Additive manufacturing technology has brought unprecedented opportunities for the manufacture of complex parts of high-strength aluminum alloys, but the current additive manufacturing aluminum alloy system is still limited to castable and weldable Al-Si alloys, which restricts the rapid development of high-performance additive manufacturing aluminum alloys. In recent years, calculation methods at different scales have been gradually used to assist in the development of high-performance additive manufacturing aluminum alloys.
The paper "Research Status and Prospect of the Development of Compute-Aided High-Performance Additive Manufacturing Aluminum Alloys" published in the journal Acta Metallurgica summarizes in detail the research results of domestic and foreign scholars in the field of design and preparation of compute-aided additive manufacturing aluminum alloys, lists representative cases of atomic, mesoscopic and macro-scale computational simulations and machine learning and other computational methods to assist the design of aluminum alloys, analyzes the strategies of different calculation methods to assist the design of alloys, and points out their shortcomings. Finally, how to promote the application of multi-scale computing in the development of high-performance additive manufacturing aluminum alloys is prospected, and its development direction is pointed out.
© 3D Science Valley White Paper
Paper Links:
https://www.ams.org.cn/CN/10.11900/0412.1961.2022.00430
A selection of pictures and texts
Figure: Design flow of a novel crack-free Ti-modified Al-Cu-Mg alloy for SLM
Summary and outlook
In this paper, the current research ideas and status quo of different scale calculation methods (atomic-scale, meso-scale and macro-scale simulation and machine learning) to assist the development of additive manufacturing aluminum alloys at home and abroad are reviewed. Different scale calculation methods have been successfully used in the design and development of additive manufacturing aluminum alloys, effectively reducing the time and cost of material design and development, and realizing the efficient design of high-performance additive manufacturing aluminum alloys: atomic-scale first-principles calculations can provide guidance for the selection and addition of solid solution strengthening elements in additive manufacturing aluminum alloys. Atomic-scale molecular dynamics simulations predicted the local melting, rapid solidification process and mechanical response of additively manufactured aluminum alloys, which provided guidance for process selection and microstructure optimization in the alloy preparation process. Mesoscopic scale computational thermodynamics to realize the composition design and process optimization of crack-free additive manufacturing aluminum alloy from the aspects of alloy printability, grain refinement, solution strengthening, precipitation strengthening and heat treatment. The mesoscale phase field simulation can study the microstructure evolution during the preparation of additive aluminum alloys, explore the influence of process parameters on the microstructure, and provide guidance for the process optimization and microstructure optimization of the alloy preparation process. The macro-scale finite element simulation can predict and study the thermal action, forming control, defect formation and so on in the preparation process of additive manufacturing aluminum alloys, which is easy to guide the process optimization. Based on a large number of experimental data-driven machine learning methods, the process parameters and process parameters that affect the forming quality and performance can be identified and classified, and the quantitative relationship between different alloy systems from process to performance can be established, so as to realize the quality monitoring and process optimization of metal additive manufacturing alloys, and improve the comprehensive performance of products.
However, the current calculation methods of different scales are often only for part of the "composition-process-structure-performance" of additive manufacturing aluminum alloys, which seriously restricts the application of multi-scale calculations in the field of additive manufacturing. To this end, the efficient development of high-performance additively manufactured aluminum alloys can be realized in the following two aspects in the future.
First, an integrated computational materials engineering framework for additive manufacturing of aluminum alloys was established. The aim is to integrate the calculation methods of different scales into a whole system, and establish the quantitative relationship of "composition-process-structure-performance" of additive manufacturing aluminum alloys.
Mishra and Thapliyal propose an integrated computational materials engineering framework for additive manufacturing alloy design and the application of different scale calculations to it. The combination of computational thermodynamics and first-principles calculations establishes the quantitative relationship between the "composition-process-microstructure" of the alloy, which is conducive to the composition design of high-performance additive manufacturing aluminum alloys without cracks. The combination of molecular dynamics, phase field simulation and finite element simulation can quantitatively simulate the temperature field, stress field and microstructure evolution in the additive manufacturing process, establish the quantitative relationship between the "process and microstructure" of the alloy, and provide accurate and efficient guidance for the optimization of the alloy preparation process. The combination of machine learning and experiments establishes the quantitative relationship between "process-microstructure-performance" of alloys, which provides guidance for quality monitoring, process optimization, and product performance improvement of metal additive manufacturing. Finally, the combination of different scale calculations realizes the overall design of additive manufacturing aluminum alloy from "composition-process-structure-performance".
It should be pointed out that an efficient and reliable integrated computational materials engineering framework strongly relies on more rational computational methods and the support of high-throughput calculations. On the one hand, there are still needs to be developed for different scale calculation methods, for example, in the improved Scheil-Gulliver model mentioned above, the influence of rapid solidification on solute segregation has been considered, but there is still a bias in the prediction of partial stable phase inhibition in the actual multiphase multiphase system under rapid solidification conditions, so it is necessary to further improve the model. In terms of phase field models, the multiphase field model with finite interface dissipation can quantitatively describe the solute retention effect under extreme non-equilibrium conditions (rapid solidification process), but to achieve quantitative simulation of the microstructure evolution of the additive manufacturing process, it is necessary to couple reliable temperature and flow field equations. On the other hand, high-throughput calculations will further improve the efficiency of alloy design. High-throughput computation is the basis for realizing "on-demand material design", which can effectively narrow the scope of experiments and provide scientific basis for experiments. At present, high-throughput computation is time-consuming, and the efficient management of computing tasks and the post-processing of computing results are still challenging. The methods to achieve high-throughput computing mainly include parallel computing and distributed computing, and the authors of this paper have developed a distributed task management system (Malac-Distmas) accelerated by machine learning in the early stage to realize high-throughput computing and storage of various data. The system is embedded with machine learning technology, which can densify the output data, reduce the amount of computation, and accelerate high-throughput computing. By coupling Malac-Distmas with different thermodynamic calculation software, high-throughput calculations of Gibbs free energy, phase diagrams, Scheil-Gulliver simulations, diffusion simulations, precipitation simulations, and thermophysical parameters are realized. In addition, Malac-Distmas is not limited to high-throughput calculations of thermodynamic, kinetic, and thermophysical properties, but can also be coupled with other computational/simulation software/code to achieve high-throughput calculations/simulations.
Second, develop multi-objective design methods and optimization strategies/technologies for high-performance additive manufacturing aluminum alloys. Based on integrated materials engineering, the quantitative relationship of "composition-process-structure-performance" of additive manufacturing aluminum alloy is established, and the corresponding multi-objective design methods (such as crack-free high strength and high conductivity, crack-free high strength and high toughness, etc.) are developed according to different application backgrounds and material performance requirements, so as to realize the efficient design and development of new high-performance aluminum alloys for additive manufacturing.
Comprehensive performance is the premise to measure whether the material can meet the engineering application. However, due to the many factors influencing material properties, the interaction between properties is complex, such as the strength and plasticity/toughness, strength and conductivity of materials, which often conflict with each other, showing a contradictory relationship between one and the other. Therefore, the design and development of materials that balance the optimal values of various properties of materials and achieve the best comprehensive performance has always been a difficult problem in the field of materials. In terms of multi-objective design method, Yi et al. established the quantitative relationship of "composition-process-structure-performance" of rare earth/alkaline earth modified cast aluminum alloy by combining computational thermodynamics, experiment and machine learning, and carried out multi-objective design of the strength and plasticity of the alloy from various aspects such as castability, grain refinement, eutectic modification, solid solution strengthening, precipitated phase characteristics and heat treatment, and successfully developed a rare earth/alkaline earth modified cast aluminum alloy with high strength and toughness. In terms of multi-objective optimization strategies, it mainly includes layer-by-layer screening optimization, multi-objective to single-order target optimization, Pareto frontier collaborative optimization and other optimization strategies. Recently, Dai et al. combined 3D quantitative phase field simulation and hierarchical multi-objective optimization strategy to obtain the parameter relationship between model parameters, microstructure and various coating properties by conducting a large number of 3D phase field simulations of the growth process of TiN coatings in the physical vapor deposition (PVD) process. Based on quantitative phase field simulation and key experimental data, a hierarchical multi-objective method was proposed to design multiple coating properties. Subsequently, marginal utility was studied based on the identification of Pareto fronts for various target combinations. Model/process parameters are filtered in a hierarchical manner to find the optimal TiN coating performance window that is consistent with the experimental results. According to different application backgrounds and material performance requirements, the above methods are reasonably applied to the design of new aluminum alloys for additive manufacturing, which is expected to realize the efficient development of aluminum alloys with excellent comprehensive properties.
Quote text
Gao Jianbao, Li Zhicheng, Liu Jia, Zhang Jinliang, Song Bo, Zhang Lijun. Research Status and Prospect of Computing-Aided Development of High-Performance Additive Manufacturing Aluminum Alloys[J]. Acta Metallurgica Sinica, 2023, 59(1): 87-105
GAO Jianbao, LI Zhicheng, LIU Jia, ZHANG Jinliang, SONG Bo, ZHANG Lijun. Current Situation and Prospect of Computationally Assisted Design in High-Performance Additive Manufactured Aluminum Alloys: A Review[J]. Acta Metallurgica Sinica, 2023, 59(1): 87-105
doi:10.11900/0412.1961.2022.0043
l Valley Column l
Website Submissions l Send to [email protected]