Skip to main content
Log in

A microarchitecture design methodology to achieve extreme isotropic elastic properties of composites based on crystal symmetries

  • Research Paper
  • Published:
Structural and Multidisciplinary Optimization Aims and scope Submit manuscript

Abstract

The present contribution describes an optimization-based design technique of elastic isotropic periodic microarchitectures with crystal symmetries aiming at the realization of composites with extreme properties. To achieve this goal, three consecutive procedures are followed: (i) a series of inverse homogenization problems with symmetry constraints, (ii) a correlation analysis between symmetries and effective elastic properties of the attained microarchitectures, and (iii) the pattern resemblance recognition of these topologies and their redesign, by adopting microstructures with two length-scales, through optimized parametric geometries. This paper is devoted to assessing the third procedure because the first two procedures have been evaluated in previous works of the authors, and here they are only summarized. By applying the methodology, two plane group symmetries are assessed to define two families of 2D periodic parameterized microarchitecture. Once the parameters have been optimized, the resulting composites achieve elastic isotropic properties close to the whole range of the theoretically estimated bounds. Particularly, an unprecedented microstructure attaining the theoretical maximum stiffness is reported. Starting from these parameterized topologies, simple, one-length scale, and easily manufacturable geometries are defined. One of the so-designed microarchitectures has been manufactured and tested, displaying an effective Poisson’s ratio of − 0.90 simultaneously with a high shear modulus.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9

Similar content being viewed by others

Replication of results

The homemade codes, including the inverse homogenization of Section 3, the geometrical description of the parameterized microstructures with conformal meshes of Section 4, and the multi-objective optimization using a genetic algorithm of Section 5, have been implemented in Matlab. Numerical data characterizing the presented results, for instance, material properties of the phases, and optimal parameters are displayed in the paper. The full set of numerical and experimental data, as well as the codes, are freely available under request.

Data availability

A video showing the transition of the microstructures attained along the four Pareto fronts, and a .csv file describing the values of the parameters and properties of all the microstructures lying on the Pareto fronts, can be found in the open-source online data repository hosted at Mendeley Data (Rossi et al. 2020a).

Code availability

The homemade code is written in Matlab and is freely available under request.

Notes

  1. It is worth remarking that void is needed as the soft phase to achieve null eigenvalues of the constitutive tensor, achieving strict bimode materials.

  2. It is commonly observed that solutions of the inverse homogenization problem, in step 1 of item (i) may provide microarchitectures with higher symmetry to the imposed one. In these cases, the procedure in item (iii) adopts this higher symmetry.

  3. Note that orthogonality of the laminate axis and the stiff material on the interphase is lost when d is increased once laminate with γ3 vanishes.

References

Download references

Acknowledgments

The authors are grateful to Juan Marcos Banegas and Javier Andrés Acosta of CIESE-UTN FRSF for their help with the experimental study.

Funding

CONICET and ANPCyT - Argentina (grants PICT 2016-2673).

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Alfredo E. Huespe.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Additional information

Responsible Editor: Zhen Luo

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

(CSV 60.3 KB)

(MP4 2.74 MB)

Appendix. Equations used in the multi-objective optimization problems

Appendix. Equations used in the multi-objective optimization problems

In the mathematical formulation of the multi-objective problem (10), the volume fraction constraints are written as follows.

In the case of p31m microstructure, replacing (11) into (10):

$$ \gamma \ {\varOmega}_{\text{lam}}(\alpha,t) + {\varOmega}_{1}(\alpha,t) - {f_{1}^{t}} \ {\varOmega}_{\mu} = 0 , $$
(13)

where:

$$ \begin{array}{@{}rcl@{}} {\varOmega}_{\text{lam}} &=& 6 \times (\frac{\sqrt{3}/2}{\sin(120^{\circ} - \alpha/2) + \sin(\alpha/2)} \times L \times t - \frac{\sin(\beta_{1}+60^{\circ}-\alpha/2)}{\sin(\beta_{1}+90^{\circ}-\alpha/2)} \\ &&\times t^{2} - \frac{\sin(\beta_{2}-60^{\circ}+\alpha/2)}{\sin(\beta_{2}-30^{\circ}+\alpha/2)} \times t^{2}) , \\ &&{\varOmega}_{1} = \frac{3}{2} \sin(30^{\circ} + \alpha/2) \times (\frac{1}{\sin(\alpha/2)} + \frac{1}{\cos(30^{\circ} - \alpha/2)}) \times t^{2} ,\\ &&{\varOmega}_{\mu} = \frac{\sqrt{3}}{2} L^{2} . \end{array} $$
(14)

The angles β1 and β2 depend on α and are calculated from the following expressions:

$$ \begin{array}{@{}rcl@{}} \tan(\beta_{1}) = \frac{3 - 2\times \sin(30^{\circ} + \alpha)}{\sqrt{3} - 2 \times \cos(30^{\circ} + \alpha)} , \\ \tan(\beta_{2}) = \frac{3 + 2\times \cos(\alpha)}{\sqrt{3} + 2 \times \sin(\alpha)} . \end{array} $$
(15)

In the case of the p3m1, by replacing (12) into (10):

$$ \gamma_{1} \ {\varOmega}_{\text{lam}1}(d,t) + \gamma_{2} \ {\varOmega}_{\text{lam}2}(d,t) + \gamma_{3} \ {\varOmega}_{\text{lam}3}(d,t) + {\varOmega}_{1}(d,t) - {f_{1}^{t}} \ {\varOmega}_{\mu}= 0 , $$
(16)

where γ1, γ2, γ3, and \({f_{1}^{t}}\) are parameters defined in Fig. 5 and:

$$ \begin{array}{@{}rcl@{}} {\varOmega}_{\text{lam}1} = {\varOmega}_{\text{lam}2} = 3 \times d \times L \times t - \sqrt{3} \times t^{2} ,\\ {\varOmega}_{\text{lam}3} = 3 \times (L - 2 \times d \times L)\times t - \sqrt{3} \times t^{2} ,\\ {\varOmega}_{1} = \frac{3 \sqrt{3}}{2} t^{2} ,\\ {\varOmega}_{\mu} = \frac{\sqrt{3}}{2} L^{2} . \end{array} $$
(17)

The d parameter may increase beyond the vanishing of the γ3 laminate region (see for example microstructures P4 and P5 in Fig. 7). In this situation, the relation between the parameters to respect a given volume fraction comes from:

$$ \gamma_{1} \ {\varOmega}_{\text{lam}1}(d,t) + \gamma_{2} \ {\varOmega}_{\text{lam}2}(d,t) + {\varOmega}_{1}(d,t) - {f_{1}^{t}} \ {\varOmega}_{\mu} = 0 , $$
(18)

where:

$$ \begin{array}{@{}rcl@{}} {\varOmega}_{\text{lam}1} &=& {\varOmega}_{\text{lam}2} = 3 \times (L - d \times L)\times t - 2\times \sqrt{3} \times t^{2} ,\\ &&{\varOmega}_{1} = 2 \times \sqrt{3} \times t^{2} - 6 \times \sqrt{3} \times (L/2 - d \times L)^{2} . \end{array} $$
(19)

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Rossi, N., Podestá, J.M., Bre, F. et al. A microarchitecture design methodology to achieve extreme isotropic elastic properties of composites based on crystal symmetries. Struct Multidisc Optim 63, 2459–2472 (2021). https://doi.org/10.1007/s00158-020-02823-w

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00158-020-02823-w

Keywords

Navigation