Data-driven and topological design of structural metamaterials for fracture resistance
Introduction
It is well-known that stress concentration will cause premature fracture of any bearing structural materials. Avoiding such a phenomenon is among the most important criteria to design a structure in engineering. However, controlling the stress distribution from a multi-scale perspective has not yet been explored. Besides, establishing inverse design optimization frameworks to support the creation of structural metamaterials with optimal architectures and fracture resistance is still in its infancy.
Topological optimization is an inverse design method that tailors the morphology of the structure by distributing materials inside a given domain under constraints like equilibrium and boundary conditions. Initially flourishing in stiffness optimization, it has been extended to tailor material microstructures to achieve prescribed or extreme constitutive properties [1], [2]. Recently, structural fracture resistance design has attracted some attention. In this direction, early works combine explicit crack initiation and propagation with structural design for fracture resistance [3], [4]. Later on, gradient-based topology optimization frameworks are proposed, where fracture simulation accounts for the entire failure process involving multiple crack types, while distributions of the constituent materials are tailored to enhance mechanical fracture properties [5], [6], [7], [8], [9]. For instance, designing the architecture of the soft constituent in biomimic composites [10] enables energy dissipation ahead of the crack tip to prevent crack propagation and toughen the composite.
On the other hand, fracture-related design of structural materials has also been addressed by introducing stress constraints to prevent material failure [11], [12]. Thereafter, the stress-based topological optimization for specific objectives (e.g. minimal mass, maximal stiffness or minimal overall stress) have been extensively investigated (see, e.g., [13]). This approach is similar to fracture resistance design from the perspective of alleviating the stress concentration around the crack tip such that the crack initiation is prevented, enhancing the fracture performance. Ref. [14] validated that the stress-based optimal structures achieve improved strength and toughness compared with other designs such as the stiffness-oriented solutions. However, both stress and failure process-based fracture resistance optimization have thus far been locked in a single-scale framework. Topological design of the material microstructures to enhance the fracture mechanical properties is, to our best knowledge, investigated here for the first time. Moreover, the ability of optimized non-periodic porous structures to achieve even further improvement in fracture properties is demonstrated.
Fracture resistance design from a multiscale perspective is a challenging topic, mainly due to the lack of robust multiscale methods for fracture simulation in the presence of complex heterogeneous media. In addition, numerical optimization is difficult with such a high-dimensional problem where heterogeneous materials and structures acting at different scales need to be optimized simultaneously. Under this stalemate, data-driven methods came into being, providing novel and prosperous ways for multi-scale design optimization, e.g. in reducing dimensionality [15] and accelerating the design of nanocomposites [16] and multiscale structures [17]. For the first time, this work combines data-driven and topology design methods to optimize the microstructure of materials by maximizing the uniform distribution of the macrostructural stress, thus toughening and strengthening the global (macroscopic) porous structure composed of brittle constituent materials.
Three steps are implemented to validate the enhancement of the fracture toughness of the porous structures. Firstly, a multiscale inverse optimization framework is established to precisely control the global stress by tailoring the distribution of the components of the homogenized stiffness tensor. Thereafter, the architecture of each unit cell is identified from a large database via data-driven techniques while guaranteeing the connectivity between neighboring cells. Finally, the fracture resistance performance of the assembled structure is validated through the phase field method [18], [19], [20], [21].
Section snippets
Stress control
For a given global macroscopic structure under specific boundary conditions, e.g., Fig. 1, its stress distribution and fracture resistance are highly dependent on its microscopic geometries. The first goal is to establish an inverse topology optimization framework to control the stress distribution by tailoring the geometries of material microstructures. The final morphology of microstructures will be pulled from a database which contain both the information of the geometries and the
Database
Next, we will find the target microscopic unit cell in each finite element based on the data-driven method. The targeted unit cells are selected from the database to achieve the optimal distribution of the stiffness tensor and then assembled into the macroscopic global structure. The database of more than 160,000 unit cells, which can be found at https://github.com/Daicong-Da/2D-Orthotropic-Unit-Cell-Dataset.git, is illustrated in Fig. 3, displaying that the constructed database covers a large
Fracture resistance validation
For the fracture resistance validation, the number of the microscopic unit cells in the macrostructure of Fig. 1 is first selected as 5 by 10, in the - and - directions, respectively. As baselines, two classical shapes, structures A and B, are chosen, and they are depicted in Fig. 4(a) and (b), respectively. The non-periodic structure (c) is designed via the proposed data-driven and topological optimization framework to minimize the maximum stress in Eq. (5), where the objective function is
Conclusions
In summary, we proposed a data-driven design framework for brittle porous structure with the purpose to enhance fracture resistance. The topological-like optimization on distribution of the stiffness tensor components is first established to control the global homogenized stress. Thereafter, data-driven techniques are used to find the target unit cells with the geometries stratifying the optimal distribution of the stiffness tensor. Then, the macroscopic structure is assembled by the target
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.
Acknowledgments
We gratefully acknowledge the financial support of the NSF CSSI program (Grant No. OAC 1835782). Yu-Chin Chan thanks the NSF Graduate Research Fellowship (Grant No. DGE-1842165).
References (28)
Materials with prescribed constitutive parameters: an inverse homogenization problem
Int. J. Solids Struct.
(1994)Tailoring materials with prescribed elastic properties
Mech. Mater.
(1995)- et al.
Topology optimization for maximizing the fracture resistance of quasi-brittle composites
Comput. Methods Appl. Mech. Engrg.
(2018) - et al.
Fracture resistance design through biomimicry and topology optimization
Extrem. Mech. Lett.
(2020) - et al.
A framework for data-driven analysis of materials under uncertainty: Countering the curse of dimensionality
Comput. Methods Appl. Mech. Engrg.
(2017) - et al.
Revisiting brittle fracture as an energy minimization problem
J. Mech. Phys. Solids
(1998) - et al.
Numerical experiments in revisited brittle fracture
J. Mech. Phys. Solids
(2000) - et al.
A phase field model for rate-independent crack propagation: Robust algorithmic implementation based on operator splits
Comput. Methods Appl. Mech. Engrg.
(2010) - et al.
A review of homogenization and topology optimization I—homogenization theory for media with periodic structure
Comput. Struct.
(1998) - et al.
Regularized formulation of the variational brittle fracture with unilateral contact: Numerical experiments
J. Mech. Phys. Solids
(2009)
Optimization of composite fracture properties: method, validation, and applications
J. Appl. Mech.
Optimization of carbon black polymer composite microstructure for rupture resistance
J. Appl. Mech.
Topology optimization of particle-matrix composites for optimal fracture resistance taking into account interfacial damage
Internat. J. Numer. Methods Engrg.
Topology Optimization Design of Heterogeneous Materials and Structures
Cited by (32)
Topology optimization of irregular multiscale structures with tunable responses using a virtual growth rule
2024, Computer Methods in Applied Mechanics and EngineeringStructural design against brittle fracture: Optimizing energy release rate and experiment
2024, Computer Methods in Applied Mechanics and EngineeringDesign and mechanical properties analysis of hexagonal perforated honeycomb metamaterial
2024, International Journal of Mechanical SciencesComplex uncertainty-oriented robust topology optimization for multiple mechanical metamaterials based on double-layer mesh
2024, Computer Methods in Applied Mechanics and EngineeringTwo-scale data-driven design for heat manipulation
2024, International Journal of Heat and Mass TransferA Poisson's ratio sign-switching mechanical metamaterial with tunable stiffness
2023, International Journal of Mechanical Sciences