Mechanism and application of a newly developed pressure casting process: horizontal squeeze casting

Special Report CHINA FOUNDRY Celebrating the Mechanism and application of a newly developed pressure casting process: horizontal squeeze casting *Li Peijie 1, Huang Xiusong 1, He Liangju 1, Liu Xiangshang 2, and Wang Benci 2 1. Department of Mechanical Engineering, Tsinghua University, Beijing 100084, China; 2. L. K. Technology Shanghai Yi Da Machinery Co. Ltd., Shanghai 200000, China Abstract: Compared to traditional high-pressure die casting (HPDC), horizontal squeeze casting (HSC) is a more promising way to fabricate high-integrity castings, owing to a reduced number of gas and shrinkage porosities produced in the casting. In this paper, the differences between HSC and HPDC are assessed, through which it is shown that the cavity filling velocity and the size of the gating system to be the most notable differences. Equipment development and related applications are also reviewed. Furthermore, numerical simulation is used to analyze the three fundamental characteristics of HSC: slow cavity filling, squeeze feeding and slow sleeve filling. From this, a selection principle is given based on the three related critical casting parameters: cavity filling velocity, gate size and sleeve filling velocity. Finally, two specific applications of HSC are introduced, and the future direction of HSC development is discussed. Key words: horizontal squeeze casting; high-integrity; cavity filling; squeeze feeding; sleeve filling CLC numbers: TG249.2 Document code: A Article ID: 1672-6421(2014)04-232-07 H igh pressure die casting (HPDC) is widely used to produce complex and thinwalled light metal parts, especially in the automotive industry [1], due to its high productivity, good dimensional accuracy and surface quality of castings. The high cavity filling velocity and thin gate used in HPDC, however, result in a large number of porosity defects in the casting, which limit its application in load-bearing parts. Thus, improving the internal quality of parts made by pressure casting has become a significant research focus [2]. The recent application of high-integrity pressure casting processes such as semisolid die casting [3], vacuum-assisted HPDC [4], horizontal squeeze casting (HSC) [5] and double control forming [6] has brought some hope to the idea of producing load-bearing parts by pressure casting. Among these processes, HSC offers the advantage of both a low porosity and a fine microstructure, which is similar to the casting achieved with traditional vertical squeeze casting (VSC) [7]. Although VSC is not sensitive to gas entrapment and flow turbulence due to its filling sequence being from bottom to top, it does have its disadvantages in terms of the large volume and complex operation required. In contrast, the use of a horizontal shot system with HSC means that the equipment needed is much simpler and cheaper, and can be easily manufactured from existing HPDC machinery [8]. Furthermore, recent reports of the successful fabrication of thick-walled engine bracket mountings by HSC prove the potential of this process [9]. * Li Peijie Born in 1962, Professor. He received his B.S., M.S. and Ph.D. degrees from Harbin Institute of Technology in 1984, 1989 and 1995, respectively. The topic of his Ph.D. dissertation is the heredity of liquid structure of aluminum alloy. After graduation, he worked at Harbin Institute of Technology until 2000, and since then has worked at Tsinghua University. Li Peijie is the director of National Centre of Novel Materials for International Research, Tsinghua Universtiy. His research interests mainly focus on the heredity of liquid structure of light metals, solidifying structure and performance control of light metals, and equipments for processing of aluminum and magnesium alloys. He is the author and co-author of more than 100 papers and 30 patents. E-mail: lipj@mail.tsinghua.edu.cn Received: 2014-05-27 Accepted: 2014-06-25 232

Celebrating the Special Report CHINA FOUNDRY 1 Differences between HSC and HPDC The process of HSC is similar to HPDC, in that the melt is first poured into the shot sleeve, and then fills the shot sleeve and die cavity when the plunger advances. After filling, the melt solidifies under the pressure from the plunger and the shrinkage in the casting is fed by the alloy through the gate. However, there are two notable differences between HPDC and HSC processes. First of all, the cavity filling velocity in HPDC is extremely high (more than 10 m s -1 ), so the highly turbulent flow can become partially atomized [10]. While in HSC, the melt steadily fills the cavity at a low velocity (10-1 -10 0 m s -1 ) and ideally from bottom to top, which prevents the entrapment of gas and oxides [11]. This difference in filling velocity is shown in Fig. 1(a). The second difference lies in the fact that the gate size used in HPDC is typically quite small so as to facilitate its quick removal for subsequent cleaning. As a result, feeding through the gate is weakened and a large number of shrinkage pores are generated in the casting. But, in HSC, both the runner and gate are quite large, which helps to strengthen the feeding effect through the gate and ensure sound castings being produced [11]. Figure 1(b) shows this difference in gating system design. (a) v HPDC (b) HPDC HSC Die Sleeve Cavity HSC 0.1 ms -1 Sleeve filling Cavity filling t Fig. 1: Filling velocities (a) and gating systems (b) of HPDC and HSC 2 Equipment and application The most widely used squeeze casting machines were those developed by Ube Industries, which includes the VSC machines popularized in the 1970s [12] and the horizontal vertical die casting machines popularized in the 1980s [13]. Since the 1990s, HSC techniques were further developed, which, compared to HPDC, offer a strict limit on the gas entrapment created during the sleeve filling stage. Prince Machine Corporation adapted a full-sleeve technique to a conventional HPDC machine, which uses a cross chamber to close the pour hole, allowing the shot sleeve to be totally filled prior to injection, and thus the gas entrapment in the sleeve is totally avoided [14]. It was further proved by fullsleeve technique that castability can be improved by as much as 20% compared to conventional HPDC techniques [15]. There is another method that can be taken to avoid gas entrapment, and that is to develop a super-slow sleeve filling technique. For example, in the Toyo 3S die casting system (Fig. 2), the shot sleeve temperature is controlled to create a melt with good fluidity, then the melt is injected into the die cavity through a multi-stage process in which the velocity can be as low as 0.05 m s -1 [16]. Other manufacturers have developed similar die casting machines, such as the DEC150MT by Toshiba, which are also capable of achieving velocities as low as 0.05 m s -1 [17]. The Chinese manufacturer L. K. Technology also offers HSC Moving half Partial squeeze Thermo couple Air Fixed half Cavity Short ceramic sleeve Heating unit Fig. 2: Toyo 3S die casting system [16] Ceramic ladle High melt fraction machines ranging in size from 280 to 800 t (Fig. 3) that can not only achieve an injection velocity as low as 0.03 m s -1, but also can be switched between HSC/HPDC functionality as needed. Several examples of components successfully fabricated by HSC have been reported since 2000, and it has been shown that castings made by the Toyo 3S die casting system can be heat treated [16]. Furthermore, through careful die design and process parameter control, Youn et al. [9] were able to use a sleeve filling velocity of 0.08 m s -1 with HSC for the 233

Special Report CHINA FOUNDRY Celebrating the Fig. 3: 800 t S-type squeeze machine by L. K. Technology squeeze feeding, and minimal gas entrapment in the shot sleeve. The commercial software ProCAST was used to simulate the filling and solidification process of a thick-walled plate by HSC, and the selection principles of these three critical parameters were discussed. The three-dimensional simulation model is shown in Fig. 4, which consists of a plate with a thickness of 10 mm, gating system and a shot sleeve with an internal diameter of 50 mm and a length of 300 mm. The die and casting materials selected were H13 steel and A356 alloy, respectively, and the major parameters in simulation are shown in Table 1. fabrication of an engine bracket mounting. Ji et al. [18] further studied the influence of casting parameters on the quality of ADC12 bars, and suggested that the filling velocity through the gate should be kept lower than 0.6 m s -1. Sleeve Cavity Melt 3 Selection principles of casting parameters From the statement above, it is clear that the three main characteristics required in the design of a HSC casting process are slow cavity filling, Fig. 4: Three-dimensional model for simulation Table 1: Major parameters in simulation (filling velocity is expressed by plunger velocity) Constants Variables Melt temperature Die temperature Initial melt fraction Cavity filling Gate thickness Sleeve filling ( C) ( C) in sleeve (m s -1 ) (mm) (m s -1 ) 0.1, 0.13, 0.4 10 0.1 670 200 0.5 0.1 4, 7, 10 0.1 0.1 10 0.1, 0.2, 0.33, 0.38, 0.4 3.1 Slow cavity filling With a high filling velocity, the melt surface easy enfolds or the exhaust exit is blocked by the melt, then the entrapment of gas and oxides occur. In such situations, the cavity should ideally be filled from bottom to top, which means the melt flows without jetting. It has been demonstrated that a critical velocity exists for almost all liquids, above which jetting of the melt occurs and casting quality dramatically decreases [19]. This critical velocity can be calculated by [19] : V crit = 2(γg/ρ) 1/4 (1) where γ is surface tension, g is the acceleration of gravity and ρ is the density of the melt. In the case of A356, when ρ is 2,420 kg m -3 and γ is 0.889 N m -1 [20], V crit is 0.49 m s -1. Figure 5 shows the cavity filling process with three different plunger velocities. It can be seen from Fig. 5 that when the plunger velocity is 0.1 m s -1, corresponding to the maximum melt velocity of 0.38 m s -1 at the gate [Fig. 5(a)], the melt flows from the bottom to the top and no jetting occurs. When the plunger velocity is 0.4 m s -1, corresponding to the maximum melt velocity of 1.07 m s -1 at the gate [Fig. 5(c)], jetting of the melt occurs, then the melt front drops and the surface enfolds, resulting in the entrapment of gas and oxides in the melt. When the plunger velocity is 0.13 m s -1, jetting is about to occur, and the maximum melt velocity of 0.52 m s -1 at the gate is considered to be the critical velocity [Fig. 5(b)], which is close to the preceding theoretical calculation. Although the critical velocity varies with material and gate structure, it has been suggested by Campbell to be close to 0.5 m s -1 in most engineering alloys [19]. Thus, the gate design and velocity setting in HSC should ensure that the gate velocity is below 0.5 m s -1. Moreover, this velocity is close to the turning point of casting quality of 0.6 m s -1 derived from experiments [18]. 234

Celebrating the Special Report CHINA FOUNDRY Fluid velocity magnitude (m s -1 ) 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0 Total time: 1.8750 s Fluid velocity magnitude (m s -1 ) Fluid velocity magnitude (m s -1 ) 0.6 Total time: 1.8528 s 2 Total time: 1.8375 s 0.54 1.8 0.48 1.6 1.4 0.42 1.2 0.36 1 0.3 0.8 0.24 0.6 0.18 0.4 0.06 (a) (b) (c) 0 0.2 0 Fig. 5: Cavity filling with plunger velocities of 0.1 (a), 0.13 (b) and 0.4 m s -1 (c) 3.2 Squeeze feeding The feeding mechanism in HSC can be expressed by the Poiseuille equation [19] : dp/dx = 8vη/πR 4 (2) where dp/dx is the pressure gradient, v is the interdendritic flow velocity, η is the viscosity, and R is the radius of the channel. Using this equation, it is found that even with a high pressure gradient, interdendritic flow stops when the interdendritic channel is blocked. Consequently, it should be possible to achieve directional temperature field in HSC. Figure 6 shows the temperature field associated with solidification using a gate with a thickness of 4, 7, or 10 mm. It shows that when the temperature of the 4 and 7 mm gates [Figs. 6(a) and (b)] falls to the solidus, a large area of the casting still remains above this temperature, resulting in a high degree of shrinkage porosity in the final casting. However, when the gate thickness is 10 mm [Figs. 6(c)], which is equal to the plate thickness, directional temperature field is achieved and the lower temperature of the casting allows it to be fed through the gate. Many factors influence the achievement of directional temperature field, such as sizes of the equivalent hot spot of gate and casting, geometry of runner, temperature field of die. Among them, sizes of equivalent hot spot of gate and casting are the two most important parameters, which can be approximately represented as the gate thickness and plate thickness in this simulation. Thus, an approximate conclusion that the gate thickness in HSC should be designed as large as the thickness of the casting, is made. Temperature ( C) 620Tliq 616.0 610 600 590 580 570 560 Tsol 556.0 550 540 530 520 Temperature ( C) Temperature ( C) Total time: 18.6213 s 620 Tliq 616.0 Total time: 15.8288 s 620Tliq 616.0 Total time: 13.1011 s 610 610 600 600 590 590 580 580 570 570 560 560 Tsol 556.0 550 550 Tsol 556.0 540 540 530 (a) (b) (c) 520 530 520 Fig. 6: Temperature field during solidification with a gate thickness of 4 (a), 7 (b), and 10 mm (c) 3.3 Slow sleeve filling Compared to vertical squeeze casting, it is quite difficult to totally eliminate gas entrapment during the horizontal injection of HSC. Garber [21] developed a mathematical model to describe the flow in the sleeve when the plunger advances (Fig. 7), indicating that the gas entrapment can be avoided if the melt advances with a stable front. If the plunger velocity is higher than the critical velocity, the melt will reach the ceiling of the sleeve and roll over, thereby leading to gas entrapment. Air Melt Air θ R Melt Fig. 7: Critical velocity filling of a shot sleeve h When the plunger velocity is lower than the critical velocity, the melt will hit the far end of the sleeve and move backward, thereby blocking the gate and also causing gas entrapment to occur. Tszeng and Chu [22] have proposed a simple equation to calculate the critical velocity, v c : v c = 2((2gR) 1/2 -(gh) 1/2 ) (3) where R is the internal radius of the shot sleeve and h is the initial height of the melt in the shot sleeve. Given an R of 25 mm and an h of 25 mm, the critical value v c is therefore found as 0.41 m s -1. The simulated results of the shot sleeve flow process with different plunger velocities are shown in Fig. 8. It demonstrates that with a plunger velocity of 0.1 or 0.2 m s -1, the melt near the plunger does not reach the ceiling of the sleeve. As a result, when the melt front reaches the far end of the sleeve it reflects back, blocking the gate and trapping the gas (indicated by the 235

Report CHINA FOUNDRY Special Fluid velocity magnitude (m s) Fig. 8: Flowing process in sleeve with different plunger velocities arrow) remaining in the sleeve. With the lower velocity of 0.1 m s-1, the gate is blocked until the sleeve filling is almost finished, and thus only a small quantity of gas is entrapped. When the plunger velocity is increased to 0.33 m s-1, a small quantity of gas is entrapped; when the plunger velocity is 0.4 m s-1, the melt front rolls over and gas entrapment occurs; and yet, no rolling or reflecting occurs with an injection velocity of 0.38 m s -1. This indicates that 0.38 m s-1 is the optimal value for sleeve filling, which is close to the theoretical calculation given earlier (0.41 m s-1). However, there may be errors between this theoretical calculation and reality, which cannot be compensated for by experimental testing due to a lack of validation methods. Most HSC manufacturers therefore tend to adopt a super-slow velocity in sleeve filling, and it is suggested that it only needs to be sufficient to ensure that the melt has a good fluidity. A typical value would be in the order of 0.1 m s-1. Celebrating the 4 Examples Figure 9 shows an engine connecting rod that was fabricated by a L. K. DCC280S machine. The design of this rod incorporated a large gate, with a plunger velocity of 0.1 m s-1 during the sleeve and cavity filling stages. This casting could be heat treated, and there was no visible evidence of gas porosity in cut sections. Figure 10(a) shows an experimental motorcycle wheel, which was fabricated by a L. K. DCC800S machine. In this instance, a large gate was used in order to achieve the required squeeze effect, resulting in several shrinkage porosities in the wheel hub when a 20 mm gate diameter was used [Fig. 10(c)]. When the gate diameter was subsequently enlarged to 30 mm, only a very small degree of porosity was observed, as shown in Fig. 10(d). This demonstrates the importance of gate size in ensuring the integrity of HSC castings. Furthermore, the fact that the gate is not positioned at the lowest point of the casting means that a bottom-to-top filling sequence cannot be achieved, which would also have an effect on the casting quality. Fig. 9: Parameter curves (a) and casting of an engine connecting rod with gating system (b) fabricated by HSC Fig. 10: Plan view (a) and section view (b) of a wheel fabricated by HSC, and porosity in casting with a gate diameter of 20 mm (c) and 30 mm (d) 236

Celebrating the Special Report CHINA FOUNDRY Investigation of the microstructure of the wheel revealed clear segregation band containing a large volume of Al-Si eutectic. In Fig. 11, the three segregation bands located near the gate can be seen to be broad and diffuse along the direction of flow. It is suggested that these bands were formed by a dilatant shear mechanism [23] and hot tearing mechanism [24] during solidification. These bands contain a large amount of liquid phase during solidification, which accelerates the movement of the central material and strengthens the feeding effect[25]. Segregation bands were also observed in a cross section of the wheel spoke, as shown in Fig. 12, including a skin-related band (i) and dilatant shear band (ii). The characteristics of segregation bands at other locations have also been studied in the previous research [24]. It was found that segregation bands tend to appear near the concave corner rather than the convex corner of the cross section, and in the thin section rather than the thick section where the section thickness changes. It was deduced that from the wall to the centre of the section, a first decreasing and then increasing shear stress gradient forms in the solidifying alloy and when the shear stress reaches the critical value at one position, the grain network collapses and a dilatant band forms by the dilatant shear mechanism[26]. 5 Development directions The fact that HSC has been successfully used to fabricate several thick-walled parts indicates that it has further potential to produce high-integrity parts. However, as the development of the HSC technique is still in its relative infancy, considerable work is still needed to extend the applicability of this method to other components. First, the equipment used for HSC needs to be further developed to incorporate the full-sleeve technique, a good combination of super-slow velocity and good fluidity in the sleeve, and die heating/cooling techniques. Secondly, the mechanism of the HSC process should be further studied, including the effects of flow type in the shot sleeve and cavity at slow velocity, solidification characteristics in the shot sleeve during filling, and solidification characteristics in the cavity under high pressure. Of particular concern is the fact that the segregation behavior in thick parts appears to be more pronounced in HSC than HPDC, which could cause the casting to fail in service. It is therefore important to fully understand this segregation behavior, and the mechanisms through which it occurs, and especially give a suitable mathematical description. Finally, further study is also needed to ascertain the effect of casting parameters on the quality of the final casting, including surface quality, macro porosity and microstructure. 6 Conclusions Fig. 11: Microstructure of gate i ii (1) Compared to HPDC, the HSC process has a lower filling velocity and larger gate size, making it suitable for producing highintegrity castings. Two techniques have been developed for HSC, the full-sleeve technique and super-slow injection technique, both of which have been mainly aimed at minimizing gas entrapment in the shot sleeve. As a result, HSC has been successfully used to produce a bracket mounting and connecting rod. (2) Through theoretical calculation and numerical simulation, it was found that the gate velocity in HSC should be below 0.5 m s-1, the gate thickness should be designed as large as the thickness of the casting, and the filling velocity should be the minimum required to ensure the melt retains good fluidity in the sleeve (a typical value is 0.1 m s-1). (3) The fabrication of a motorcycle wheel by HSC revealed that a large gate helps to eliminate shrinkage porosity in the casting. However, segregation bands, which play a significant role in feeding, were observed in the wheel, and could cause the casting to fail in service. (4) The future development of HSC should be focused on improving the equipment used, the understanding of the process itself, and studying the effect of casting parameters on the final product quality. References Fig. 12: Segregation bands in a cross-section of wheel spoke [1] Laukli H I. High pressure die casting of aluminium and magnesium alloys - grain structure and segregation 237

Special Report CHINA FOUNDRY Celebrating the characteristics. Trondheim, Norway: Norwegian University of Science and Technology, 2004. [2] Kaufmann H, Uggowitzer P J. Metallurgy and processing of high-integrity light metal pressure castings. Berlin: Fachverlag Schiele & Schoen GmbH, 2007. [3] Kang C G, Lee S M, Kim B M. A study of die design of semisolid die casting according to gate shape and solid fraction. Journal of Materials Processing Technology, 2008, 204(1-3): 8-21. [4] Wen W, Luo A A, Zhai T, et al. Improved bending fatigue and corrosion properties of a Mg-Al-Mn alloy by super vacuum die casting. Scripta Materialia, 2012, 67(11): 879-882. [5] Huang X S, Lv Z G, He L J, et al. Computer-aided designs of die and thermal control for fabrication of A356 wheels. Materials Transactions, 2013, 54(8): 1491-1495. [6] Jiang J F, Wang Y, Chen G, et al. Comparison of mechanical properties and microstructure of AZ91D alloy motorcycle wheels formed by die casting and double control forming. Materials & Design, 2012, 40: 541-549. [7] Ghomashchi M R, Vikhrov A. Squeeze casting: an overview. Journal of Materials Processing Technology, 2000, 101(1-3): 1-9. [8] Iyer A. Squeeze casting: the future. Melbourne, Australia: International Specialized Skills Institute, 2011. [9] Youn S W, Kang C G, Seo P K. Thermal fluid/solidification analysis of automobile part by horizontal squeeze casting process and experimental evaluation. Journal of Materials Processing Technology, 2004, 146(3): 294-302. [10] Gourlay C M, Laukli H I, Dahle A K. Defect band characteristics in Mg-Al and Al-Si high-pressure die castings. Metallurgical and Materials Transactions A, 2007, 38A(8): 1833-1844. [11] Luo S J, Chen B G, Qi P X. Liquid forging and squeeze casting technology. Beijing: Chemical Industry Press, 2007. (in Chinese) [12] Yamaguchi T U, Uchida M. Vertical die casting machines. US Patent, 4088178, 09-05-1978. [13] Dannoura S. Horizontal mold clamping and vertical injection type die cast machine. US Patent, 4655274, 07-04-1987. [14] Damian B, Nelson C. Full-sleeve squeeze casting process using a conventional horizontal die casting machine. In: Proceedings of the Second China International Diecasting Congress, Shenyang, China, 2000: 195-197. [15] Konopka Z, Zyska A, łągiewka M, et al. The 'full sleeve' application in the horizontal cold-chamber machine for pressure die casting of aluminium alloys. Archives of Foundry Engineering, 2008, 8(1): 65-70. [16] Kawawuchi Y, Wang L. Super slow shot diecasting system (Japanese 3S casting system). In: Proceedings of the Second China International Diecasting Congress, Shenyang, China, 2000: 195-197. (In Chinese) [17] Hybrid die casting machine. http://www.toshiba-machine.co.jp/ en/product/diecast/lineup/hybrid.html. [18] Ji L Q, Xiong S M, Masayuki M, et al. Effects of the super slow speed die casting process parameters on properties of ADC12 aluminum alloy. Foundry, 2007, 56(10): 1057-1061. (In Chinese) [19] Campbell J. Casting. Oxford, UK: Butterworth-Heinemann, 2003. [20] Anson J P, Drew R A L, Gruzleski J E. The surface tension of molten aluminum and Al-Si-Mg alloy under vacuum and hydrogen atmospheres. 1999, 30(6): 1027-1032. [21] Garber L W. Theoretical analysis and experimental observation of air entrapment during cold chamber filling. Die Casting Engineer, 1982, 26(3): 14-22. [22] Tszeng T C, Chu Y L. A study of wave formation in shot sleeve of a die-casting machine. Journal of Engineering for Industry, 1994, 116(2): 175-182. [23] Gourlay C M, Meylan B, Dahle A K. Shear mechanisms at 0-50% solid during equiaxed dendritic solidification of an AZ91 magnesium alloy. Acta Materialia, 2008, 56(14): 3403-3413. [24] Huang X S, He L J, Mi G B, et al. Characteristics of defect bands and their formation mechanisms in A356 wheel fabricated by horizontal squeeze casting. Materials Science and Technology, 2014, DOI: http://dx.doi.org/10.1179/1743284 714Y.0000000542. [25] Otarawanna S, Laukli H I, Gourlay C M, et al. Feeding mechanisms in high-pressure die castings. Metallurgical and Materials Transactions A, 2010, 41(7): 1836-1846. [26] Gourlay C M, Dahle A K. Dilatant shear bands in solidifying metals. Nature, 2007, 445: 70-73. The present work was financially supported by L. K. Technology. 238