Journal of University of Science and Technology Beijireg Volume 13, Number 1, February 2006, Page 60 Materials Evaluation of the mold-filling ability of alloy melt in squeeze casting Haiying Zhang), Shuming Xing), Qinghua Zhang, Jianbo and Wen Lid) 1) Semisolid Forming Research Center, Beijing Jiaotong University, Beijing 100044, China 2) Shijiazhuang Locomotive Depot, Shijiazhuang 050000, China 3) Department of Material and Engineering, Hebei University of Science and Technology, Shijiazhuang 050018, China (Received 2004-12-13) Abstract: The mold-filling ability of alloy melt in squeeze casting process was evaluated by means of the maximum length of Ar- chimedes spiral line. A theoretical evaluating model to predict the maximum filling length was built based on the flowing theory of the incompressible viscous fluid. It was proved by experiments and calculations that the mold-filling pressure and velocity are prominent influencing factors on the mold-filling ability of alloy melt. The mold-filling ability increases with the increase of the mold-filling pressure and the decrease of the proper mold-filling velocity. Moreover, the pouring temperature relatively has less ef- fect on the mold-filling ability under the experimental conditions. The maximum deviation of theoretical calculating values with ex- perimental results is less than 15%. The model can quantitatively estimate the effect of every factor on the mold-filling ability. Key words: squeeze casting; process parameters; mold-filling ability; theoretical calculation; experimental evaluation 1. Introduction The mold-filling ability of alloy melt can directly affect the inner quality and surface roughness of the products, and is regarded as a basic problem in the field of material forming process, especially in the re- searches of new materials and processes. In the tradi- tional casting technology, the evaluation of the mold- filling ability is carried on by measuring the maximum filling length of a spiral-line specimen; however, this evaluation is limited for casting materials, mould ma- terials, pouring technologies and forming methods. Hence in special casting process such as squeeze casting, semisolid metal casting, lost foam casting and so on, it is a new problem to evaluate the mold-filling ability of alloy melt. S.M. Xing investigated the mold- filling ability of semisolid alloy in the mold with many holes of different diameters under a pressure, and proposed a theoretical formula of the filling length 11. Recently, Julio Aguilar 2 designed a meander sample of AZ91 alloy with a length of 2420 mm to evaluate the filling ability of alloys in the die-casting process. Because there is not an available formula to predict filling length, this evaluation method is not convenient in engineering. T.J. Zhang 3 investigated the relationship between magnetic flux density and input voltage as well as distance, and found that the mold-filling length of the melt increases rapidly with the increase of the average magnetic-flux density, and Corresponding author: Haiying Zhang, E-mail: yingying-zh 126.com that the upper steel mold is superior to the upper gyp- sum mold. Moreover, in researching on high pressure die casting, S. Kulasegaram 4, Z. Liu 5, and Cleary Paul 6 studied the mold-filling ability of alloys by simulations. E.N. Pan 7 studied the effects of pour- ing temperature, coating thickness etc. on the mold- filing ability of lost foam A356 alloy, the calculated flow length based on a modified EPC Flow- Solidification equation showed a good agreement with the experimental results. The evaluation method of mold-filling ability mainly includes three types, ex- perimental evaluation, theoretical calculation, and simulation. The experimental evaluation method is a basic one because of its intuitional effects, but it is limited for the experimental conditions and is difficult for popular use. The simulation method is not con- venient in engineering though it is more exact than the experimental method in given conditions. However, the theoretical calculation method can be used con- veniently to predict the filling ability of various alloys in various processes without any special equipment. Squeeze casting is a new forming technology that is expected to be widely used in the future. However, the filling ability of alloy melt in the process is not clear. Studies evaluating the mold-filling ability of alloy melt in squeeze casting are few. In this article, ex- perimental evaluation and theoretical calculation are combined, and a mathematical model for calculating mold-filling ability is obtained. H.Y. Zhng et al., Evaluation of the mold-filling ability of alloy melt in squeeze casting 61 2. Experimental evaluation of the mold- filling ability in squeeze casting Based on spiral-line specimen in gravity casting, a special spiral-line-specimen mold is designed (shown in Fig. l), which is used for experimental evaluation of the mold-filling ability of alloy melt in squeeze casting. The mold is made of H13 steel. Its construc- tion mainly includes three parts, the upper mold, the lower mold, and the punch. Fitted on the glide-piece of the forming device, the upper mold has an Ar- chimedes spiral-line cavity with a total length of 1350 mm. The lower mold is fitted on the worktable of the forming device; its mold-cavity is the cylindrical pressing chamber. In the sidewall of the pressing chamber, there is a runner that connects with the Ar- chimedes spiral-line cavity. The punch connects with mold-filling piston of the forming device. The forming device is the double-action hydraulic presses which has the functions such as mold-locking, mold-filling, ejecting, and mold-filling velocity transiting. The process parameters such as mold-filling pressure, pressing velocity (mold-filling velocity) etc. can be set and the experimental data can be saved automatically by a computer controlling system. The forming type is inverted-extrusion in which the alloy melt in the cyl- indrical chamber is pressed into the Archimedes spi- ral-line cavity through the runner inverted to the punch moving direction. 3 Fig. 2. Archimedes spiral-line specimen. After melting, alloy melt is poured into the cylin- drical chamber of the lower mold, the upper mold moves down too close to the lower mold under a hy- draulic press. Then, the punch drops near the alloy surface with a fast velocity and then presses the alloy melt with a slow velocity into the Archimedes spiral- line cavity. Keeping a set time, the upper mold de- taches with the lower mold, and the Archimedes spi- ral-line specimen is obtained (shown in Fig. 2). Fi- nally, the lengths of the Archimedes spiral lines are measured. 3. Mathematical evaluation of the mold- filling ability 3.1. Data preparation The geometrical dimensions of the pressing cham- ber, the runner and the spiral-line cavity are shown in Fig. 3. Fig. 1. Spiral-line-specimen mold: 1-the punch: 2-the upper mold: 3-the lower mold: 4-the pressing chamber. d i Fig. 3. Schematic diagram of the mold. 1-the punch: 2 -the pressing chamber: 3-the runner; 4-the spiral line During the mold-filling process in squeeze casting, there exists two types of forces: one is the driving powers that force alloy melt to flow to the Archimedes 62 J. Univ. Sci. Technol. Beijing, Voh13, No.1, Feb 2006 spiral-line cavity; the other is the resistant powers which block off flowing and cause the pressure losing. When the driving powers are less than the resistant powers, alloy melt will stop filling, and the filling length reaches the maximum length L, simultane- ously. The driving powers mainly include the mold- filling pressure provided by the punch and the gravity of alloy melt. It is considered that the gravity is far less than mold-filling pressure that can be ignored. The resistant powers mainly include the friction re- sistance and the partial resistance. Hence the conditions for keeping alloy melt filling can be described by the following equation: P 2 F (1) where P is the mold-filling pressure and F is the total resistance of the friction resistance and the partial resistance. In squeeze casting, alloy melt can be regarded as a stable, incompressible viscous fluid. That is to say, the alloy density p is a constant in the filling process. Suppose the viscosity coefficient of alloy melt v keeps a constant in the whole mold-filling process. The pressure provided by the punch in the pressing chamber is named as mold-filling pressure P . The moving velocity of the punch is regarded as the aver- age flowilig velocity V of alloy melt in the pressing chamber (named as mold-filling velocity). Based on the continuous equation of fluid, the average velocity of alloy melt in the runner is Vo = AV /Ao and the av- erage velocity of alloy melt in the spiral-line cavity is Vl = A V / A l , where A , Ao, Al are the cross- section area of the pressing chamber, the runner, and the spiral-line cavity, respectively. Suppose the whole pressing chamber is poured with alloy melt at the beginning, the moving height of the punch (the descend height of alloy melt in the pressing chamber) is H and consequently the length filled spiral line is L . Based on the mass conservation law, the relationship of H and L is H Z - A1L A 3.2. Friction resistance (1) Friction resistance generated from the alloy melt flow in the pressing chamber. When the Reynolds number is less than 2300, the flowing state of alloy melt is regarded as a laminar flow, and when the Reynolds number is more than 4000, the flowing state is regarded as a turbulent flow. Otherwise, the flowing state is a transitional flow state intervening laminar flow with turbulent flow. The Reynolds number of the alloy melt flow in the press- ing chamber is Re = VD l v , where D is the diameter of the pressing chamber. If it is a laminar flow, the friction resistance coefficient is t1 = 641 Re 8. Based on the momentum transmit equation of the in- compressible viscous fluid, the friction resistance in the pressing chamber is H pV2 02 Re D 02 313.6 HpV2 - 313.6HpvV Fl = 9.8t1 - = - - (34 If it is a turbulent flow, the coefficient of friction S, so the fric- 0.129 - 0.129 resistance is cl = Re0.2 - f VD ) O (UJ tion resistance is I H pV2 0.632 HpV2 - 0.632HpV2 F1 = 9.8& - = - - D 2 D .( ,),.12 (3b) (2) Friction resistance generated from the flow of alloy melt in the runner. Similarly, the Reynolds number of the alloy melt flow in the runner is Re0 = VoDo /v = AVDo /Aov, while Do is shown in the Fig. 3. The Reynolds number varies with the difference of the flow velocity and the runner dimension. So the flowing state is a laminar flow or a turbulent flow under different conditions. If it is a laminar flow, the coefficient of friction resis- tance is t2 = 641 Re0 = 64Aov/ AVDo , and hence the friction resistance is H pVo2 - 313.6 Hp(AV)2 - - 313.6HApvV F2 = 9.852 - - - Do 2 Reo DoAo2 AoDo2 (44 If it is a turbulent flow, the coefficient of friction , therefore - I 0.129 resistance is l2 =- Reoo.12 A V D . the friction resistance is - F2 = 9.852 , - H pVo2 = _ 0.632 Hp(AV) - DO 2 Reo0.12 DoAo2 0.6328 p A 2V 0.12 AVDo (3) Friction resistance generated from the alloy melt flow in the Archimedes spiral-line cavity. H. Y. Zhang et aL, Evaluation of the mold-filling ability of alloy melt in squeeze casting 63 According to the dimension of the Archimedes spi- ral-line cavity, the maximum of d / D1 is only 0.1, d and D, are shown in Fig. 3, so the coefficient of partial resistance is very little, the alloy melt flow in the Ar- chimedes spiral-line cavity is regarded as the flow in the direct pipe and the partial resistance is ignored. The Reynolds number is denoted as Rel = Vld l v = AVd / Alv . The coefficient of friction resistance for a laminar flow is t3 = /Rel = 64A1v/ AVd , therefore the friction resistance is L pV12 d 2 Re, dA12 Ald 313.6 L P ( A V ) - 313.6LApvV F3 = 9.853 - = - - (54 If it is a turbulent flow, the coefficient of friction - resistance is t3 = - - - 0129 , and hence the Relo. fAVd)12 friction resistance is - F3 = 9.853 I - LpV12 = - 0.632 Lp(AV)2 - d 2 Re10.2 dl 0.632LA pV 3.3. Partial resistance When alloy melt flows from the pressing chamber to the runner and then to the spiral-line cavity, the partial resistance is generated from the momentum losing caused by changes in flow velocity direction and in flow space. When the flowing state is a turbu- lent flow, the resistance losing increases proportion- ally with the square of average velocity and is shown in the following equation 8: where KPart is the coefficient of partial resistance. When the melt flow from the pressing chamber to the runner, the coefficient of partial resistance is as shown in the following equation: K = k(1-) A 2 A (71 where k is the amending coefficient in the range of 0.5-1.75 8. Hence the partial resistance is Similarly, when the melt flow from the runner to the spiral-line cavity, the partial resistance is (9) 3.4. Total resistance If the flowing states of alloy melt are all the laminar flow in the pressing chamber, the runner, and the spi- ral-line cavity, the total resistance is 313.6HpvV + 313.6HApvV + 313.6LApvV + 4.9k ( %+ F = F ; . = i=l 0 2 AoDo Aid Bring the Eq. (2) into (10) and simplify F = 313.6C1pvVL+4.9kC2pA2V2 (11) where C1, C2 are the constants related with the mold dimensions which are determined as in the following: A 1 A 1 A AD AoDo2 Ald2 C 1 = : + - + - If the flowing states of alloy melt are the turbulent flow in the pressing chamber, the runner and the spi- ral-line cavity, the total resistance is F = F3,+F2+F3+F4+F5 = 0.632pV 2L A1 + A1A + A2 I+ ADRe. Ao2DoReo0.12 A12dRe10.12 4.9kC2pA2V (14) 3.5. Critical conditions of filling mold and the fill- ing length of alloy melt Based on Eq. (l), if the flowing states of alloy melt are the laminar flow in the pressing chamber, the run- ner and the spiral-line cavity, then the critical filling conditions of alloy melt is as shown in the following equation: P 2 313.6ClpvVL+4.9kC2pA2V2 (15) So the filling length of alloy melt is limited by P-4.9kC2pA2V L I 3 I3.6ClpvV The maximum filling length is determined by P - 4.9kC2pA2V Lmax = 313.6ClpV 64 J. Univ. Sci. Technol. Beuing, V01.13, No.1, Feb 2006 If the flowing states of alloy melt are the turbulent P - 4.9kC2pA2V L A2 1 flow in the pressing chamber, the runner and the spi- 0.632pV2 + A1A + ral-line cavity, then the filling length of alloy melt is ADRe0.12 Ao2DoReo0.2 A12dRe10.12 The maximum filling length is P- 4.9kC2pA2V A. Do Reo A2 1 Lln, = AlA + A, 2dRel 0.12 According to Eqs. (17) and (19), the effect factors of the maximum filling length are mainly pressure system, parameters of the mold, alloy melt viscosity etc. Pressure system includes the pressing velocity of the punch and the mold-filling pressure. Parameters of the mold include the geometry dimensions of the pressing chamber, the runner, and the spiral-line cav- ity. Their effects on the maximum filling length are different in different flowing states. When the flowing states are all the laminar flow, the maximum filling length has a direct ratio to the mold-filling pressure and has ad inverse ratio to the mold-filling velocity and the alloy melt viscosity. When the flowing states are the turbulent flow in the runner and the spiral-line cavity, the maximum filling length has also a direct ratio to the mold-filling pressure and has an inverse ratio to the Reynolds number which integrates the mold-filling velocity and the viscosity of alloy melt with the parameters of mold. Whatever be the Rowing state, the maximum filling length increases with the decrease of the mold-filling velocity. One of the rea- sons is that the friction resistance increases with the increase of the mold-filling velocity and the viscous power. The other reason being that the flowing state is a turbulent flow under the faster mold-filling velocity, and there exists pulse velocity and the additional stress generated in the normal direction and the tangent di- rection of the flow path, the resistance in a turbulent flow is larger than that in a laminar flow. Similarly, from Eqs. (17) and (19), the parameters of the mold, the area of the pressing chamber, the run- ner, and the spiral-line cavity have important effects on the mold-filling ability in squeeze casting. The amend coefficient k related with the connections be- tween the pressing chamber, the runner and the spiral- line cavity has a negative effect on the maximum fill- ing length. Therefore, avoiding connections of the right angles and smooth transition in the mold cavity design are beneficial to improve the mold-filling abil- ity. 4. Experiments and discussion 4.1. Experiments In order to verify the availability of the theoretical model mentioned earlier, two orthogonal experiments of three factors and two levels with A356 Aluminum and 25 Steel were designed and carried out. Experi- mental parameters, levers, and results are shown in Tables 1 and 2. Table 1. Mold-filling ability experiments with A356 aluminum in squeeze casting Test No. 1 2 3 4 5 6 7 T/C 620(620) 620(621) 620( 62 1 ) 620(620) 670(670) 670(673) 670(67 1) p / MPa v/ (ms-1) 50(49.5) 20(68) 50(49.0) 40(87) 95(94.0) 20(74) 95(95.0) 40(91) 50(49.0) 20(63) 50(49.9) 40(81) 95(95.0) 20(71) PV Lmax - - 660 - 457 - 1023 880 - 842 - 613 - 1083 - 865 8 670(670) 9X96.1) 40(86) - Klj 3020 2572 3608 3247 T=6423 K2i 3403 385 1 2815 3176 Si 18336 204480 78606 630 S,=314358 Note: T is the pouring temperature; p is the average mold-filling pressure; V is the mold-filling velocity; the data of outer ( ) are setting values; the data of inner ( ) are practical values; Kv is the experimental results summation when the level number is i of the j column; Sj is the sum of deviation squares of the j column; S, is the total sum of deviation squares; T is the total length of this ex- periment. H . Y. Zhung et al., Evaluation of the mold-filling ability of alloy melt in squeeze casting 65 Table 2. Mold-filling ability experiments with 25 steel in squeeze casting Test No. p I MPa V/(mmX) PV t l s L-lmm 1 65(65.3) 15(55) - 0 398 2 65(65.3) 15(53) - 5 460 3 65(65.4) 40(67) - 0 225 4 65(65.2) 40(69) - 5 273 5 120(120.2) 15(62) - 0 585 6 120( 120.1) 15(67) - 5 520 7 120(120.4) 40(75) - 0 460 8 120(120.2) 40( 80) - 5 352 T=3273 Klj 1356 1963 1670 1668 KZi 1917 1310 1603 1605 S; 39340 53301 561 496 S F 104220 Note: t is the delayed time of the punch extrusion. From the values of KIj, KZj, it is shown that the fill- ing length increases with the increase of the mold- filling pressure and the decrease of the mold-filling velocity. These influencing tendencies are consistent with the results of the theoretical model. Moreover, from the values of Sj, it is shown that there exist great differences in the influencing extents of various fac- tors on the maximum filling length. The most promi- nent factors are the mold-filling pressure and the mold-filling velocity, and their interrelation has a little effect on the filling length. It is also shown that the pouring temperature and the delayed time of the punch extrusion are unimportant factors. It is indicated that the viscosity change caused by the pouring tempera- ture and the delayed time of the punch extrusion has less effect than the mold-filling pressure and the mold- filling velocity. The experimental results are also re- flected by the theoretical formula. Calculated with the practical values of parameters (shown in Table 3) in the experiments, the Reynolds numbers are larger than 3700 in the pressing chamber, the runner or the spiral-line cavity, so the flowing states are all regarded as the turbulent flow. According to Eq. (19), the theoretical filling lengths are shown in Table 4. Table 3. Values of additional parameters Table 4. Comparison of the maximum filling lengths by experiment with those by theoretical calculation A356 Aluminum alloy 25 Steel Items llmm 660 457 1023 880 460 273 520 352 L /mm 675 399 1133 745 47 1 260 570 389 1 2 3 4 2 4 6 8 SI % 2 13 11 15 2 5 10 11 100 1 L - I1 Note: - t h e experimental lengths; L - t h e theoretical calculating lengths; &-the relative deviation, 6 = 1 From Table 4 it is to be noted that whatever be the value for A356 or 25 steel, the relative deviations of experimental results with theoretical values are not more than 15%. It is difficult to determine the actual reasons that lead to the difference. the theoretical model, the partial resistance in the spiral-line cavity is ignored, and the amend coefficient k is an experiential value which is different due to the difference of the practical joint forms; in the experiments, the collection and transition of signals are delayed in the experi- menh the mold temperature, the Pouring temperature, and the uniformity of the Coating thickness are diffi- cult to accurately Control. The theoretical model Can Provide a good guidance for understanding accurately the effect Of VXiOUS factors O n the mold-filling ability. 4.2. Effect of the pouring mass of alloy melt on the filling length Eqs. (17) and (19) are obtained by supposing the whole pressing chamber being filled with alloy melt. Actually, it is impossible that the chamber is fully 66 J. Univ. Sci. Technol. Beijing, V01.13, No.1, Feb 2006 poured with alloy melt, hence Eq. (19) must be amended. Suppose there is a height of h that is still not been filled to the top of the pressing chamber. Based on the conservation of mass, AH = Aoh + Al L , then O632Aoh 0.632Ah ) ADRe0.12 AoDoReo0.12 ADRe0.12 Ao2DoReo0.12 A12dRe10.12 4.9kCzA2 +:-+ A2 1 AlA + - - 0.632pV A1 + From Eq. (21), it is shown that the filling length by theoretical calculation increases with the increase of the poured alloy melt mass. So it is also one of the reasons that cause the deviation between the experi- mental results and theoretical calculation values. 5. Conclusions (1) In squeeze casting, both the mold-filling pres- sure and the mold-filling velocity have prominent ef- fects on the mold-filling ability of alloy melt, and the pouring temperature has less effect under the experi- mental conditions. The mold-filling ability increases with the increase of the mold-filling pressure and the decrease of the proper mold-filling velocity. (2) The maximum filling length of alloy melt in squeeze casting is calculated by Eqs. (17) and (19). Comparing the theoretical calculating values with the experimental results, the maximum relative deviation is not more than 15%. References l S.M. Xing, L.Z. Zhang, D.B. Zeng, et al., Mold-filling Aoh+AIL A H = Therefore, Eq. (19) is amended as ability of semisolid alloy, J. Univ. Sci. Technol. Beijing, 9(2002), No.4, p.253. Julio Aguilar, Martin Fehlbier, Tilman Grimmer, et al., Processing of semisolid Mg alloys, in Transaction of 8th S2P, Cyprus, 2004, p.113. T.J. Zhang, J.J. Guo, Y.Q. Su, et al., Effect of traveling magnetic on mould-filling length of the A357 melt during casting thin walled plate, J. Muter. Sci. Technol., 19(2003), No.1, p.43. S. Kulasegaram, J. Bonet, J. Lewis, et al., High pressure die casting simulation using a lagrangian particle method, Commun. Numer. Methods Eng., 19(2003), No.9, p.679. Z. Liu, Z.G. Wang, Y. Wang, et al., Study on mould fill- ing and solidification of high pressure magnesium die casting process, Chin. J. Muter. Res., 13(1999), No.6, p.641. Cleary Paul, Prakash Mahesh, Ha Joseph, et al., Modeling of cast systems using smoothed-particle hydrodynamics, JOM, 56(2004), No.3, p.67. E.N. Pan and K.Y. Liao, Analysis of the flow and solidifi- cation patterns of the lost foam (EX) A356 A1 cast alloy, J. Chin. SOC. Mech. Eng., 21(2000), No.3, p.283. X.Y. Zhang, Metallurgy Transition Theory (in Chinese), Metallurgical Industry Press, Beijing, 1981, p.71.