Research ArticleMATERIALS SCIENCE

Discretely assembled mechanical metamaterials

See allHide authors and affiliations

Science Advances  18 Nov 2020:
Vol. 6, no. 47, eabc9943
DOI: 10.1126/sciadv.abc9943
  • Fig. 1 Discrete mechanical metamaterial subsystem description and characterization.

    (A) A 3 × 3 × 3 lattice consists of 27 individual voxels. (B) Voxels consist of six individual faces. (C) Faces consist of beams and joints. (D) Experimental results for subsystem characterization, where we see that joints (rivets + nodes) are individually stiffer and stronger than voxels, which are governed by beam properties. (E) Subsystem testing setups. Photo credit: Benjamin Jenett, MIT.

  • Fig. 2 Four types of discretely assembled mechanical metamaterials.

    Left to right: Rigid, compliant, auxetic, and chiral. (A) As-molded face parts. (B) Single voxel, front view. (C) A 2 × 2 × 2 cube, front view. (D) Single voxel, oblique view. (E) A 2 × 2 × 2 oblique view. Scale bars, 10 mm (A), 25 mm (B and D), and 50 mm (C and E). Photo credit: Benjamin Jenett, MIT.

  • Fig. 3 Rigid mechanical metamaterial.

    (A) Characteristic unit cell voxel demonstrating beam buckling and positive transverse strain in response to compressive load. (B) Experimental test setup for n = 1 to 4, undeformed (left), and at initial beam failure (right). (C) Geometric parameters for simulations, where beam thickness t is a function of lattice pitch P. (D) Effective stiffness for reduced order beam model simulation and experimental results demonstrating asymptotic behavior approaching continuum value at increasing voxel count. (E) Reduced order beam model simulation results for rigid and compliant lattice of 10 × 10 × 10 cube. Observable are modulus-density scaling values being linear for rigid and near quadratic for compliant. Photo credit: Benjamin Jenett, MIT.

  • Fig. 4 Compliant mechanical metamaterial.

    (A) Characteristic unit cell voxel demonstrating flexure spring-beam deformation and small transverse strain in response to compressive load. (B) Experimental test setup for n = 1 to 4, undeformed (left), and at onset of nonlinearity (right). (C) Geometric parameters for simulations, where spring-beam amplitude a is a function of lattice pitch P. (D) Effective stiffness simulation and experimental results, which show near continuum value at low voxel count for all but the smallest spring-beam amplitude designs. (E) Simulation results for effective Poisson’s ratio for rigid and compliant lattice, with large spring-beam amplitudes having a value of near zero. Photo credit: Benjamin Jenett, MIT.

  • Fig. 5 Auxetic mechanical metamaterial.

    (A) Characteristic unit cell voxel demonstrating reentrant mechanism action resulting in negative transverse strain in response to compressive load. (B) Experimental test setup for n = 1 to 4, undeformed (left), and deformed to 0.2 strain (right), with measured points on side faces circled in red. (C) Reduced order beam model simulation results recreating experiments, with out-of-plane reentrant behavior highlighted. (D) Geometric parameters for simulations, where reentrant distance d is a function of lattice pitch P. (E) Effective Poisson’s ratio simulation and experimental results. (F) 3D contour plot demonstrating effect of boundary conditions resulting in near-zero Poisson’s ratio at edges. Photo credit: Benjamin Jenett, MIT.

  • Fig. 6 Chiral mechanical metamaterial.

    (A) Characteristic unit cell voxel demonstrating out-of-plane coordinated rotation in response to compressive load. (B) Simulation and experimental results for odd and even column cross sections in combination with design rules 1 and 2 (R1 and R2). (C) Two chiral part types allow internal frustration to be avoided, thus enabling scalable chiral architecture. (D) Design rules 1 (left) and 2 (right), which emerge from odd and even columns, respectively. (E) Experimental and reduced order beam model simulation results of n = 1, 2, and 3, showing total twist increases as column voxel width increases, but normalized twist per strain is lower for n = 2. Photo credit: Benjamin Jenett, MIT.

Supplementary Materials

Stay Connected to Science Advances

Navigate This Article