sustranu71

Bio-inspired core structuresTitanium alloys

itanonaturre s

nite element model validates the structural performance and can further optimise the unit cell design.

ful forand haeros

mann et al. [8] highlighted the importance of additive manufactur-ing for aerospace industry, to produce bionic structuralcomponents with complex surfaces and internal features. SLMmanufacture allows the use of aerospace grade titanium alloyTi-6Al-4V for the core material. The excellent corrosion resistant

found that face-t (f2cc) ex] studied pand out-o

compression and shear. They compared analytical, numerimechanical test results for diamond and corrugated congurFurthermore they suggested that these structures are equivasquare honeycomb in longitudinal shear strength but lower intransverse shear and compression. George et al. [12] used CFRP lat-tice structures to enhance the performance of foam cores. Com-pression and shear results showed increase in strength andmodulus, while a major gain was achieved for the energy absorp-tion characteristic. Wicks and Hutchinson [13] demonstrated the

Corresponding author at: Joining Technology Group, Singapore Institute ofManufacturing Technology (SIMTech), 71 Nanyang Drive, Singapore 638075,Singapore. Tel: +65 6793 8378

E-mail address: feihs@simtech.a-star.edu.sg (S. Feih).

Composite Structures 118 (2014) 294302

Contents lists availab

Composite S

sevbre skin material [7].In recent years, the advancement of additive manufacture and

in particular the selective laser melting (SLM) technology has ledto alternative design possibilities for the sandwich core. Emmel-

lattice structures produced from SLM. It wascentred unit cell structures with vertical supporthe highest specic yield strength. Cote et al. [11ties of prismatic lattice structures for in-planehttp://dx.doi.org/10.1016/j.compstruct.2014.07.0360263-8223/ 2014 Elsevier Ltd. All rights reserved.hibitedroper-f-planecal andations.lent tomotive, a variety of cores including honeycomb, foam and balsawood cores are generally employed [13].

Advanced metallic core design has in the past been limited bytraditional manufacturing processes such as investment casting,forming and welding [4,5]. Aerospace sandwich structures gener-ally utilise aluminium honeycomb cores, despite of problems withpoor interface bonding between core and composite skins [6] andpossible problems with galvanic corrosion in the case of carbon

3D structural core category. Key performance parameters of a corematerial are its compressive and shear strength in relation to itsweight. The structural performance of truss and lattice structuresand their potential as a metallic core in sandwich panels has beenstudied by various research groups. The deformation behaviour ofa large variety of lattice structures was researched by Emmelmannet al. [8], Mines et al. [9], Rehme and Emmelmann [10] and Coteet al. [11]. Rehme and Emmelmann compared specic strength of1. Introduction

Sandwich structures are very userequiring low weight, high strengthFor standard industry applications inDesign charts show that the performance of the proposed titanium core in both compression and shear issuperior (strength) or equal (stiffness) to honeycomb cores for aerospace applications.

2014 Elsevier Ltd. All rights reserved.

structural applicationsigh energy absorption.pace, marine and auto-

properties and exceptionally high strength to density ratio of tita-nium alloys make them a prominent choice for high performanceaerospace applications. Titanium alloys can also easily be bondedto carbon bre composites.

Truss and lattice type core structures belong to the innovativeSelective laser meltingSandwich materials

internal truss angles of less than 60 without requiring additional support structures. Mechanical testingis conducted to determine the deformation and failure of the core structure in compression and shear. APerformance of bio-inspired Kagome trusunder compression and shear loading

I. Ullah a,b, J. Elambasseril a, M. Brandt a, S. Feih a,b,c,aAdvanced Manufacturing Precinct, RMIT University, GPO Box 2476, Melbourne 3001, Ab Sir Lawrence Wackett Aerospace Research Centre, School of Aerospace, Mechanical & Mc Joining Technology Group, Singapore Institute of Manufacturing Technology (SIMTech),

a r t i c l e i n f o

Article history:Available online 2 August 2014

Keywords:

a b s t r a c t

Additive manufacture of tachievable through traditi4V Kagome truss core structures. These bio-inspired co

journal homepage: www.elcore structures

aliafacturing Engineering, RMIT University, GPO Box 2476, Melbourne 3001, AustraliaNanyang Drive, Singapore 638075, Singapore

ium structures allows the realisation of advanced design strategies notl manufacturing methods. This work analyses the performance of Ti-6Al-es produced by selective laser melting (SLM) for composite sandwich struc-tructures can be manufactured for truss diameters larger than 0.6 mm and

le at ScienceDirect

tructures

ier .com/locate /compstruct

This study focuses on the deformation and failure of Kagomestructures made from Ti-6Al-4V under the two main loading con-ditions of compression and shear. The demonstration of superiorproperties for titanium Kagome truss structures manufacturedwith a technique capable of large scale panel production foraerospace applications has not yet been demonstrated in theliterature. Furthermore, while deformation characteristics havebeen successfully predicted, past studies have not included the fail-ure validation of Kagome truss structures under various loadingconditions. SLM is utilised to produce the structural Kagome coresfor sandwich composites. The design concept for a typical

ite sandwich structures with SLM core.

I. Ullah et al. / Composite Structures 118 (2014) 294302 295Fig. 1. Design of advanced compossuperior crush resistance of truss cores compared to honeycombmaterials. In addition to a superior high strength-to-weight ratio,truss and lattice type core structures have the advantage of beingmultifunctional: they enable superior blast wave mitigation, andenhanced thermal management [14] or uid ow [15].

A special class of 3D truss structure is the Kagome structure. TheKagome truss geometry is originally inspired from the rod-likeinternal structure of cancellous bone, and is therefore often classi-ed as a bio-inspired core material. Lee and Kang [16] analysedthe compressive behaviour of wire-woven Kagome structures withdifferent diameters, lengths andmultiple truss layers. It was shownthat Kagome structures performed better than aluminium foam andegg box structures in compressive strength and energy absorptioncapacity [16] and also had better energy absorbing characteristicsthan tetragonal truss structures, such as resistance to plastic buck-ling and deferred susceptibility to softening [17]. Wadley et al. [18]manufactured Kagome structures through investment casting andconcluded that these structures perform better than tetrahedral,pyramid and honeycomb structures with higher specic strength.

Fig. 2. Kagome truss cell as used for analysis with design parameters of diameter,truss length and cell height and internal angle.

Table 1SLM build parameters.

Parameter Value

Laser power 175 WLayer thickness 30 lmArea scan speed 710 mm/sEnergy density 68.5 J/mm3

Hatch type CheckerboardHatch spacing 120 lmSpot size 80 lmPlatform temperature 200 Caerospace sandwich component is shown in Fig. 1. The papersummarises the outcomes of the complete design process frommanufacture to testing, including numerical validation of thedeformation and failure processes. Kagome structures are com-pared against commonly used core structures with the help ofdesign charts. The optimum Kagome design parameters are identi-ed for the static loading cases of compression and shear.Fig. 3. (a) Preprocessing of test specimen for SLM production (b) specimen panelproduced from SLM showing Kagome truss and faces.

Fig. 4. Shear test set-up.

Fig. 5. (a) FE mesh of Kagome cell, locating three types of elements (b) radial cross-section (c) axial cross-section, locating three types of elements (Type 1:C3D4, Type2:C3D6, Type 3:C3D8R).

296 I. Ullah et al. / Composite Structures 118 (2014) 294302

chamber is lled with Argon to avoid material oxidation. Processparameters (laser power, scan type, and layer thickness) used inthis study are listed in Table 1. These parameters were recom-mended by the machine manufacturer SLM solutions and werefound to produce structures with low porosity of less than 0.3%as established through tomography. The bulk material has a mod-ulus of 110GPa and yield strength of 1028 MPa. The properties areweakly dependent on the build direction. An average failure strainof 9% is observed at failure. While the modulus and tensile strength

tructures 118 (2014) 294302 297I. Ullah et al. / Composite S2. Design and manufacturing

2.1. Kagome truss core design

The Kagome cell existing in a sandwich panel is illustrated inFig. 2. The Kagome structure has the advantage of isotropic proper-ties in the horizontal plane due to the evenly-distributed arrange-ment of the three trusses. A Kagome truss may be geometricallyparameterised by truss diameter d, internal angle of the trussstructure h, length of truss members l and cell height h. For thisstudy, the height h of the cell is xed to a height of 11.5 mm, whichis a typical core height in sandwich panels. The diameter andinternal angle of the truss structure play an important role in char-acterising the mechanical properties and are considered keyparameters in optimisation studies. The length l of the truss mem-bers is a dependent variable and will be controlled only by theinternal angle of the truss and cell height.

2.2. SLM processing

The SLM system used in this research is the SLM250HL with abuild volume of 250 250 350 mm equipped with a YLR-Fibre-Laser of 400 W maximum power in the continuous mode. The

Fig. 6. (a) Fracture locus limit curve and (b) tensile stressstrain data.compare well to forged Ti-6Al-4V components, the strain-to-fail-ure is reduced signicantly due to the change of the typical abmicrostructure to a martensitic structure due to the faster coolingrates obtained during SLM manufacture [19].

2.3. Manufacturing constraints for truss structures

In full scale panel construction the core structures are ideallyproduced with incorporated metallic faces as shown in Fig. 3(a)to allow easy joining of the core to lightweight face sheets typicallymade from carbon bre composites. In order to enable manufac-ture of a panel with face plates through SLM, these faces must beconstructed normal to the build platform. Fig. 3(b) shows thepre-processing in AutoFab software of Kagome compression spec-imens with support structure under face sheets.

The SLM manufacturing method has some limitation for theminimum diameter and internal angle for constructing the trussstructures in the horizontal conguration: The minimum diameterof trusses that can be accurately produced with the SLM systemand the current build parameters was found to be 0.5 mm; belowthis value the circular truss cross-section distorted from its originalshape. For the internal angle (h) of the Kagome truss structure, anupper limit of construction in SLM was found as truss membersbecome almost horizontal when the internal angle increases morethan 60.

3. Mechanical testing

Aerospace sandwich structures are prone to combinations ofcompression and shear loading conditions (either static or as aresult of impact scenarios). The performance of Kagome trussstructures was therefore evaluated for these two static loadingconditions. Six different sample congurations were built withtwo truss diameters (d = 0.6 mm and 1.2 mm) and three internalangles (h = 50o, 55o and 60o). For each sample conguration, threespecimens were tested.

3.1. Compression testing

Flatwise compression testing of single Kagome truss cells wasundertaken on a 10kN Instron mechanical testing system. Dis-placement controlled compression test speed was maintained at0.5 mm/min. Specimens were placed between at compressionplates, and an extensometer was attached between the two facesheets of the specimen to determine the compressive strain.

Table 2The coefcients of failure locus curve of Ti-6Al-4V.

Coefcients Ti-6Al-4V Billet Ti-6Al-4V SLM

D1 0.164 0.1D2 0.292 0.1D3 1.376 1.3D4 0.052 0.03D5 0.461 0.2D 1.89 1.156D7 1.853 2.0gT 0.49 0.458

was 0.05 s for up to 6 mmof displacement (50% compressive strain),and automatic mass scaling was used to maximise the stable timeincrements of 1e7s with minimal dynamic effects.

The mesh quality and element type was found to be a criticalfactor for accurate results. Linear elements with lumped massmatrix were required for the explicit analysis. A change of elementtype along the outer surface of the truss towards the centre junc-tion was found to lead to minor stress jumps at the transitionpoints, which triggered failure at that location. The mesh of thetruss structure was therefore modelled with three different ele-ment types shown in Fig. 5 to achieve mesh-independent failurepredictions. The truss members were modelled with tetrahedralelements (C3D4) on the outer surface and triangular prism(C3D6) and brick (C3D8R) elements at the centre of the trusscross-section. In the case of a compression FE model for 0.5 mmtruss diameter and 55o angle a total of 200,000 elements weregenerated for converged results. A similar meshing technique isutilised for the shear model.

The failure of ductile metals is strongly dependent on the stresstriaxiality as established by previous research studies [20,21]. Bao

tructures 118 (2014) 294302298 I. Ullah et al. / Composite S3.2. Shear testing

A unit cell with two truss structures separated by a face sheetwas designed to allow the symmetric transfer of shear load. Thetest conguration is shown in Fig. 4(a). The specimen face sheetson either end were glued to steel plates and mounted on the testxture (Fig. 4(b)). The centre plate was loaded downward, whileside plates were kept xed in horizontal and vertical directions.This set-up introduces horizontal movement constraints duringthe shear test, but the constraint forces remain small. The down-ward displacement was measured with an attached LVDT.

4. Finite element model

An explicit model with nonlinear material and geometry is usedfor the following predictions in Abaqus 6.11. The explicit dynamicsolution was found to run in a more stable manner when predictingfailure of the truss structures. An elasticplastic material model isused in the nite elementmodelwith the properties given in Section2.2. Additionally, ductile fracture and progressive damage materialbehaviour is used in the analysis, hence requiring solid continuumelements. The time period used for simulations in Abaqus/Explicit

and Wierzbicki [20] demonstrated that failure in metal specimen

Fig. 7. Deformation behaviour of single Kagome truss under compression for (a)1.2 mm truss diameter and 55 angle and (b) 0.5 mm diameter and 60 angle.Circles on load-displacement trace indicate the point of truss failure in the test andnumerical model.Fig. 8. (a) Deformation behaviour of 0.5 mm diameter and 50 angle double cellKagome truss comparison of test and FE under shear (b) test specimen (c) FE model.

tructI. Ullah et al. / Composite Scan occur in three main modes: ductile void growth, pure shear ora mixed mode void growth. The strain-to-failure depends on thestress triaxiality. The distinctive regions are illustrated inFig. 6(a). Giglio et al. [22] used a set of three equations with sevendifferent coefcients (D1D7) to completely describe the failurestrain under multi-axial stress conditions:

ef

D113g D2 1=3 < g 6 0 Shear failure

D3g2 D4g D5 0 < g 6 gT Mixed modeD6 D7eD6g gT 6 g Ductile failure

8>>>>>:

1

Fig. 9. Ti-6Al-4V Kagome trusses as compared to honeycomb and foam core family obenchmark honeycomb and closed-cell foam materials is obtained from CES EduPack20ures 118 (2014) 294302 299where g represents the stress triaxiality, and gT is the transitionpoint between mixed mode and ductile failure stress triaxiality.The graph with dashed red curve shown in Fig. 6(a) illustrates themodel presented for conventionally produced Ti-6Al-4V as estab-lished by Giglio et al. [22]. Specimens manufactured with SLM havea signicantly lower ductility due to their martensitic microstruc-ture. In order to construct a failure envelope for the SLM producedTi-6Al-4V, model calibrations were conducted in a similar mannerto Giglio et al. [22] with the available tensile test results and oneset of Kagome truss shear specimens (0.5 mm, 50). The resultingcoefcients are summarised in Table 2. A ductile damage modelwith damage evolution was applied in Abaqus to capture thisbehaviour accurately.

f structures for (a) compressive strength and (b) compressive modulus. Data for12.

truct300 I. Ullah et al. / Composite S5. Results and discussions

Force and displacement were recorded for each compressionand shear test. To obtain results for strength and moduli for shearand compression loading, the following data analysis wasundertaken:

rmax Fmax=Aface 2E r= with e Dh=h: 3

For the given structures, the face area, Aface, was equal to thecross-sectional area of Kagome unit cell and depended only onthe truss angle. For the linear strain conversion, the current coreheight during deformation was divided by the original height ofthe truss core structures, which for all congurations was11.5 mm as outlined previously.

Fig. 10. Ti-6Al-4V Kagome trusses as compared to honeycomb and foam core familyhoneycomb and closed-cell foam materials is obtained from CES EduPack2012.ures 118 (2014) 294302Fig. 7 shows the stressstrain graphs for the compression testsfor examples of both thick and thin truss structures. The predic-tions were very accurate for all test cases for peak load, deforma-tion characteristics and failure location. The experimental elasticstiffness was within 510% of the numerical predictions. In thepresented work, the experimental stiffness under compressionloading was found to be very sensitive to non-parallel face skinsand deviations from the cylindrical geometry of truss members,both of which can occur during manufacturing. Additional machin-ing of top and bottom skin surfaces was undertaken followingmanufacture to minimise the effect of non-parallel face skins.

As expected, the maximum strength and stiffness increase sig-nicantly with an increase in truss diameter from 0.5 mm to1.2 mm (7.4 times for strength and 4.5 times for stiffness). Failurewas found to occur consistently in the regions of highest bending

of structures for (a) shear strength and (b) shear modulus. Data for benchmark

tructdeformation in the individual truss elements as highlighted in theimages of the deformed specimens. The numerical simulationsreveals that the plastic strains in this part of the structure weresignicantly higher than in the surrounding structure during com-pression. The failure mode is ductile void growth as determined bythe failure model. It is also of interest to note that failure for thinstructures occurs at a larger compressive strain (0.25 as comparedto 0.14), indicating that thinner structures are able to withstand ahigher compressive displacement as the resulting maximumsurface strains in thin trusses are lower during bending. This is val-idated by the experimental results. This nding is important whenconsidering the optimum structural parameters for energy absor-bance (area under forcedisplacement curve) during impactscenarios.

Fig. 8(a) shows the shear behaviour for truss structures undershear deformation. Numerical shear load simulations can also cap-ture the load bearing capacity and deformation accurately for thisload type when compared to mechanical tests. The maximumstress under shear is signicantly lower than under compression.

The three key deformation and failure characteristics are high-lighted in Fig. 8(b). Location A shows buckling of the rst truss atpeak load, while location B shows buckling of the second trussmember. Location C in Fig. 8(a) corresponds to the location of ten-sile fracture of the rst truss on the stressstrain curve, which isfollowed by an abrupt load drop. The general trend of this curveis captured well with the numerical model. The increase in stressin mechanical testing prior to reaching point C is observed due toa slight tilting of the positioning xture, which was not incorpo-rated in the model. Increasing the truss diameter from 0.6 to1.2 mm increased the stiffness by a factor of 4 and the strengthby a factor of 7.4.

During the experimental tests, various failure locations werefound, such as (1) joint failures, (2) failures close to the joints,and (3) failure of truss/plate connections (see Fig. 8(b)). Failurepredictions in the numerical model occurred at the joint connec-tion of the truss elements (see Fig. 8(c)); however, the other exper-imental failure locations also showed high plastic strains in thenumerical simulation, indicating a potential failure location. Thecentre joint thickness most likely increased slightly during manu-facture, which is not considered in the numerical model. Similarly,the connection between the truss elements and face sheets will beprone to defects (pores). The stress triaxiality at failure is close tozero (near pure shear onset) in the failure model.

The good agreement between numerical and experimentalresults indicates the usefulness of numerical simulations for theexploration of the Kagome truss design space. For additional designresponses, truss diameters between 0.5 and 1.3 mm diameter andangles between 45 and 60 were explored and compared to con-ventional aluminium honeycomb core structures for aerospaceapplications as well as foam and metallic core structures. Thedesign limits are derived through manufacturing constraints(internal angle) and engineering considerations for diameter to cellheight ratio.

The results for compression loading can be seen in Fig. 9. For thesame core density, the Kagome truss core structure outperformsthe conventional core structures signicantly for peak strength.The stiffness for the same density is comparable to honeycombstiffness, but signicantly higher than for polymeric and metallicfoam core structures. The geometric parameter limits are indicatedin the design charts. Kagome structures under compression per-form best for the largest diameters and largest internal angles.

The results for shear deformation can be seen in Fig. 10. Forshear strength, the Kagome truss core again outperforms honey-

I. Ullah et al. / Composite Scomb structures, and metallic and polymeric closed cell foammaterials. Regarding shear stiffness, the Kagome truss structureoutperforms all types of foam, and shows a similar stiffness to[1] Hall DJ, Robson BL. A review of the design and materials evaluation programfor the grp/foam sandwich composite hull of the ran minehunter. Composites1984;15:26676.

[2] Johnson AF, Sims GD. Mechanical properties and design of sandwich materials.Composites 1986;17:3218.

[3] Herrmann A, Zahlen P, Zuardy I. Sandwich structures technology incommercial aviation. In: Thomsen OT, Bozhevolnaya E, Lyckegaard A, editors.Sandwich structures 7: advancing with sandwich structures andmaterials. Netherlands: Springer; 2005. p. 1326.most honeycomb cores. For shear optimisation, Kagome trussstructures perform best under shear for smallest internal anglesand largest diameters. In general proposed Kagome core structuresare located in upper left area of the compressive and shear strengthcharts (Fig. 9(a) and Fig. 10(a)), which show that they have highestspecic strengths among the competitors.

6. Conclusions

Bio-inspired Kagome truss core structures were tested to failurein compression and shear, and the deformation characteristics andfailure could be successfully predicted with nonlinear explicit niteelement analysis incorporating a ductile failure metal criterion.

Numerical analysis was then utilised to create design charts for awide range of truss core parameters (diameters between 0.5 and1.3 mm and internal angles between 45 and 60). For these param-eters, core structures could be built without including additionalsupport structures. Minimising support structures is important to(a) minimisemanufacturing cost and (b) for the viable manufactureof core structural panels combining a large number of unit cells.

The Kagome truss core structures were found to perform betterthan conventional honeycomb aerospace core structures in termsof their specic strength for compression and shear for all param-eters analysed. The core stiffness under these loading conditionsremained similar to honeycomb structures.

It is shown that the optimisation of the Kagome truss parame-ters depends on the load case. The internal angle of the truss struc-ture can be tailored to perform best for shear (close to 45) orcompressive loading (close to the 60 limit due to manufacturingconstraints). The internal angle of a larger sandwich panel withnumerous truss structures may also be changed depending onthe trusses location within the panel. The diameter of the trussescan be maximised locally or uniformly for a given weight require-ment of the panel.

These ndings make SLM manufactured titanium truss corestructures a viable replacement for honeycomb cores in futureaerospace applications. SLM manufacture of core structuresfurthermore has the added advantage of better bonding to compos-ite skins due to the integrated face sheets and the possibility ofincorporating additional through-thickness reinforcement for thecomposite skins in the forms of z-pins.

Acknowledgments

The authors acknowledge the efforts of Mr. Aaron Pateras, formanufacturing of SLM structures and Mr. Peter Tkatchyk, for pro-viding support during mechanical testing. The CT-Scanning for thiswork was done at the South Australian node of the AustralianNational Fabrication Facility under the National CollaborativeResearch Infrastructure Strategy to provide nano- and micro-fabri-cation facilities for Australias researchers. One of the authors, I.Ullah, acknowledges the support of the Australian Governmentthrough the Endeavour Scholarship for this research project.

References

ures 118 (2014) 294302 301[4] Chiras S, Mumm DR, Evans AG, Wicks N, Hutchinson JW, Dharmasena K, et al.The structural performance of near-optimized truss core panels. Int J SolidsStruct 2002;39:4093115.

[5] Lee YH, Kang KJ. A wire-woven cellular metal: part-I, optimal design forapplications as sandwich core. Mater Des 2009;30:443443.

[6] Balasko M, Svab E, Molnar G, Veres I. Classication of defects in honeycombcomposite structure of helicopter rotor blades. Nucl Instrum Meth Phys ResSect A-Accelerators Spectrometers Detectors Associated Equipment2005;542:4551.

[7] Fischer P, DeLuccia J. Effects of graphite-epoxy composite materials on thecorrosion behavior of aircraft alloys. In: Symposium on Environmental Effectson Advanced Composite Materials 1975, Montreal, Quebec, p. 95.

[8] Emmelmann C, Petersen M, Kranz J, Wycisk E. Bionic lightweight design bylaser additive manufacturing (LAM) for aircraft industry. 2011, p. 80650L-80650L-12.

[9] Mines RAW, Tsopanos S, Shen Y, Hasan R, McKown ST. Drop weight impactbehaviour of sandwich panels with metallic micro lattice cores. Int. J. ImpactEng. 2013;60:12032.

[10] Rehme O, Emmelmann C. Rapid manufacturing of lattice structures withselective laser melting. Laser-based Micropackaging 2006;6107. art. no.61070K, K1070-K1070.

[11] Cote F, Deshpande VS, Fleck NA, Evans AG. The compressive and shearresponses of corrugated and diamond lattice materials. Int J Solids Struct2006;43:622042.

[12] George T, Deshpande VS, Sharp K, Wadley HNG. Hybrid core carbon bercomposite sandwich panels: fabrication and mechanical response. ComposStruct 2014;108:696710.

[13] Wicks N, Hutchinson JW. Performance of sandwich plates with truss cores.Mech Mater 2004;36:73951.

[14] Wadley HNG. Multifunctional periodic cellular metal. Philos Trans R Soc A-Math Phys Eng Sci 2006;364:3168.

[15] Joo JH, Kang KJ, Kim T, Lu TJ. Forced convective heat transfer in all metallicwire-woven bulk Kagome sandwich panels. Int J Heat Mass Transfer2011;54:565862.

[16] Lee BK, Kang KJ. A parametric study on compressive characteristics of wire-woven bulk Kagome truss cores. Compos Struct 2010;92:44553.

[17] Hyun S, Karlsson AM, Torquato S, Evans AG. Simulated properties of Kagomeand tetragonal truss core panels. Int J Solids Struct 2003;40:698998.

[18] Wadley HNG, Fleck NA, Evans AG. Fabrication and structural performance ofperiodic cellular metal sandwich structures. Compos Sci Tech2003;63:233143.

[19] Brandt M, Sun S, Leary M, Feih S, Elambasseril J, Liu Q. High-value SLMaerospace components: from design to manufacture. In: Subic A, Advances inEngineering Materials, Product and Systems Design 2013, 633, p. 13547.

[20] Bao Y, Wierzbicki T. On fracture locus in the equivalent strain and stresstriaxiality space. Int J Mech Sci 2004;46:8198.

[21] Bai YL, Wierzbicki T. A new model of metal plasticity and fracture withpressure and Lode dependence. Int J Plast 2008;24:107196.

[22] Giglio M, Manes A, Vigano F. Ductile fracture locus of Ti-6Al-4V titanium alloy.Int J Mech Sci 2012;54:12135.

302 I. Ullah et al. / Composite Structures 118 (2014) 294302

Performance of bio-inspired Kagome truss core structures under compression and shear loading1 Introduction2 Design and manufacturing2.1 Kagome truss core design2.2 SLM processing2.3 Manufacturing constraints for truss structures

3 Mechanical testing3.1 Compression testing3.2 Shear testing

4 Finite element model5 Results and discussions6 ConclusionsAcknowledgmentsReferences