Design of periodic elastoplastic energy dissipating microstructures

Abstract

The design of periodic elastoplastic microstructures for maximum energy dissipation is carried out using topology optimization. While the topology optimization of elastic microstructures has been performed in numerous studies, microstructural design considering inelastic behavior is relatively untouched due to a number of reasons which are addressed in this study. An RVE-based multiscale model is employed for computational homogenization with periodic boundary constraints, satisfying the Hill-Mandel principle. The plastic anisotropy which may be prevalent in materials fabricated through additive manufacturing processes is considered by modeling the constitutive behavior at the microscale with Hoffman plasticity. Discretization is done using enhanced assumed strain elements to avoid locking from incompressible plastic flow under plane strain conditions and a Lagrange multiplier approach is used to enforce periodic boundary constraints in the discrete system. The design problem is formulated using a density-based parameterization in conjunction with a SIMP-like material interpolation scheme. Attention is devoted to issues such as dependence on initial design and enforcement of microstructural connectivity, and a number of optimized microstructural designs are obtained under different prescribed deformation modes.

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

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14

References

  1. Alberdi R, Khandelwal K (2017) Topology optimization of pressure dependent elastoplastic energy absorbing structures with material damage constraints. Finite Elem Anal Des 133:42–61. https://doi.org/10.1016/j.finel.2017.05.004

    MathSciNet  Article  Google Scholar 

  2. Alberdi R, Zhang G, Khandelwal K (2018a) A framework for implementation of rve-based multiscale models in computational homogenization using isogeometric analysis. Int J Numer Methods Eng 114(9):1018–1051. https://doi.org/10.1002/nme.5775

    MathSciNet  Article  Google Scholar 

  3. Alberdi R, Zhang G, Li L, Khandelwal K (2018b) A unified framework for nonlinear path-dependent sensitivity analysis in topology optimization. International Journal for Numerical Methods in Engineering. https://doi.org/10.1002/nme.5794

  4. Amstutz S, Andra H (2006) A new algorithm for topology optimization using a level-set method. J Comput Phys 216(2):573–588. https://doi.org/10.1016/j.jcp.2005.12.015

    MathSciNet  MATH  Article  Google Scholar 

  5. Amstutz S, Giusti SM, Novotny AA, de Souza Neto EA (2010) Topological derivative for multi-scale linear elasticity models applied to the synthesis of microstructures. Int J Numer Methods Eng 84(6):733–756. https://doi.org/10.1002/nme.2922

    MathSciNet  MATH  Article  Google Scholar 

  6. Aymeric M, Allaire G, Jouve F (2018) Elasto-plastic shape optimization using the level set method. SIAM J Control Optim 56(1):556–581. https://doi.org/10.1137/17M1128940

    MathSciNet  MATH  Article  Google Scholar 

  7. Bendsøe MP, Sigmund O (1999) Material interpolation schemes in topology optimization. Arch Appl Mech 69(9):635–654. https://doi.org/10.1007/s004190050248

    MATH  Google Scholar 

  8. Bendsøe MP, Sigmund O (2003) Topology optimization: theory, methods and applications, 2nd edn. Springer Science & Business Media, Berlin

    Google Scholar 

  9. Bensoussan A, Lions JL, Papanikolau G (1978) Asymptotic analysis for periodic structures. North Holland, Amsterdam

    Google Scholar 

  10. Blanco PJ, Sánchez PJ, de Souza Neto EA, Feijóo RA (2016) Variational foundations and generalized unified theory of rve-based multiscale models. Arch Comput Meth Eng 23(2):191–253. https://doi.org/10.1007/s11831-014-9137-5

    MathSciNet  MATH  Article  Google Scholar 

  11. Cadman JE, Zhou S, Chen Y, Li Q (2013) On design of multi-functional microstructural materials. J Mater Sci 48(1):51–66. https://doi.org/10.1007/s10853-012-6643-4

    Article  Google Scholar 

  12. Cantrell J, Rohde S, Damiani D, Gurnani R, DiSandro L, Anton J, Young A, Jerez A, Steinbach D, Kroese C, Ifju P (2017) Experimental characterization of the mechanical properties of 3d printed abs and polycarbonate parts. In: Yoshida S, Lamberti L, Sciammarella C (eds) Advancement of optical methods in experimental mechanics, vol 3. Springer International Publishing, Cham, pp 89–105

    Google Scholar 

  13. Carstensen JV, Lotfi R, Guest JK, Chen W, Schroers J (2015) Topology optimization of cellular materials with maximized energy absorption. International Design Engineering Technical Conferences and Computers and Information in Engineering Conference Volume 2B: 41st Design Automation Conference. https://doi.org/10.1115/DETC2015-47757

  14. Chen W, Xia L, Yang J, Huang X (2018) Optimal microstructures of elastoplastic cellular materials under various macroscopic strains. Mech Mater 118:120–132. https://doi.org/10.1016/j.mechmat.2017.10.002

    Article  Google Scholar 

  15. Christensen J, Kadic M, Kraft O, Wegener M (2015) Vibrant times for mechanical metamaterials. MRS Commun 5(3):453–462. https://doi.org/10.1557/mrc.2015.51

    Article  Google Scholar 

  16. Crisfield M (1991) Non-linear finite element analysis of solids and structures. Wiley, West Sussex

    Google Scholar 

  17. De Borst R, Feenstra PH (1990) Studies in anisotropic plasticity with reference to the hill criterion. Int J Numer Methods Eng 29(2):315–336. https://doi.org/10.1002/nme.1620290208

    MATH  Article  Google Scholar 

  18. De Souza Neto EA, Feijoo RA (2006) Variational foundations of multi-scale constitutive models of solid: small and large strain kinematical formulation. LNCC Research & Development Report 16

  19. De Souza Neto EA, Amstutz S, Giusti SM, Novotny AA (2010) Topological derivative-based optimization of micro-structures considering different multi-scale models. CMES: Comput Model Eng Sci 62(1):23–56. https://doi.org/10.3970/cmes.2010.062.023

    MathSciNet  MATH  Google Scholar 

  20. De Souza Neto EA, Peric D, Owen DRJ (2011) Computational methods for plasticity: theory and applications. Wiley, West Sussex

    Google Scholar 

  21. Deaton JD, Grandhi RV (2014) A survey of structural and multidisciplinary continuum topology optimization: post 2000. Struct Multidiscip Optim 49 (1):1–38. https://doi.org/10.1007/s00158-013-0956-z

    MathSciNet  Article  Google Scholar 

  22. Fleck NA, Deshpande VS, Ashby MF (2010) Micro-architectured materials: past, present and future. Proc R Soc London Math Phys Eng Sci 466(2121):2495–2516. https://doi.org/10.1098/rspa.2010.0215

    Article  Google Scholar 

  23. Frazier WE (2014) Metal additive manufacturing: a review. J Mater Eng Perform 23(6):1917–1928. https://doi.org/10.1007/s11665-014-0958-z

    Article  Google Scholar 

  24. Geers M, Kouznetsova V, Brekelmans W (2010) Multi-scale computational homogenization: trends and challenges. J Comput Appl Math 234(7):2175–2182. https://doi.org/10.1016/j.cam.2009.08.077

    MATH  Article  Google Scholar 

  25. Gibiansky LV, Sigmund O (2000) Multiphase composites with extremal bulk modulus. J Mech Phys Solids 48(3):461–498. https://doi.org/10.1016/S0022-5096(99)00043-5

    MathSciNet  MATH  Article  Google Scholar 

  26. Giusti S, Novotny A, de Souza Neto E, Feijoo R (2009) Sensitivity of the macroscopic elasticity tensor to topological microstructural changes. J Mech Phys Solids 57(3):555–570. https://doi.org/10.1016/j.jmps.2008.11.008

    MathSciNet  MATH  Article  Google Scholar 

  27. Guo N, Leu M C (2013) Additive manufacturing: technology, applications and research needs. Front Mech Eng 8(3):215–243. https://doi.org/10.1007/s11465-013-0248-8

    Article  Google Scholar 

  28. Hill R (1948) A theory of the yielding and plastic flow of anisotropic metals. Proc R Soc London Math Phys Eng Sci 193(1033):281–297. https://doi.org/10.1098/rspa.1948.0045

    MathSciNet  MATH  Article  Google Scholar 

  29. Hill R (1963) Elastic properties of reinforced solids: Some theoretical principles. J Mech Phys Solids 11 (5):357–372. https://doi.org/10.1016/0022-5096(63)90036-X

    MATH  MathSciNet  Article  Google Scholar 

  30. Hill R (1967) The essential structure of constitutive laws for metal composites and polycrystals. J Mech Phys Solids 15(2):79–95. https://doi.org/10.1016/0022-5096(67)90018-X

    MathSciNet  Article  Google Scholar 

  31. Kasper EP, Taylor RL (1997) A mixed enhanced strain method: Linear problems. Department of Civil and Environmental Engineering, University of California at Berkeley; Report No: UCB/SEMM-97/02

  32. Kasper EP, Taylor RL (2000) A mixed-enhanced strain method: Part i: Geometrically linear problems. Comput Struct 75(3):237–250. https://doi.org/10.1016/S0045-7949(99)00134-0

    Article  Google Scholar 

  33. Kato J, Hoshiba H, Takase S, Terada K, Kyoya T (2015) Analytical sensitivity in topology optimization for elastoplastic composites. Struct Multidiscip Optim 52(3):507–526. https://doi.org/10.1007/s00158-015-1246-8

    MathSciNet  Article  Google Scholar 

  34. Lee JH, Singer JP, Thomas EL (2012) Micro-/nanostructured mechanical metamaterials. Adv Mater 24(36):4782–4810. https://doi.org/10.1002/adma.201201644

    Article  Google Scholar 

  35. Lefebvre LP, Banhart J, Dunand DC (2008) Porous metals and metallic foams: Current status and recent developments. Adv Eng Mater 10(9):769–774. https://doi.org/10.1002/adem.200890025

    Article  Google Scholar 

  36. Li J, Wu B, Myant C (2016) The current landscape for additive manufacturing research. ICL AMN report

  37. Li L, Zhang G, Khandelwal K (2017a) Topology optimization of energy absorbing structures with maximum damage constraint. Int J Numer Methods Eng 112(7):737–775. https://doi.org/10.1002/nme.5531

    MathSciNet  Article  Google Scholar 

  38. Li L, Zhang G, Khandelwal K (2017b) Topology optimization of energy absorbing structures with maximum damage constraint. Int J Numer Methods Eng 112(7):737–775. https://doi.org/10.1002/nme.5531

    MathSciNet  Article  Google Scholar 

  39. Mandel J (1966) Contribution theorique a l’etude de l’ecrouissage et des lois de l’ecoulement plastique. Springer, Berlin

    Google Scholar 

  40. Maute K, Schwarz S, Ramm E (1998) Adaptive topology optimization of elastoplastic structures. Struct Optim 15(2):81–91. https://doi.org/10.1007/BF01278493

    Article  Google Scholar 

  41. Miehe C, Koch A (2002) Computational micro-to-macro transitions of discretized microstructures undergoing small strains. Arch Appl Mech 72(4):300–317. https://doi.org/10.1007/s00419-002-0212-2

    MATH  Article  Google Scholar 

  42. Neves M, Rodrigues H, Guedes J (2000) Optimal design of periodic linear elastic microstructures. Comput Struct 76(1):421–429. https://doi.org/10.1016/S0045-7949(99)00172-8

    Article  Google Scholar 

  43. Osanov M, Guest JK (2016) Topology optimization for architected materials design. Annu Rev Mater Res 46(1):211–233. https://doi.org/10.1146/annurev-matsci-070115-031826

    Article  Google Scholar 

  44. Pavliotis G, Stuart A (2008) Multiscale methods: averaging and homogenization. Springer-Verlag, New York

    Google Scholar 

  45. Rashed M, Ashraf M, Mines R, Hazell P J (2016) Metallic microlattice materials: A current state of the art on manufacturing, mechanical properties and applications. Mater Des 95:518–533. https://doi.org/10.1016/j.matdes.2016.01.146

    Article  Google Scholar 

  46. Saeb S, Steinmann P, Javili A (2016) Aspects of computational homogenization at finite deformations: a unifying review from reuss’ to voigt’s bound. Appl Mech Rev 68(5): 050801-33. https://doi.org/10.1115/1.4034024

    Article  Google Scholar 

  47. Schaedler TA, Ro CJ, Sorensen AE, Eckel Z, Yang SS, Carter WB, Jacobsen AJ (2014) Designing metallic microlattices for energy absorber applications. Adv Eng Mater 16(3):276–283. https://doi.org/10.1002/adem.201300206

    Article  Google Scholar 

  48. Schellekens J, De Borst R (1990) The use of the hoffman yield criterion in finite element analysis of anisotropic composites. Comput Struct 37 (6):1087–1096. https://doi.org/10.1016/0045-7949(90)90020-3

    Article  Google Scholar 

  49. Sigmund O (1994) Materials with prescribed constitutive parameters: An inverse homogenization problem. Int J Solids Struct 31(17):2313–2329. https://doi.org/10.1016/0020-7683(94)90154-6

    MathSciNet  MATH  Article  Google Scholar 

  50. Sigmund O (1995) Tailoring materials with prescribed elastic properties. Mech Mater 20(4):351–368. https://doi.org/10.1016/0167-6636(94)00069-7

    MathSciNet  Article  Google Scholar 

  51. Sigmund O (2000) A new class of extremal composites. J Mech Phys Solids 48(2):397–428. https://doi.org/10.1016/S0022-5096(99)00034-4

    MathSciNet  MATH  Article  Google Scholar 

  52. Simo JC, Rifai MS (1990) A class of mixed assumed strain methods and the method of incompatible modes. Int J Numer Methods Eng 29(8):1595–1638. https://doi.org/10.1002/nme.1620290802

    MathSciNet  MATH  Article  Google Scholar 

  53. Strang G (2007) Computational science and engineering. Wellesley-Cambridge Press, Wellesley

    Google Scholar 

  54. Svanberg K (1987) The method of moving asymptotes—a new method for structural optimization. Int J Numer Methods Eng 24(2):359–373. https://doi.org/10.1002/nme.1620240207

    MathSciNet  MATH  Article  Google Scholar 

  55. Swan CC, Arora JS (1997) Topology design of material layout in structured composites of high stiffness and strength. Struct Optim 13(1):45–59. https://doi.org/10.1007/BF01198375

    Article  Google Scholar 

  56. van Dijk NP, Maute K, Langelaar M, van Keulen F (2013) Level-set methods for structural topology optimization: a review. Struct Multidiscip Optim 48(3):437–472. https://doi.org/10.1007/s00158-013-0912-y

    MathSciNet  Article  Google Scholar 

  57. Xu Q, Lv Y, Dong C, Sreeprased TS, Tian A, Zhang H, Tang Y, Yu Z, Li N (2015) Three-dimensional micro/nanoscale architectures: fabrication and applications. Nanoscale 7:10883–10895. https://doi.org/10.1039/C5NR02048D

    Article  Google Scholar 

  58. Zhang G, Li L, Khandelwal K (2017) Topology optimization of structures with anisotropic plastic materials using enhanced assumed strain elements. Struct Multidiscip Optim 55(6):1965–1988. https://doi.org/10.1007/s00158-016-1612-1

    MathSciNet  Article  Google Scholar 

Download references

Funding

The presented work is supported in part by the US National Science Foundation through Grant CMMI-1762277. Any opinions, findings, conclusions, and recommendations expressed in this paper are those of the authors and do not necessarily reflect the views of the sponsors.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Kapil Khandelwal.

Additional information

Publisher’s Note

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

Responsible Editor: Jianbin Du

Appendices

Appendix A: Sensitivity derivatives for Hoffman model

This appendix presents the derivatives needed for sensitivity analysis using the Hoffman plasticity model which were originally derived in Ref. Zhang et al. (2017). These derivatives are based on the discrete form of the evolution equations which come from the numerical implementation using an elastic predictor/plastic return-mapping algorithm. As this implementation has been detailed in prior studies (De Borst and Feenstra 1990; Schellekens and De Borst 1990; De Souza Neto et al. 2011), it is not presented herein. Interested readers are directed to Appendix A in Zhang et al. (2017) for details on the elastic predictor/plastic return-mapping algorithm.

For the choice of ck shown in (45), the corresponding local constraint set Hk is

$$ \boldsymbol{H}^{k} = \left[\begin{array}{cc} \boldsymbol{H}^{k}_{1} \\ {\vdots} \\ \boldsymbol{H}^{k}_{n_{ele}} \end{array}\right] = \boldsymbol{0} \quad \text{with} \quad \boldsymbol{H}^{k}_{e} = \left[\begin{array}{llll} \boldsymbol{H}^{k}_{e_{1}} \\ \boldsymbol{H}^{k}_{e_{2}} \\ \boldsymbol{H}^{k}_{e_{3}} \\ \boldsymbol{H}^{k}_{e_{4}} \\ \tilde{\boldsymbol{H}}^{k}_{e} \end{array}\right] $$
(57)

Here,

$$ \tilde{\boldsymbol{H}}^{k}_{e} = \sum\limits_{r = 1}^{n_{ipt}} \boldsymbol{G}^{T}_{e_{r}} \boldsymbol{\sigma}^{k}_{e_{r}} w_{r} = \boldsymbol{0} $$
(58)

and for an elastic step

$$ \boldsymbol{H}^{k}_{e_{r}} = \left\{\begin{array}{llll} \boldsymbol{h}^{k}_{e_{r1}} = \hat{\boldsymbol{\sigma}}^{k}_{e_{r}} - \hat{\boldsymbol{C}}^{e} \left( \hat{\boldsymbol{\varepsilon}}^{k}_{e_{r}} - \hat{\boldsymbol{\varepsilon}}^{p^{k}}_{e_{r}} \right) = \boldsymbol{0} \\ \boldsymbol{h}^{k}_{e_{r2}} = \hat{\boldsymbol{\varepsilon}}^{p^{k}}_{e_{r}} - \hat{\boldsymbol{\varepsilon}} ^{p^{k-1}}_{e_{r}} = \boldsymbol{0} \\ h^{k}_{e_{r3}} = \alpha^{k}_{e_{r}} - \alpha^{k-1}_{e_{r}} = 0 \\ h^{k}_{e_{r4}} = {\Delta}\gamma^{k}_{e_{r}} = 0 \end{array}\right. $$
(59)

while for a plastic step

$$ \boldsymbol{H}^{k}_{e_{r}} = \left\{\begin{array}{llll} \boldsymbol{h}^{k}_{e_{r1}} = \hat{\boldsymbol{\sigma}}^{k}_{e_{r}} - \hat{\boldsymbol{C}}^{e} \left( \hat{\boldsymbol{\varepsilon}}^{k}_{e_{r}} - \hat{\boldsymbol{\varepsilon}}^{p^{k}}_{e_{r}} \right) = \boldsymbol{0} \\ \boldsymbol{h}^{k}_{e_{r2}} = \hat{\boldsymbol{\varepsilon}}^{p^{k}}_{e_{r}} - \hat{\boldsymbol{\varepsilon}} ^{p^{k-1}}_{e_{r}} - {\Delta}\gamma^{k}_{e_{r}} \left( \boldsymbol{P}\hat{\boldsymbol{\sigma}}^{k}_{e_{r}} + \boldsymbol{q} \right) = \boldsymbol{0} \\ h^{k}_{e_{r3}} = \alpha^{k}_{e_{r}} - \alpha^{k-1}_{e_{r}} - \sqrt{\frac{2z^{k}_{e_{r}}}{3}} {\Delta}\gamma^{k}_{e_{r}} = 0 \\ h^{k}_{e_{r4}} = \frac{1}{2} \hat{\boldsymbol{\sigma}}^{k^{T}}_{e_{r}} \boldsymbol{P} \hat{\boldsymbol{\sigma}} ^{k}_{e_{r}} + \boldsymbol{q}^{T} \hat{\boldsymbol{\sigma}}^{k}_{e_{r}} - \zeta^{2}(\alpha^{k}_{e_{r}}) = 0 \end{array}\right. $$
(60)

with \(z^{k}_{e_{r}} = \left ( \boldsymbol {P}\hat {\boldsymbol {\sigma }}^{k}_{e_{r}} + \boldsymbol {q} \right )^{T} \boldsymbol {Z} \left ( \boldsymbol {P}\hat {\boldsymbol {\sigma }}^{k}_{e_{r}} + \boldsymbol {q} \right )\). The derivative Hk/uk− 1 = 0 while

$$\begin{array}{@{}rcl@{}} \frac{\partial{\boldsymbol{H}^{k}}}{\partial{\boldsymbol{u}^{k}}} &=& \left[\begin{array}{cc} \frac{\partial{\boldsymbol{H}^{k}_{1}}}{\partial{\boldsymbol{u}^{k}}} \\ {\vdots} \\ \frac{\partial{\boldsymbol{H}^{k}_{n_{ele}}}}{\partial{\boldsymbol{u}^{k}}} \end{array}\right] \text{with } \\ \frac{\partial{\boldsymbol{H}^{k}_{j}}}{\partial{\boldsymbol{u}^{k}}}&=& \overset{n_{ele}}{\underset{e = 1}{\mathcal{A}}} \left( \frac{\partial{\boldsymbol{H}^{k}_{j}}}{\partial{\boldsymbol{u}^{k}_{e}}} \right), \quad j = 1,2,...,n_{ele} \end{array} $$
(61)

The terms \(\partial {\boldsymbol {H}^{k}_{j}} / \partial {\boldsymbol {u}^{k}_{e}}\) are nonzero only when j = e and are calculated as

$$\begin{array}{@{}rcl@{}} \frac{\partial{\boldsymbol{H}^{k}_{e}}}{\partial{\boldsymbol{u}^{k}_{e}}} &=& \left[\begin{array}{llllllll} \frac{\partial{\boldsymbol{H}^{k}_{e_{1}}}}{\partial{\boldsymbol{u}^{k}_{e}}} &\frac{\partial{\boldsymbol{H}^{k}_{e_{2}}}}{\partial{\boldsymbol{u}^{k}_{e}}} & \frac{\partial{\boldsymbol{H}^{k}_{e_{3}}}}{\partial{\boldsymbol{u}^{k}_{e}}} &\frac{\partial{\boldsymbol{H}^{k}_{e_{4}}}}{\partial{\boldsymbol{u}^{k}_{e}}} & \frac{\partial{\tilde{\boldsymbol{H}}^{k}_{e}}}{\partial{\boldsymbol{u}^{k}_{e}}} \end{array}\right]^{T} \\ \frac{\partial{\boldsymbol{H}^{k}_{e_{r}}}}{\partial{\boldsymbol{u}^{k}_{e}}} &=& \left[\begin{array}{ccc} -\hat{\boldsymbol{C}}^{e} \boldsymbol{T}_{2} \boldsymbol{B}_{r} \\ \boldsymbol{0} \\ \boldsymbol{0} \\ \boldsymbol{0} \end{array}\right] \quad \text{and} \quad \frac{\partial{\tilde{\boldsymbol{H}}^{k}_{e}}}{\partial{\boldsymbol{u}^{k}_{e}}} = \boldsymbol{0} \end{array} $$
(62)

The derivatives Hk/ck and Hk/ck− 1 have the same structure, i.e.

$$\begin{array}{@{}rcl@{}} &\frac{\partial\boldsymbol{H}^{k}}{\partial\boldsymbol{c}^{k}}=\left[\begin{array}{cccc} \frac{\partial\boldsymbol{H}_{1}^{k}}{\partial\boldsymbol{c}_{1}^{k}} & \boldsymbol{0} & {\dots} & \boldsymbol{0}\\ \boldsymbol{0} & \frac{\partial\boldsymbol{H}_{2}^{k}}{\partial\boldsymbol{c}_{2}^{k}} & {\dots} & \boldsymbol{0}\\ {\vdots} & {\vdots} & {\ddots} & \vdots\\ \boldsymbol{0} & \boldsymbol{0} & {\dots} & \frac{\partial\boldsymbol{H}_{n_{ele}}^{k}}{\partial\boldsymbol{c} _{n_{ele}}^{k}} \end{array}\right]\\ &\frac{\partial\boldsymbol{H}^{k}}{\partial\boldsymbol{c}^{k-1}}=\left[\begin{array}{cccc} \frac{\partial\boldsymbol{H}_{1}^{k}}{\partial\boldsymbol{c}_{1}^{k-1}} & \boldsymbol{0} & {\dots} & \boldsymbol{0}\\ \boldsymbol{0} & \frac{\partial\boldsymbol{H}_{2}^{k}}{\partial\boldsymbol{c}_{2}^{k-1}} & {\dots} & \boldsymbol{0}\\ {\vdots} & {\vdots} & {\ddots} & \vdots\\ \boldsymbol{0} & \boldsymbol{0} & {\dots} & \frac{\partial\boldsymbol{H}_{n_{ele}}^{k}}{\partial\boldsymbol{c} _{n_{ele}}^{k-1}} \end{array}\right] \end{array} $$
(63)

since \(\boldsymbol {c}^{k}_{i}\) and \(\boldsymbol {c}^{k}_{j}\) are independent and \(\boldsymbol {H}^{k}_{i}\) and \(\boldsymbol {H}^{k}_{j}\) are uncoupled when ij. The nonzero sub-matrices have the form

$$\begin{array}{@{}rcl@{}} &\frac{\partial\boldsymbol{H}^{k}_{e}}{\partial\boldsymbol{c}^{k}_{e}}=\left[\begin{array}{cccccc} \frac{\partial\boldsymbol{H}_{e_{1}}^{k}}{\partial\boldsymbol{c}_{e_{1}}^{k}} & \boldsymbol{0} & \boldsymbol{0} & \boldsymbol{0} & \frac{\partial\boldsymbol{H}_{e_{1}}^{k}}{\partial\tilde{\boldsymbol{\alpha}}_{e}^{k}}\\ \boldsymbol{0} & \frac{\partial\boldsymbol{H}_{e_{2}}^{k}}{\partial\boldsymbol{c}_{e_{2}}^{k}} & \boldsymbol{0} & \boldsymbol{0} & \frac{\partial\boldsymbol{H}_{e_{2}}^{k}}{\partial\tilde{\boldsymbol{\alpha}}_{e}^{k}}\\ \boldsymbol{0} & \boldsymbol{0} & \frac{\partial\boldsymbol{H}_{e_{3}}^{k}}{\partial\boldsymbol{c}_{e_{3}}^{k}} & \boldsymbol{0} & \frac{\partial\boldsymbol{H}_{e_{3}}^{k}}{\partial\tilde{\boldsymbol{\alpha}}_{e}^{k}}\\ \boldsymbol{0} & \boldsymbol{0} & \boldsymbol{0} &\frac{\partial\boldsymbol{H}_{e_{4}}^{k}}{\partial\boldsymbol{c}_{e_{4}}^{k}} &\frac{\partial\boldsymbol{H}_{e_{4}}^{k}}{\partial\tilde{\boldsymbol{\alpha}}_{e}^{k}} \\ \frac{\partial\tilde{\boldsymbol{H}}_{e}^{k}}{\partial\boldsymbol{c}_{e_{1}}^{k}} & \frac{\partial\tilde{\boldsymbol{H}}_{e}^{k}}{\partial\boldsymbol{c}_{e_{2}}^{k}} &\frac{\partial\tilde{\boldsymbol{H}}_{e}^{k}}{\partial\boldsymbol{c}_{e_{3}}^{k}} & \frac{\partial\tilde{\boldsymbol{H}}_{e}^{k}}{\partial\boldsymbol{c}_{e_{4}}^{k}} & \boldsymbol{0} \end{array}\right]\\ \end{array} $$
$$\begin{array}{@{}rcl@{}} &\frac{\partial\boldsymbol{H}^{k}_{e}}{\partial\boldsymbol{c}^{k-1}_{e}}=\left[\begin{array}{ccccc} \frac{\partial\boldsymbol{H}_{e_{1}}^{k}}{\partial\boldsymbol{c}_{e_{1}}^{k-1}} & \boldsymbol{0} & \boldsymbol{0} & \boldsymbol{0} & \frac{\partial\boldsymbol{H}_{e_{1}}^{k}}{\partial\tilde{\boldsymbol{\alpha}}_{e}^{k-1}} \\ \boldsymbol{0} & \frac{\partial\boldsymbol{H}_{e_{2}}^{k}}{\partial\boldsymbol{c}_{e_{2}}^{k-1}} & \boldsymbol{0} & \boldsymbol{0} & \frac{\partial\boldsymbol{H}_{e_{2}}^{k}}{\partial\tilde{\boldsymbol{\alpha}}_{e}^{k-1}} \\ \boldsymbol{0} & \boldsymbol{0} & \frac{\partial\boldsymbol{H}_{e_{3}}^{k}}{\partial\boldsymbol{c}_{e_{3}}^{k-1}} & \boldsymbol{0} & \frac{\partial\boldsymbol{H}_{e_{3}}^{k}}{\partial\tilde{\boldsymbol{\alpha}}_{e}^{k-1}} \\ \boldsymbol{0} & \boldsymbol{0} & \boldsymbol{0} &\frac{\partial\boldsymbol{H}_{e_{4}}^{k}}{\partial\boldsymbol{c}_{e_{4}} ^{k-1}} & \frac{\partial\boldsymbol{H}_{e_{4}}^{k}}{\partial\tilde{\boldsymbol{\alpha}}_{e}^{k-1}} \\ \frac{\partial\tilde{\boldsymbol{H}}_{e}^{k}}{\partial\boldsymbol{c}_{e_{1}}^{k-1}} & \frac{\partial\tilde{\boldsymbol{H}}_{e}^{k}}{\partial\boldsymbol{c}_{e_{2}}^{k-1}} &\frac{\partial\tilde{\boldsymbol{H}}_{e}^{k}}{\partial\boldsymbol{c}_{e_{3}}^{k-1}} & \frac{\partial\tilde{\boldsymbol{H}}_{e}^{k}}{\partial\boldsymbol{c}_{e_{4}}^{k-1}} & \boldsymbol{0} \end{array}\right] \end{array} $$
(64)

The terms in \(\partial {\boldsymbol {H}^{k}_{e}} / \partial {\boldsymbol {c}_{e_{1}}^{k-1}}\) are the same for both elastic and plastic steps, i.e.

$$ \frac{\partial{\boldsymbol{H}_{e_{r}}^{k}}}{\partial{\boldsymbol{c}_{e_{r}}^{k-1}}} = \left[\begin{array}{cccc} \boldsymbol{0} & \boldsymbol{0} & \boldsymbol{0} & \boldsymbol{0} \\ \boldsymbol{0} & -\boldsymbol{T}_{2} & \boldsymbol{0} & \boldsymbol{0} \\ \boldsymbol{0} & \boldsymbol{0} & -1 & 0 \\ \boldsymbol{0} & \boldsymbol{0} & 0 & 0 \end{array}\right], \quad \frac{\partial\boldsymbol{H}_{e_{r}}^{k}}{\partial\tilde{\boldsymbol{\alpha}}_{e}^{k-1}} = \boldsymbol{0}, \quad \frac{\partial\tilde{\boldsymbol{H}}_{e}^{k}}{\partial\boldsymbol{c}_{e_{r}}^{k-1}} = \boldsymbol{0} $$
(65)

The terms \(\partial \boldsymbol {H}_{e_{r}}^{k} / \partial \tilde {\boldsymbol {\alpha }}_{e}^{k}\) and \(\partial \tilde {\boldsymbol {H}}_{e}^{k} / \partial \boldsymbol {c}_{e_{r}}^{k}\) in \(\partial \boldsymbol {H}^{k}_{e} / \partial \boldsymbol {c}^{k}_{e}\) are

$$ \frac{\partial\boldsymbol{H}_{e_{r}}^{k}}{\partial\tilde{\boldsymbol{\alpha}}_{e}^{k}} = \left[\begin{array}{cccc} -\hat{\boldsymbol{C}}^{e} \boldsymbol{T}_{2} \boldsymbol{G}_{e_{r}} \\ \boldsymbol{0} \\ 0 \\ 0 \end{array}\right], \quad \frac{\partial\tilde{\boldsymbol{H}}_{e}^{k}}{\partial\boldsymbol{c}_{e_{r}}^{k}} = \left[\begin{array}{llll} w_{r} \boldsymbol{G}_{e_{r}}^{T} & \boldsymbol{0} & \boldsymbol{0} & \boldsymbol{0} \end{array}\right] $$
(66)

regardless of whether the step is elastic or plastic. Finally, the term \(\partial \boldsymbol {H}_{e_{r}}^{k} / \partial \boldsymbol {c}_{e_{r}}^{k}\) at an elastic step is

$$ \frac{\partial\boldsymbol{H}_{e_{r}}^{k}}{\partial\boldsymbol{c}_{e_{r}}^{k}} = \left[\begin{array}{cccc} \boldsymbol{T}_{1} & \hat{\boldsymbol{C}}^{e} \boldsymbol{T}_{2} & \boldsymbol{0} & \boldsymbol{0} \\ \boldsymbol{0} & \boldsymbol{T}_{2} & \boldsymbol{0} & \boldsymbol{0} \\ \boldsymbol{0} & \boldsymbol{0} & 1 & 0 \\ \boldsymbol{0} & \boldsymbol{0} & 0 & 1 \end{array}\right] $$
(67)

and at a plastic step is

$$\begin{array}{@{}rcl@{}} &&{}\frac{\partial\boldsymbol{H}_{e_{r}}^{k}}{\partial\boldsymbol{c}_{e_{r}}^{k}} = \left[\begin{array}{llll} \boldsymbol{M}_{11} & \boldsymbol{M}_{12} & \boldsymbol{M}_{13} & \boldsymbol{M}_{14} \\ \boldsymbol{M}_{21} & \boldsymbol{M}_{22} & \boldsymbol{M}_{23} & \boldsymbol{M}_{24} \\ \boldsymbol{M}_{31} & \boldsymbol{M}_{32} & M_{33} & M_{34} \\ \boldsymbol{M}_{41} & \boldsymbol{M}_{42} & M_{43} & M_{44} \end{array}\right]\\ &&\boldsymbol{{}M}_{11} = \boldsymbol{T}_{1}, \quad \boldsymbol{M}_{12} = \hat{\boldsymbol{C}}^{e} \boldsymbol{T}_{2}, \quad \boldsymbol{M}_{13} = \boldsymbol{0}, \quad \boldsymbol{M}_{14} = \boldsymbol{0} \\ &&\boldsymbol{{}M}_{21} = -{\Delta}\gamma^{k}_{e_{r}} \boldsymbol{P} \boldsymbol{T}_{1}, \boldsymbol{M}_{22} = \boldsymbol{T}_{2}, \boldsymbol{M}_{23} = \boldsymbol{0}, \boldsymbol{M}_{24} = -\left( \boldsymbol{P}\hat{\boldsymbol{\sigma}}^{k}_{e_{r}} + \boldsymbol{q} \right) \\ &&\boldsymbol{{}M}_{31} = -\sqrt{\frac{1}{6z^{k}_{e_{r}}}} {\Delta}\gamma^{k}_{e_{r}} \frac{dz^{k}_{e_{r}}}{d\hat{\boldsymbol{\sigma}}^{k}_{e_{r}}} \boldsymbol{T}_{1}, \boldsymbol{M}_{32} = \boldsymbol{0}, M_{33} = 1, M_{34} = -\sqrt{\frac{2z^{k}_{e_{r}}}{3}} \\ &&\boldsymbol{{}M}_{41} = \hat{\boldsymbol{\sigma}}^{k^{T}}_{e_{r}} \boldsymbol{P} \boldsymbol{T}_{1} + \boldsymbol{q}^{T} \boldsymbol{T}_{1}, \boldsymbol{\!M}_{42} = \boldsymbol{0}, \!M_{43} = \!-2\zeta \left( \alpha^{k}_{e_{r}} \right) \!\frac{d\zeta}{d\alpha^{k}_{e_{r}}}, M_{44} = 0\\ \end{array} $$
(68)

The derivative with respect to the element density variables is

$$ \frac{\partial\boldsymbol{H}^{k}}{\partial\boldsymbol{\rho}} = \left[\begin{array}{ccccc} \frac{\partial\boldsymbol{H}_{1}^{k}}{\partial\rho_{1}} & \boldsymbol{0} & {\dots} & \boldsymbol{0} \\ \boldsymbol{0} & \frac{\partial\boldsymbol{H}_{2}^{k}}{\partial\rho_{2}} & {\dots} & \boldsymbol{0} \\ {\vdots} & {\vdots} & {\ddots} & {\vdots} \\ \boldsymbol{0} & \boldsymbol{0} & {\dots} & \frac{\partial\boldsymbol{H}_{n_{ele}}^{k}}{\partial\rho _{n_{ele}}} \end{array}\right] $$
(69)

The nonzero entries \(\partial \boldsymbol {H}_{e}^{k} / \partial \rho _{e}\) are

$$\begin{array}{@{}rcl@{}} &&\frac{\partial\boldsymbol{H}_{e}^{k}}{\partial\rho_{e}} = \left[\begin{array}{ccccc} \frac{\partial\boldsymbol{H}_{e_{1}}^{k}}{\partial\rho_{e}} &\frac{\partial\boldsymbol{H}_{e_{2}}^{k}}{\partial\rho_{e}} & \frac{\partial\boldsymbol{H}_{e_{3}}^{k}}{\partial\rho_{e}} & \frac{\partial\boldsymbol{H}_{e_{4}}^{k}}{\partial\rho_{e}} & \frac{\partial\tilde{\boldsymbol{H}}_{e}^{k}}{\partial\rho_{e}} \end{array}\right] \\ &&\frac{\partial\tilde{\boldsymbol{H}}_{e}^{k}}{\partial\rho_{e}} = \boldsymbol{0} \end{array} $$
(70)

For an elastic step, \(\partial \boldsymbol {H}_{e_{r}}^{k} / \partial \rho _{e}\) is

$$ \frac{\partial\boldsymbol{H}_{e_{r}}^{k}}{\partial\rho_{e}} = \left[\begin{array}{cccc} -\frac{\partial{\hat{\boldsymbol{C}}^{e}}}{\partial{\rho_{e}}} \left( \hat{\boldsymbol{\varepsilon}}^{k}_{e_{r}} - \hat{\boldsymbol{\varepsilon}}^{p^{k}}_{e_{r}} \right) \\ \boldsymbol{0} \\ 0 \\ 0 \end{array}\right] $$
(71)

and for a plastic step it is

$$ \frac{\partial\boldsymbol{H}_{e_{r}}^{k}}{\partial\rho_{e}} = \left[\begin{array}{cccc} -\frac{\partial{\hat{\boldsymbol{C}}^{e}}}{\partial{\rho_{e}}} \left( \hat{\boldsymbol{\varepsilon}}^{k}_{e_{r}} - \hat{\boldsymbol{\varepsilon}}^{p^{k}}_{e_{r}} \right) \\ -{\Delta}\gamma^{k}_{e_{r}} \left( \frac{\partial\boldsymbol{P}}{\partial\rho_{e}} \hat{\boldsymbol{\sigma}}^{k}_{e_{r}} + \frac{\partial\boldsymbol{q}}{\partial\rho_{e}} \right) \\ -\sqrt{\frac{1}{6z^{k}_{e_{r}}}} \frac{\partial z^{k}_{e_{r}}}{\partial\rho_{e}} {\Delta} \gamma^{k}_{e_{r}} \\ \frac{1}{2} \hat{\boldsymbol{\sigma}}^{k^{T}}_{e_{r}} \frac{\partial\boldsymbol{P}}{\partial\rho_{e}} + \frac{\partial\boldsymbol{q}^{T}}{\partial\rho_{e}} \hat{\boldsymbol{\sigma}}^{k}_{e_{r}} - 2\zeta \frac{\partial\zeta}{\partial\rho_{e}} \end{array}\right] $$
(72)

The derivatives \(\partial {\hat {\boldsymbol {C}}^{e}} / \partial {\rho _{e}}\), P/ρe, q/ρe, \(\partial z^{k}_{e_{r}} / \partial \rho _{e}\) and ζ/ρe come from the material interpolation.

Appendix B: Sensitivity verification

The sensitivity analysis outlined in Section 3.3 is verified in this Appendix by comparing the values computed using the analytical adjoint method with those obtained numerically using the central difference method (CDM). The CDM calculates the sensitivity of a response function F with respect to a design variable xe as

$$ \frac{\partial F(\boldsymbol{x})}{\partial x_{e}} \approx \frac{F(\boldsymbol{x} + {\Delta}\boldsymbol{h}) - F(\boldsymbol{x} - {\Delta}\boldsymbol{h})} {2{\Delta} h} $$
(73)

where x is the full vector of design variables and Δh is a vector with entries of zero except at the index corresponding to xe, where the entry is the perturbation value Δh. The perturbation value for CDM used in the following verification study is Δh = 10− 5.

The example used for sensitivity verification is a square RUC domain discretized into a 10 × 10 mesh of plane strain Q1/E4 EAS elements with a thickness of 1. The discretized domain and element numbering are shown in Fig. 15a. The size of the domain is set to 1000 by 1000 and the elastic material parameters are E = 2500 and ν = 0.38. The hardening coefficient Kh = 125 and is normalized by the yield stress σy = 20.

Fig. 15
figure15

Square RUC domain with element numbering and density distribution

The density distribution used for this example is shown in Fig. 15b, where the density values within the square inclusion are ρin = 0.2 while the outer density values are ρout = 0.7. The penalization parameters p0 and p1 are set to 3 and 2.5, respectively. No filter is utilized so that in this case F/x = F/ρ. The RUC is subject to a simple shear deformation by prescribing the macroscopic strain tensor as \(\hat {\overline {\boldsymbol {\varepsilon }}} = \left [\begin {array}{llll} 0 \quad 0 \quad 0.2 \end {array}\right ]\). Two verification cases are run using this example. In case 1, the scaling factors χ and ω are set to 1 and the angle θ = 0, resulting in the von Mises model without rotation. In case 2, ω = 4, χ = 2 and θ = π/4. Figure 16 shows the sensitivity values calculated using the two cases of material parameters. From here it can be seen that the sensitivity values match closely, verifying the correct implementation of the path-dependent adjoint sensitivity analysis from Section 3.3.

Fig. 16
figure16

Comparison of sensitivity values obtained using the adjoint method and CDM with Δh = 10− 5 for two cases of Hoffman plasticity under simple shear

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Alberdi, R., Khandelwal, K. Design of periodic elastoplastic energy dissipating microstructures. Struct Multidisc Optim 59, 461–483 (2019). https://doi.org/10.1007/s00158-018-2076-2

Download citation

Keywords

  • Architectured microstructures
  • Nonlinear topology optimization
  • RVE-based multiscale models
  • Computational homogenization
  • Anisotropic plasticity