MICROSTRUCTURE SIMULATION OF HIGH PRESSURE DIE CAST MAGNESIUM ALLOY BASED ON MODIFIED CA METHOD

This introduction paper is based on the paper "MICROSTRUCTURE SIMULATION OF HIGH PRESSURE DIE CAST MAGNESIUM ALLOY BASED ON MODIFIED CA METHOD" published by "ACTA METALLURGICA SINICA".

1. Overview:

• Title: MICROSTRUCTURE SIMULATION OF HIGH PRESSURE DIE CAST MAGNESIUM ALLOY BASED ON MODIFIED CA METHOD
• Author: WU Mengwu, XIONG Shoumei
• Year of publication: 2010
• Journal/academic society of publication: ACTA METALLURGICA SINICA
• Keywords: magnesium alloy, high pressure die casting, dendrite growth, nucleation model, microstructure simulation

2. Abstract:

As the lightest structural material, magnesium alloy has been widely used in the automotive, aerospace and electronic industries. High pressure die casting (HPDC) process is the dominant process for magnesium alloy products. The microstructure of die cast magnesium alloy has a great influence on the final performance of the castings. Numerical simulation provides a way to predict the solidification structure and the corresponding mechanical properties. However, as one of the most widely used methods in microstructure simulation, the cellular automaton (CA) method has difficulties in simulating the solidification structure of magnesium alloy with hcp crystal structure, though simulations of solidification structure for bcc and fcc metals have been widely reported. Besides, for the microstructure simulation of magnesium alloys by HPDC process, accurate nucleation model has to be considered, and by far little report was found on it.
In the present paper, based on the accurate temperature field of die castings obtained by an inverse heat transfer model, analysis of the temperature curves during solidification was made to establish a nucleation model that correlated the cooling rate with the nucleation density of magnesium alloys during solidification of HPDC process. A modified CA model was also developed to simulate the crystal growth of magnesium alloys. It takes account of the solute diffusion, constitutional undercooling, curvature undercooling, and anisotropy etc. Validations were made to the model, and the results show that the model has the capability to simulate the dendrite growth of magnesium alloy with different growth orientations. Besides, the model can also reveal the dendrite morphology with features of secondary and ternary dendrite branches, the dendrite competition growth under different temperature gradients and solidification rates, and the three dimensional morphology of the dendrite growth. To validate the nucleation and growth model established for magnesium alloy under HPDC process, “step-shape” die castings of AM50 magnesium alloy were produced at different process parameters. The average grain size prediction results are in good agreement with the experimental ones.

3. Introduction:

Magnesium alloy, the lightest structural material, sees widespread use in automotive, aerospace, and electronic industries due to advantages like low density, high specific strength and stiffness, and good formability. High pressure die casting (HPDC) is the primary and most widely used forming process for magnesium alloys, enabling the production of high-precision, lightweight, durable, and energy-efficient components, often suitable for thin-walled complex cavities. The microstructure of die-cast magnesium alloys significantly impacts their final performance. Therefore, understanding and controlling the microstructure formation during HPDC is crucial. Numerical simulation, particularly microstructure simulation, has become a valuable tool for predicting solidification behavior and optimizing process parameters. The cellular automaton (CA) method is widely applied for microstructure simulation due to its ability to model dendrite morphology and grain structure based on physical principles. However, simulating the hexagonal close-packed (hcp) structure of magnesium alloys presents challenges compared to the more commonly reported simulations for bcc and fcc metals, as standard CA models struggle to capture the specific anisotropy of hcp growth. Furthermore, accurate nucleation models specifically for the rapid solidification conditions inherent in the HPDC process of magnesium alloys are essential for predictive simulations but have received limited attention in previous research.

4. Summary of the study:

Background of the research topic:

The performance of HPDC magnesium alloy components is strongly linked to their microstructure. Predicting and controlling this microstructure through simulation is key to optimizing the HPDC process and ensuring desired component properties.

Status of previous research:

While the CA method is established for microstructure simulation, its application to hcp magnesium alloys faces difficulties in accurately representing anisotropic dendrite growth. Additionally, there is a lack of specific and validated nucleation models for magnesium alloys under the rapid cooling conditions characteristic of the HPDC process.

Purpose of the study:

This study aimed to develop and validate improved models for simulating the microstructure evolution of magnesium alloys during the HPDC process. Specifically, it focused on establishing an accurate nucleation model based on cooling rate and a modified CA growth model capable of simulating the anisotropic dendrite growth of hcp magnesium alloys.

Core study:

The core of the study involved:

1. Obtaining accurate temperature fields during HPDC using an inverse heat transfer model.
2. Analyzing solidification cooling curves to develop a nucleation model correlating nucleation density with cooling rate.
3. Developing a modified CA model for hcp magnesium alloy dendrite growth, incorporating factors like solute diffusion, constitutional undercooling, curvature undercooling, and interface anisotropy.
4. Validating the developed nucleation and growth models by simulating the microstructure of AM50 magnesium alloy "step-shape" castings produced under different HPDC process parameters and comparing the results with experimental observations.

5. Research Methodology

Research Design:

The research combined numerical simulation with experimental validation. An inverse heat transfer method was used to determine the thermal conditions during HPDC. Based on this, a nucleation model was formulated. A modified CA method was developed to simulate dendrite growth. The combined nucleation and growth models were then used to simulate the microstructure of an AM50 alloy casting. Finally, HPDC experiments were conducted using an AM50 alloy and a "step-shape" die to produce castings under varying process conditions, and their microstructures were characterized experimentally to validate the simulation predictions.

Data Collection and Analysis Methods:

• Temperature Field: An inverse heat transfer algorithm based on measured thermocouple data from the die was used to calculate the temperature field and cooling curves within the casting.
• Nucleation Model: Cooling curves were analyzed to determine nucleation temperatures and cooling rates. Experimental metallography (using the intercept method on polished and etched samples) was used to determine average grain size, which was converted to nucleation density. Relationships between cooling rate, undercooling, and nucleation density were established and fitted with mathematical functions.
• Growth Model: A modified CA model was implemented, incorporating equations for solute transport (diffusion), interface kinetics (considering solute, curvature, and thermal undercooling), and anisotropic growth specific to hcp structures. A specific neighbor definition and capture rule were used to simulate growth with different orientations.
• Simulation: The CA model, coupled with the nucleation model and macroscopic temperature fields (interpolated to the CA grid), was used to simulate dendrite growth, competitive growth, 3D morphology, and final grain structure in the AM50 casting.
• Experimental Validation: AM50 "step-shape" castings were produced on a TOYO650t HPDC machine under different pouring temperatures and mould preheating temperatures. Samples were sectioned, polished, and etched for metallographic analysis (Optical Microscopy implied by figures). Average grain sizes were measured experimentally and compared with simulation predictions.

Research Topics and Scope:

The study focused on:

• Developing a cooling-rate-dependent nucleation model for AM50 alloy in HPDC.
• Modifying the CA method to simulate anisotropic dendrite growth (including secondary/ternary arms, preferred orientations) of hcp magnesium alloys.
• Simulating dendrite competition growth under directional solidification conditions.
• Simulating 3D dendrite morphology.
• Simulating the microstructure (specifically average grain size) of an AM50 "step-shape" casting under various HPDC process conditions (different pouring and mould temperatures).
• Validating the simulation model against experimental results for average grain size.

6. Key Results:

Key Results:

1. An inverse heat transfer model provided accurate temperature field data for the HPDC process.
2. A nucleation model for HPDC magnesium alloys was established, correlating nucleation density with cooling rate. The relationship was found to be approximately linear at high cooling rates (> 350 K/s) and exponential at lower cooling rates (< 350 K/s). Nucleation undercooling generally increased with cooling rate.
3. A modified CA growth model was successfully developed and demonstrated its capability to simulate key features of hcp magnesium alloy dendrite growth, including growth with different orientations, the characteristic 60° branching angle for secondary and ternary arms, competitive growth under directional solidification (consistent with theoretical predictions λ₁∝G<0xE2><0x81><0xBB>⁰·⁵V<0xE2><0x81><0xBB>⁰·²⁵), and 3D dendrite morphology.
4. Simulations of the AM50 "step-shape" casting using the developed nucleation and growth models predicted microstructures (e.g., finer grains near the surface/thin sections, coarser grains towards the center) consistent with HPDC characteristics.
5. Validation showed good agreement between the simulated average grain sizes and those measured experimentally from AM50 castings produced under different pouring temperatures and mould temperatures. The model correctly predicted the trend of increasing grain size with higher pouring temperature and higher mould temperature.

Figure Name List:

• Fig.1 Thermal analysis of the temperature profile obtained by the inverse heat transfer model
• Fig.2 Relationships of nucleation density (a) and undercooling (b) with cooling rate
• Fig.3 Fitting between cooling rate with nucleation density at the lower cooling rate part (a) and the higher cooling rate part (b)
• Fig.4 Dendrite growth of magnesium alloy with different growth orientations (WAl—mass fraction of Al) (a) simulated result in a cooling rate of 10 K/s (b) OM photograph of casting of AZ91D with sand mould by using polarized light [15]
• Fig.5 Dendrite competition growth of magnesium alloy in direct solidification with different temperature gradients and solidification rates (a) GL=10 K/mm, V=2.5×10⁻⁵ m/s (b) GL=10 K/mm, V=2.5×10⁻⁴ m/s (c) GL=50 K/mm, V=2.5×10⁻⁴ m/s
• Fig.6 3D simulated results and SEM image [15] of casting of magnesium alloy (a) simulated result of single crystal dendrite growth with a cooling rate of 10 K/s (b) simulated result of columnar dendrite growth with a temperature gradient of 10 K/mm and a cooling rate of 1 K/s (c) SEM image of casting of AZ91D with permanent mould [15]
• Fig.7 Dimensions of the “step shape” casting [25]
• Fig.8 Comparison between simulated results (a—c) and metallographs (d—f) of the “step shape” castings (located at center of the Step 4 in Fig.7) of magnesium alloy AM50 with different pouring temperatures of 933 K (a, d), 953 K (b, e) and 983 K (c, f)
• Fig.9 Comparison between simulated results (a—c) and metallographs (d—f) of the “step shape” casting (located at surface of the Step 2 in Fig.7) of magnesium alloy AM50 with different mould preheating temperatures of 453 K (a, d), 423 K (b, e) and 393 K (c, f)
• Fig.10 Comparison between measured and simulated average grain size of the “step shape” casting of magnesium alloy AM50 with different mould preheating temperatures (located at surface of Step 2 in Fig.7) and pouring temperatures (located at center of Step 4 in Fig.7)

7. Conclusion:

1. Based on the temperature field obtained by the inverse heat transfer method, analysis of the solidification cooling curve allowed the establishment of a nucleation model for HPDC relating nucleation density to cooling rate. Near the casting surface and in thin sections where cooling rates are high, rapid chilling occurs, and the nucleation density is approximately linearly related to the cooling rate. In thicker sections near the center, where heat transfer is reduced and latent heat is released, the overall cooling rate is lower; additionally, considering the potential effect of externally solidified crystals (ESCs) transported from the shot sleeve increasing grain density in the center, the nucleation density follows an approximately exponential relationship with the cooling rate in this regime.
2. The developed growth model successfully simulates the growth of magnesium alloy dendrites with different orientations. The simulated morphology is similar to experimental observations, showing hexagonal symmetry on the basal plane. The model captures the competitive growth of columnar dendrites during directional solidification, with primary arm spacing following theoretical trends (decreasing with increasing temperature gradient and solidification rate). It also simulates 3D dendrite growth.
3. The developed nucleation and growth models were applied to simulate the microstructure of an AM50 "step-shape" casting under different pouring and mould temperatures. The predicted average grain sizes showed good agreement with experimental measurements, validating the applicability of the proposed models for simulating microstructure evolution in practical HPDC processes.

8. References:

• [1] Friedrich H, Schumann S. J Mater Process Technol, 2001; 117: 276
• [2] Mordike B L, Eert T. Mater Sci Eng, 2001; A302: 37
• [3] Li R D, Yu HP, Yuan X G. Foundry, 2003; 52:597 (李荣德,于海朋,袁晓光.铸造,2003;52:597)
• [4] Boettinger WJ, Coriell SR, Greer A L, Karma A, Kurz W, Rappaz M, Trivedi R. Acta Mater, 2000; 48: 43
• [5] Gandin C A, Rappaz M. Acta Metall Mater, 1994; 42: 2233
• [6] Rappaz M, Gandin C A, Desbiolles JL, Thévoz P. Metall Mater Trans, 1996; 27A: 695
• [7] Nastac L. Acta Mater, 1999; 47: 4253
• [8] Beltran-Sanchez L, Stefanescu D M. Metall Mater Trans, 2004; 35A: 2471
• [9] Wang W, Lee PD, Mclean M. Acta Mater, 2003; 51: 2971
• [10] Zhu MF, Hong C P. ISIJ Int, 2001; 41: 436
• [11] Zhu MF, Kim J M, Hong C P. ISIJ Int, 2001; 41: 992
• [12] Böttger B, Eiken J, Ohno M, Klaus G, Fehlbier M, Schmid-Fetzer R, Steinbach I, Buhrig-Polaczek A. Adv Eng Mater, 2006; 8: 241
• [13] Liu ZY, Xu Q Y, Liu B C. Acta Matall Sin, 2007; 43: 367 (刘志勇,许庆彦,柳百成.金属学报,2007;43:367)
• [14] Huo L, Han Z Q, Liu B C. Acta Matall Sin, 2009; 45: 1414 (霍亮,韩志强,柳百成.金属学报,2009;45:1414)
• [15] Fu Z N, Xu Q Y, Xiong S M. Chin J Nonferr Met, 2007; 17: 1567 (付振南,许庆彦,熊守美.中国有色金属学报,2007; 17: 1567)
• [16] Stefanescu D M, Upadhya G, Bandyopadhyay D. Metall Trans, 1990; 21A: 997
• [17] Oldfield W. Trans ASM, 1966; 59: 945
• [18] Thévoz P, Desbiolles JL, Rappaz M. Metall Trans, 1989; 20A: 311
• [19] Ohsasa K, Matsuura K, Kurokawa K, Watanabe S. Mater Sci Forum, 2008; 575: 154
• [20] Guo Z P. PhD thesis, Tsinghua University, Beijing, 2009 (郭志鹏.清华大学博士学位论文,北京,2009)
• [21] Sasikumar R, Sreenivasan R. Acta Metall Mater, 1994; 42: 2381
• [22] Beltran-Sanchez L, Stefanescu D M. Int J Cast Met Res, 2002; 15: 251
• [23] Laukli H I, Lohne O, Sannes S, Gjestland H, Arnberg L. Int J Cast Met Res, 2003; 16:515
• [24] Hunt J D, Lu S Z. Metall Mater Trans, 1996; 27A: 611
• [25] Guo Z P, Xiong S M, Cho S H, Choi JK. Acta Matall Sin, 2007; 43: 103 (郭志鹏,熊守美,曺尚铉,崔正吉.金属学报,2007; 43: 103)

9. Copyright:

• This material is a paper by "WU Mengwu, XIONG Shoumei". Based on "MICROSTRUCTURE SIMULATION OF HIGH PRESSURE DIE CAST MAGNESIUM ALLOY BASED ON MODIFIED CA METHOD".
• Source of the paper: https://doi.org/10.3724/SP.J.1037.2010.00279

This material is summarized based on the above paper, and unauthorized use for commercial purposes is prohibited.
Copyright © 2025 CASTMAN. All rights reserved.