Full length articleMechanical properties, failure mechanisms, and scaling laws of bicontinuous nanoporous metallic glasses
Graphical abstract
Introduction
Benefiting from a disordered atomic arrangement, metallic glasses (MGs) have many unique properties, such as high hardness, strength, and wear resistance [1], [2], [3]. By introducing pores in their structure, porous MGs (PMGs) synergize the qualities of porous structures and MGs. That enables strong and lightweight materials, which are suitable for various structural and functional applications, such as catalysis [4,5], sensing [6], dye degradation [7], membrane filters [8], and gas absorption [9]. PMGs can be obtained by various methods such as water vapor release [10], salt dissolution [11], and low-pressure infiltration of hollow carbon microspheres [12]. Some possible applications of PMGs benefit from a high surface to volume ratio, i.e., high specific area. PMGs with pore size in the nanometer scale maximize the specific area and are gaining more and more interest from the community. Tanaka et al. [13] fabricated Pd-based nanoporous MGs (NPMGs) by chemical dealloying a Pd30Ni50P20 MG. The 30–60 nm sized nanopores were shown to enhance remarkably the catalytic activity and reusability. Jiao et al. [14] fabricated 3-dimensional bicontinuous NPMGs by selectively dealloying of spinodally decomposed MG precursors and passivation. The underlying dealloying process can be adjusted for tuning the pore size in such nanoporous structures. The stochastic nature of the spinodal decomposition process and the nano-length scale of the porous features result in a high specific surface area making such material a promising candidate for gas absorption.
While MGs have attracted considerable interest due to their mechanical properties, their significant brittleness, which can lead to catastrophic failure, notably limits their possible applications. Surprisingly, recent studies of the mechanical properties of NPMGs have suggested that the introduction of pores and nanopores is able to alleviate the brittleness inherent to MGs. Atomistic investigations, e.g., using molecular dynamics (MD) simulations, have been increasingly used to complement experiments and unveil the underlying deformation mechanisms of NPMG during tensile loading [15], [16], [17], [18], [19]. Sopu et al. [15] employed MD to study Cu64Zr36 NPMGs and concluded that a structure with nanopores with optimized size and distribution could achieve homogenous plastic deformation while maintaining the strength close to that of monolithic bulk MGs (BMGs). Liu et al. [16] investigated the mechanical behavior of Ta NPMGs and found that the spatial distribution of pores and pore sizes have a strong influence on the displayed ductility. Lin et al. [17] simulated the tensile loading mechanical response of Cu50Zr50 NPMGs with well-defined cylindrical ligaments and observed a slenderness ratio dependent failure by necking. Liu et al. [18] investigated the tensile loading behavior of stochastic Cu64Zr36 bicontinuous NPMGs showing an anomalous ductile behavior, displaying an extended plastic regime and a gracious failure by collective ligaments necking. They described the deformation mechanism as a synergistic combination of delocalized necking of ligaments mostly aligned with the loading direction and concurrent progressive reorientation of remaining ligaments. In a similar study, Zhang et al. [19] investigated the sensitivity of the mechanical properties of Cu50Zr50 bicontinuous NPMG to porosity and temperature. They also attributed the plasticity found to the bending of the ligaments. Experimentally, it is challenging to directly apply tensile loading simulations on nanoporous samples. Instead, nanoindentation experiments are often conducted to evaluate the tensile behavior of nanoporous samples. Zhang et al. [20] employed nanoindentation, tension, and compression loading to evaluate the deformation behavior of Cu55.4Zr35.2Al7.5Y1.9 NPMG prepared using selective dissolution and found bending of ligaments and collapse of ligament nodes in the early stage of deformation, followed by plastic deformation and failure of ligaments.
While monolithic MGs have well defined mechanical properties, those of NPMGs depend on their porous structure topology. Efforts have been spent in developing scaling laws for NPMGs based on experimental results and theoretical calculations [19], [20], [21], [22], [23], [24], [25], [26]. However, current scaling laws are developed empirically or based on simplified topology models that consider only porosity as a variable. Functional relationships depending on multiple variables, such as porosity and composition, can be developed using Artificial Intelligence (AI) based methods. The emergence of AI methods is bringing a new dawn to the development of materials science [27,28]. Machine Learning (ML) and Genetic Programming (GP) algorithms have been experiencing resurgence in recent years within the materials science community [29,30]. AI technologies can provide researchers with tools to overcome the barriers between designing, synthesizing, and processing materials by accelerating simulations [31,32], prediction of properties [26,[33], [34], [35], [36]], design of synthetic routes [37,38], optimization of experimental parameters [39,40], and enhancement of characterization methods [41,42]. As for nanoporous materials, it is challenging to predict the mechanical properties due to their complex geometry. A suitable AI method, such as GP, which mimics natural selection, can be employed to describe the functional relationship between bicontinuous NPMGs mechanical properties and system variables.
Motivated by Darwin's theory of natural selection, GP is a powerful evolutionary technique commonly used to automatically generate programs suitable for user-defined tasks [43]. AI technologies are currently enabling the uncovering of physical laws. Among the AI methods, GP is a promising technique to acomplish symbolic regression, allowing one to find suitable mathematical models describing data, when little knowledge of the data structure or distribution is available [44,45]. Within the GP-based symbolic regression (GPSR), expressions are randomly generated and evolve in successive generations, which improve the description of the relationships of interest [46]. GPSR has already been applied to numerous problems [47]. In contrast to other AI methods [48], [49], [50], one can derive physical insights from the expressions obtained with the GP method. Chopra et al. [51] developed GP models to predict the compressive strength of concrete based on in situ data from literature. Cai et al. [52] used GPSR to derived heat transfer correlations, including the equation functional and its parameters, from experimental data on heat transfer measurements, which were used to predict the performance of thermal components. Langdon and Barrett [52] applied GP in drug discovery by evolving simple, biologically interpretable, in silico models of human oral bioavailability. Barmpalexis et al. [52] found a function mapping levels of four polymers to three different properties of a pharmaceutical release tablet with help of GPSR. Recently, Im et al. [53] applied the GP methodology to identify governing equations in non-linear multi-physics systems.
In the present work, we use MD simulations to study the mechanical properties of CuZr bicontinuous NPMGs. We then apply GPSR to derive scaling laws describing the mechanical behavior as a function of system variables and compare them against existing scaling relationships. The results indicate that the GPSR-derived models are able to predict accurately both the Young's modulus, E, and ultimate tensile strength (UTS) as a function of relative density and alloy composition. This research demonstrates that the GPSR is able to uncover expressions that can predict accurately the mechanical properties and also provide physical insights into complex systems.
Section snippets
Model generation and morphology characterization
Initially, cubic CuZr BMG simulation cells with 5.4 nm sides are prepared following the same procedure as reported previously [54]. Five different compositions, 0.25, 0.36, 0.50, 0.64, and 0.72 Cu are selected to study the effect of compositions on the mechanical behavior of bicontinuous NPMG. We chose this composition range because it is reported that the experimental CuZr BMGs can be synthesized within this range [55]. Periodic boundary conditions are applied along the three cartesian
Geometrical and morphological features of bicontinuous NPMG
The initial NPMG structures for 5 different porosities, , , , and , are shown in Fig. 1. The structures are self-similar to each other due to the level set method, i.e., the structures are generated from a common topology by scaling the porosity level. These bicontinuous structures consist of connected solid and porous phases. Fig. 1 also shows the normalized curvature distribution for the different initial NPMG structures. The two principal curvatures and are calculated
System size effect on mechanical properties
As we change the system size, we found that the ligament size has negligible influence on the mechanical properties of bicontinuous NPMG for all compositions when the porosity and other morphological features are kept fixed. Similar results have been found in the study of open cell porous Cu50Zr50 MG structures with by Zhang et al. [23]. The results obtained from tensile loading simulations indicate the absence of dependence between E, the UTS, and the yield strength with the ligament
Conclusions
In this work, we have investigated the mechanical properties of bicontinuous nanoporous CuZr MGs with different relative densities, system sizes, and compositions during tensile loading using MD simulations. While the samples have self-similar structures, the relative density dictates the surface morphology and ligament size. We found that the system size of bicontinuous NPMG has little impact on the mechanical properties while the relative density and the Cu concentration have strong effects.
Data availability
The data that support the findings of this study are available from the corresponding author on request.
Disclaimer
This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Award Number DE-SC0020295. Computation for the work described in this paper was supported by the University of Southern California Center for Advanced Research Computing (carc.usc.edu).
References (78)
- et al.
Metallic glasses as structural materials
Scr. Mater
(2006) - et al.
Atomic structure evolution in metallic glasses under cyclic deformation
Comput. Mater. Sci.
(2019) Manufacture, characterisation and application of cellular metals and metal foams
Prog. Mater. Sci.
(2001)- et al.
Recent development and application products of bulk glassy alloys☆
Acta Mater
(2011) - et al.
Towards easily tunable hydrogen storage via a hydrogen-induced glass-to-glass transition in Mg-based metallic glasses
Acta Mater
(2016) - et al.
Structure-property relationships in nanoporous metallic glasses
Acta Mater
(2016) - et al.
Brittle-to-ductile transition in monatomic Tantalum nanoporous metallic glass
J. Non. Cryst. Solids.
(2019) - et al.
Mechanical properties of nanoporous metallic glasses: insights from large-scale atomic simulations
Int. J. Plast.
(2020) - et al.
Bicontinuous nanoporous design induced homogenization of strain localization in metallic glasses
Scr. Mater.
(2021) - et al.
Deformation behavior of a nanoporous metallic glass at room temperature
Int. J. Plast.
(2022)