Determination of the metal/die interfacial heat transfer coefficient of high pressure die cast B390 alloy

Transcription

1 IOP Conference Series: Materials Science and Engineering Determination of the metal/die interfacial heat transfer coefficient of high pressure die cast B390 alloy To cite this article: Yongyou Cao et al 2012 IOP Conf. Ser.: Mater. Sci. Eng View the article online for updates and enhancements. Related content - Section thickness-dependant interfacial heat transfer in squeeze casting of aluminum alloy A443 Zhizhong Sun, Xuezhi Zhang, Henry Hu et al. - Modeling of microstructure evolution of magnesium alloy during the high pressure die casting process Mengwu Wu and Shoumei Xiong - The improvement of aluminium casting process control by application of the new CRIMSON process X Dai, M Jolly and B Zeng Recent citations - An Initial Study of a Lightweight Die Casting Die Using a Modular Design Approach Sebastian Müller et al - Simultaneous Effect of Plunger Motion Profile, Pressure, and Temperature on the Quality of High-Pressure Die-Cast Aluminum Alloys Elena Fiorese and Franco Bonollo - Cao Yongyou et al This content was downloaded from IP address on 20/08/2018 at 03:36

2 Determination of the metal/die interfacial heat transfer coefficient of high pressure die cast B390 alloy Yongyou Cao 1,2, Zhipeng Guo 1,2 and Shoumei Xiong 1,2 1 Department of Mechanical Engineering, Tsinghua University, Beijing , China 2 State Key Laboratory of Automobile Safety and Energy, Tsinghua University, Beijing , China smxiong@tsinghua.edu.cn Abstract. High-pressure die cast B390 alloy was prepared on a 350 ton cold chamber die casting machine. The metal/die interfacial heat transfer coefficient of the alloy was investigated. Considering the filling process, a finger -shaped casting was designed for the experiments. This casting consisted of five plates with different thicknesses (0.05 inch or 1.27 mm to 0.25 inch or 6.35 mm) as well as individual ingates and overflows. Experiments under various operation conditions were conducted, and temperatures were measured at various specific locations inside the die. Based on the results, the interfacial heat transfer coefficient and heat flux were determined by solving the inverse heat transfer problem. The influence of the moldfilling sequence, sensor locations, as well as processing parameters including the casting pressure, die temperature, and fast/slow shot speeds on the heat transfer coefficient were discussed. 1. Introduction High-pressure die casting (HPDC) is one of the most economical casting processes for mass producing net-shaped parts. Due to the excellent properties of die castings, an increasing number of die casting products are currently used in the automotive, aerospace, medical, electronic, and other industries. Techniques such as computer-assisted design and engineering, which have rapidly developed in recent years, are applied in the modeling and simulation of the filling and solidification processes in HPDC. The use of these techniques significantly optimizes processes and saves costs. However, computerbased techniques are only beneficial when the material properties as well as the boundary and initial conditions used as inputs are correct [1]. The interfacial heat transfer coefficient (IHTC), characterizing the thermal resistance between the metal and the mold, is believed to be the one of most important parameters during the solidification process for computer simulations [2-3]. Numerous studies [1-5] on determining the IHTC under various casting conditions have been conducted. In the current paper, a die casting experiment was conducted using a finger -shaped casting of B390 alloy. The metal/die IHTC was determined according to the temperature readings obtained at different locations inside the die by solving the inverse heat transfer problems. The influence of the mold filling sequence, sensor locations, as well as processing parameters including the casting pressure, die temperature, and fast/slow shot speeds on the heat transfer coefficient were discussed. 2. Experiments 2.1. Finger-shaped casting Published under licence by Ltd 1

3 A specially designed casting, namely, a finger-shaped casting, was used in the current study. As shown in Figure 1, this casting has five plates with different thicknesses from T1 (0.05 inch or 1.27 mm) to T5 (0.25 inch or 6.35 mm), with an interval of 0.05 inch (or 1.27 mm). The biscuit of the casting was designed with a diameter of 60 mm and a thickness of about 20 mm. Each plate is 203 mm long (along the metal filling direction) and 19 mm wide. At the interface between the casting and die of each plate, one-dimensional heat transfer was assumed. Figure 1. Finger-shaped casting: (a) configuration, and (b) actual casting showing the location of sensors. Commercial aluminum (Al) alloy B390 (Al-17Si-4Cu) was poured into a TOYO 350 ton cold chamber high-pressure die casting machine. The chemical compositions of the alloy and the die material H13 steel is given in Table 1. Table 1. Chemical composition of B390 alloy (Al-17Si-4Cu) and H13 steel. Element wt.% Si Cu Mg Fe Zn Mn Ni Sn Al B Bal. Element wt.% C Mn Si S P Cr Mo V Fe H <0.005 < Bal Die configuration and sensor installation To gain a sufficiently rapid response time to follow the HPDC process and accurately measure the temperatures inside the die, a special temperature sensor unit (TSU) was designed. As shown in Figure 2, at each distance (1, 3, and 6 mm) from the front wall of the TSU, two thermocouples were adjusted. The thermocouples were grounded sheathed K-type thermocouples with 0.5 mm outside diameter and mm wire diameter. These thermocouples were inserted into 1.1 mm diameter holes and vacuum nicro-brazed to the TSU body of H13 steel. To investigate the influence of filling, two TSUs were located at both ends of each finger plate close to the gate (G) and overflow (F), as illustrated in Figure 1(b). Figure 3 displays the 10 TSUs containing 60 thermocouples embedded inside the stationary die until the front wall approached the cavity surface. The cooling as well as heating lines were present in the die at the distance of 25 mm from the parting face in order to control the mold temperature during HPDC. Real-time temperature data were then recorded using a data acquisition system manufactured by the Integrated Measurement Corporation (Berlin Germany) with a sampling rate of 500 Hz. 2

4 Figure 2. Configuration of the temperature sensor unit. Figure 3. Graphical installations of temperature sensor units (TSUs) and data acquisition system in the cold-chamber die caster. Table 2 lists the thermal properties of B390 alloy and H13 steel. The processing parameters, including the casting pressure (P), die temperature (T d ), fast shot speed (v H ), slow shot speed (v L ), and pouring temperature (T P ) were varied in each set of experiment. Others were fixed at the same condition shown in Table 3. Nearly 240 shots were performed; the first 20 shots were conducted to preheat the dies to thermal equilibrium. Table 2. Thermal properties of related materials. Table 3. Values of related processing parameters. Thermal properties B390 H13 Processing parameters Thermal conductance λ (W m -1 C T a Pouring temperature ) T P ( C ) Specific heat C (J kg -1 C Die temperature T ) T d ( C ) Density ρ (kg m -3 ) T Casting pressure Solidus temperature T S ( C) P (MPa) Liquidus temperature T L ( C) Latent heat L s (J kg -1 ) a T stands for temperature ( C). Basic condition Variable and Slow shot speed v L (m s -1 ) 0.2 Fast shot speed v H (m s -1 ) 1.75 a Min stands for no intensification pressure , 83.81, 67.39, and Min a 0.1, 0.3, 0.4, and , 1.5, 2, and Results and discussion 3.1. Heat transfer estimation Figure 4 shows a sample of the measured temperature profiles of the 30 cycles at the T4G position under the basic condition. The temperature profiles of the same position illustrate cyclic characteristics in shape while the die casting phases were sequentially performed. Given that these cycles were performed under the same operation condition, the results showed that the process was quite reproducible. The temperature curves measured at the same distance, such as A1 and A2 or C1 and C2 almost overlapped. The maximum difference between the two temperature profiles measured at the 3

5 same distance never exceeded 4 C, indicating that the heat transfer process can be reasonably assumed to be one dimensional. Figure 4. Sequential temperatures of temperature sensor unit at the T4G position for 30 cycles under the basic condition. The IHTC cannot be directly calculated using the measured die temperatures because the interfacial heat flux density (IHFD) cannot be directly measured during casting and solidification. However, the computer program used an inverse method [4-6] based on the principle of Beck [7-8]. The average of measured temperatures at 1 and 6 mm from the cavity surface (T 1 and T 6 ) were used to evaluate IHFD, IHTC, and the die surface temperature. The inversely calculated temperature at 3 mm from the cavity surface, namely T 3c, was compared to the average temperature at B1 and B2 measured at the same distance to validate the inverse modeling. Once the shot was performed, the IHTC abruptly increased until reaching the peak value, maintained its value at a higher level, and then sharply decreased. An analysis using the inverse method at the T4G position with respect to the last cycle of the measured temperatures (T 1 and T 6 ) was subsequently performed, as shown in Figure 4. The curves of IHFD (q) and IHTC (h), as well as the die surface temperature, casting center, and surface temperatures designated by T ds, T cc, and T cs were then obtained, as shown in Figure 5. There was a very good fit between the measured (T 3m ) and calculated (T 3c ) temperatures at 3 mm, indicating that the inverse estimation results were quite reliable. The casting surface temperature (T cs ) abruptly dropped after the shot was performed, only taking 72 ms to drop below the liquidus temperature. This result indicated that the molten alloy immediately lost its superheat after it made contact with the die cavity surface, and after a fast heat transfer at the metal/die interface. This phenomenon can be attributed to the prompt rise in the die surface temperature (T ds ). The casting surface temperature then continuously decreased, indicating a much smaller cooling rate, as also evidenced in Figure 5 from the smaller slope of the curve. Corresponding to the rapid decrease in the casting surface temperature, the IHTC abruptly increased immediately after the shot was performed. This abrupt increase was also associated with the rapid increase in the IHFD until the peak value of W m -2 was reached. The IHTC kept growing until reaching W m -2 K -1 when the IHTC started to fluctuate, rising and falling rapidly. The abrupt decrease in the IHTC was due to the fact that the close contact previously achieved between the casting and the die deteriorated. This deterioration was probably caused by the lack of the 4

6 required pressure transferred from inside as the solidification process proceeded. An analysis of the IHFD curve revealed that after the peak value was reached, the heat flux exponentially decayed until the value was at a much lower level. Figure 5. Typical results at the T4G position during the 30 th cycle under the basic condition. T 1, T 6, and T 3m denote the average measured temperatures at 1, 6, and 3 mm from the cavity surface, respectively. T 3c is the inversely calculated temperature at 3 mm from the cavity surface. T ds is the die surface temperature. T cs is the casting surface temperature. T cc is the casting center temperature. q is the interfacial heat flux density. h is the interfacial heat transfer coefficient Influence of interfacial heat transfer Casting geometry and sensor location. Figure 6 details the related heat transfer curves of IHTC and IHFD for all positions during the 30 th cycle under the basic condition, except for the T4F position due to data recording failure. The peak values of IHFD and IHTC are listed in Table 4. Similar trends in the profiles of both IHTC and IHFD were found. However, upon careful considerations of the different sensor locations and thicknesses, obvious differences between the corresponding interfacial heat transfer behaviors of the IHTC and IHFD curves (mainly, the shape and peak) were observed. First, the peak values of IHTC and IHFD varied with the different positions and casting thicknesses. The IHTC peak of T3G near the gate reached the peak value of W m -2 K -1, the second peak value was W m -2 K -1 of T5G, and the lowest peak was W m -2 K -1 of T1G near the gate of the thinnest plate T1. Similar patterns existed in the IHFD. Second, the high retention time of the IHTC varied with the different casting thicknesses. Figure 6 clearly shows that this time gradually extends with increased casting thickness. Finally, the mold filling sequence of each plate highly depended on the shape of the runner system. The different horizontal distance from each plate to the vertical runner and different size of each ingate could directly influence the interfacial heat transfer behavior. The interfacial heat transfer behavior of T1F and T3G could be good examples for well explaining this point. By comparing the heat transfer profiles of the four positions in two plates (T1 and T3) to others, different trends and special features could be found. For example, due to the ingate nearest the vertical runner in T3G, the headmost filling during the fast shot phase caused the overheating of the alloy impacting the surface of the die cavity. This phenomenon created the distinguishing feature of the filling, leading to the maximum IHTC. As for the thinnest plate T1, given the farthest distance to the vertical runner and the smallest ingate, the molten alloy abruptly sprayed into the cavity with the simulated velocity of around 110 m s -1 5

7 compared with the average ingate velocity of 50 m s -1 during the fast shot phase. Then, the metal jet first hit the sensor surface of the die cavity near the overflow (T1F) instead of that near the gate. Consequently, the IHFD near the overflow (T1F) reached its peak value W m -2 higher and 76 ms earlier than that near the gate (T1G), even compared with other positions. T1 plate was the thinnest (only 0.05 inch or 1.27 mm); hence, the quickly occurring solidification led to the slimmest IHTC profile. Helenius et al. [9] have similarly proposed that the peak value of the IHTC is greatly dependent on the status of the contact between the molten alloy and the die surface where the thermocouples are installed. A high IHTC peak value is observed when the melt alloy directly hits the location below which the thermocouples are adjusted during their initial contact. Figure 6. Profiles of IHTC and IHFD at all positions during the 30 th cycle under the basic condition. However, if the effect of the mold filling is ignored, the general trend of the IHTC peaks near the gate is higher than that near the overflow. The high retention time of the IHTC gradually extends with increased casting thickness. Sensor locations Near overflow (F) Near gate (G) Table 4. Values of q max and h max in Figure 6. Time b (s) q max values q 10 6 (W m -2 ) Time b (s) h max values h 10 3 (W m -2 K -1 ) T1F T2F T3F T4F a T5F T1G T2G T3G T4G T5G a Data at the T4G position failed. b Time of prompt rise point of q and h was about s, as shown in Figure 6. 6

8 Figure 7 shows the distribution of the IHTC and IHFD peaks at all positions of the finger-shaped casting for 30 cycles under the basic condition, except the T4F position. Considering the particular filling feature for plate T3, wherein the average of IHTC peaks was about W m -2 K -1, the heat transfer behavior was ignored in the following discussion. As for the positions near the gate, the peak range of the IHTC was from W m -2 K -1 to W m -2 K -1. A thicker casting corresponded to a higher average of IHTC peaks. The average of IHTC peaks at the T5G position of the thickest plate was about W m -2 K -1, whereas that at T1G was about W m -2 K -1. As for the positions near the overflow, a similar pattern was found, but the range of the averages was from W m -2 K -1 to W m -2 K -1, which was lower than that near the gate. The fastest ingate velocity of the thinnest plate T1 severely vibrates the IHTC peaks at the T1F position. This vibration is greater than that on the peaks near the gate. Figure 7. All peak values of IHTC and IHFD at all positions for 30 cycles under the basic condition. T1 to T5 are the five plates from the thinnest to the thickest, respectively Initial die surface temperature. After a great deal of data analysis of the finger-shaped casting of B390, the following findings were obtained. (1) The casting pressure had nearly no effect on the IHTC peak value. With increased intensification pressure, the distribution of IHTC peaks was more discrete because of the molten alloy more sharply impacting the die surface. (2) The peak value of IHTC near the gate increased with increased slow shot speed, especially for the plate T5. (3) The peak value of IHTC near the gate mildly decreased with increased fast shot speed. During HPDC process, the die temperature changes as the casting is sequentially performed and the initial die surface temperature (the die surface temperature before injection) on other hand could characterize the variation of the heat transfer between the metal and the die to some degree. After the thermal equilibrium inside the die was achieved, the initial die surface temperatures at different locations changed in a similar manner: the initial die surface temperature gradually grew as the castings were sequentially performed. Considering previous research on metal/die interfacial heat transfer behavior using the step shape casting for AM50 and ADC12 alloys, Guo et al. [5-6, 10] have found that the influence of processing parameters on the IHTC was mainly on the peak value. The shape of the IHTC profile was not significantly affected. The initial die surface temperature (T IDS ) had 7

9 the dominant influence on the IHTC peak value (h max ) out of all the processing parameters. By correlation analysis, the following relationship was found to have the best fit: 2 Ah max exp h IDS h B T (1) Figure 8. IHFD peaks as a function of the initial die surface temperature under all conditions. Figure 9. IHTC peaks as a function of the initial die surface temperature under all conditions. Figures 8 and 9 illustrate the IHFD and IHTC peaks at 9 positions of 5 plates for about 240 cycles under all conditions as a set of functions of the initial die surface temperature. One of the prominent characteristics was that all data followed a negative slope versus the initial die surface temperature. 8

10 The other parameters did not show such a large influence, including the casing pressure, the slow and fast shot speeds, as well as the pouring temperature. The peak values of the IHFD and IHTC were dominated by the initial die surface temperature, and changed according to: q max ln ln A lnt B 2 TIDS q q IDS q (2) h max ln ln A lnt B 2 TIDS h h IDS h where A q, B q, A h, and B q are the coefficients of fitting. Equations (2) and (3) show that the peak values of IHFD and IHTC decrease with increased initial die surface temperature. The coefficients of the fitting the peak value of IHFD and IHTC varied with the sensor location. By considering the IHTC as a function of the temperature gap at the interface, the effect of the initial die surface temperature can be easily understood. When the influence of the initial die surface temperature on the IHTC is considered, the explanation becomes easier. However, the possibility of the initial temperature of the die influencing the temperature of the liquid metal at the interface is not without merit, and should therefore influence the surface tension. As a consequence, the micro-contact conditions may change, resulting in the modifications of the heat resistance and interfacial heat transfer behavior. According to the relationship between the peak value of IHTC and the initial die surface temperature, the proper peak value of IHTC could be used for solidification simulation, as well as the function of temperature or solidification fraction considering the casting thickness. 4. Summary and conclusions A detailed method for measuring the heat transfer of B390 alloy during HPDC has been established using a finger-shaped casting. The IHFD and IHTC have been successfully determined based on the inverse method. Based on the results, the following conclusions are drawn. (1) The IHTC quickly increases right after the die casting shot until reaching the peak, and then slowly decreases. The peak range of the IHTC between the B390 alloy and H13 steel is from W m -2 K -1 to W m -2 K -1. (2) The mold-filling sequence of each plate directly influences the interfacial heat transfer behavior. The general value of the IHTC peak near the gate (from W m -2 K -1 to W m -2 K -1 ) is higher than that near the overflow (ranging from W m -2 K -1 to W m -2 K -1 ). The high retention time of the IHTC gradually extends with increased casting thickness. (3) The initial die surface temperature has the most dominant influence on the IHTC peak value among all the processing parameters. With increased initial die surface temperature, the IHTC peak decreases. The coefficients of fitting the IHFD and IHTC peaks vary with the sensor location. Acknowledgments The current research was funded by the Ministry of Science and Technology (MOST) of China under contract nos. 2011ZX , 2011BAE21B00, 2011ZX , and 2010DFA The die casting experiments were conducted with the aid of engineers from TOYO Machinery Co., Ltd. References [1] Hamasaiid A, Dour G, Dargusch M, Loulou T, Davidson C and Savage G 2006 Heat transfer at the casting/die interface in high pressure die casting - Experimental results and contribution to modeling Modeling of Casting, Welding and Advanced Solidification Processes - XI (Opio, France, 28 May-2 June 2006) (TMS) ed C A Gandin and M Bellet pp (3) 9

11 [2] Pehlke R D 1964 Unidirectional analysis of heat transfer during continuous casting Met. Eng. Q [3] Pehlke R D and Berry J T 2002 Heat transfer at the mold/metal interface in permanent mold casting of light alloys Proc. from the 2nd Int. Aluminum Casting Technology Symp. (Columbus, USA, 7-9 October) (ASM International) Ed M Tiryakioğlu and J Campbell pp [4] Lau F, Lee W B, Xiong S M and Liu B C 1998 Study of the interfacial heat transfer between an iron casting and a metallic mould J. Mater. Process. Technol [5] Guo Z P, Xiong S M, Liu B C Li M and Allison J 2008 Effect of process parameters, casting thickness, and alloys on the interfacial heat-transfer coefficient in the high-pressure diecasting process Metall. Mater. Trans. A 39A 2896 [6] Guo Z P 2009 Study on metal die interfacial heat transfer behaviour during high pressure die casting process Doctor of Engineering (Beijing: Tsinghua University) [7] Beck J V, Blackwell B and Haji-Sheikh A 1996 Comparison of some inverse heat conduction methods using experimental data Int. J. Heat Mass Tran [8] Helenius R, Lohne O, Arnberg L and Laukli H I 2005 The heat transfer during filling of a highpressure die-casting shot sleeve. Mater. Sci. Eng A 52 [9] Guo Z P, Xiong S M, Liu B C Li M and Allison J 2009 Understanding of the influence of process parameters on the heat transfer behavior at the metal/die interface in high pressure die casting process. Sci. in China Ser. E-Tech. Sci