry) and utilized in the simulations. The inner sleeve was 1045 steel, whose liquidus temperature was 1493 °C and solidus temperature was 1415 °C.

All ingredients were melted in a medium-frequency induction-melting furnace. Then, the elements and alloy contents were adjusted. At the same time, the 1045 steel sleeve was preheated and the coating formed by centrifugal spraying. The uphill casting started when the end temperature and components of molten metal meet the requirements. Temperature was measured with an infrared thermometer at different times and at the A and B positions of the mold part shown in Figure 2, and then the average of each set of data was taken as the actual measured temperature at each location.

There were 20 interfaces between ten parts, and the interface heat transfer coefficient at each interface was determined according to the interface type (metal–metal, metal– sand and sand–sand). An interface heat transfer coefficient of 5000 W/m2 K was applied to liquid metal and metal;

Figure 3 shows the simulated temperature field of the duplex roller sleeve at different times from actual trials. Because the pouring completed at about 65 s in the actual process, when the cast pouring time is 30 s, the

International Journal of Metalcasting

Results and Discussion

Figure 3. Temperature distribution simulation results: (a) at 30 s; (b) at 106 s; (c) at 2800 s.

temperature field only appears in the half of the casting shown in Figure 3a. The temperature of the outer surface of the steel sleeve near the molten metal increases significantly; in the meanwhile, the temperature of the molten metal near the steel sleeve outer surface decreases considerably, while the temperature of molten metal near the mold changes a little. The results indicate that a large amount of heat transferred from the metal to the steel sleeve, and the heat transferred outward is not significant because the heat transfer coefficient of the coating–metal interface is small. At 106 s, the temperature of the molten metal decreases, and the temperature of the steel sleeve increases to a higher temperature, and the distribution of temperature becomes more uniform. However, the liquid core at the top point of the casting is wider because the high-temperature liquid finally reaches the top of the casting (Figure 3b). At 2800 s, the temperature of the duplex roller sleeve decreases remarkably, while both ends of top and bottom have a higher temperature and the middle part has the lowest temperature (Figure 3c). The simulated temperature of point A and B in Figure 2 is then verified with the temperature measured at corresponding locations. From the simulated result in Figure 3,

Figure 4. Comparison between the actual measured and the predicted temperature.

the temperature of the mold has significant correlation with temperature of the casting; therefore, the change in temperature of mold could reflect the temperature variation of the casting. The temperatures of point A and B in Figure 2 corresponding to the upper sand box and the middle cold mold were measured, respectively, as shown in Figure 4. The calculated results based on the appropriate parameters obtained through multiple iterations are also illustrated in Figure 4. The results demonstrate that the calculated result are in good agreement with the measured temperature, and the parameters defined in the simulation process simulation are reasonable. In addition, the casting temperature of point C shown in Figure 2 was 108 °C when opening flask, and the simulated temperature of this point was 110.7 °C. The small difference suggests good agreement between the computed and measured results. From the results stated in Figure 3, the temperature change in the entire casting directly influences the temperature distribution of 1045 steel sleeve and mold. The temperature of the boundary between the molten metal and the steel sleeve strongly reflects the bonding strength, and the temperature of the mold could reflect the precision of the modulated parameters. The ideal metallurgical bond formed when the high-temperature liquid metal reached the surface of the 1045 steel and the surface was melting. Then the cast iron solidified on the mixed layer. The surface of the 1045 steel would not melt if the temperature was too low, and the low temperature would effect the bonding layer. The crack and the peeling would appear on the interface zone during the deformation or temperature change because of a great difference of properties. Therefore, the temperature distribution along the cross section of all components at the same height of point B is illustrated in Figure 5a, and the variation of temperature with time at the boundary between the molten metal and the steel sleeve is illustrated in Figure 5b. From Figure 5a, as the time is 65 s, just after the completion of pouring, the temperature of the steel sleeve outside (D point in Figure 2) increases quickly from 600 to 1268 °C, and the temperature of the steel sleeve inside is still lower. Meanwhile, the

International Journal of Metalcasting

Figure 5. Temperature distribution of cross section at different time (a) and temperatures of the 1045 steel sleeve and the casting (D point) (b).

temperature of the cold mold changes a little. Thus, the high-temperature gradient is formed in the steel sleeve and the mold. However, the highest temperature of the casting is near the cold mold, suggesting that more heat was transferred from the metal to the steel sleeve. At 248 s, the temperature distribution of the steel sleeve becomes uniform and the outside temperature of it increases to 1310 °C. In the meanwhile, the temperature of the casting decreases and the temperature of cold mold increases. The heat mainly transfers to the cold mold. At 1052 s, the temperature of the steel sleeve decreases less, but the temperature of the casting decreases remarkably, and the temperature of the cold mold still continuously increases. The heat transmission becomes mainly outward. As an important parameter, the high temperature and its holding time at the boundary between the steel sleeve and the casting determines the microstructure and strength of the bonding layer; thus, the variations of the temperatures of the steel sleeve outside and the casting inside with time are illustrated in Figure 5b. At the completion of pouring, the temperature of the steel sleeve outside quickly raises from 600 to 1268 °C, while the temperature of the steel sleeve outside and the casting inside decreases to 1227 and 1254 °C, respectively, because a large amount of heat is transferred and assimilated by the steel sleeve. Then, the temperature increases again and reaches 1303 °C for the steel sleeve outside and 1314 °C for the casting inside, at 305 s. Subsequently, the temperature decreases continuously. From the results stated above, it is not difficult to understand that weak bonding strength appears in the duplex roller sleeve under these process conditions (Figure 1). For 1045 steel, the solidus temperature is 1415 °C, and the liquidus temperature is 1493 °C. However, as the molten metal is poured, the temperature of the steel sleeve surface only gets to 1268 °C and then decreases to 1227 °C. In the meanwhile, the temperature of the molten metal is also decreased to 1254 °C, which is lower than 1320 °C, the

International Journal of Metalcasting

liquidus temperature of HCCI, and even lower than 1275 °C, the solidus temperature of HCCI. Although the temperatures of the steel sleeve outside surface and the casting inside rise again to 1303 and 1314 °C, respectively; both are much lower than the solidus temperature of 1045 steel, and slightly higher than the solidus temperature of HCCI. There is not sufficient heat to melt the steel sleeve surface. In the meanwhile, the cast HCCI near the steel sleeve surface quickly solidify. A stratified microstructure is formed between 1045 steel and high-chromium cast iron, compared with typical high-chromium cast iron group, there is no carbide in the microstructure and is called carbide-free zone. It is difficult to form the metallurgical bonding between the steel sleeve and the cast HCCI. Hence, increasing the boundary temperature for the steel would effectively improve the microstructure and the strength of the bonding layer. The results stated above indicate that the bonding strength of the duplex roller sleeve strongly depends on the temperature of the steel sleeve outer surface and the HCCI alloy inner surface during the casting and cooling processes, while the quick decrease in temperature of the bonding layer depends on the heat absorbed by and transferred to the steel sleeve, and the heat absorbed by and transferred from the mold. Therefore, the pouring temperature, preheating temperature of the steel sleeve and casting sand may play important roles in determining the temperature of the bonding layer of the duplex sleeve. Hence, the effects of the three factors on the temperature of the steel sleeve outer-surface were simulated and the results are shown in Figure 6. Figure 6 shows the simulated results of the effects of the pouring temperature, preheating temperature of the steel sleeve and mold on the temperature of the steel sleeve outer-surface (at point D in Figure 2). From the results stated above, a large temperature gradient forms on the cross section of the steel sleeve, and the heat from the casting is quickly absorbed and transferred because of the

Figure 6. Influence of parameters: preheating temperature of the 1045 steel sleeve (a); pouring temperature (b); preheating temperature of the casting sand (c) on the temperature of inner sleeve (D point).

lower preheating temperature of the steel sleeve. Therefore, increasing the preheating temperature of the steel sleeve temperature does remarkably increase the temperature of steel sleeve outer-surface. More importantly, increasing the temperature of the steel sleeve outer-surface decreases the thermal gradient and this gives greater time and thermal energy to partially melt the outside of the steel as shown in Figure 6a. When the pouring temperature is 1530 °C and the preheating temperature of the mold is 60 °C, the temperature of the outer-surface of steel sleeve increases notably when the pouring is just completed. In addition, the decreasing of temperature accompanied by the holding time reduces is shown in Figure 6a. The temperature rise and drop depend mainly on the heat absorption and transmission from the molten metal to the steel sleeve and then from the steel sleeve to the environment. With the increase in the preheating temperature, the temperature of the outersurface of the steel sleeve increases more notably, and the peak temperature increases from 1310 to 1327, 1377 and 1428 °C when the preheating temperature is increased from 600 to 800, 1000 and 1200 °C, respectively. The results indicate that the temperature of the outer-surface of steel sleeve increases to 1377 °C when the preheating temperature is 1000 °C and then the time the boundary temperature is above the liquidus temperature of the HCCI

is also prolonged and near to the solidus temperature of 1045 steel, favoring the formation of metallurgical bonding layer. The temperature of the steel sleeve boundary depends mainly on the heat absorbed by and transmitted to the steel sleeve, so increasing the heat input may increase the temperature of the outer-surface of the steel sleeve. Figure 6b shows the effects of the pouring temperature on the temperature of the outer-surface of the steel sleeve as the preheating temperatures of the steel sleeve and the mold are 1000 and 60 °C, respectively. Although increasing the pouring temperature strongly raises the temperature of the steel sleeve outer-surface, the higher pouring temperature can result in coarsening microstructure and decreasing the mechanical properties of the HCCI. Therefore, the appropriate pouring temperature is considered to be in the range of 1530–1550 °C. In addition, for the casting, large heat transmission occurs from the casting to the mold, thus the preheating temperature of the casting sand can also influence the temperature of the molten metal and further affects the variation of the temperature of steel sleeve outer surface. Figure 6c shows the effects of the preheating temperature of casting sand on

International Journal of Metalcasting

Figure 7. Metallographic image (a); scanning electron microscopy (SEM) (b). Table 2. EDS Analysis of the Point E, F, G, H and I in Figure 7b/wt% Point

C

Cr

Fe

Si

Ni

Mo

E

0.54

–

99.22

0.24

–

–

F

1.35

2.34

95.88

0.43

–

–

G

3.30

7.26

85.95

0.81

1.11

0.50

H

3.30

11.07

82.04

0.98

1.26

0.59

I

1.78

11.61

81.94

0.82

1.72

–

the temperature of the steel sleeve outer surface for 1530 °C pouring temperature and 1000 °C preheating temperature of steel sleeve. As the casting sand preheating temperature increases from 60 to 160 °C, the temperature of steel the sleeve outer surface increases from 1366 to 1416 °C, while further increasing the preheating temperature causes little increase in the temperature of the steel sleeve outer surface. In this work, the casting sand is located in the runner channel, upper sand box and lower sand box. Increasing the casting sand preheating temperature can decrease the heat loss during the pouring; as a result, the temperature of the steel sleeve outer-surfaces increases. But the rise of the target temperature is small because its heat dissipation is increased as sand temperature increases. From the simulated results stated above, increasing the pouring temperature, preheating temperature of the steel sleeve and the casting sand can increase the temperature of the steel sleeve outer-surface and prolong the time the HCCI remains liquid, which would favor the formation metallurgical bonding layer of the duplex roller sleeve. Therefore, appropriate pouring temperature in the range of 1530–1550 °C, steel sleeve preheating temperature in the range of 800–1000 °C and casting sand preheating temperature at about 160 °C were considered to improve the microstructure and bonding strength of the duplex roller sleeves. In order to verify the feasibility of the processes, the 1550, 1000 and 160 °C pouring temperature, preheating

International Journal of Metalcasting

temperature of steel sleeve and casting sand, respectively, are adopted in casting a roll sleeve under the conditions of the other parameters remain the same as the original process. Figure 7 shows the microstructure of the bonding layer of the trial. The uniform black boundary layer forms between the HCCI and 1045 steel in Figure 7a, and the interface zone is about 35–40 lm. Compared with Figure 1, the boundary near 1045 steel appears irregular, which maybe strongly increases the bonding strength. From the SEM observation, the interface zone includes the molten layer and the diffusion layer. The molten layer formed when the high-temperature liquid metal reached the surface of 1045 steel and the surface was melting. The high temperature improved the element diffusion and the diffusion layer formed. The uniformly distributed, finegrained carbides dispersed in the matrix. The good metallurgical bonding attributes to the melting of steel and diffusion joint between casting HCCI and 1045 steel, which results in the formation of continuous transition layer. The results can be confirmed by chemical elements scanning analysis, as shown in Table 2. Point E was located in the 1045 steel side and included mainly C, Fe and Si elements. Point I was located in matrix of the HCCI, and its C content is the smaller than the point G or H because the large amounts of (Fe,Cr)7C3 formed. From point F to Point H, the C, Cr content increase continuously, this attributed higher C, Cr content in the HCCI. The Fe content decreases continuously from 1045 steel to the HCCI. The tensile strength of sample is 340 MPa, which is much higher than that of the original sample of 135 MPa. The results verify that proper casting process can improve the microstructure and bonding strength of the duplex roller sleeve. From the results stated above, increasing the surface temperature for the steel would effectively improve the strength of the bonding layer. Increasing the pouring temperature, preheating temperature of steel sleeve and casting sand could strongly increase the temperature of steel sleeve outer-surface and its holding time at liquid. The high temperature would melt the 1045 steel surface and improved the element diffusion, so the ideal metallurgical bond formed.

Conclusions 1.

2.

3.

For a duplex roller sleeve consisting of HCCI in the outer-layer and 1045 steel sleeve in the inside, the cracking or peeling of the HCCI layer during subsequent heat treatment or application is mainly caused by poor bonding strength between HCCI and 1045 steel, and the metallurgical bonding layer is not formed because there is not enough heat to melt the steel sleeve surface. The simulation model of the duplex cast roller sleeves during uphill casting process was established with the computer-aided engineering software, and the validity of the model was verified by comparing the calculated results with the experimental data. Increasing the preheating temperature of the steel sleeve, the pouring temperature of HCCI and the preheating temperature of casting sand can increase the boundary temperature between the cast HCCI and 1045 steel sleeve. The preheating temperature of the steel sleeve or the pouring temperature of HCCI has a significant influence. Furthermore, the microstructure is improved and the bonding strength of the duplex roller sleeve is enhanced. Simulation of each kind of situation, 1550, 1000 and 160 °C for three temperatures, respectively, were adopted in casting a roll sleeve. Increasing the temperature of the steel sleeve outer-surface decreases the thermal gradient, and this gives greater time and thermal energy to partially melt the outside of the steel, so a qualified duplex roller sleeve with a good metallurgical bonding layer is obtained.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

T. Noguchi, K. Asano, S. Hiratsuka, H. Miyahara, Int. J. Cast Metal. Res. 21(4), 219–225 (2008) T. Noguchi, N. Horikawa, H. Nagate, T. Nakamura, K. Sato, Int. J. Cast Metal. Res. 18(4), 214–220 (2005) S. Ghadimi, M. Sedighi, F. Djavanroodi, A. Asgari, Mater. Manuf. Processes 30(10), 1256–1261 (2015) H.F. Liu, Ph.D. Thesis of Jilin University, 2000, pp. 108–110 (chinese) J.W. Shi, D.X. Yang, F. Ni, R. Long, Hot Working Technology, 2005, 6, pp 6–8 (chinese) J. Liu, C. Liu, Mater. Manuf. Processes 30(4), 563–568 (2015) J. Roucka, M. Kova´c, M. Jarosˇ, O. Sˇikula, K. Hrba´cek, B. Podhorna´, Int. J. Metalcast. 10, 71–80 (2010) Y.X. Yan, Q. Sun, J.J. Chen, H.K. Pan, Mater. Manuf. Processes 30(10), 1174–1178 (2015) S.L. Lu, F.R. Xiao, S.J. Zhang, W.N. Mao, B. Liao, Appl. Therm. Eng. 73, 512–521 (2014) N.N. Song, Y.K. Luan, Y.L. Bai, Z.A. Xu, X.H. Kang, D.Z. Li, J. Mater. Sci. Technol. 28, 147–154 (2012) S. Inthidech, P. Sricharoenchai, Y. Matsubara. Int. J. Metalcast. 25–34 (2012) Ali Hadji, Khedidja Bouhamla, Hichem Maouche, Int. J. Metalcast. 10(1), 45–55 (2016) ProCast User Manual Version 2009. 1. ESI group. The virtual try-out space company (2009) S.L. Lu, F.R. Xiao, Z.H. Guo, L.J. Wang, H.Y. Li, B. Liao, Appl. Therm. Eng. 93, 518–528 (2016)

Acknowledgments This research is supported by Hebei Science and Technology Support Program (13274202D) and Hebei University of Science and Technology fund (XL201002).

International Journal of Metalcasting