Crack monitoring and failure investigation on inkjet printed sandwich structures under quasi-static indentation test

Accepted Manuscript Crack monitoring and failure investigation on inkjet printed sandwich structures under quasi-static indentation test

Vishwesh Dikshit, Arun Prasanth Nagalingam, Yee Ling Yap, Swee Leong Sing, Wai Yee Yeong, Jun Wei PII: DOI: Reference:

S0264-1275(17)30934-6 doi:10.1016/j.matdes.2017.10.014 JMADE 3412

To appear in:

Materials & Design

Received date: Revised date: Accepted date:

13 June 2017 21 September 2017 6 October 2017

Please cite this article as: Vishwesh Dikshit, Arun Prasanth Nagalingam, Yee Ling Yap, Swee Leong Sing, Wai Yee Yeong, Jun Wei , Crack monitoring and failure investigation on inkjet printed sandwich structures under quasi-static indentation test. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Jmade(2017), doi:10.1016/j.matdes.2017.10.014

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Crack Monitoring and Failure Investigation on Inkjet Printed Sandwich Structures Under Quasi-Static Indentation Test

PT

Vishwesh Dikshit 1, Arun Prasanth Nagalingam 1, Yee Ling Yap 1, Swee Leong Sing 1, Wai Yee Yeong 1,* and Jun Wei 2 1 Singapore Centre for 3D Printing, School of Mechanical and Aerospace Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Singapore; [email protected] (V.D.); [email protected] (A.P.N.); [email protected] (Y.L.Y.); [email protected] (S.L.S.) 2 Singapore Institute of Manufacturing Technology, 71 Nanyang Drive, Singapore 638075, Singapore; [email protected] * Correspondence: [email protected]; Tel.: +65-6790-4343

Abstract

D

MA

NU

SC

RI

In this research contribution, effort is taken to monitor the crack initiation and crack propagation of three-dimensional (3D) printed corrugated sandwich structures using acoustic emission technique. Vertical pillars were introduced in between the existing sinusoidal wavelike corrugations to improve the load bearing capacity of these structures. The vertical pillared corrugated structures were 3D printed with single and multi-material combinations in the facesheet and tested for their indentation resistance. To monitor the exact invisible crack initiation and crack propagation in the 3D printed corrugated structures, a highly-sensitive acoustic emission (AE) testing method was introduced. The resulting AE data points during testing illustrated a cluster of low amplitude data points from 40 to 65 dB indicating invisible crack initiations. High amplitude points up to 95 dB indicated visible cracks propagating until the end of specimen failure. Prevalent failure mechanism for single material (type A) specimens was found to be shear cracking of facesheets with micro steps and failure mechanism of multimaterial (type B) specimens were found to be delamination and shear cracking of multimaterial layers. Load bearing capacity was maximum at 2.14 ± 0.3 kN for type A specimens under a flat indenter with a displacement of 2.12 ± 0.45 mm.

1. Introduction

PT E

Keywords: Quasi-static indentation, 3D printing, Material printing, Sandwich structures, Indenters, Acoustic Emission.

AC

CE

Sandwich structures and corrugated panels have very high stiffness and load bearing capacity in the out-of-plane loading directions [1]. These structures are mostly manufactured using corrugation techniques. Corrugated structures are manufactured from various materials like cardboards, aluminum sheets and fiber sheets which include Kevlar, Nomex and aramid fibers. By including suitable facings on the top and bottom of these corrugated cores a new classification known as corrugated panels can be obtained. Top and bottom facesheets can be made from materials like aluminum sheets or fiber facesheets with different fiber layup configurations. These corrugated sandwich panels have high load bearing capacity and high stiffness due to the facings at top and bottom [1, 2]. These structures are used in marine, aerospace and heavy vehicle industries due to their light weight properties. Even though these structures have high stiffness properties under compressive loading, they are prone to damage when loaded locally. Optimized lightweight sandwich structures can be obtained by proper selection of core design, core material, facesheet layup configuration and facesheet material [3]. Apart from localized bending, the failure mode of corrugated sandwich structures include core crushing, core shear, facesheet delamination, facesheet wrinkling, fiber breaking in

ACCEPTED MANUSCRIPT

NU

SC

RI

PT

facesheet and fiber cracking due to improper stacking [4]. Upon impact of foreign objects on sandwich structures, their load bearing capacity and stiffness reduces until the point of failure. By proper understanding of the failure in sandwich structures, there is a high possibility to prevent the damage foreseen. Therefore, damage initiation maps were developed and tested experimentally, numerically [5] and by theoretical methods [6]. The predicted failure modes matched satisfactorily with experimental results. Reports show that the failure modes of sandwich structures under low-velocity impact tests are found to be similar under quasi-static indentation tests [7, 8]. Failure modes in relation to facesheet failure, facesheet wrinkling, and core failure were developed by analytical methods [9]. To overcome the major failure modes, polymeric foam cores were used to control the core crushing. However, these type of foam models showed shear failure [10-12]. Proper selection and layup of facesheet reduce the risk of shear failure by increasing the inertia of the whole sandwich structure [13, 14]. Apart from the selection of core and facesheet materials, the impact of foreign objects (indenter geometry in case of quasi-static indentation test) has an effect on the failure mode of sandwich structures. Sandwich structures show different failure modes under various indenter geometries. Tests were conducted apart from standard hemispherical indenter to study the failure modes under other indenter geometries like conical sharp tip and flat faced indenters [12, 15-17]. Thus, studying the effect of impacting object on sandwich structures is also considerably important.

AC

CE

PT E

D

MA

Core design plays major role in the load bearing capacity of sandwich structures [18]. Various core design like trusses, cellular cores, corrugated cores, Kirigami [19] and kagome cores [20] are developed to improve the load bearing capacity. However, these structures have their own advantages and limitations. Hence, designing an optimized core that satisfies all the requirements like high load bearing capacity, high energy absorption, stiffness properties is very difficult due to manufacturing process limitations. Despite the manufacturing limitations, efforts were taken to modify the core design to improve the mechanical properties of sandwich structures. Aluminum alloyed trapezoidal corrugated structures with facings on top and bottom were investigated under different indenter geometries. Matrix cracking with delamination was the major failure mode of these structures [21]. Investigations by modifying the kirigami structures based on egg box tessellated patterns resulted in 41 % energy absorption improvement compared to conventional corrugated structures [22]. Facesheet lamination using vacuum assisted resin transfer molding (VARTM) showed improvements in compressive strength of the structure [23]. Bi-directional corrugated cores were developed to reduce the anisotropic properties and to improve the bending properties. The results showed that these bi-directional corrugated sandwich structures showed quasi-isotropic bending in three-point bending tests [24]. Other core modifications such as corrugated Y-frames [25], variation in cell configuration [26], variation in core materials were done to improve the mechanical performance of corrugated sandwich structures. However, still there exists a need for an efficient manufacturing process to fabricate cellular cores for achieving the desired optimized configuration that suits the requirements. 3D printing also known as additive manufacturing (AM) has a greater advantage of manufacturing complex geometries at much lower costs than traditional manufacturing processes. A review on AM in unmanned aerial vehicles was done to prove the efficacy of AM components in aerospace industries [27]. Metal additive manufactured components can

ACCEPTED MANUSCRIPT

SC

RI

PT

be produced in net shape or near net shape. These components can be directly used for functional operations without further processing. Optimized sandwich structures having complex core designs that cannot be produced using conventional techniques have started to emerge into aerospace industries. Metrological benchmark studies on the printing capability of inkjet printing machine showed that the structures printed using PolyJet technology can have intricate structures without warpage or distortion [28]. Investigations on PolyJet printed sandwich structures showed shape recovery effects. The results show that these structures have high energy absorption characteristics and can recover their shape after bearing the loading [29]. Quasi-static indentation test of complex core structures like truncated pyramids was performed and verified using finite element analysis [30]. Introducing vertical pillars in trapezoidal sandwich structures improves their mechanical performance. These structures were subjected to quasi-static indentation tests under various indenter geometries. Structures printed using multi material combinations showed excellent load bearing capacity rather than components printed using a single material. Indenter geometry also plays a major influence on the failure mechanism of these structures [17]. Therefore, 3D printed sandwich structures are gaining popularity due to the manufacturing advantages and light weight properties.

CE

PT E

D

MA

NU

Damage to sandwich panels during operation is inevitable. Due to the complexity of core produced using 3D printing, there is a need for a novel technique to determine and monitor the crack initiation and crack propagation in these structures. Without a novel sandwich panel damage monitoring technique, it is very difficult to determine the damage occurred due to morphing actions or impact of foreign objects. In this contribution, efforts are taken to monitor the crack initiation and crack propagation of vertical pillared sinusoidal wave-like corrugated sandwich structures. This technique will help in monitoring even unobservable and hard to predict minor cracks in the structures. Vertical pillared sinusoidal wave-like corrugated sandwich structures were 3D printed using 3D Systems’ ProJet® MJP 5500X (3D systems, Rock Hill, CA, USA) using single material and multi-material combinations for the facesheet. The 3D printed structures were then subjected to quasi-static indentation tests under standard hemispherical, conical and flat indenters to determine the failure modes under each impacting object. Furthermore, results were discussed on load bearing capacity, the effect of facesheet material and failure mechanism in relation to acoustic emission monitoring data points.

2. Design, manufacturing and experiment methodology

AC

2.1 Specimen description

To improve the mechanical strength of the corrugated structure, vertical pillars are introduced at regular intervals of 10 mm. Design specification of the new vertical pillared corrugated structure is given in Table 1. These structures are like sandwich panels, the vertical pillared sinusoidal wave-like corrugated structure acts as the core, and top and bottom facings act as facesheets. The new vertical pillared sinusoidal wave-like corrugated structure will be analyzed for its indentation behavior.

ACCEPTED MANUSCRIPT Table 1: Design specification of the vertical pillared sinusoidal wave-like corrugated structure Nomenclature

Dimensions (mm)

A

100

B

16

C

1

PT

Geometry

E

SC

F

RI

D

12 100

MA

2.2 Manufacturing method and materials

10

NU

G

1

PT E

D

In this work, 3D Systems’ ProJet® MJP 5500X an inkjet 3D printing machine was used for fabricating the specimens. The advantage of this inkjet printing is that the accuracy of printing is about 27 [31]. Therefore, the specimens fabricated using inkjet printing will have higher dimensional accuracy compared to conventional sandwich structures. ProJet® MJP 5500X also offers VisiJet® CF-BK material, and VisiJet® CR-WT acrylonitrile butadiene styrene (ABS)-like material. These materials can be printed even as composites by inkjet fusion printing method. One such material combination obtained from fusing is VisiJet® RWT-FBK 400 flexible material. Material properties are tabulated in Table 2.

Material/Property

VisiJet® CF-BK VisiJet® RWT-FBK 400

Tensile Strength

Elongation at Break

Flexural Strength

(MPa)

(%)

(MPa)

1.04

56

8.1

74

1.04

2.2

290

0.5

1.04

10

25

7.3

Density (g/cc)

AC

VisiJet® CR-WT

CE

Table 2: Material properties of the fabricated specimens

Therefore, Sinusoidal wave-like corrugated structures with vertical pillars are 3D printed in the following types Type A: Top and bottom facesheet along with core are 3D printed using VisiJet® CR-WT material as a single structure. Type B: Top and Bottom facesheet is built in three layers, First layer-VisiJet® CR-WT, second layer-VisiJet® RWT-FBK 400, and third layer-VisiJet® CR-WT material

ACCEPTED MANUSCRIPT respectively. The core is built using VisiJet® CR-WT material. Even though it has material variation, the facesheet and core remain as a single structure.

SC

RI

PT

Initially, the specimens were modeled using SolidWorks software (Dassault Systèmes SOLIDWORKS Corp., Waltham, Massachusetts, USA). Figure 1 shows the modeling done in CAD software with a variation of material (white as CR-WT and Black as RWT-FBK) on the facesheet and the corresponding 3D printed specimens. The specimen’s dimensions are of 100 × 100 × 16 mm. The facings consist of 3 layers each of 1 mm in dimension at the top and bottom.

Type B - 3 layers

Type A - 3 layers

Layer 1- VisiJet® CR-WT material

VisiJet® CR-WT material

NU

Layer 2- VisiJet® RWT-FBK material

MA

Layer 3- VisiJet® CR-WT material

(b)

D

(a)

PT E

Figure 1: CAD model and 3D printed (a) Type-A specimen (b) Type-B Specimen

AC

CE

Figure 2 shows the process chain of 3D printing. The next step in 3D printing process chain is the data transmission and part building. The CAD models are imported and assigned the required materials and printed. As inkjet printing uses wax as support material, it must undergo a post processing step. The printed specimens were placed in an oven at 70 oC until all the wax was completely removed. The specimens were then removed, washed and dried off.

Figure 2: Process chain of 3D printing.

2.3 Experiment Methodology Indentation test was performed to analyze the localized deformation as the specimen is subjected to compressive loading locally at the indenter contact area. The type of loading is a localized out-of-plane load. Damage geometry and damage resistance under different indenter shapes can be studied by this indentation test. ASTM D6264/6264M standard test

ACCEPTED MANUSCRIPT procedure was followed for indentation testing [32]. In this test procedure damage resistance of the specimen is quantified in terms of critical contact force required to induce specific size and type of damage in the specimen. The test procedure for indentation consists of two conditions: (1)-testing the specimens in an edge supported condition and (2)-the rigidly backed condition. Edge supported condition is followed throughout the test to allow specimen deflection during testing.

NU

SC

RI

PT

Quasi-static indentation tests were carried out as shown in Figure 3. Three different types of indenters were used in the study. The standard hemispherical indenter is insufficient to study the damage characteristics on brittle 3D printed sandwich structures. Later two more indenters, namely flat end and conical sharp tip indenters, were introduced to study the effect of indenter geometries as shown in Table 3.

Figure 3: Test plan

MA

Therefore, after 3D printing, both the type A and type B specimens were subjected to quasistatic indentation under three indenters. Five specimens were tested under each indenter to ensure repeatability of the experiments and reliability of the results. Acoustic emission (AE) sensors were mounted along with the specimens to predict the exact cracking time.

D

Table 3: Indenter description and geometry

Tip

Diameter (mm)

2

3

Hemispherical

Conical sharp tip

Flat

CE

AC

Geometry (mm)

1

PT E

Indenter

13

50

13

Quasi-static indentation testing was done using SHIMADZU AG-X (SHIMADZU Corp., Kyoto, Japan) machine with loading capacity up to 10 kN. A custom-made aluminum fixture with 76.4 mm diameter opening and thickness of 40 mm was used for the edge supported condition as shown in Figure 4. Loading was done at a rate of 1.25 mm/min as per the standard.

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

MA

Figure 4: (a) Quasi-static indentation machine setup, (b) AE sensor mountings, (c) CAD model of fixture-edge supported test condition and (d) Cross section showing position of indentation.

CE

PT E

D

Four AE sensors were used to capture the invisible crack initiation and crack propagation mechanism during testing. The arrangement of AE sensors (two on top and two on bottom) on the setup are shown in Figure 4 (b). A semisolid lubricant grease is used underneath the sensors to ensure better acoustic wave reception. Acoustic emission sensors receive acoustic signals at the rate of 1 MHz from the cracks. The working principle of AE sensors are similar to non-destructive testing (NDT). Changes in the received signals will be recorded. The amplitude of data points recorded vary with each damage criteria. Therefore, crack initiation and crack propagation can be detected using the AE signals recorded. To establish the failure mechanism of specimens, video recording was done throughout the experiment. The specimens were cut into two halves to analyze the failure in the indented region along with frame by frame analysis of the video recorded and a representative model is provided.

AC

3. Result and Discussion Quasi-static indentation tests were performed on the 3D printed corrugated structures using displacement as a controlled phenomenon. The specimens were subjected to loading until failure by cracking or bottom facesheet indentation. Results are discussed based on the load bearing capacity and effect of 3D printed material on sinusoidal corrugated sandwich structures. Crack initiation and crack propagation signals were obtained from AE sensors. Experiment results Three possible modes of failure were noted in the specimens during indentation with conical indenter. These three modes of failures are the reasons for conical indenter displacement values in Table 4. (1) During bottom facesheet indentation – the indenter has penetrated through the bottom sheet (displacement ≥ 16 mm); (2) Before bottom facesheet indentation

ACCEPTED MANUSCRIPT (indenter is between top and bottom facesheet), as cracks propagated the specimens failed (crack propagation) at this position the indenter displacement will be (3 mm < displacement < 16 mm); and (3) During top facesheet indentation - after the indenter penetrated the top facesheet a crack developed reducing force value to zero, at this position displacement is less than or equal to the top facesheet thickness (displacement ≤ 3 mm).

Bottom facesheet damage area

(kN)

(mm)

mm2

mm2

Exposed area damage percentage (bottom facesheet) (%)

1.45 ±0.2

2.54 ± 0.4

133.3

793.7

52.8

1.21 ± 0.2

3.33 ± 0.5

134.1

1550.9

33.8

1.0 ± 0.1

11.44 ± 3.3a

140.7

1864.8

40.8

0.76 ± 0.2

16.70 ± 1.4a

155.1

215.4

4.6b

2.1 ± 0.3

2.12 ±0.4

132.5

1244.4

27.1

1.77 ± 0.1

2.57 ± 0.6

133.3

1520.0

33.1

SC

MA

A-Hemi Average ± SD B-Hemi Average ± SD A-Coni Average ± SD B-Coni Average ± SD A-Flat Average ± SD B-Flat Average ± SD

RI

Max force at crack initiation

Top facesheet damage area

NU

Specimen

Max displacement at crack initiation

PT

Table 4: 3D printed sandwich structure properties under indentation test.

a

CE

PT E

D

Identified modes of failure under conical indentation (1) Bottom facesheet failure; (2) crack propagation leading force values to reach zero; and (3) Sudden crack propagation during top facesheet indentation (Neither the specimen cracked into two halves nor the bottom facesheet was indented). b In this case, there were only cracks left over in the specimen’s bottom facesheet. Material elimination was not observed due to high energy absorption capacity of the specimens. It is to also note that the conical indenter penetrated an average of 16.70 mm, representing that indenter displacement is higher than the specimen total height of 16 mm. This represents to local bending in the specimen. As the force values reached zero, it was considered that specimen has failed and the experiment was terminated. This is the reason for very less damage percentage of type B specimen under conical indenter.

AC

3.1 Effect of indenter geometry in load bearing capacity Load bearing capacity of the structures are discussed based on indentation under various indenter geometries. Thus, a comparison of the load bearing capacity of sinusoidal wave-like corrugated sandwich structures for type A and type B specimens under three different indenter geometries are shown in Figure 5.

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

D

Figure 5: (a) Force-displacement curves for type-A specimen under different indenter geometries, (b) Force-displacement curves for type-A specimen type-B specimens, (c) Maximum load bearing capacity of type A specimens and (d) Maximum load bearing capacity of type B specimens

AC

CE

PT E

Quasi-static indentation curves for type A specimens under different indenters show that the load bearing capacity of the specimen is high under a flat faced indenter. The force acting on the specimen under a flat faced indenter is similar to pressure distributed over a surface area. Since force is distributed over the surface area of the indenter, the specimen can withstand a higher average force of 2.14 kN with local core bending. Under a conical sharp tip indenter, the specimen’s top facesheet failed immediately with a maximum average force of 1.06 kN with crack initiation and this crack propagated as the indenter penetrated further downwards. High sliding frictional resistance in-between the walls of indenter and specimen was noted during indenter penetration due to the conical geometry. The obtained average load bearing capacity of the specimen under a hemispherical indenter of 1.45 kN was lower than flat faced and higher than sharp tip conical indenters respectively. Therefore, from the difference in average force values obtained, it can be concluded that indenter geometry has a significant effect on the load bearing capacity of the structures. A similar response was observed for specimens with multi-material facesheet configuration. The load bearing capacity under hemispherical indenter seemed to be between flat faced and conical indenters. It should also be noted that the load bearing capacity of type A specimens is higher than type B specimens for the same indenter geometries. This difference is due to the elastic behavior of VisiJet® RWT-FBK 400 which is a slightly flexible material in the facesheet for type B specimens. As the indenters penetrated the second layer of type B specimens, the elasticity of results in a localized stretching in the form of bending. This, in turn, led to layer delamination and top facesheet failure. Hence, the resistance to indenter

ACCEPTED MANUSCRIPT penetration was low, leading to lower load bearing capability. The specimens could withstand a maximum average force of 1.21 kN, 0.76 kN and 1.77 kN when using hemispherical, conical and flat indenters respectively. Therefore, it is concluded that the facesheet material also plays a major role in the load bearing capability. 3.2 Crack initiation and crack propagation from acoustic wave signals.

SC

RI

PT

ProJet® MJP 5500X uses rigid plastic materials as shown in Table 2. These materials under mechanical loading fail suddenly without elastic deformation. Thus, it becomes highly difficult to predict their exact failure behavior. Therefore, a novel approach to monitor the exact crack initiation and crack propagation during a failure is required. AE data loggers are highly sensitive to detect even minor changes in their monitoring environment. Examples include geometrical changes, external entity interference, structural health monitoring of structures etc. The following sections describe the crack initiation and crack propagation signals obtained by coupling AE data loggers during quasi-static indentation testing.

AC

CE

PT E

D

MA

NU

Both quasi-static indentation testing and data loggers were turned on at the same time and experiments were carried out. Preliminary tests were conducted to determine the damage mechanism on top facesheet during indenter penetration. The highly sensitive sensors can detect even weak signals that indicate invisible crack initiation and crack propagation. Following observations were noted – A cluster of low amplitude data points from 40 to 45 dB represent unobservable cracks (crack initiation) and delamination in the top facesheet layers or visible localized indentation due to indenter penetration. Data points ranging from 50 to 75 dB represents visible crack initiation in the local indentations formed. These cracks were branching from the outer surface of localized indentation. During these crack initiations, AE data loggers recorded data points at 50 to 75 dB and 40 to 45 dB signals were accompanied with it after few microseconds. Lastly, large amplitude data points from 75 to 95 dB were recorded during crack propagation. At these data points, there was audible cracking noise from the specimens along with visible material elimination from the top facesheet. These high amplitude data points were accompanied with low amplitude data points from 40 to 75 dB after few microseconds. Thus, by matching the force vs. time response curves from quasistatic indentation loading and AE data signals, exact crack initiation and propagation in the specimen are mapped in Figure 6.

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

MA

Figure 6: Force vs. Time and AE data points

Type A structure

AC

CE

PT E

D

It is possible to confirm the specimen failure using peak amplitude mapping obtained from AE data loggers. In Figure 6(a) the clusters of low amplitude level (40 to 45 dB) signals denote that minor unobservable cracks were developing in the specimen due to indenter penetration. Therefore, at this stage, there was high resistance offered by the specimen to hemispherical indenter progression. As the crack propagated, force values decreased drastically denoting top facesheet failure. At this point, it should also be noted that amplitude levels increased for about 90 to 95 dB respectively denoting crack propagation. This increase in AE signals are due to the noise levels detected by AE data loggers during crack initiation and crack propagation. In Figure 6(c) the low amplitude data points in the range of 40-60 dB under conical indenter indicates small local failure at the sharp tip. As the indenter progressed and penetrated further, there was high resistance to indenter penetration around the circumference of the conical face. This led to continuous minor cracks in the corrugated structure followed by huge cracks. Therefore, the clustered data points in the range of 60 to 95 dB support that there was continuous internal cracking the specimen due to indenter penetration from 320th s until the specimen failed. In Figure 6(e) the low amplitude data points ranged from 45 to 50 dB and represent unobservable internal cracking in the core. Around 80th s, a high 95 dB noise level denotes top facesheet failure with crack propagation. Mostly all the specimens failed by cracking in two halves because of distributed pressure under the indenter surface. Type B structure The data points in Figure 6(b) shows a difference in AE data points from type A structure. Even before the top facesheet fully failed (full indenter penetration in top facesheet), the

ACCEPTED MANUSCRIPT

RI

PT

amplitude values started to increase until 95 dB. This denotes that the RWT-FBK layer inbetween the rigid CR-WT layers failed internally due to local bending and delamination. This was followed by top CR-WT layer cracking at 150th s followed by top facesheet failure at 200th s with a drop in force values from 1 kN to 0.4 kN. Figure 6(d) shows that in the case of multi material facesheet and sharp tip conical indentation, a similar clustering effect was obtained. Low amplitude clusters were obtained only after 40th s from the start of the experiment. This was due to the elastic deformation in the specimen as the indenter penetrated further, after which top CR-WT layers started to show visible crack and the cracks propagated throughout the specimen. This crack propagation is supported by the high amplitude noise level ranging from 60 to 95 dB from 180 to 200th s until the specimen fails. Figure 6(f) shows the AE data points obtained under flat indentation for type B material. The observations are similar corresponding to facesheet and core failure like previous cases, supporting that AE signal recognition can be used as a standardized crack monitoring technique.

SC

3.3 Indentation surface morphology

NU

Type A Structure

PT E

D

MA

Due to rigid material behavior, the top facesheet failed at an average indenter displacement of 2.54 mm. Percentage of damage observed during a failure is very important in determining the replacement of panel size after failure. Hence, only a low percentage of failure is desired during impact or indentation loading so that replacement cost can be reduced. Damage geometry in type-A specimens is presented in Figure 7(a). Damage shapes on the top facesheet were found to be circular and bottom facesheet damage shapes were circular, square and rectangular for hemispherical indenter. Images were calibrated using a 0.1 mm microscope calibration objective lens. Damage area was obtained using ImageJ (open source java image processing software)[33, 34] and percentage of damage is calculated and shown in Table 4.

AC

CE

Cracking of bottom facesheet was dominant under conical indenter among all other damage geometries like rectangular, square and circular failure geometry after indentation as shown in Figure 7(b). Bottom facesheet cracking was purely due to the influence of indenter geometry (sharp tip). Comparing to percentage damage under a hemispherical indenter the damage is low under a conical indenter, but it should also be noted that the visible cracks in the specimen is higher. Damage geometries under a flat faced indenter were mostly of square and triangular shapes as shown in Figure 7(c). Percentage of damage is least under a flat indenter.

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 7: Observed damage geometry for type A specimens

AC

CE

PT E

D

Field emission scanning election microscopy (FESEM) was performed on the cross section of a cracked specimen under standard hemispherical indenter. The failure was found to be faceted curved surfaces and river patterns as shown in Figure 8(a). The macroscopic crack growth direction due to indenter penetration and the faceting pattern due to micro stepped shearing in the material is clearly visible. These types of cracks are generally found in macroscopic shear conditions in brittle solids [35]. Higher magnification on the faceted steps in Figure 8(c), showed the microscopic failure was river line patterns. These river line directions were parallel to the general direction of crack propagation.

Figure 8: (a) FESEM analysis of failure mechanism in type A specimens, (b) riverline crack patterns, (c) faceted step in the fracture surface.

ACCEPTED MANUSCRIPT Type B Structure

AC

CE

PT E

D

MA

NU

SC

RI

PT

Damage geometry of type-B specimens are show in Figure 9. Under hemispherical indenter as shown in Figure 9(a) the damage geometry observed on the top facesheet was of circular shape, this was due to the indenter geometry. It should also be noted that unlike CR-WT material RWT-FBK materials bottom facesheet damage geometry was only circular, this may be due to the elastic stretching in the bottom facesheet during indenter penetration. The elastic stretching in the facesheet is supported by the results obtained such that for RWT-FBK materials, the average displacement at failure was about 3.33 mm higher than for rigid CRWT material structures respectively.

Figure 9: Observed damage geometry for type B specimens.

Similar to type-A specimens, under conical indenter as shown in Figure 9(b), cracking of bottom facesheet was the failure geometry observed in the case of multi material facesheet too. This proves that indenter geometry has a major role in the damage geometry observed during impact or indentation loading. A damage of only 4.69 % was noted for the damaged

ACCEPTED MANUSCRIPT specimen, as the dominant failure was by cracks in the bottom facesheet. This is due to the elastic stretching in the specimen which in turn led to the absorption of energy with a higher displacement of 11.44 mm in an average and resulting in lower damage percentage.

CE

PT E

D

MA

NU

SC

RI

PT

Damage geometry on the specimens under flat indenter was found to be of rectangular, triangular and circular shapes as shown in Figure 9(c). A damage percentage of 33.1% was noted in the exposed area. It is to note that the percentage damage is less in the case of flat faced indenter compared to other indenter geometries.

AC

Figure 10: (a) FESEM analysis of failure mechanism in type B specimens, (b) riverline pattern in layer 1 of facesheet, (c) interlaminar shear fracture in layer 2, (d) riverline pattern in layer 3 of facesheet.

Fracture cross-section of type B specimens were analyzed under FESEM and images are shown in Figure 10. CR-WT layers had faceted stepped cracks same as type A specimens. The microscopic damage was riverline patterns as shown in Figure 10(b) and Figure 10(d). These patterns were along the parallel direction of crack propagation. RWT-FBK 400 layer had an array of echelon cracks. These types of cracks due to shear stress during failure. This type of crack is particularly found in interlaminar shear fracture in laminations (facesheet laminated). These cracks were oriented in 45 degrees to the direction of shear failure as shown in Figure 10(c). Apart from these failure mechanism investigations, crack prediction technique was simultaneously performed during the experiment.

ACCEPTED MANUSCRIPT 3.4 Failure mechanism

PT

Identifying the failure mechanism is important to trace the drawbacks in the component and rectify it after failure. Failure mechanism during quasi-static indentation test will help to identify the limitation of a loaded structure. Effort is taken to identify the stages of failure in the newly design sandwich structure with vertical pillars. Video recording is done throughout the experiments to analyses the specimen failure. Rough trails were conducted to identify the failure mechanism after obtaining each data point in the AE screen. Stages of failure in the specimen are represented based on the video recording, AE data recorded and from the damage area morphology obtained from FESEM analysis.

AC

CE

PT E

D

MA

NU

SC

RI

Type A Structure

Figure 11: Stages of failure in type A specimens under three indenter geometries

ACCEPTED MANUSCRIPT

SC

RI

PT

The failure of type A sinusoidal wave-like corrugated sandwich structures is differentiated into three stages as shown in Figure 11. In case of hemispherical indenter, as it penetrated the top facesheet, shear failure was noted in the top facesheet along with center vertical pillar cracking and wall buckling in the nearby walls (Figure 11(a)). Noticeable crack was found in the sinusoidal wave-like corrugated structures and vertical pillars with further indenter penetration (Figure 11(b)). Finally, the specimen’s bottom facesheet failed by shear cracks along with local bending under the indenter contact area (Figure 11(c)). The specimen under conical indenter showed shear failure on top facesheet, but no noticeable wall buckling or cracking was found (Figure 11(d)). With further indenter penetration, sinusoidal wave and vertical pillars started to crack. During this stage, there was resistance to indenter penetration from the top facesheet in the form of sliding friction (Figure 11(e)). This was due to the small damage area created on top facesheet and the increase in indenter’s cross-section geometry along the vertical direction. The specimens failed with shear failure of bottom facesheet without any local bending (Figure 11(f)). Thus, the damage area on bottom facesheet created by sharp conical indenter is less compared to hemispherical indentation.

MA

NU

Under Flat indenter, during stage 1, there was local bending of top facesheet along with center vertical pillar cracking (Figure 11(g)). As the indenter penetrated further, top facesheet failed and noticeable cracking was found in the vertical pillars and sinusoidal wave structures along with bottom facesheet local bending (Figure 11(h)). Finally, the specimen failed with a sudden brittle crack in the vertical pillars and bottom facesheet (Figure 11(i)). On investigation, the bottom facesheet failed with shear cracking showing same behavior as that of hemispherical and conical indentation failures. Type B Structure

PT E

D

Type B structures showed slightly different behavior during failure. Due to the better elastic properties of RWT-FBK 400 Material in the facesheet, the specimen deflected entirely during stage 1 as that of beam deflection (Figure 12(a)). During this stage, the vertical pillars started to buckle without any fracture of cracks.

CE

With further indenter penetration, top facesheet failed with elongation and shear failure (Figure 12(b)). Sinusoidal wave, vertical pillars cracked and bottom facesheet failure was noticed with shear failure (Figure 12(c)). Under conical indenter, only the top facesheet showed deflection, there was no change in bottom facesheet configuration (Figure 12(d)).

AC

As the conical indenter penetrated further, the deflection of the structured increased and top facesheet failed by shear cracking. There was also necking and minor cracks noticed in the vertical pillars (Figure 12(e)). As the indenter made point contact with bottom facesheet, the sliding friction increased along with specimen deflection and bottom facesheet failed (Figure 12(f)). The indenter created only very less damage area during bottom facesheet indentation. Failure stages under flat indenter was similar to hemispherical indenter except in stage 2, there was cracking in bottom facesheet before even the indenter made contact with it. Material was removed and specimen failed as the indenter contacted bottom facesheet (Figure 12(i)).

CE

PT E

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

Figure 12: Stages of failure in type B specimens under three indenter geometries

4. Conclusion

A specially designed vertical pillared sinusoidal wave-like corrugated sandwich structure is inkjet printed using 3D systems’ InkJet printing technology with varying facesheet material combinations and tested for their load bearing capacity under quasi-static indentation loading. Some of the major findings in this research work include. 1.) Damage behavior of 3D printed corrugated structures varies under different indenter geometries. These results in the different load bearing capacity of the structures. Hence, corrugated structures have higher load bearing capacities under a flat faced indenter and poor load bearing capacities under conical sharp tip indenters. 2.) Material variation in the facesheet of corrugated structures has a major effect on the damage percentage of the structure. Type A specimens built with CR-WT rigid ABS-like

ACCEPTED MANUSCRIPT material showed brittle failure resulting in higher percentage of damage, while type B specimens with multi-material facesheet configuration have higher energy absorption characteristics leading to lower damage percentage. 3.) Crack monitoring approach using acoustic emission technique was used to monitor the crack initiation and crack prediction in 3D printed corrugated structures. A cluster of low amplitude data points indicated unobservable minor crack initiation and high amplitude data points levels indicated visible crack propagation in the specimen.

RI

PT

Thus, crack monitoring approach will overcome the existing limitation of failure detection/ crack prediction in intricate components manufactured by 3D printing. Future works on the 3D mapping of failure points using AE sensors will even aid in detecting the exact failure point in the sandwich structure making it easier for repair operations and prevent catastrophic failures.

SC

Acknowledgments: This work was supported under the A*STAR TSRP—Industrial Additive Manufacturing Programme by the A*STAR Science & Engineering Research Council (SERC) (Grant No. 1325504105).

[4]

[5]

[6]

[7]

[8]

[9]

[10]

MA

D

PT E

[3]

CE

[2]

B.T. Cao, B. Hou, Y.L. Li, H. Zhao, An experimental study on the impact behavior of multilayer sandwich with corrugated cores, International Journal of Solids and Structures 109 (2017) 33-45. C. Kılıçaslan, M. Güden, İ.K. Odacı, A. Taşdemirci, Experimental and numerical studies on the quasi-static and dynamic crushing responses of multi-layer trapezoidal aluminum corrugated sandwiches, Thin-Walled Structures 78 (2014) 70-78. C. Yuan, O. Bergsma, S. Koussios, L. Zu, A. Beukers, Optimization of Sandwich Composites Fuselages Under Flight Loads, Applied Composite Materials 19(1) (2010) 47-64. D. Sun, W. Zhang, Y. Zhao, G. Li, Y. Xing, G. Gong, In-plane crushing and energy absorption performance of multi-layer regularly arranged circular honeycombs, Composite Structures 96 (2013) 726-735. L. Gornet, S. Marguet, G. Marckmann, Numerical modelling of Nomex® honeycomb cores: Failure and effective elastic properties, in III European Conference on Computational Mechanics, (2006) 509-509. S. Zhu, G.B. Chai, Damage and failure mode maps of composite sandwich panel subjected to quasi-static indentation and low velocity impact, Composite Structures 101 (2013) 204-214. F. Zhu, G. Lu, D. Ruan, Z. Wang, Plastic Deformation, Failure and Energy Absorption of Sandwich Structures with Metallic Cellular Cores, International Journal of Protective Structures 1(4) (2010) 507-541. P. Chen, Z. Shen, J. Xiong, S. Yang, S. Fu, L. Ye, Failure mechanisms of laminated composites subjected to static indentation, Composite structures 75(1) (2006) 489495. E.W. Andrews, N.A. Moussa, Failure mode maps for composite sandwich panels subjected to air blast loading, International Journal of Impact Engineering 36(3) (2009) 418-425. P. Bollen, N. Quievy, C. Detrembleur, J.M. Thomassin, L. Monnereau, C. Bailly, I. Huynen, T. Pardoen, Processing of a new class of multifunctional hybrid for

AC

[1]

NU

REFERENCE

ACCEPTED MANUSCRIPT

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

PT

RI

SC

[17]

NU

[16]

MA

[15]

D

[14]

PT E

[13]

CE

[12]

AC

[11]

electromagnetic absorption based on a foam filled honeycomb, Materials & Design 89 (2016) 323-334. G. Caprino, M. Durante, C. Leone, V. Lopresto, The effect of shear on the local indentation and failure of sandwich beams with polymeric foam core loaded in flexure, Composites Part B: Engineering 71 (2015) 45-51. S.L. Cheng, X.Y. Zhao, Y.J. Xin, S.Y. Du, H.J. Li, Quasi-static localized indentation tests on integrated sandwich panel of aluminum foam and epoxy resin, Composite Structures 129 (2015) 157-164. F. Meraghni, F. Desrumaux, M. Benzeggagh, Mechanical behaviour of cellular core for structural sandwich panels, Composites Part A: Applied Science and Manufacturing 30(6) (1999) 767-779. F. Scarpa, G. Burriesci, F. Smith, B. Chambers, Mechanical and electromagnetic behaviour of auxetic honeycomb structures, Aeronautical Journal 107(1069) (2003) 175. M.H. G. Zhou, J. Loughlan, and N. Hookham, Damage characteristics of composite honeycomb sandwich panels in bending under quasi-static loading., Journal of Sandwich Structures and Materials 8 (2006) 55–90. C.D.M. Muscat–Fenech, J. Cortis, C. Cassar, Characterizing QSLVII Damage of Composite Sandwich Hulls, Procedia Engineering 88 (2014) 141-148. V. Dikshit, A.P. Nagalingam, Y.L. Yap, S.L. Sing, W.Y. Yeong, J. Wei, Investigation of Quasi-Static Indentation Response of Inkjet Printed Sandwich Structures under Various Indenter Geometries, Mater. 10(3) (2017) 290. B. Han, Z.-J. Zhang, Q.-C. Zhang, Q. Zhang, T.J. Lu, B.-H. Lu, Recent advances in hybrid lattice-cored sandwiches for enhanced multifunctional performance, Extreme Mechanics Letters 10 (2017) 58-69. Y. Hou, R. Neville, F. Scarpa, C. Remillat, B. Gu, M. Ruzzene, Graded conventionalauxetic Kirigami sandwich structures: Flatwise compression and edgewise loading, Composites Part B: Engineering 59 (2014) 33-42. B. Niu, B. Wang, Directional mechanical properties and wave propagation directionality of Kagome honeycomb structures, European Journal of Mechanics A/Solids 57 (2016) 45-58. J. Liu, W. He, D. Xie, B. Tao, The effect of impactor shape on the low-velocity impact behavior of hybrid corrugated core sandwich structures, Composites Part B: Engineering 111 (2017) 315-331. R.K. Fathers, J.M. Gattas, Z. You, Quasi-static crushing of eggbox, cube, and modified cube foldcore sandwich structures, International Journal of Mechanical Sciences 101-102 (2015) 421-428. H. Norouzi, Y. Rostamiyan, Experimental and numerical study of flatwise compression behavior of carbon fiber composite sandwich panels with new lattice cores, Construction and Building Materials 100 (2015) 22-30. D.Y. Seong, C.G. Jung, D.Y. Yang, K.J. Moon, D.G. Ahn, Quasi-isotropic bending responses of metallic sandwich plates with bi-directionally corrugated cores, Materials & Design 31(6) (2010) 2804-2812. L. St-Pierre, N.A. Fleck, V.S. Deshpande, The dynamic indentation response of sandwich panels with a corrugated or Y-frame core, International Journal of Mechanical Sciences 92 (2015) 279-289. X. Zhang, H. Zhang, Z. Wen, Experimental and numerical studies on the crush resistance of aluminum honeycombs with various cell configurations, International Journal of Impact Engineering 66 (2014) 48-59.

ACCEPTED MANUSCRIPT

[34] [35]

PT

RI

SC

[33]

NU

[32]

MA

[31]

D

[30]

PT E

[29]

CE

[28]

G.D. Goh, S. Agarwala, G.L. Goh, V. Dikshit, S.L. Sing, W.Y. Yeong, Additive manufacturing in unmanned aerial vehicles (UAVs): Challenges and potential, Aerospace Science and Technology 63 (2017) 140-151. Y.L. Yap, C. Wang, S.L. Sing, V. Dikshit, W.Y. Yeong, J. Wei, Material jetting additive manufacturing: An experimental study using designed metrological benchmarks, Precision Engineering 50 (2017) 275-285. Y.L. Yap, W.Y. Yeong, Shape recovery effect of 3D printed polymeric honeycomb, Virtual and Physical Prototyping 10(2) (2015) 91-99. V. Dikshit, Y.L. Yap, G.D. Goh, H. Yang, J.C. Lim, X. Qi, W.Y. Yeong, J. Wei, Investigation of out of plane compressive strength of 3D printed sandwich composites, in 37th Riso International Symposium on Materials Science, 139, (2016). H. Yang, J.C. Lim, Y. Liu, X. Qi, Y.L. Yap, V. Dikshit, W.Y. Yeong, J. Wei, Performance evaluation of projet multi-material jetting 3D printer, Virtual and Physical Prototyping 12(1) (2017) 95-103. ASTM D6264 / D6264M-12, Standard Test Method for Measuring the Damage Resistance of a Fiber-Reinforced Polymer-Matrix Composite to a Concentrated Quasi-Static Indentation Force, ASTM International, West Conshohocken, PA, 2012, www.astm.org. J. Schindelin, C.T. Rueden, M.C. Hiner, K.W. Eliceiri, The ImageJ ecosystem: An open platform for biomedical image analysis, Molecular Reproduction and Development 82(7-8) (2015) 518-529. C.A. Schneider, W.S. Rasband, K.W. Eliceiri, NIH Image to ImageJ: 25 years of image analysis, Nat Methods 9(7) (2012) 671-5. D. Hull, Fractography: observing, measuring and interpreting fracture surface topography, Cambridge University Press1999.

AC

[27]

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

SC

RI

Graphical abstract

ACCEPTED MANUSCRIPT

Crack Monitoring and Failure Investigation on Inkjet Printed Sandwich Structures Under Quasi-Static Indentation Test

Highlights

PT

RI

SC NU MA D PT E CE

 

Acoustic emission technique to monitor the invisible crack initiation during quasistatic loading. Highly scalable design of single and multimaterial inkjet printed sandwich structures. Failure mechanism and fracture morphology of inkjet printed sandwich structures under three different indenters.

AC

