Pattern Switching in Soft Cellular Structures and

polymers Article

Pattern Switching in Soft Cellular Structures and Hydrogel-Elastomer Composite Materials under Compression Jianying Hu 1 , Yu Zhou 1 , Zishun Liu 1, * and Teng Yong Ng 2 1

2

*

International Center for Applied Mechanics, State Key Laboratory for Strength and Vibration of Mechanical Structure, Shaanxi Engineering Research Center of Nondestructive Testing and Structural Integrity Evaluation, Xi’an Jiaotong University, Xi’an 710049, China; [email protected] (J.H.); [email protected] (Y.Z.) School of Mechanical and Aerospace Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Singapore; [email protected] Correspondence: [email protected]; Tel.: +86-029-8266-4354

Academic Editors: Rui Xiao, Jinlian Hu and Chih-Feng Huang Received: 8 May 2017; Accepted: 13 June 2017; Published: 16 June 2017

Abstract: It is well known that elastic instabilities induce pattern transformations when a soft cellular structure is compressed beyond critical limits. The nonlinear phenomena of pattern transformations make them a prime candidate for controlling macroscopic or microscopic deformation and auxetic properties of the material. In this present work, the novel mechanical properties of soft cellular structures and related hydrogel–elastomer composites are examined through experimental investigation and numerical simulations. We provide two reliable approaches for fabricating hydrogel–elastomer composites with rationally designed properties and transformed patterns, and demonstrate that different geometries of the repeat unit voids of the periodic pattern can be used to influence the global characteristics of the soft composite material. The experimental and numerical results indicate that the transformation event is dependent on the boundary conditions and material properties of matrix material for soft cellular structures; meanwhile, the deformation-triggered pattern of matrix material affects the pattern switching and mechanical properties of the hydrogel–elastomer material, thus providing future perspectives for optimal design, or serving as a fabrication suggestion of the new hydrogel–elastomer composite material. Keywords: soft periodic structures; composite hydrogel–elastomer materials; pattern switching; mechanical properties

1. Introduction Soft porous materials have been widely investigated for their remarkable mechanical properties, including isotropic foam [1–3] and loading-bearing cancellous bones of human shins [4]. When loaded beyond critical value, a small region or the whole structure may collapse as a result of elastic instabilities, thus leading to pattern switching and a negative Poisson’s ratio. Interest in such auxetic materials lies in the fact that a negative Poisson’s ratio may significantly improve many of the effective mechanical properties of a material (for example, the flexural rigidity and plane strain fracture toughness [5–7]). The research of these porous material ranges from the natural world to synthetic systems. Besides single crystals of arsenic and cadium in nature, it is found that several types of skin do show auxetic behavior, including catskin, cow teat skin, and salamander skin [8]. Shan et al. perforated a sheet with elongated cuts (arranged to form a periodic pattern with either six-fold or three-fold symmetry), to realize the 2D auxetic materials [9]. Man-made auxetic framework inspired by cubic

Polymers 2017, 9, 229; doi:10.3390/polym9060229

www.mdpi.com/journal/polymers

Polymers 2017, 9, 229

2 of 13

crystals has been tailored and designed with specific Poisson’s ratios, tensile, and shear moduli [10]. Anti-tetrachiral anisotropic lattice structures with negative Poisson’s ratios have been investigated through theoretical, numerical, and experimental approaches, for their in-plane and out-of-plane elastic behaviors (which may find applications in aerospace sandwich panel cores [11]). Subsequent research models were developed for central nodules, which were connected to eight others via corner fibrils [12] or 3-, 4-, and 6-connected chiral and anti-chiral honeycombs [13], and so on. Recently, the design of the complex microstructures in 2D lattice or 3D periodic cube structures has posed exciting challenges; increasing number of research groups are now focusing on this topic. Meanwhile, advances in fabrication technologies are enabling the realization of such special functional structures at nano- and micro-scales through the microfabrication processes, interference lithography, and 3D printing technology, where the ordered or irregular structures can easily be produced. 3D auxetic lattice via Bucklicrystal were designed and fabricated through assembling the 6-hole, 12-hole, or 24-hole units in the cubic crystal systems of the body center cubic (BCC), the solid center (SC), or the face center cubic (FCC) [14,15]. An elastic cube consisting of two sets of cylindrical voids which permeate in orthogonal directions has also been reported [16]. Exploiting the phase-transforming and switchable properties of soft materials under pressure, soft machines for actuation have been developed [17,18]. In this paper, we investigate the possibility of exploiting deformation characteristics to trigger dramatic homogeneous pattern transformations in certain classes of simple periodic elastomeric structures. Similar soft cellular structures were first studied by Bertoldi and Mullin for their mechanical properties through theory, simulations, and experiments [19–22]. There is very limited literature on the relationship between pattern transformation and material properties of the matrix material. He et al. [23] discovered that the transformation phenomenon for the shape memory polymer (SMP) can be triggered even by the stress relaxation process, through numerical simulation. In addition, Jang et al. [24] and Li et al. [25] combined pattern instability and shape-memory hysteresis of these structural materials for photonic switching. Apart from the research on the pattern switching of the cellular solids, another major target of this work is the hydrogel–elastomer composite material by assembling a pre-shaped cellular elastomers pattern and hydrogels together. Research on gel composites have focused on the deformation of the multilayer gel structures [26] or the fiber–reinforced hydrogels [27], toward revealing the physical mechanisms behind the designed structures. Further, surface pattern or pattern transformations can endow solid materials with some special optical, electronic, and acoustic properties due to the effects of wave interference [28–30]. In the present work, the hydrogel–elastomer composite material provides interesting pattern transformation and novel mechanical properties. The comparisons between the experimental and numerical results clearly show the mechanism of the pattern switch for each skeleton microstructure to be a form of local elastic instability, enabling reversible and repeatable transformation events. 2. Pattern Switching in Soft Cellular Solids In order to address the issue of pattern switching in soft cellular solids, we have manufactured and performed experiments on two ordered arrays of cellular specimens which possess the geometry shown in Figure 1. Here, we explore its mechanical behavior under uniaxial compression through both experimental and numerical approaches. The size of the specimens (as shown in Figure 1a) is 100 mm × 100 mm, comprising the microstructure of a 10 × 10 mm2 square (arrays of circular voids of 8.67 mm diameter with 10 mm center-to-center spacing); the size of the model in Figure 1b is 100 mm × 100 mm with 10 × 10 mm2 square (arrays of square voids of 7.7 mm length with 10 mm center-to-center spacing). The volume fractions of voids in the two specimens are the same at 59%.

Polymers 2017, 9, 229 Polymers 2017, 9, 229 Polymers 2017, 9, 228

3 of 13 3 of 13 3 of 13

(a) (a)

(b) (b)

Figure 1. 1. Full Full specimens specimens(the (thewhole wholemodel) model)and and periodic representative volume elements (units in Figure periodic representative volume elements (units in red Figure 1. line). Full specimens (the whole model) and periodic representative volume elements (units in red dash dash line). red dash line).

2.1. Materials for Soft Cellular Solids 2.1. Materials Materials for for Soft Soft Cellular Cellular Solids Solids 2.1. 2.1.1. TangoBlackPlus®®-3D Printing Material for the Cellular Structure 2.1.1. TangoBlackPlus TangoBlackPlus®-3D -3D Printing Printing Material Material for for the the Cellular Cellular Structure Structure 2.1.1. The first material used for the cellular structure was a 3D-printing material called ® The first usedused for thefor cellular was a 3D-printing called TangoBlackPlus The firstmaterial material the structure cellular structure was amaterial 3D-printing material called ® (TBP for ® Inc., Eden TangoBlackPlus short, Stratasys Prairie, MN, USA), for which the Young’s ® Inc., Eden Prairie, MN, USA), for which the Young’s modulus was 0.5 MPa (TBP for short, Stratasys ® ® TangoBlackPlus Stratasys Prairie, USA), which the Young’s modulus was 0.5 (TBP MPa for andshort, tensile strength Inc., was Eden 1.2 MPa [31].MN, These two for properties of this base and tensile strength was 1.2 MPa [31]. These two properties of this base two material was determined by modulus was 0.5 MPa and tensile strength was 1.2 MPa [31]. These properties of this base material was determined by dynamic mechanical analysis (DMA). Meanwhile, the same mechanical dynamic mechanical analysis (DMA). Meanwhile, the same mechanical behavior of this rubber-like material determined by dynamic mechanical analysisthrough (DMA).the Meanwhile, the same mechanical behaviorwas of this rubber-like material was characterized dog-bone-shape samples after material was characterized through the dog-bone-shape samples after they had been stretched inafter the behavior of this rubber-like material was characterized through the dog-bone-shape samples they had been stretched in the uniaxial tension mode in a SHIMADZU (AGS-X1kN + 250 mm) uniaxial tension mode in a SHIMADZU (AGS-X1kN + 250 mm) testing machine (Shimadzu Corp., they had been stretched in Corp., the uniaxial in a SHIMADZU + 250 mm) testing machine (Shimadzu Kyoto,tension Japan). mode A constant strain rate of (AGS-X1kN 5 mm/min was applied Kyoto, Japan). A(Shimadzu constant strain rate of 5 mm/min was applied during loading. Dog-bone-shaped testing machine Corp., Kyoto, Japan). A constant strain rate of 5 mm/min was applied during loading. Dog-bone-shaped samples (refer to Figure 2)—which follow the ASTM/D638 samplesloading. (refer to Dog-bone-shaped Figure 2)—which follow the(refer ASTM/D638 standard (typefollow III)—were prepared by during samples to Figure 2)—which ASTM/D638 standard (type III)—were prepared by 3D-printing technology. For simplicity, the the strain and stress 3D-printing technology. For simplicity, the strain and stress discussed here are the nominal strain standard prepared byand 3D-printing technology. For simplicity, the strain and can stress discussed(type here III)—were are the nominal strain stress. Thus, the elastomeric stress-strain behavior be and stress. Thus, the elastomeric stress-strain behavior can be modeled as a linear elastic model discussed here are the nominal strain and stress. Thus, the elastomeric stress-strain behavior can be modeled as a linear elastic model with Young’s modulus of 592.92 kPa and Poisson’s ratio of 0.44, with Young’s modulus of 592.92 kPa and Poisson’s ratio ofof0.44, the values ofPoisson’s which weratio will of use for modeled linear elastic with Young’s modulus 592.92 0.44, the valuesasofa which we willmodel use for modeling and simulation in thekPa nextand Section. Figure 3 shows a modeling and simulation in the next Section. Figure 3 shows a comparison between the finite element the values ofbetween which we use for modeling simulation in the Section. Figure shows a comparison thewill finite element methodand (FEM) results and thenext experimental output3data. method (FEM) resultsthe andfinite the experimental output data. comparison between element method (FEM) results and the experimental output data.

(a) (a)

(b) (b)

Figure 2. (a) Specimen dimensions (in mm; thickness is 3 mm) in the ASTM D638 standard, type III; Figure (a) thickness (b) the 2. dog-bone-shaped tensile bar(in formm; tensile testing.is Figure 2. (a) Specimen Specimen dimensions dimensions (in mm; thickness is 33 mm) mm) in in the the ASTM ASTM D638 D638 standard, standard, type type III; III; (b) the dog-bone-shaped tensile bar for tensile testing. (b) the dog-bone-shaped tensile bar for tensile testing.

Polymers 2017, 9, 229 Polymers Polymers2017, 2017,9,9,228 228

4 of 13 4 4ofof1313

Figure The stress-strain curve forfor dog-bone-shaped tensile bar—FEM (finite(finite element method) results 3.3. The stress-strain curve tensile bar—FEM method) Figure3. The stress-strain curve fordog-bone-shaped dog-bone-shaped tensile bar—FEM (finiteelement element method) (open blue symbols), and experimental results (black solid line). results (open blue and results (black solid results (open bluesymbols), symbols), andexperimental experimental results (black solidline). line).

2.1.2. 2.1.2.PDMS PDMSfor forthe theCellular CellularStructure Structure

also used PDMS (polydimethylsiloxane) asasthe the matrix material. We Wealso alsoused usedPDMS PDMS(polydimethylsiloxane) (polydimethylsiloxane)as thematrix matrixmaterial. material.First, First,we weemployed employed3D 3D printing technology to build a PLA (polylactic acid) mold and inject PDMS to fabricate the cellular printing technology to build a PLA (polylactic acid) mold and inject PDMS to fabricate the cellular 4a). structures The PDMS system ofofchoice structures(as (asshown shownininFigure Figure4a). 4a).The ThePDMS PDMSsystem systemof choicefor forthe thepresent presentprocess processisisSylgard Sylgard (DOW Corning, which is applied in a 25.1:1 ratio (w/w) 184 (DOW Corning, Midland, MI, USA), which is applied in a 25.1:1 ratio (w/w) of siloxane 184 (DOW Corning, Midland, MI, USA), which is applied in a 25.1:1 ratio (w/w) of siloxaneand and curing step atat70 7070◦°C three hours. C°Cfor curingagent, agent,and andrequires requiresa athermal thermalcuring curingstep stepat forthree threehours. hours.The Themechanical mechanicalbehaviors behaviors characterized strip-shape samples (as shown ininFigure Figure 4b). for forPDMS PDMSwere werecharacterized characterizedthrough throughthe thestrip-shape strip-shapesamples samples(as (asshown shownin Figure4b). 4b).Once Oncethe the Young’s modulus of the PDMS is obtained through uniaxial tension testing, the elastomeric Young’s modulus of the is obtained through through uniaxial tension testing, the elastomeric Young’s modulus of PDMS the PDMS is obtained uniaxial tension testing, thestress-strain elastomeric stress-strain can modeled a acompressible Neo-Hooken model with the behavior can behavior be modeled as be abe compressible Neo-Hooken model with the initial Young’s modulus of stress-strain behavior can modeledasas compressible Neo-Hooken model with theinitial initial Young’s modulus ofof0.2 MPa 0.2 MPa and Poisson’s ratio ofand 0.48. Young’s modulus 0.2 MPa andPoisson’s Poisson’sratio ratioofof0.48. 0.48.

(a) (a)

(b) (b)

Figure Figure4.4.Schematic Schematicdiagram diagramofofthe thefabrication fabricationprocess processofofthe thePDMS PDMS(polydimethylsiloxane) (polydimethylsiloxane)cellular cellular Figure 4. Schematic diagram of the fabrication process of the PDMS (polydimethylsiloxane) structures. structures. cellular structures.

2.2. 2.2.Experiments Experimentsand andModeling Modelingfor forSoft SoftCellular CellularSolids Solids 2.2. Experiments and Modeling for Soft Cellular Solids 2.2.1. 2.2.1.Experimental ExperimentalProtocol Protocol 2.2.1. Experimental Protocol Following construction the soft cellular Figure 1), we then used two Followingthe theconstruction constructionof thesoft softcellular cellularsolids solids(shown (shownin Figure1), 1),we wethen thenused usedtwo two Following the ofofthe solids (shown ininFigure different rubber-like materials totofabricate the continuum version ofofthe porous periodic lattice different rubber-like materials fabricate the continuum version the porous periodic lattice different rubber-like materials to fabricate the continuum version of the porous periodic lattice structures. quantitative experimental results, we devised the test system for structures.In Inorder ordertoto toachieve achieve test system forthe the structures. In order achieve quantitative quantitativeexperimental experimentalresults, results,we wedevised devisedthe the test system for soft material. The testing device can bebedivided into four parts: Main Testing Platform with Ball soft material. The testing device can divided into four parts: Main Testing Platform with Ball the soft material. The testing device can be divided into four parts: Main Testing Platform with Ball Screw Linear Actuators; Force Sensor; Integrated Circuit; Power Supply (as shown inin ScrewLinear LinearActuators; Actuators;Force ForceSensor; Sensor;Integrated IntegratedControl ControlCircuit; Circuit;and andPower PowerSupply Supply(as (asshown shownin Screw Control and Figure 5). Each specimen was subjected to uniaxial compression in the plane of the sheet, at Figure5).5). Each specimen subjected to uniaxial compression in theofplane of the ata a Figure Each specimen waswas subjected to uniaxial compression in the plane the sheet, at asheet, constant constant nominal strain rate using screw-driven testing The constantstrain nominal rateofof1 1mm/min mm/min usinga ahandmade handmade screw-driven testingmachine. machine. The nominal rate strain of 1 mm/min using a handmade screw-driven testing machine. The nominal nominal stress vs. nominal strain behavior was recorded, based on computed data from the nominal stress vs. nominal strain recorded, based data on computed data from the stress vs. nominal strain behavior wasbehavior recorded,was based on computed from the measured force measured force and displacement. measured force and displacement. and displacement.

Polymers 2017, 9, 229

5 of 13

Polymers 2017, 9, 228

5 of 13

Polymers 2017, 9, 228

5 of 13

Figure 5. Component diagram for the handmade screw-driven testing machine.

Figure 5. Component diagram for the handmade screw-driven testing machine.

2.2.2. Modeling

2.2.2. Modeling

Figure 5. Component diagram for the handmade screw-driven testing machine.

Using the commercial software ABAQUS (Dassault Systèmes Simulia Corp., Providence, RI, USA), the commercial cellular structures wereABAQUS simulated(Dassault through two approaches: first, by considering theRI, fullUSA), Using software Systèmes Simulia Corp., Providence, 2.2.2. Modeling specimen of dimension 100 mm through × 100 mm capturingfirst, the by geometry differences the the cellular structures were simulated twothus approaches: considering the fullatspecimen Using the commercial software ABAQUS (Dassault Systèmes Simulia Corp., Providence, RI,on boundaries as well as the displacement applied on top surfaces as force, while the displacement of dimension 100 mm × 100 mm thus capturing the geometry differences at the boundaries as well USA), the cellular structures were simulated through approaches: by considering the full a bottom surface was on constrained in the same two direction; and first, secondly, bythe considering as thethe displacement applied top surfaces as force, while the displacement on bottom surface specimen of dimension 100 mm × of 100themm thuswith capturing the geometry differencesconditions at the representative volume element (RVE) domain appropriate periodic boundary was constrained in well the same and secondly, by considering a representative volumeon element boundaries as asinternal the direction; displacement ontotop as force, while the displacement and containing small defects in applied the mesh aidsurfaces the initiation of instabilities. (RVE)the of the domain with appropriate periodic boundary conditions and containing small internal bottom surface was constrained in the same direction; and secondly, by were considering a Numerical simulations of the deformation of the two periodic structures conducted defects in the nonlinear meshvolume to aid the initiation of the instabilities. representative element (RVE) domain with appropriate boundary conditions utilizing finite element analysis (we represented the solids periodic with C3D8 elements). For TBP and containing small internal defects inbehavior the mesh to aid theperiodic initiation of instabilities. Numerical simulations ofstress-strain the deformation of the two were conducted utilizing material, the elastomeric was modeled as astructures linear elastic one with Young’s Numerical simulations of the deformation of the two periodic structures were conducted nonlinear finite analysis (we represented the (the solids with C3D8 elements). For TBPofmaterial, modulus ofelement 592.92 kPa and Poisson’s ratio of 0.44 elastomeric stress-strain behavior the utilizing nonlinear element analysis (we represented the solids withone C3D8 elements). TBP PDMS material canfinite be modeled as was a compressible model, with the initialFor Young’s the elastomeric stress-strain behavior modeled asNeo-Hooken a linear elastic with Young’s modulus of material, the elastomeric stress-strain behavior was modeled as nominal a linear elastic one with Young’s modulus of 0.2 MPa and Poisson’s ratio of 0.48). To obtain stress vs. nominal strain 592.92 kPa and Poisson’s ratio of 0.44 (the elastomeric stress-strain behavior of the PDMS material can modulus of 592.92 kPa and Poisson’s ratio of 0.44 (the elastomeric stress-strain behavior of the behavior, total force onNeo-Hooken load edge (top surface) was as a function of the0.2 applied be modeled as the a compressible model, with themonitored initial Young’s MPa and PDMS material can be modeled as a compressible Neo-Hooken model, withmodulus the initialofYoung’s displacement. Poisson’s ratioofof0.2 0.48). obtain nominal vs. nominal behavior, thenominal total force on load modulus MPaToand Poisson’s ratiostress of 0.48). To obtainstrain nominal stress vs. strain edge (top surface) was monitored as edge a function of the applied displacement. behavior, the total force on load (top surface) was monitored as a function of the applied 2.3. Results and Discussion for Soft Cellular Solids displacement. Experimental andfor numerical results for the nominal stress vs. nominal strain behavior of 2.3. Results and Discussion Soft Cellular Solids specimens two types offor materials (up to a compressive strain of 0.1) are shown in Figures 6 and 7. 2.3. Resultsof and Discussion Soft Cellular Solids Experimental andprovide numerical results for the nominal stress response vs. nominal strain behavior The results shown a typical characteristic of a stress-strain and observed pattern of Experimental and numerical (up results for the nominalstrain stressofvs. nominal straininbehavior of specimens of two types of materials to a compressive 0.1) are shown Figures 6 and 7. transformation for TBP and PDMS, respectively. specimens of two types of materials (up to a compressive strain of 0.1) are shown in Figures 6 and 7. The results shown provide a typical characteristic of a stress-strain response and observed pattern The results shown provide a typical characteristic of a stress-strain response and observed pattern transformation for TBP and PDMS, respectively. transformation for TBP and PDMS, respectively.

(a)

(b)

Figure 6. The mechanical behavior for TangoBlackPlus cellular structures is as follows: (a) Nominal (a)strain curves (red solid line with open symbols and(b) stress vs. nominal black solid line with open ® cellular symbols, respectively, represent with square voids and circular voids. dot (a) dash dot line Figure 6. The mechanical behaviorTBP for TangoBlackPlus structures is as Red follows: Nominal ® cellular structures Figure 6. red Thedot mechanical behavior for TangoBlackPlus is aswith follows: (a)voids; Nominal and line depict full model andline RVEwith simulations for theand TBP model stress vs. nominal strainthe curves (red solid open symbols black solid linesquare with open stressblack vs. nominal strain curves (red solid line with open symbols and black solid line with solid line and black dash line the fullvoids model and RVE simulations for the symbols, respectively, represent TBP show with square and circular voids. Red dot dashTBP dotmodel line open symbols, respectively, represent TBP with square voids and circular voids. Red dot dash dot line and with circular voids); (b) Experimental images of structures at a macroscopic strain of 0.1. and red dot line depict the full model and RVE simulations for the TBP model with square voids; ®

red dot linesolid depict model RVE simulations the TBP model with square black linethe andfull black dash and line show the full modelfor and RVE simulations for the TBP voids; model black withand circular voids); Experimental images of structures a macroscopic of 0.1. solid line black dash(b) line show the full model and RVE at simulations for strain the TBP model with circular voids); (b) Experimental images of structures at a macroscopic strain of 0.1.

Polymers 2017, 9, 229 Polymers 2017, 9, 228

6 of 13 6 of 13

(a)

(b)

Figure 7. The Themechanical mechanicalbehavior behavior PDMS cellular structures: (a) Nominal vs. nominal Figure 7. forfor PDMS cellular structures: (a) Nominal stressstress vs. nominal strain strain curves (blue solid line with open symbols and black solid line with open symbols represent curves (blue solid line with open symbols and black solid line with open symbols represent PDMS PDMS structures withvoids square voids andvoids. circular voids. Blue dot dot line dashand dotblue linedot andline blue dot cellularcellular structures with square and circular Blue dot dash depict line depict the full model and RVE (representative volume element) simulations for the PDMS the full model and RVE (representative volume element) simulations for the PDMS model with square model with solid square blackdash solidline line and black line still the full model andPDMS RVE holes; black lineholes; and black still show thedash full model andshow RVE simulations for the simulations for the PDMS model with circular voids); (b) Experimental images of structures a model with circular voids); (b) Experimental images of structures at a macroscopic strain of 0.1 andat 0.4. macroscopic strain of 0.1 and 0.4.

It can be seen in Figures 6a and 7a that a plateau occurs in the stress-strain curve when the strains It can be seen in Figures 6a and 7a that a plateau occurs in the stress-strain curve when the reach above εc , i.e., the total stress will become independent of strain after the critical strain εc From strains reach above εc, i.e., the total stress will become independent of strain after the critical strain the stress vs. strain curves, it can be observed that regardless of the kind of material property, the εc From the stress vs. strain curves, it can be observed that regardless of the kind of material stress of the cellular material is approximately proportional to the strain for small displacements, and property, the stress of the cellular material is approximately proportional to the strain for small compression within the material manifests as foreshortening of interstitial vertical ligaments in the displacements, and compression within the material manifests as foreshortening of interstitial structure. The departures from linearity are the result of elastic instabilities in the microstructures that vertical ligaments in the structure. The departures from linearity are the result of elastic instabilities trigger the pattern transformations. in the microstructures that trigger the pattern transformations. Interestingly, after the transformation to the new configuration, various patterns show different Interestingly, after the transformation to the new configuration, various patterns show mechanical properties in their cellular structures. In the structure with ordered circular voids the mode different mechanical properties in their cellular structures. In the structure with ordered circular of deformation is characterized by a critical eigenmode consisting of mutually orthogonal ellipses. voids the mode of deformation is characterized by a critical eigenmode consisting of mutually The simulations for RVE and full specimen are both consistent with the experiments. As shown in orthogonal ellipses. The simulations for RVE and full specimen are both consistent with the Figures 6a and 7a, if switching the pattern of the RVE model is the same as the full model calculation, experiments. As shown in Figures 6a and 7a, if switching the pattern of the RVE model is the same it is clearly observed that the RVE results exhibit an earlier departure from linearity than the full as the full model calculation, it is clearly observed that the RVE results exhibit an earlier departure model calculation, since the former does not capture the influence of the boundary conditions on the from linearity than the full model calculation, since the former does not capture the influence of the initiation of the instability. In the structure with ordered square voids, the phenomenon is totally boundary conditions on the initiation of the instability. In the structure with ordered square voids, different in that much of the macroscopic deformation is observed to be accommodated by the the phenomenon is totally different in that much of the macroscopic deformation is observed to be critical buckling of the side struts. For the elastic range in Figure 6, the images for RVE and the full accommodated by the critical buckling of the side struts. For the elastic range in Figure 6, the specimen simulations match well with the experimental results. Only the experimental result of images for RVE and the full specimen simulations match well with the experimental results. Only the TBP specimen with circular voids does not match well with the full model and RVE simulation the experimental result of the TBP specimen with circular voids does not match well with the full results; the reason for this discrepancy is that compression testing leads to the fracture of the inter-hole model and RVE simulation results; the reason for this discrepancy is that compression testing leads ligaments in the matrix material. However, for the hyperelastic range in Figure 7, the RVE results to the fracture of the inter-hole ligaments in the matrix material. However, for the hyperelastic capture a transformation which is a totally new and different pattern (triggered by the rotation of range in Figure 7, the RVE results capture a transformation which is a totally new and different the matrix domains diagonally bridging neighboring hole), which thus results in the maximum pattern (triggered by the rotation of the matrix domains diagonally bridging neighboring hole), critical force to pattern switching. Hence, the experimental and numerical results indicate that the which thus results in the maximum critical force to pattern switching. Hence, the experimental and transformation event is highly dependent on the boundary edge conditions and material properties of numerical results indicate that the transformation event is highly dependent on the boundary edge the matrix material. conditions and material properties of the matrix material. 3. Novel Mechanical Behavior of the Composite Gel Material under Compression 3. Novel Mechanical Behavior of the Composite Gel Material under Compression The composite material that is composed of two components (hydrogel and elastomer) has been The composite material that is composed of two components (hydrogel and elastomer) has proposed in our previous work [32]. In the earlier study, we can easily understand that if the holes been proposed in our previous work [32]. In the earlier study, we can easily understand that if the holes are filled with a more rigid material than that of the matrix, the cellular structure will not lead to any pattern switch [33]. Therefore, a much softer material needs to be placed in the holes to

Polymers 2017, 9, 229

7 of 13

are filled with a Polymers 2017, 9, 228

more rigid material than that of the matrix, the cellular structure will not lead to 7 ofany 13 pattern switch [33]. Therefore, a much softer material needs to be placed in the holes to achieve the achieve pattern for amaterial. composite material. et al. [34] that discovered that the inclusion pattern the switch for aswitch composite Mullin et al.Mullin [34] discovered the inclusion which has a which has a Young’s modulus of1% greater of the bulk would suppress theBased pattern Young’s modulus of greater than of thethan bulk1% would suppress the pattern switch. onswitch. present Based on present assumptions finite(FEA), element analysis here (FEA), we two propose to use two kinds assumptions and finite element and analysis we propose to use kindshere of hydrogel-elastomer of hydrogel-elastomer composite material which experimentally exhibit the pattern switch in composite material which experimentally exhibit the pattern switch in composite gel material. In the composite gel material. In the the modeling the experiment, the materials specific combinations two8. modeling and the experiment, specificand combinations of the two are illustratedof in the Figure materials are illustrated in Figure 8. The specimen shown in Figure 8a is a 100 mm × 100 mm square The specimen shown in Figure 8a is a 100 mm × 100 mm square model of circular gel inclusions of model of circular inclusions of 8.67 mm diameter with 10 mm center-to-center spacing in both 8.67 mm diametergel with 10 mm center-to-center spacing in both vertical and horizontal distributions; 2 vertical and horizontal themm specimen shown in Figure mm × 100 arrays mm model the specimen shown indistributions; Figure 8b is 100 × 100 mm model with 108b×is10100 mm square of gel 2 square arrays of gel inclusions of 7.7 mm length. The volume fractions of gel in with 10 × 10 mm inclusions of 7.7 mm length. The volume fractions of gel in the two types of composite materials have the types of composite materials have the same percentage of 59%. thetwo same percentage of 59%.

(a)

(b)

Figure Figure 8.8. Schematic Schematicillustration illustrationfor forfull fullspecimens specimens(the (thewhole wholemodel) model)and andperiodic periodic representative representative volume elements (units in red dash line) for hydrogel–elastomer composites. volume elements (units in red dash line) for hydrogel–elastomer composites.

3.1. Material Characterization of Polyacrylamide Hydrogel 3.1. Material Characterization of Polyacrylamide Hydrogel In the composite material, two different materials are used for the matrix, i.e., the elastomer for In the composite material, two different materials are used for the matrix, i.e., the elastomer for the matrix and the hydrogel for inclusions. We have already discussed the material properties of the matrix and the hydrogel for inclusions. We have already discussed the material properties of matrix material in Section 2.1. The inclusion that adheres with the matrix material is matrix material in Section 2.1. The inclusion that adheres with the matrix material is polyacrylamide polyacrylamide hydrogel. For an ideal elastomeric gel, the effect of mixing the polymer and the hydrogel. For an ideal elastomeric gel, the effect of mixing the polymer and the solvent is solvent is represented by the osmotic pressure as a function of the swelling ratio J . We prepared the represented by the osmotic pressure as a function of the swelling ratio J. We prepared the polyacrylamide hydrogels using two different procedures. For the first one (in order to obtain polyacrylamide hydrogels using two different procedures. For the first one (in order to obtain robust robust hydrogel–PDMS interfaces), acrylamide was dissolved in distilled water to 1.9 M hydrogel–PDMS interfaces), acrylamide was dissolved in distilled water to 1.9 M concentration, along concentration, along with N,N-methylenebisacrylamide (MBAA, 0.001× weight of acrylamide) as a with N,N-methylenebisacrylamide (MBAA, 0.001× weight of acrylamide) as a crosslinker (we used crosslinker (weacid used α-×Ketoglutaric acid (0.041× weight ofpolymerization acrylamide) as light-induced α-Ketoglutaric (0.041 weight of acrylamide) as light-induced initiator). The gels polymerization initiator). The gels were cured for 2 h under UV irradiation. Using the other were cured for 2 h under UV irradiation. Using the other approach (to obtain weak hydrogel–TBP approach (to obtain weak hydrogel–TBP interfaces), the same concentration of acrylamide and interfaces), the same concentration of acrylamide and MBAA-to-acrylamide ratio was used to form MBAA-to-acrylamide ratio was used to form gels with ammonium persulfate (0.0079× weight of gels with ammonium persulfate (0.0079× weight of acrylamide) as polymerization initiators, and acrylamide) as polymerization initiators, N,N,N′,N′-tetramethylethylenediamine (0.0025× N,N,N 0 ,N 0 -tetramethylethylenediamine (0.0025and × weight of acrylamide) as a crosslinker. The gels were weight of acrylamide) as a crosslinker. The gels were cured for 2 h at a temperature of 60 °C. ◦ cured for 2 h at a temperature of 60 C. ItItisisknown determined stress-stretch curves of polyacrylamide gelsgels fit knownthat thatthe theexperimentally experimentally determined stress-stretch curves of polyacrylamide the Gaussian-chain model well. TheThe freefree energy function forfor thethe hydrogel fit the Gaussian-chain model well. energy function hydrogeltakes takesthe theform formofofthe the Flory–Huggins model by Hong et al. [35], Flory–Huggins model by Hong et al. [35],

1 kT  J   W 1 NkT (I  3  2log J )  kT J 1 log J  χ W = NkT 2 ( I − 3 − 2 log J ) −   ( J − 1) log J 1 J + 2 ν J−1  J

(1) (1)

where N is the number of polymeric chains per reference volume, χ the dimensionless parameter of the enthalpy of mixing, and kT the temperature in the unit of energy [36]. A representative value of the volume per molecule is ν = 10−28 m3, so kT/ν = 4 × 107 Pa. Thus, a gel is fully characterized by a scalar NkT, and the isotropic stretch λ0 = J01/3 for the initial free-swelling state. According to Li’s experimental results [37], the two dimensionless material properties of freely swollen gels (Nν and

Polymers 2017, 9, 229

8 of 13

where N is the number of polymeric chains per reference volume, χ the dimensionless parameter of the enthalpy of mixing, and kT the temperature in the unit of energy [36]. A representative value of the volume per Polymers 2017, 9, 228molecule is ν = 10−28 m3 , so kT/ν = 4 × 107 Pa. Thus, a gel is fully characterized 8 ofby 13 Polymers 9, 228 8 of 13 a scalar2017, NkT, and the isotropic stretch λ0 = J0 1/3 for the initial free-swelling state. According to Li’s χ) can be appropriately the numerical simulation. Here of wefreely adoptswollen the values Nν =and 0.001 experimental results [37],chosen the twofor dimensionless material properties gels (Nν χ) χ) can be appropriately chosen for the numerical simulation. Here we adopt the values Nν = 0.001 and χ = 0.5. can be appropriately chosen for the numerical simulation. Here we adopt the values Nν = 0.001 and and χ = 0.5. χ = 0.5. 3.2. Weak or Robust Interface of the Hydrogel–Elastomer Composite Material 3.2. 3.2. Weak Weak or or Robust Robust Interface Interface of of the the Hydrogel–Elastomer Composite Material In Section 3.1, we have mentioned two approaches to assemble pre-shaped patterned cellular In we have mentioned two approaches to assemble pre-shaped patterned cellular In Section Section 3.1, mentioned approaches pre-shaped patterned cellular elastomers and3.1, hydrogels together for the fabrication of hydrogel–elastomer composite material, as elastomers and together of of hydrogel–elastomer composite material, as elastomers andhydrogels hydrogels togetherfor forthe thefabrication fabrication hydrogel–elastomer composite material, shown in Figure 9. shown in Figure 9. 9. as shown in Figure

Figure 9. A schematic illustration of the fabrication of robust micro-structured hydrogel–elastomer Figure 9. A schematic of of robust Figure Amaterial: schematic illustration of the fabrication fabrication micro-structured hydrogel–elastomer composite (a)illustration The pre-shaped hydrogel and elastomer are assembledhydrogel–elastomer together; (b) After composite material: (a) The pre-shaped hydrogel and elastomer are assembled together; composite material: The pre-shaped hydrogel and elastomer assembled together; (b) (b) After After ultraviolet or thermal(a) irradiation, the resultant hydrogel compositeare material forms. ultraviolet ultraviolet or or thermal irradiation, the resultant hydrogel composite material forms.

To obtain robust interfaces for composite hydrogel–PDMS solids, we adopt the simple yet To obtain robust interfaces for composite hydrogel–PDMS solids, weaddress adopt the simple yet versatile method proposed byfor the Zhao group [38]. The key point to PDMS’ oxygen To obtain robust interfaces composite hydrogel–PDMS solids, weisadopt the simple yet versatile versatile method proposed by the Zhao group [38]. The key point is to address PDMS’ oxygen inhibition. We spray solution diphenyl ketone onaddress the elastomer use an method proposed by the the acetone Zhao group [38].ofThe key point is to PDMS’surface oxygenand inhibition. inhibition. Weacetone spray the acetone solution of diphenyl ketone onof the andsolution useThe an ethanol to wash it several times ketone for surface absorption theelastomer benzophenone solution. We spraysolution the solution of diphenyl on the elastomer surface and usesurface an ethanol ethanol solution to wash it several times for surface absorption of the benzophenone solution. The benzophenone acts as for an surface ultraviolet-assisted agent for solution. covalent The crosslinking hydrogel to wash it several times absorption of grafting the benzophenone benzophenone acts benzophenone acts as an ultraviolet-assisted grafting agentmaterial for hydrogel covalent crosslinking hydrogel polymers on elastomer surfaces. The assembled composite is exposed under as an ultraviolet-assisted grafting agent for covalent crosslinking polymers on ultraviolet elastomer polymers on elastomer The assembled composite material is hydrogel–TangoBlackPlus exposed under ultraviolet irradiation together withsurfaces. the synthesis of polyacrylamide hydrogel. The surfaces. The assembled composite material is exposed under ultraviolet irradiation together with® ® ® irradiation together with the synthesis of polyacrylamide hydrogel. The hydrogel–TangoBlackPlus composite material can be simplified by filling the pre-shaped hydrogel into the porous material, the synthesis of polyacrylamide hydrogel. The hydrogel–TangoBlackPlus composite material can composite material can the be filling the pre-shaped hydrogel into the means porous material, which means thatfilling there is simplified no bondingbybetween the matrix material and inclusions. This that is mainly be simplified by pre-shaped hydrogel into the porous material, which there which means that there is no bonding between the matrix material and inclusions. This is mainly due to the elastomer surfaces’ inhibition of polymer crosslinking and grafting, as well as fluidic is no bonding between the matrix material and inclusions. This is mainly due to the elastomer due to theinhibition elastomer surfaces’ of block polymer and grafting, well of as fluidic characters of pre-gelofsolutions (which will the crosslinking interfacial connection ifcharacters noasspecial chemical surfaces’ polymerinhibition crosslinking and grafting, as well as fluidic pre-gel characters of pre-gel (which will block composite theifinterfacial connection if no are special chemical treatment(which is carried out). The hydrogel–elastic materials fabricated in solutions will solutions block the interfacial connection no special chemical treatment isillustrated carried out). treatment is carried out). The hydrogel–elastic composite materials fabricated are illustrated in Figure 10. The hydrogel–elastic composite materials fabricated are illustrated in Figure 10. Figure 10.

Figure 10. fabricated through ultraviolet (a,b) or 10. Specimens Specimens of ofhydrogel–elastomer hydrogel–elastomercomposite compositematerial material fabricated through ultraviolet (a,b) Figure 10. Specimens of hydrogel–elastomer composite material fabricated through ultraviolet (a,b) thermal (c,d) irradiation. or thermal (c,d) irradiation. or thermal (c,d) irradiation.

3.3. Results and Discussion of Composite Material 3.3. Results and Discussion of Composite Material When a periodic elastomeric cellular solid is compressed, the array of pores undergoes an When a periodic elastomeric cellular is compressed, the array of poresalso undergoes an unstable transformation at a critical pointsolid [39–42]. Similar instability processes trigger the unstable at a critical point composite [39–42]. Similar instability processes also trigger and the change totransformation new configurations in the novel gel materials [32,34]. The experimental

Polymers 2017, 9, 229

9 of 13

3.3. Results and Discussion of Composite Material When a periodic elastomeric cellular solid is compressed, the array of pores undergoes 9an Polymers 2017, 9, 228 of 13 Polymers 2017,transformation 9, 228 9 of 13 unstable at a critical point [39–42]. Similar instability processes also trigger the change to new configurations in the composite gel materials [32,34]. The experimental and numerical numerical results capture thenovel mechanical behavior of hydrogel–elastomer composite material, as numerical results the capture the mechanical behavior of hydrogel–elastomer composite material, as results capture mechanical behavior of hydrogel–elastomer composite material, as shown in shown in Figures 11 and 12. shown in Figures 11 and 12. Figures 11 and 12.

(a) (a)

(b) (b)

Figure 11. The mechanical behavior behavior for hydrogel–TangoBlackPlus hydrogel–TangoBlackPlus (a) Nominal stress ® ® composites: Figure 11.11. The stress Figure Themechanical mechanical behaviorfor for hydrogel–TangoBlackPlus® composites: composites: (a) (a) Nominal Nominal stress vs. nominal strain curves (red solid line with full symbols and black solid line with full symbols, vs.vs. nominal strain curves (red solid symbols, nominal strain curves (red solidline linewith withfull fullsymbols symbolsand andblack black solid solid line line with with full full symbols, respectively, represent hydrogel–TBP composites with square inclusions and circular inclusions. respectively, represent hydrogel–TBP composites with square inclusions and circular inclusions. respectively, represent hydrogel–TBP composites with square inclusions and circular inclusions. Red Red dot dash dot line red dot line depict the full model and RVE simulations for hydrogel-TBP Red dot dash and red dot line depict the full model and RVE simulations for hydrogel-TBP dot dash dot line and red dot line depict the full model and RVE simulations for hydrogel-TBP model model inclusions; black solid lineand andblack black dash line showthe thefull fullmodel model and RVE model with inclusions; black line dash line and RVE with with squaresquare inclusions; black solid linesolid and black dash line show the fullshow model and RVE simulations for simulations for model hydrogel-TBP model with(b) circular inclusions); Experimental imagesof of simulations hydrogel-TBP model with circular inclusions); Experimental images hydrogel-TBP with circular inclusions); Experimental images (b) of(b) specimens from experiments specimens from andat full modelsimulation simulation result macroscopic strain of 0.1, pattern specimens from experiments experiments and full model atatmacroscopic strain 0.1, pattern and full model simulation result macroscopic strain of result 0.1, pattern transformation of of RVE model at macroscopic strain of model 0.12. at transformation of RVE model at macroscopic macroscopicstrain strainofof0.12. 0.12. transformation

(a) (a)

(b) (b)

Figure 12. The mechanical behavior for hydrogel–PDMS composites: (a) Nominal stress vs. nominal Figure 12.12. The mechanical composites:(a) (a)Nominal Nominal stress nominal Figure The mechanicalbehavior behaviorfor for hydrogel–PDMS hydrogel–PDMS composites: stress vs.vs. nominal strain curves for hydrogel–PDMS composites specimens (blue solid line with full symbols and black composites specimens (blue solid lineline with fullfull symbols andand black strain curves for for hydrogel–PDMS strain curves hydrogel–PDMS composites specimens (blue solid with symbols solid line with full symbols respectively represent hydrogel–PDMS composites with square solid line with respectively represent compositeswith withsquare square black solid linefull withsymbols full symbols respectively representhydrogel–PDMS hydrogel–PDMS composites inclusions and circular inclusions. Blue dot dash dot line and blue dot line depict the full model and inclusions and circular inclusions. Blue dash line and blue line depict full model inclusions and circular inclusions. Blue dotdot dash dotdot line and blue dotdot line depict thethe full model and RVE simulations for hydrogel–PDMS model with square inclusions; black solid line and black dash and RVE simulations for hydrogel–PDMS square inclusions; black solid black RVE simulations for hydrogel–PDMS modelmodel with with square inclusions; black solid lineline andand black dash line show the full model and RVE simulations for hydrogel–PDMS model withmodel circular inclusions); line the full model RVE simulations for hydrogel–PDMS withinclusions); circular linedash show theshow full model and RVE and simulations for hydrogel–PDMS model with circular (b) Experimental images of specimens from experiments and full simulation result at (b) Experimental of specimens from experiments andmodel full model simulation result (b) inclusions); Experimental images of images specimens from experiments and full model simulation result at macroscopic strain of 0.1, pattern transformation of RVE model at macroscopic strain of 0.12. at macroscopic strain of 0.1, pattern transformation RVE modelatatmacroscopic macroscopic strain strain of macroscopic strain of 0.1, pattern transformation of of RVE model of 0.12. 0.12.

As expected, the mechanical behavior of the hydrogel–elastomer composite material displays expected, themechanical mechanicalbehavior behavior of of the hydrogel–elastomer displays AsAs expected, the hydrogel–elastomercomposite compositematerial material displays almost linear elastic behavior at the initial stage with a change to a different elastic relationship at almost linear elasticbehavior behavior at at the initial stage with a change to atodifferent elastic relationship at the almost linear elastic the initial stage with a change a different elastic relationship the critical strain of εc. The patterns for the composite materials at the nominal strain of 0.1 are at critical strain of ε . The patterns for the composite materials at the nominal strain of 0.1 are shown c the critical strain of The12b. patterns the 11a, composite materialsproperties at the nominal of 0.1 are shown in Figures 11bεc.and From for Figure the mechanical of the strain hydrogel–TBP shown in Figures From Figure thewith mechanical properties of theHowever, hydrogel–TBP composite material11b in and the 12b. experiments match11a, well the numerical analysis. the composite material in the experiments match well with the numerical analysis. experimental results for hydrogel–PDMS composite material are not consistent withHowever, numericalthe experimental hydrogel–PDMS composite are not consistent with numerical analysis. The results reason isforthat material property of PDMSmaterial has changed when it was exposed to UV analysis. The reason is that material property of PDMS has changed when it was UV irradiation [43,44]. It should be noted that this often occurs in experiments and should exposed be taken to into irradiation [43,44]. It should be noted that this often occurs in experiments and should be taken into consideration in future research simulation work.

Polymers 2017, 9, 229

10 of 13

in Figures 11b and 12b. From Figure 11a, the mechanical properties of the hydrogel–TBP composite material in the experiments match well with the numerical analysis. However, the experimental results for hydrogel–PDMS composite material are not consistent with numerical analysis. The reason is that material property of PDMS has changed when it was exposed to UV irradiation [43,44]. It should be noted that this often occurs in experiments and should be taken into consideration in future research simulation work. Polymers 2017, 9, 228 10 of 13 Figure 13 also provides a comparison between the matrix materials and hydrogel–elastomer composite material. In stress vs. strain curves, soft cellular material and hydrogel–elastomer composites before the the pattern patternswitching. switching. Following Followingpattern pattern compositesboth bothexhibit exhibit elastic elastic properties properties before transformation, the stress plateaus emerge in soft cellular material; while the stress is dependent transformation, the stress plateaus emerge in soft cellular material; while the stress is dependenton strain in hydrogel–elastomer composites with thetheinclusions noted that thatthe the on strain in hydrogel–elastomer composites with inclusions plugging. plugging. ItIt is is noted deformation-triggered patterns of the composite materials depend on the deformed pattern of the deformation-triggered patterns of the composite materials depend on the deformed pattern of the matrix material; cantemporarily temporarily repair fractured ligaments and cracks matrix material;while whilethe thegel gel inclusion inclusion can repair fractured ligaments and cracks on theon thematrix matrixmaterial. material.

(a)

(b)

® Figure Comparison cellular structure structure and and Figure13.13. (a) (a) Comparisonbetween betweenthe the TBP TBP (TangoBlackPlus (TangoBlackPlus®)) cellular ® ® composites (red solid line with full symbols and black solid line with full hydrogel-TangoBlackPlus hydrogel-TangoBlackPluscomposites (red solid line with full symbols and black solid line with symbols respectively represent hydrogel–TBP composites with square inclusions and circular full symbols respectively represent hydrogel–TBP composites with square inclusions and circular inclusions; solidline line with open symbols and thesolid black line symbols with open symbols inclusions;the thered red solid with open symbols and the black line solid with open respectively represent the TBP cellular structure withstructure square voids circular voids; thecircular red dot voids; dash dot and respectively represent the TBP cellular withand square voids and theline red dot thedot black solid linethe depict thesolid full model simulation for the TBP model with square circular dash line and black line depict the full model simulation for thevoids TBPand model with voids); (b) Comparison cellularbetween structure the andPDMS the hydrogel–PDMS composites square voids and circularbetween voids); the (b) PDMS Comparison cellular structure and the (blue solid linecomposites with full symbols andline black solid with and full symbols, respectively, represent hydrogel–PDMS (blue solid with fullline symbols black solid line with full symbols, hydrogel–PDMS composites with square inclusions with and circular solid inclusions; line with respectively, represent hydrogel–PDMS composites square inclusions; inclusions the andblue circular open symbols and the black solid line with open symbols, respectively, represent the PDMS cellular the blue solid line with open symbols and the black solid line with open symbols, respectively, structure with square voids and circular voids; the blue dot dash dot line and the black solid line depict represent the PDMS cellular structure with square voids and circular voids; the blue dot dash dot the full model simulation for the PDMS model with square voids and circular voids). line and the black solid line depict the full model simulation for the PDMS model with square voids and circular voids).

In the case of hydrogel–PDMS composite material (shown in Figure 14a), the material is again deformed because there is covalent crosslinking between the PDMS and14a), hydrogels in the robust In the case of hydrogel–PDMS composite material (shown in Figure the material is again interface. As the gel inclusions deswell at room temperature, the side of the specimen buckles and deformed because there is covalent crosslinking between the PDMS and hydrogels in the robust the buckling mode is evident in Figure 14b. As there is no covalent attachment between TBP and interface. As the gel inclusions deswell at room temperature, the side of the specimen bucklesthe and hydrogel, the phenomena thus did not exist in hydrogel–TBP composites.

the buckling mode is evident in Figure 14b. As there is no covalent attachment between TBP and the hydrogel, the phenomena thus did not exist in hydrogel–TBP composites.

In the case of hydrogel–PDMS composite material (shown in Figure 14a), the material is again deformed because there is covalent crosslinking between the PDMS and hydrogels in the robust interface. As the gel inclusions deswell at room temperature, the side of the specimen buckles and Polymers 2017, 9,mode 229 of 13 the buckling is evident in Figure 14b. As there is no covalent attachment between TBP11and the hydrogel, the phenomena thus did not exist in hydrogel–TBP composites.

(a)

(b)

Figure Figure 14. 14. Hydrogel–PDMS Hydrogel–PDMS composite compositematerial materialbuckles buckleswith withthe thedeswelling deswellingof ofthe the gel gel inclusions inclusions at at room temperature. room temperature.

4. Conclusions The aim of this paper was to progress the investigations into the mechanical properties of soft cellular material and the associated designed composite material through experimental and numerical analyses. We first investigated the deformation-triggered patterns on soft cellular materials with different material properties. The experimental and numerical results indicate that the transformation event is highly dependent on the boundary edge conditions and material properties of the matrix material. Next, we fabricated novel hydrogel–elastomer composite materials using two approaches by assembling the gel inclusions into the periodic elastomeric cellular structures under ultraviolet or thermal irradiation. From our experiments, we fortuitously found that gel inclusion can temporarily repair fractured ligaments and cracks on the matrix material, such that the initial rupture of the matrix material does not affect the mechanical properties of the hydrogel–elastomer material. This finding may provide future perspectives for optimal design or serve as a fabrication suggestion for new gel composite materials. Acknowledgments: This work was supported by the National Natural Science Foundation of China (11372236 and 11572236). Author Contributions: Jianying Hu carried out modeling/simulation and experimental works, analyzed the data and drafted manuscript. Yu Zhou carried out modeling/simulation, and experimental works, analyzed the data and drafted manuscript. Zishun Liu initiated the project, guided the project, analyzed data, drafted manuscript and finally approved the manuscript. Teng Yong Ng analyzed data and drafted manuscript. Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3. 4. 5. 6.

Friis, E.A.; Lakes, R.S.; Park, J.B. Negative Poisson’s ratio polymeric and metallic foams. J. Mater. Sci. 1988, 23, 4406–4414. [CrossRef] Grima, J.N.; Alderson, A.; Evans, K.E. An Alternative Explanation for the Negative Poisson’s Ratios in Auxetic Foams. J. Phys. Soc. Jpn. 2005, 74, 1341–1342. [CrossRef] Pozniak, A.A.; Smardzewski, J.; Wojciechowski, K.W. Computer simulations of auxetic foams in two dimensions. Smart Mater. Struct. 2013, 22, 084009. [CrossRef] Williams, J.L.; Lewis, J.L. Properties and an anisotropic model of cancellous bone from the proximal tibial epiphysis. Trans. ASME J. Biomech. Eng. 1982, 104, 5–56. [CrossRef] Evans, K.E.; Alderson, K.L. The static and dynamic moduli of auxetic microporous polyethylene. J. Mater. Sci. Lett. 1992, 11, 1721–1724. [CrossRef] Caddock, B.D.; Evans, K.E. Microporous materials with negative Poisson’s ratios. I. Microstructure and mechanical properties. J. Phys. D 1989, 22, 1877–1882. [CrossRef]

Polymers 2017, 9, 229

7. 8. 9. 10. 11. 12. 13.

14. 15. 16. 17. 18.

19. 20. 21.

22. 23. 24. 25. 26. 27.

28.

29. 30.

12 of 13

Wang, L.; Lau, J.; Thomas, E.L.; Boyce, M.C. Co-Continuous Composite Materials for Stiffness, Strength, and Energy Dissipation. Adv. Mater. 2011, 23, 1524–1529. [CrossRef] [PubMed] Evans, K.E.; Alderson, K.L. Auxetic materials: The positive side of being negative. Eng. Sci. Educ. J. 2000, 9, 148–154. [CrossRef] Shan, S.; Kang, S.H.; Zhao, Z.; Fang, L.; Bertoldi, K. Design of planar isotropic negative Poisson’s ratio structures. Extreme Mech. Lett. 2015, 4, 96–102. [CrossRef] Hughes, T.P.; Marmier, A.; Evans, K.E. Auxetic frameworks inspired by cubic crystals. Int. J. Solids Struct. 2010, 47, 1469–1476. [CrossRef] Chen, Y.J.; Scarpa, F.; Liu, Y.J.; Leng, J.S. Elasticity of anti-tetrachiral anisotropic lattices. Int. J. Solids Struct. 2013, 50, 996–1004. [CrossRef] Gaspar, N.; Smith, C.W.; Alderson, A.; Grima, J.N.; Evans, K.E. A generalised three-dimensional tethered-nodule model for auxetic materials. J. Mater. Sci. 2011, 46, 372–384. [CrossRef] Alderson, A.; Alderson, K.L.; Attard, D.; Evans, K.E.; Gatt, R.; Grima, J.N.; Miller, W.; Ravirala, N.; Smith, C.W.; Zied, K. Elastic constants of 3-, 4- and 6-connected chiral and anti-chiral honeycombs subject to uniaxial in-plane loading. Compos. Sci. Technol. 2010, 70, 1042–1048. [CrossRef] Babaee, S.; Shim, J.; Weaver, J.C.; Chen, E.R.; Patel, N.; Bertoldi, K. 3D soft metamaterials with negative Poisson’s ratio. Adv. Mater. 2013, 25, 5044–5049. [CrossRef] [PubMed] Yuan, S.; Shen, F.; Bai, J.; Chua, C.K.; Wei, J.; Zhou, K. 3D soft auxetic lattice structures fabricated by selective laser sintering: TPU powder evaluation and process optimization. Mater. Des. 2017, 120, 317–327. [CrossRef] Willshaw, S.; Mullin, T. Pattern switching in two and three-dimensional soft solids. Soft Matter 2012, 8, 1747. [CrossRef] Yang, D.; Jin, L.; Martinez, R.V.; Bertoldi, K.; Whitesides, G.M.; Suo, Z. Phase-transforming and switchable metamaterials. Extreme Mech. Lett. 2016, 6, 1–9. [CrossRef] Yang, D.; Mosadegh, B.; Ainla, A.; Lee, B.; Khashai, F.; Suo, Z.; Bertoldi, K.; Whitesides, G.M. Buckling of Elastomeric Beams Enables Actuation of Soft Machines. Adv. Mater. 2015, 27, 6323–6327. [CrossRef] [PubMed] Bertoldi, K.; Reis, P.M.; Willshaw, S.; Mullin, T. Negative Poisson’s Ratio Behavior Induced by an Elastic Instability. Adv. Mater. 2010, 22, 361–366. [CrossRef] [PubMed] Bertoldi, K.; Boyce, M.C. Mechanics of the hysteretic large strain behavior of mussel byssus threads. J. Mater. Sci. 2007, 42, 8943–8956. [CrossRef] Bertoldi, K.; Boyce, M.C.; Deschanel, S.; Prange, S.M.; Mullin, T. Mechanics of deformation-triggered pattern transformations and superelastic behavior in periodic elastomeric structures. J. Mech. Phys. Solids 2008, 56, 2642–2668. [CrossRef] Mullin, T.; Deschanel, S.; Bertoldi, K.; Boyce, M.C. Pattern Transformation Triggered by Deformation. Phys. Rev. Lett. 2007, 99, 084301. [CrossRef] [PubMed] He, Y.; Guo, S.; Liu, Z.; Liew, K.M. Pattern transformation of thermo-responsive shape memory polymer periodic cellular structures. Int. J. Solids Struct. 2015, 71, 194–205. [CrossRef] Xia, Z.; Zhang, Y.; Ellyin, F. A unified periodical boundary conditions for representative volume elements of composites and applications. Int. J. Solids Struct. 2003, 40, 1907–1921. [CrossRef] Li, J.; Shim, J.; Deng, J.; Overvelde, J.T.B.; Zhu, X.; Bertoldi, K.; Yang, S. Switching periodic membranes via pattern transformation and shape memory effect. Soft Matter 2012, 8, 10322. [CrossRef] Xiao, R. Modeling Mismatch Strain Induced Self-Folding of Bilayer Gel Structures. Int. J. Appl. Mech. 2016, 8, 164004. [CrossRef] Liu, Y.; Zhang, H.; Wang, J.; Zheng, Y. Anisotropic Swelling in Fiber-Reinforced Hydrogels: An Incremental Finite Element Method and Its Applications in Design of Bilayer Structures. Int. J. Appl. Mech. 2016, 8, 1640003. [CrossRef] Zhang, Y.; Matsumoto, E.A.; Peter, A.; Lin, P.-C.; Kamien, R.D.; Yang, S. One-step nanoscale assembly of complex structures via harnessing of an elastic instability. Suppl. Inf. 2008, 8, 1192–1196. [CrossRef] [PubMed] Lee, J.-H.; Wang, L.; Boyce, M.C.; Thomas, E.L. Periodic Bicontinuous Composites for High Specific Energy Absorption. Nano Lett. 2012, 12, 4392–4396. [CrossRef] [PubMed] Bertoldi, K.; Boyce, M.C. Mechanically triggered transformations of phononic band gaps in periodic elastomeric structures. Phys. Rev. B 2008, 77, 052105. [CrossRef]

Polymers 2017, 9, 229

31. 32. 33. 34. 35. 36. 37. 38. 39.

40. 41.

42. 43.

44.

13 of 13

Wang, K.; Chang, Y.-H.; Chen, Y.; Zhang, C.; Wang, B. Designable dual-material auxetic metamaterials using three-dimensional printing. Mater. Des. 2015, 67, 159–164. [CrossRef] Hu, J.; He, Y.; Lei, J.; Liu, Z.; Swaddiwudhipong, S. Mechanical behavior of composite gel periodic structures with the pattern transformation. Struct. Eng. Mech. 2014, 50, 605–616. [CrossRef] Michel, J.; Lopezpamies, O.; Pontecastaneda, P.; Triantafyllidis, N. Microscopic and macroscopic instabilities in finitely strained porous elastomers. J. Mech. Phys. Solids 2007, 55, 900–938. [CrossRef] Mullin, T.; Willshaw, S.; Box, F. Pattern switching in soft cellular solids under compression. Soft Matter 2013, 9, 4951. [CrossRef] Hong, W.; Liu, Z.S.; Suo, Z.G. Inhomogeneous swelling of a gel in equilibrium with a solvent and mechanical load. Int. J. Solids Struct. 2009, 46, 3282–3289. [CrossRef] Flory, P.J.; Rehner, J., Jr. Statistical Mechanics of Cross-Linked Polymer Networks I. Rubberlike Elasticity. J. Chem. Phys. 1943, 11, 512–520. [CrossRef] Li, J. Mechanical Behavior of Hydrogels for the Use of Ionic Devices; Harvard University: Cambridge, MA, USA, 2015. Yuk, H.; Zhang, T.; Parada, G.A.; Liu, X.Y.; Zhao, X.H. Skin-inspired hydrogel–elastomer hybrids with robust interfaces and functional microstructures. Nat. Commun. 2016, 7, 12028. [CrossRef] [PubMed] Kang, S.H.; Shan, S.; Noorduin, W.L.; Khan, M.; Aizenberg, J.; Bertoldi, K. Buckling-Induced Reversible Symmetry Breaking and Amplification of Chirality Using Supported Cellular Structures. Adv. Mater. 2013, 25, 3380–3385. [CrossRef] [PubMed] Jang, J.-H.; Koh, C.Y.; Bertoldi, K.; Boyce, M.C.; Thomas, E.L. Combining Pattern Instability and Shape-Memory Hysteresis for Phononic Switching. Nano Lett. 2009, 9, 2113–2119. [CrossRef] [PubMed] Singamaneni, S.; Bertoldi, K.; Chang, S.; Jang, J.-H.; Thomas, E.L.; Boyce, M.C.; Tsukruk, V.V. Instabilities and Pattern Transformation in Periodic, Porous Elastoplastic Solid Coatings. ACS Appl. Mater. Interfaces 2009, 1, 42–47. [CrossRef] [PubMed] Bertoldi, K.; Gei, M. Instabilities in multilayered soft dielectrics. J. Mech. Phys. Solids 2011, 59, 18–42. [CrossRef] Efimenko, K.; Wallace, W.E.; Genzer, J. Surface Modification of Sylgard-184 Poly(dimethyl siloxane) Networks by Ultraviolet and Ultraviolet/Ozone Treatment. J. Colloid Interface Sci. 2002, 254, 306–315. [CrossRef] [PubMed] Schnyder, B.; Lippert, T.; Kötz, R.; Wokaun, A.; Graubner, V.-M.; Nuyken, O. UV-irradiation induced modification of PDMS films investigated by XPS and spectroscopic ellipsometry. Surf. Sci. 2003, 532–535, 1067–1071. [CrossRef] © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).